Sunteți pe pagina 1din 194

petroff

defence

by A. Raetsky
& M. Chetverik

EVERYMAN CHESS
Gloucester Publishers pic www.everymanchess.com
First published in 2005 by Gloucester Publishers pic (formerly Everyman Publishers
plc), Northburgh House. 10 Northburgh Street. London EC1 V OAT

Copyright © 2005 Alexander Raetsky and Maxim Chetvcrik

The right of Alexander Raetsky and Maxim Chetverik to be identified as the authors
of this work has been asserted in accordance with the Copyrights, Designs and Pat­
ents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, electronic, electrostatic, magnetic
tape, photocopying, recording or otherwise, without prior permission of the publisher.

British Library Cataloguing-in-Publication Data


A catalogue record for this book is available from the British Library.

ISBN 1 85744 378 0

Distributed in North America by The Globe Pequot Press, P.O Box 480,
246 Goose Lane, Guilford, CT 06437-0480.

AU other sales enquiries should be directed to Everyman Chess, Northburgh House,


10 Northburgh Street, London ECtV OAT
tel: 020 7253 7887 fax: 020 7490 3708
email: info@everymanchess.com
website: www.everymanchess.com

Everyman is the registered trade mark of Random House Inc. and is used in this
work under license from Random House Inc.

EVERYMAN CHESS SERIES (formerly Cadogan Chess)


Chief advisor: Garry Kasparov
Commissioning editor: Byron Jacobs
General editor: John Emms

Typeset and edited by First Rank Publishing, Brighton.


Cover design by Horatio Monteverde.
Production by Navigator Guides.
Printed and bound in the US by Versa Press.
CONTENTS I

Bibliography 4
lntnxluction 5

1 e4 e5 2 lLif3 lLif6

1 3l£Jxe5: The Main Line with 8 c4 7


2 3l£Jxe5: The Main Line with 8 :.c1 34
3 3l£Jxe5: Black Plays 6.. ..id6 47
4 3l£Jxe5: Deviations from the Main Line 69
5 3l£Jxe5: Fourth and Fifth Move Alternatives for White 80
6 3 d4: The Main Line 104
7 3 d4: S .i.d6 and 5th Move Alternatives for Black
... 143
8 3 d4: Fourth Move Alternatives 158
9 3 d4: Black Plays 3...exd4 168
I0 Third Move Alternatives for White 178

Index of Complete Games 189


The Petroff Defen ce

BIBLIOGRAPHY I

Books
Ji'a.rhionable Varia/ion in tbe PetrrJ.ffDye11ce, IUctsky and Chetvcrik (Voronczh 1992)
Modi!Ji 1•aritmt msskoi parlii, IUetsky and Chctvcrik (Voronezh 1990)
Rltssisci.Je Ptlflie, Schwarz (Hamburg 19H6)
JI.Jtssischc f.'erfridigNI'!,. Konikowski and Sicbenhaar (Germany 1992)
The Cochrtmc G01nbil, Matsukevich (Moscow 1994)
TIJc PeiiT!ff,Janjgava (Gambit 2001)
The J>etrq!J Defence, l'orintos and Haag (Batsford 1991)
17:Je PetroffDefence, Yusupov (Oirns 1999)

Periodicals
Cl.�ess 11ifo1711ttnl 1-90
Nen' i11 Chess YeariJOok. 1-72

Software
Chess Assistant 7.1

4
INTRODUCTION I

'1'/Je e111irr lheol)• �/chm opmi11gs is diiJided i11/o lbe Pellf!lf D�Je11ce tmd the �jerted J>rllf!UDtjellce (1 d4, 1 t4
tllld other li11ts). Alexander Ractsky
-

TI1e Petroff Defence has been one of the most fashionable 'open games' in the last decade,
but it also has a long history. After 1 c4 e5 2 ltlf3, instead of defending his e5-pawn, Black
prepares a counterattack on White's e4-pawn with 2...ltlf6. J .ucena mentioned this idea as far
back as 1479, while in 1512 Damiano analysed 3 ltlxe5 ll'lxe4 and, of course, after 4 tfe2
reached conclusions that were discourah>i.ng for Black. No wonder the symmetrical opening
was forgotten for many centuries after that! However, in I H24 Russian maestro, Alexander
Petroff, found out that 3 ltlxcS could be met by 3 ...d6!, and only after that should Black cap­
ture the e4-pawn. In 1842 another Russian expert, Jaenisch, published valuable analysis in
l'ttltlfllede. The opening was given a name of the Petroff or the Russian Defence Qn Russia,
Germany and Scandinavia).
It is quite narural that two Russian masters write about the Petroff Defence for the pub­
lisher Everyman Chess. Just like ballet, the Russian Defence could be labelled as property of
the Russians. Grandmaster Arrur Yusupov is an outstanding thl:orctical expert of this opening,
while Russian champions Smyslov, Karpov and Kramnik often usc it in practice. Raetsky and
( :hctverik have not made a substantial contribution to the theory of the Petroff Defence so
far. However, it's worth pointing out that our first published articles and booklets were de­
vntL'l.l to this opening in the early 1990s. Alexander Raetsky has considered the Petroff De­
fence to be his f.1vourite opening for more than twenty years and can be proud, at least, of the
l(uantity of his b>amt."S played in the Petroff Defence (about 200, including 60 correspondence
1-�amcs), if not the quality.
Emotional players arc scared off by the S)•mmetrical trend in the Petroff Defence; they arc
afraid of the drawish aspirations of weaker opponents who play White. But what can be more
symmetrical than the initial position in chess, which has not yet been ruined by the notorious
'draw death' despite Capablanca's indications? A more skilful strategic player triumphs in the
Petroff Defence regardless of the colour of his pieces and the position's symmetry. The
healthy strategical foundation of this opening allows Black to defend a lot of systems in the
Petroff Defence even while playing against stronger opponents. Even the lines declared
'doubtful' by theory are normally better than their reputation.

5
The Petroff Defence

Apart from minor alternatives (see Chapter 10), White has to choose between 3 d4 and 3
'Oxe5. The authors believe that after 3 d4, the move 3.../0xe4 is stronger than 3...exd4. Here
the sharp variations like 4 dxe5 .i.c5!? and 4 .i.d3 d5 5 'Oxe5ltld7 6'0xd7 ..i.xd7 7 0-0 �4!?
can replace the popular, solid, but uninteresting alternatives (4 dxe5 d5 and 4 .i.d3 d5 5/0xe5
/Od7 6 'Oxd7 .i.xd7 7 0-0 .i.d6 respectively). After 3 'Oxe5 it is more difficult for Black to
initiate an open battle and the positional niceties of the fight come to the fore.
One final advantage for the fans of the Petroff Defence: you don't have to study numerous
complicated openings like the Ruy Lopez, Two Knights Defence, Italian Game and Scotch
Gamel After 2'0f3'0f6 you are home and dry!
Finally, some acknowledgements. Special thanks go to the founder of the Petroff Defence,
Mr Petroff, and many thanks to our Danish/Scottish friend Jacob Aagaard for his technical
help. Also thanks go to Zoya Nayshtut for her excellent translation into English and also to Mr
Yusupov for his great book on the Petroff Defence- the real bible for people from our 'cast'
(Petroff Defence players).
Play the Petroff Defence and be happy!

Alexander Raetsky & Maxim Chetverik,


Voronezh,
January 2005

6
CHAPTER ONE I
3ltJxe5: The Main Line
with 8 c4

1 e4 e5 2 �f3 l!Jf6 3 l!Jxe5 d6 4 l0f3 effective, but White maintains an trutJaove


li)xe4 5 d4 d5 6 .i.d3 .i.e7 7 0-0 �c6 8 after 9...tL'lb4, hassling the d3-bishop.
c4 The ...l'Llc6-b4 raid is more appropriate
i\Jter 1 e4 e5 2 tL'lf:\ tL'lf6 3 tL'lxe5 d6 4 when the other knight is still placed on c4,
li'!fJ tL'lxe4 5 d4 d5 6 .i.d3 Carl Jaenisch sug­ i.e. 8...tL'lb4 (Games 1-10). One option for
)l;ested the system 6....i.e7 7 0-0 tL'lc6 as far White is simply to ignore this attack: 8...tL'lb4
back as the 19th century. This line is still very 9 cxJ5 tL'lxd3 10 'ii'xd3 1i'xd5 II l:.c l .i.f5
popular today, especially with regard to the (Games 9-10). In these games White tries to
immediate attack on the centre with 8 c4, the take advantage of the hanging position of the
Kubject of this chapter. black pieces with tL'lb1-c3 �mmt:diately or
Let's see how this line has developed over after 12 g4 .i.g6) or 12 tL'le5.
the years. The oldest reply, 8....i.e6, is not Instead of 9 cxd5 White more often rc­
very good in view of 9 cxd5 .i.xd5 10 tL'lc3 trl..-ats his bishop to a safe place with 9 .i.e2
(hy the way, the li)wenthai-Morphy game (Games 1-8), a lint: that has bccome fashion­
ran be considered a model - sec the notes to able owing to Karpov. The line 9...dxc4 10
( lame 1l) or 9 l:te I l'Llf6 10 c5. After the .i.xc4 0-0 (Game 8) leads to a standard posi­
more agJ,rressive development of the bishop tion witl1 an isolated pawn on d4. Instead
with R ...i.g4, Whitc should prefer 9l'Llc3 to Black usually prefers to maintain pressure in
Kimplifying the play in the ccntre with 9 cxd5 the centre by means of with ...0-0 and ....i.e6
..xd5 to l'Llc3 tL'lxc3 11 bxc3 (again sec the or ....i.f5. However, the order of the moves
notes to Game 11). is important here. After 9....i.e6 the queen
With 8...tL'lf6 I31ack protects the d5-square exchange 10 ..a4+ 'ilt'd7 11 'ilt'xd7+ �xd7 12
and strengthens his kingside at the same l'Llc3 is interesting, and 10 c5!? followed by
time. In the case of 9 tL'lc3 .i.e6 (Game 1 1), an attack on the quccnside is even more so
White has a pleasant choice between 10 cxd5 (sec the notes to Game 6).
nnd to c5, so Black should refrain from the 9 ...0-0 is preferable to 9....i.e6 because af­
development of the cH-bishop. If 9...0-0 10 ter 10 l'Llc3 Black has several possible ways to
h1 Black appears to equalise by means of develop his light-squared bishop. In Game 7
IO dxc4 11 .i.xc4l'Lla5 12 .i.d3 .i.c6. If we
.•. Black plays 10 ...b6 followed by ....i.b7. One
start with 9 h3 (Game 12), castling is again of the main lines is 10 ....i.c6, against which

7
The Petroff Defence

White can play 1 1 !.DeS (Games 3-4) and 1 1 Pava.'IOvic, Istanbul Oly mpia d 2000 - and
.l.e3 (Games S-6). here Pavasovic claims an edge for Black after
The most fashionable line at the moment
16...exd3 17 Wb3 'ifd7) 1 S...d3 1 6 �3 dxc2
is 10....i.f5, which was utilised by Kramnik in
1 7 1i'xe2 .l.e6 1 8 llcl1 1Wc7, which Pavasovic
his 2004 World Championship match with
regards as unclear.
Leko. lbis move is discussed in Games 1 -2.
b) 1 1 �3 leads fairly dirccdy to a draw:
1 1 ...dxc4 12 .l.xc4/.Df6 1 3/.Des 'ifxd4 (Bolo-
Game 1 gan pointed out the error 13 ...1.Dc2? 14/.Dxf7
Leko-Kramnik llxf7 IS .i.xf7+ �f8 1 6 .l.e3 when White is
World Ch. (Game 1), Brissago 2004 much better) 14 .i.xf7+ Wh8 1 S I.Lif3 •d7 16
'------• /.Des ..d4 17 I.Lif3 'Wd7 was agreed drawn in
1 e4 e5 2 �f3 �f6 3 �xe5 d6 4 �f3 Bologan-Zamicki, Buenos Aires 2000.
Q)xe4 5 d4 d5 6 .i.d3 &6 7 0-0 J.e7 8 1 1 ...�c3 1 2 bxc3 &6
c4�9 J.e2
Of course saving the bishop is the most
popular option. Alternatives arc studied in
Games 9-1 0.
9 0-0 10 �c3 J.t5
.•.

1 3 lZ.e1
After 13 cxdS .xdS White has a few
choices:
a) 14 lle 1 llfe8 transposes to the main
game.
1 1 a3 b) 14 .i.e3 l'lfd8 1S :e1 &s 1 6/.Des f6!
This is the natural move but we should (16.....b3? is just a blunder: 1 7 .l.c41 .xd 1
also consider two others: 1 8 .i.xf7+ � fB 19l:taxdt /.Dc6 20 .l.ct i.LixeS
a) 1 1 �5 cS and now: 21 AxeS .i.g4 22 llde1 .i.f6 23 ll5e4 and
a 1) 1 2 dxcS is not a dear error but it does White has won a clear pawn, Movsesian-1
allow pressure on f2: 12..ixcS 13 I.DxciS P.Nielsen, Blmdesliga 1 999) 1 71.Dd3 i..d6 1 81
(White must be careful: if 1 3 a.1? then I.Db4 •n is unclear.
1 3...1.L!xf2! 1 4 llxf2/.Dc2 1 5 lla2 llc8 1 6/.Dd3 c) 14 c4 1i'd6 IS dSI.DeS t6/.Dd4 .i.d7 1 7
.l.xf2+ 17 I.Lixf2 dxc4 and Black has a clear a 4 c6!? 1 8 l:tb t (perhaps White shoula prefer
advantage) 13...lDxdS 14 ..xd5 'irxdS 1 5 1 8 'ifb3 cxdS 1 9 cxdS 1Wb4 20 'ifxb4 .l.xb4,
cxdSlLd8 16 i.c4 llfc8 t7/.Df3 h6 1 8 i..d2 when his d-pawn is marginal!)' more of a
I.Lixcl2 1 9 I.Lixd2 i.cl4 with an unclear posi­ strength than a weakness) 18 ...cxdS 19 cxd5
tion. ..xdS 20 I.Lif5 .c6 21 I.Lixe7+ ..xc7 22
a2) 12 a3 cxd4 13 I.Lixc4 dxe4 14 axb4 f6 llxb7 1We6 was level in Van Den Doei­
1 5 I.Dg4 (15 �t3?1 d3 1 6 .i.xd3 - J.Polgar- Schandorff, Esbjerg 200 1.

8
3 liJxe5: The Main Line with 8 c4

tl) 14 .if4lfr.ts 15 .i.xc7 b6 16 .i.f4 Lc8 22 g41 1Wf4 (even worse is 22. ..1!rd7?! 23
17 .i.t.12 .i.d6 18 llc1 h6 19 a4 :C7 20 h3 .i.xf7+! Wh8 - the point is 23...Wxf7 24
Wh3 21 .i.b5 (Krakops-Ulescas, Leon 2001). llxa51 llxa5? 25 1i'b3+ Wg6 26 llc6+ when
Now Black cook! have played 2t ...'exdl 22 Black will soon have to give up his queen -
l:lcxd1 �b3 23 L2 �xd2 24 �xd2 llxc3 24 ..ixd6 1i'xf7 25 ..tc5 and White was close
with equality. to victory in Karpov-Portisch, Torino 1982)
13 .. .1Z.e8 23 .ixd6 'irxd6 24 'ifo •d7 25 :C2lDc4 26
I\ simple, sensible developing move but llacl and White has a strong initiative, espe­
131ack has other options. cially since 26...1le8?? loses to 27 lle7!.
13....i.f6 is probably not qui te good b2) 17...i.e6 18 .i.xdi fxe6 19 �0 :aes
enough to cyualise: 14 .i.f4 �aS 1 5 cxdS 20 .ih4 1t'c6 21 1tc2 h6 and Black is just
WxdS 16 Wa4 (if 16 .i.xc7 :lacS 17 .i.xa5 active enough to hold et:juality: 22 .i.g3 1i'd5
Wxa5 18 c4 llfd8 Black has pleasant com­ 23 a4 �4 24 lle4 a6 25 1i'e2 1i'c6 26 lle1
pensation) 16...b6 t7lDd2 .i.d7 18 'irt.11 and .i.xg3 27 hxg3 tLki6 28 llxe6 1i'xc3 29 lieS
\Vhitc had an cdbJC in Short-Bologan, Skan­ 1i'b3 was now agreed drawn in 1\.dams­
tlcnborg 2003. This advantage grew after Anand, FlOE World Championship, New
IK.J:ac8 19 i.d3 .i.f5 20 �4 .i.e7 21 'irh5 Dell1i 2000.
i.xc4 22 1txd5 .i.xdS 2.l llxe7 llfe8 24 14 cxd5
Jl:ac I �f8 25 llxe8+ llxe8 26 llxe8+ Wxc8 14 .i.f4 is discussed in Game 2
27 .i.xc7 . 14 ....xd5 1 6 ..tf4
13...dxc4 14 .i.xc4 has been well tested: White has a minor alternative in 1S �e3
.i.f6 (:tlso fine is IS...�aS!? 16 �d2 'ird7 17
.to .i.d6 18 c4 c6 19 'lfa4 .i.e? with an un­
clear position) 16lDd2 .l:lad8 17 .i.c4 11d7 1 8
�b3 .i.e7 19 a4 .i.d6 20 'lfd2 (Bologan­
Degrnevc, Belfort 2002) ands here Bologa.n
gives 20...�7! 21 .i.f4 b6 as equal.
1 5.. .1Z.ac8

a) 14....i.f6 IS ..tf4 'ird7 16 ..ta2 llfe8 17


li.\g5!? (this is Ftacnik's suggestion; 17 'irt.12
bS 18 .i.gS .i.g6 19 .i.xf6 gxf6 20 �h4 Wg7
21 f4 toe7 22 ...f2 �f5 23 g4 �xh4 24
'l'xh4 f5 was unclear in Adams-Karpov,
Dortmund 1999) 17...llxet+ 181fxcl .i.xg5
IIJ i.xgS llc8 20 'ird2 h6 21 i.f4 tlX!s 22 d5
b5 and White has a nagging t:dge. 1 8 h3
h) 14. ..i.d6 1 S i.gS 'ird7 16 �h4 �a5 1 7 16 .i.d3 is a serious alternative. 16...1Wd7
.l.a2 with a further branch: 17 llbl .i.xd3 1 8 'irxd3 b6 19 d5 .i.f6 2 0 c4
b l ) Black should avoid 17... b5?1 18 a4! a6 h6 21 h3 lle7 22 llbd 1 lids (if 22...:Ce8 23
19 ••xb5 axb5 20 �f5 1fxf5 21 i.e7 llfb8 llxc7 �xe7 24 l:te1 lidS 25 .icS White has

9
The Petro ff Defence

the more comfortable game) 23 l:lxe7 lbxc7 .to •d7 20 1i'a4 lbe5 21 '1Vxd7 lbxf3+ 22
24 lbes .i.xe5 25 .i.xe5 l:le8 26 .i.g3 lbf5 27 lbxf3 ltxd7 23 lbe5 l:ld5 as in Leko­
.i.xc7! ..xc7 28 .xf5 'lrxc4 29 d6 and Bologan, Dortmund (rapid) 2004. Igor Zait­
White's strong passed pawn gave him the sev now suggests 24 c4 l:ldd8 25 lbxg6 fxg6
edge in Leko-Anand, Unares 2003. 26 d5 �fl 27 �ft l:ld7 28 l:lab 1 b6 29 ltb3
A few days after this main game Lcko as a way for White to gain an edge.
tried 16 c4 'lre4 1 7 .i.e3 and now:
a) 1 7...'ifc2!? (Kramnik's novelty) 1 8 d5
lbas 19 lbd4 •xdt 20 l:lcxdt (if 20 l:laxdt
Parkin analysed 20....i.d7 21 lbb5 b6!? 22
lbxa7 l:la8 2' lbb5 .i.xb5 24 cxb5 .i.xa3 as
equal) 20....i.d7 21 .i.d2 .i.f6 22 .i.xa5 .i.xd4
23 l:lxd4 l:lxe2 Vz-112 Leko-Kramnik, World
Championship (Game 3), Brissago 2004.
b) Also reasonable is 17 ....i.f6 18 l:la2 b6
1 9 h3lba5 with a further split
b1) 20 .i.d2 lbb7!? (20...�1?1 leads to a
very unpleasant ending: 21 'ifxb1 Lb1 22
l:lxb1 l:lxe2 23 'itft l:lce8 24lbg1 ll2e4 25
.i.xa5 bxa5 26 lle2, Kotronias-Marjanovic, 1 8 c4?!
Kallithea 2003) 21 'ifa4 .i.e6 22 d5 lbc5 23 This allows Black the chance to give up
1i'xa7 .i.d7 when White's awkwardly placed his queen for a definite equality. A better try
queen provides Black with compensation for for the advantage is 1 8 lbdz!? .i.f5
the pawn. (1 R.ixg2? may look worrying for White,
b2) 20 g4 .i.g6 21 g5 .i.c7 22 lbe5 .i.d6 23 but I. Zaitsev provided the clever refutation:
lbxg6 hxg6 24 c5 .i.ffi (Kasparov analysed 19 c4 .c6 20 d5 .g6 21 .i.hS .i.f3+ 22
24...bxc5 25 dxc5 .i.f4 26 .i.g4 llcd8 27 l:ld2 Lg6 .i.xdt 23 .i.f5) 1 9 .t.n 'lid7 20 lbb3
.i.xg5 28 .i.d7 lieS 29 l:ld4 as slighdy better lbxb3 21 1i'xb3 c6 when White has the mer�
for White) 25 .i.g4 l:lcd8 26 l:lae2 1i'c6! 27 est of edge£.
1i'c2 (27 cxb6 was immediately abandoned as 1 8...li»lc41 1 9 .i.xc4 1i'xc4 20 �d2 'ttd5
a draw in Anand-Adams, Unares 2002) 21 li»le4 ••e4 22 i.g5 1i'xe1 + 23 1i'xe1
27...bxc5 28 dxc5 •d5 29 1lrc3 lbc6 is equal .i.xg5 24 ••5 i.f& 25 •••7 c5 26 •xb7
according to Dokhoian. i.xd4 27 :S2
Brissago (3) 2004. Belov gives 27 l:tdt llb8 28 .d7 l:le2 29
1 6....i.e41? ltxd4 cxd4 30 11Vxc.J4 as equal.
Previously 1 6..if6 had been played: 1 7 27 ...c4
lbh2!? 'ifa5 1 8 .i.d2 lbc7 (or 1 8. ..l:lcd8 1 9 Black must avoid the back rank tric�1
.i. f3 h6 20 lbg4 .i.xg4 21 hxg4 .i.g5 22 27.. 1J..c7? 281le2!.
.i.xg5 hxgS - Kramnik-Anand, Wijk aan Zee 28 Jle2 lled8 29 e4?!
2003 - and here Hu?.man gives 23 .cl! White should have played 29 l:ld2 .i.e5 30
l:txet+ 24 'IVxel 'itffi 25 l:tbt lle8 26 .cl as £4! .i.f6 (not 30 ...i.xf4?! 31 llxd8+ llxd8 32
.

clearly better for White) 1 9 lbg4 .i.xg4 20 'ifc6 hS 33 'ifxc4 .i.g3 34 �ft :Cs 35 1rc3
Lg4 ltcd8 when Anand bclit.-ves White is h4 36 a4 when White has some winning
slighdy better. chances) 31 ltxd8+ l:lxd8 32 .c7 c3 33 a4 g6
1 7 .i.e3 c!lla 5 34 aS l:ld2 35 a6 lta2 36 �h2 c2 37 a7 after
Another try is 17 ...l:lcd8 18 lbd2 ..tg6 19 which the game would have been drawn.

10
3 li)xe5: The Main Line with 8 c4

46 .D.5a6 47 'tlt'c3 l:la4 48 •ce .D.8a6


•••

49 WeB+ Wg7 50 'tlt'b5 A4a5 51 1rb4


l:ld5 52 'tlt'b3 l:lad6 53 1i'c4 l:ld3 54 Wf2
l:la3 55 Wc5 l:la2+ 56 Wg3 Af6 57 Wb4
l:laa6 58 Wg2 l:lf4 59 Wb2 + l:laf6 60
'tlt'e5 Axf3 61 1i'a1 l:lf1 62 •ca A1f2+
63 Wg3 l:l2f3+ 64 'tlt'xf3 Axf3+ 65 Wxf3
Wf6 0-1

Game 2
Grischuk-Adams
Halkidiki 2002

29 . ..c3 30 We4 ..i.b6! 1 e4 e5 2 l£lf3 ll'lf6 3 ll'lxe5 d6 4 ll'lf3


Now Black can think about trying to win. ll'lxe4 5 d4 d5 6 ..td3 li)c6 7 0-0 ..i.e7 8
31 1t'c2 c4 ll'lb4 9 .te2 0-0 1 0 ll'lc3 .tf5 1 1 a3
White has to blockade: 31 Wb4?! .i.xf2+ ll'lxc3 1 2 bxc3 li)c6 13 l:le1 l:leB 14 .i.f4
12 'i&i>xf2 c2. The main altemative to 14 cxd5.
31 ...g6 32 11'b3 l:ld6
·n1e careless 32 ... fld2? would spoil cvcry­
lhin�-f. 33 llc7 llxf2 34 11'xb6 c2 35 llc71 and
White wins.
33 .D.c2 .ia5 34 g4 .D.d2 35 Wg2 .D.cdB 36
J:l.xc3! .ixc3 37 �i'xc3 .D.2d5 38 'tlt'c6 l:la5
39 'iii>g3 llda8 40 h4 ll5a6 41 Wc1 l:la5
42 ft6 .D.xa4 43 h5 .D.4a5 44 Wf47
A disastrous slip. Instead 44 hxj.,>6 hxj.,>6 45
11 g5 46 'ird6 and White hollis the draw.
44 .g51 45 'tlt'f6 h61
..

14 ...dxc4 1 5 ..i.xc4 ..i.d6 1 6 AxeS+ 'tlt'xe8


1 7 li)g5 .i.g6 18 .ixd6 cxd6 19 h4 •e7
Anand demonstrnteJ the following clever
variation: 19...M?I 20 h5 .i.xhS 21 'ti'xh5
hxg5 22 'iii'h21 llcB 23 lletlli:Je7 24 .i.d3 g6
25 WxgS llc7 26 llc3 and White is obviously
bener.
20 'tlt'g4 h6 21 ll'lh3
21 h5 has also been tried, but led only to a
draw in Movscsian-Gelfand, Bled Olympiad
2002 after 2t....i.xh5 22 WxhS hxg5 23 lldt
Now the win is just a matter of rime. J:lfB 24 .zld3 'ire 1+ 25 �h2 'lt'xf2 26 J:lh3
46 f3 Wf4+ 27 �g1 Wet+ 28 �h2 'iff4+.
The point is that 46 .xh6? ll8a6 traps the 21 1t'e4
•••

L lucen. This is okay, but an interesting possibility

11
The Petroff Defence

is 2t...1i'f6 22 l:tet �5 23 1Wf3 «i>ffl24 �f4 kingside pawns decide the game.
(Anand-Karpov, Prngue 2002). Anand now 30 ll:!f4 .i.h7 31 .ta2 �7 32 ll:!e6 llxf2
gives 24...1Wxh4!? 25 .i.xf7 1i'g4 26 'ifxg4 33 �xf2 .igS 34 �f8+ �es 35 .bg8
.i.xg4 27 .i.d.5 �e7 28 .i.xb7 l:b8 as offering �xf8 36 .idS!
Black enough compensation. White has kept the extra pawn and now
2211'g3 dominates the kniJ.,rht: Black's position is
Black can also defend against 22 �f4 �eS hopeless.
23 dxeS 1i'xc4 24 exd6 1i'xc3 25 l:tf1 .i.e4! 36...�e7 37 �e3 ¢>f6 38 ¢>f4ll:lc6 39
(not 25...D.d8? 26 �xg6 fxg6 27 d7 1i'c6 28 .i.xc6 bxc6 40 c4 �e6 41 g4 ¢>f6 42
l:tdl bS 29 hS! gxhS 30 .xhS when White g5+ hxg5+ 43 Wg4 1 -0
has a huge advantage, Mortensen­
M.Andersson, Sweden 2003) 26 �6 i.d3 27 Game 3
'ilfOI fxe6 28l:td1 1i'xa3 29l:txd3 Wet+ 30 Anand-Shirov
'it>h2 1i'c6 31 1i'xc6 bxc6 32 l:l.c3- the ruok Moscow2001
ending is drawn.
22 ll:!a5 23 .ia2 ...d3 24 llf1 •xg3 25
••• 1 e4 e5 2 ll:!f3 ll:!f6 3 ll:!xe5 d6 4 ll:!f3
fxg3 lieS 26 hS! .id3 ll:!xe4 5 d4 d5 6 .id3 ll:!c6 7 0-0 .ie7 8
Black avoids the obvious double attack: c4ll:lb4 9 .te2 0-0 1 0 �c3 .i.e&
26...i.xh5? 27 :5 b6 28 l:.xhS :Xd 29 l:lf5
when White should win.
27 .txf7 + ¢>f8
27...'it>h7?! is too passive: 28 l:.f3 ..ie4 29
l:.e3 d5 30 �f4 �4 31 l:.e2 �xa3 32 .ixd5
i.xdS 33 �d5 and White has good winning
chances.
28 llf2

1 1ll:le5
Apart from 11 .i.e3, which is studied in
Games 5-6, there arc two alternatives to note:
a) 11 .if4 cS 12 l:lc1 .i.f6 13 �bS? (this
expedition fails to some brilliant resources;
White should prefer 13 11'a4 �xc3 14 bxc3
�c6 15l:tab1 'ilfd7 with an unclear position)
13 ...dxc4 14 �7 �5! 15 �xc6 fxc6 16
28 . . .oh7? .i.g3 (Adams-Shirov, Dortmund 1998). Shi­
A fatal slip. After 28...l:.xc3 29 ..ig(rr �g8 rov now analysed 16...�xg3 17 hxg3 bS! HI
White has no more than a draw. a4 cxd4 19 �xd4 'tWb6! 20 �f3 (the lx:autiful
29 .td5 llf8 point is 20 �xbS? ..xf2+! 21 �xf2 .i.d4
The problem with 29...:Xd is 30 l:.f31 mate!) 20... .i.xb2 21 l:.bt �d! 22 ..c2 �xa4
J:ct+ 31 Wh2 i.c4 32 2:f7+ Wd8 33 .i.xc4 23 l:.xb2 �xb2 24 '1Vxb21lac8 when Black's
%lxc4 34 :Xg7 :Xd4 35 g4, when White's excellent pawns promise a clear advantage.

12
3 fDxe5: Th e Main Line with 8 c4

b) 11 a3 is safe but unthreatening. For ex­ plan. 15 f4 was tried in Anand-Lcko, Leon
nrnplc, 1 t...lL'lxc3 12 bxc3 tL'lc6 13 cxd5 2001, the game being level after 15...f6 16
.lxd5 14 c!Od2 tL'la5 15 .i.d3 b6 16 •c2 h6 tL'lg4 .i.xg4 17 .i.xg4 fS 18 .i.e2 .i.f6 19 l:tbt
17 .i.b2 .i.b7 18 lbe4 b5! 19 a4 (tf 19 .i.xb5 i.d4+ 20 Wh1 tL'lc1 21 b4 b6 22 l:tb3 lbes 23
t'5 20 "Wa4 c6 21 .i.c4+ lL'lxc4 22 •xc4+ •d5 i.e3 i.xe3 24 .l:txe3 llkt6 25 l:tb3 .f6 26
2.3 llk12 .i.g5 24 'lfd3 c5 Black's active bish­ bxc5 bxc5 27 11fa1 ..xa1 28 llxa1 Wfl.
ops provide ample compensation) 19...a6 20 15 f3!? is more enticing. Motylev analysed
uxb5 axb5 21 .i.xb5 fS 22 tL'lc5 .i.xcS 23 15...•c7 16 i.f4! i.d6 17 fxe4 l:tc8 (of
(lxcS •d5 24 c6 lL'lxc6 25 •d3 ll:le5 26 course the idea is 17..ixe5 18 d6) 18 .a4
Vxd5+ .i.xd5 with an equal position, Lcko­ l:tc7 19 tL'ld3 l:txe4 20 .i.xd6 .xd6 21 ..if3
Krnmnik, Dortmund 1999. with a slight edge for White.
1 1 ... c5
This early break gives White the chance to
ntlvance in the centre. The alternative 11...f6
is the subject of Game 4.

1 5...f6
Black threatens the knight at a moment
when it has no choice but to rctteat. Instead
15....i.d6 16 f4 f6 gives White the additional
1 2 l0xa4 option of 17 .xe4!? l:te8 (if Black takes the
If 12 .i.e3 cxd4 13 .i.xd4 lL'lxc3 14 bxc3 piece with 17... fxc5 18 fxe5 .l:te8 White has
li:k6 15 lL'lxc6 bxc6 16 .a4 cS 17 ..ie3 rl4 strong passed pawns and an attack as com­
Ulack equali.<;es comfortably acconling to pensation after 19 .i.d3 rfi 20 c6) 18 .i.h5
Yusupov. Ae7 19 i.fl+ (Morgado assessed 19 lL'lf7
12 ...dxa4 1 3 d5 .i.e& '1Vc7 20 lL'lxd6 l:txe4 21 lL'lxe4 as unclear)
Black must retreat since after 13.....id6?! 19...Wh8 20 tL'lg6+ hxrfi 21 'IVxrfi i.g4 22 fS!
14 a3 i.xeS 15 axb4 ..t£5 16 bxc5 White's l:txf7 23 Vxf7 .c7 24 ..xc7 ..ixc7 25 .tf4
pawns are too strong. Baklan-Timman, and Black's minor pieces cannot find any
Neum 2000 continued 16, ..•c7 17 g3 11fxc5 activity so White has an edge.
IK i.e3 •d6 19 c5 •f6 20 �3 ..Lb2 21 1 6�
J:a41 (Baklan pointed out that 21 lLb1 ..id4 The play is certainly complicated, but
22 'ifxb7 ..ih3 23 .l:tfd1 ..ixe3 24 fxe.l Wc3 is White seems to have the better of it.
just unclear) 21..ie5 22 11fxb7 ..th3 23 Act 1 6.....d6
aS 24 d6 and the pawns promise White an In Topalov-Shirnv, FIDE World Cham­
obvious advantage. pionship, Moscow 2001 Black tried 16..id6
14 a3 �6 1 5 •c2 17 f4 cxf3 (or 17....txg4 18 .txg4 fS 19 .i.e2
Anacking the pawn Js White's natural .te7 20 ..ie3 ..i.f6 21 l:tadt and White's ad-

13
The Petroff Defence

vantage is small but defmite) 18 i.xf3 'ifc7 tage with the calm 28 'ifa3.
19 g3 i.d7 20 liJt2 bS 21 b3 :ae8 22 'iPg2 26 .i.f4
i.eS 23 :b1 i.d4 24 i.d2 and White re­ Now this simple move is enough to con­
tained the usual nagging edge. finn a solid advantage.
1 7 f3 26 ...h6 27 .i.d6!
White tries to destroy his opponent's cen­
tre. Anand also analysed 17 g3 fS 18 i.f4
'ifb6 19liJe3 i.f6 (not 19...g5?1 20 i.eS f4 21
gxf4 gxf4 22 'iPh I! and it is White who at­
tacks along the g-file) 20 h4 ...xb2 21 'ifxb2
i.xb2 22 :a2 when White ha.c; some com­
pensation but no definite advantage.
17 f5
... 1 8 lbf2 .i.f6

Not 27 :et? ...xcl+ 28 liJxel :Xet+ 29


Wh2 gS and Black escapes with a draw.
27....:tf6 28 lbf4 •e4 29 ltle6 .:texe6 30
dxe6 •xe6 31 .i.g3 .i.xb2 32 .:l.e1 •f7
33 .i.h4 .:I.e& 34 .:Xe6 ••e6 35 •c2 .i.d4
36 •xf5 1i'xc4 37 Wh2 ••2
If Black tries 37...'ifct White wins with 38
'ifd5+Wh7 39 ...e4+Wg8 40.te71.
Black is more interested in the attack than 38 .i.g31
mere pawns. Covering eS from checks ensures the win.
19 fxe4 .i.e5 20 h3 .i.d4 21 e5!7 38 1i'd1 39 .:n "tlrb3 40 1ie4 1ib5
.•.

Similarly, White would rather return the If 40...i.f6 then 41 :xf6! gxf6 42 'ifg6+
pawn than allow the fB-rook into the game. forces mate. For example, 42...Wh8 43
21 .....xe5 22 Wh 1 .i.d7 'ifxf6+ Wh7 44 ...f5+ and the bishop will
Black's queen bishop settles for a modest soon join the attack with check.
square. 22 .f4 gains the £5-squarc but only at
.. 41 11Ve6+ Wh7 42 'iff5+ Wg8 4a
the expense of surrendering g4: 23 liJd3 i.£5 11Vc8+ �h7 44 .:tf8 1 -o
24 i.g4! (not 24liJxe5?! i.xc2 25 i.d3 i.xd3
26liJxd3 gS 27 h4 h6 28 hxgS hxgS 29 g3 bS, Game 4
when White's centre collapses) 24... g5 25 Leko-Grischuk
.id2 and White retains the initiative. Wijk aan Zee 2002
23 ltld3 .i.a4 24 •xa4 •xe2 25 .:l.f3
.:tae8 1 e4 e5 2 ltlf3 ltlf6 3 ltlxe5 d6 4 l£lf3
Again this is simple, sensible development. lbxe4 5 d4 d5 6 .i.d3 lbc6 7 0-0 .i.e7 8
Black had a tricky try in 25...b5 26 cxbS c4 lbb4 9 .i.e2 0-0 1 0 lbc3 .i.e& 1 1 ltle5
liJb4!? 27 axb4 c4. The idea is to kick the f6 1 2 ltlf3
knight and follow with ...1ict+ and ...i.eS+, This simple retreat causes Black the most
but White keeps control and a clear advan- problems.

14
3 et:Jxe5: The Main Line with 8 c4

lDc6 19 b4 axb4 20 axb4 d4 21 11'b2 :Xc4 22


i.xc4 .i.xc4 23 llfd1 when Morgado claims a
clear advantage for White.
A more worthwhile alternative is 1 2..lle8.
For example, 1 3 .i.e3 i.fB 1 4 a3 ltixc3 1 5
bxc3ltic6 16 cxd5 .i.xdS 1 7 c4 .i. f7 18 'iVc2
ltiaS 19 i.d3 g6 20 i.d2 (Morgado suggests
White can gain an edge with 20 h4!? cS 21
dxcS 'iVc7 22 .i.d4 i.xc5 23 i.xf6 1ff4 24
i.g5 1Wg4 25 ltih2 ifd) 26 'iVc3 ltixc4 27
l:lael WdS 28 �r4) 20...c5 21 dS (fimman
assesses the variation 21 1Wa4 llk6 22 d5
lDti4 2l ltixd4 cxd4 24 .i.b4 b6! as equal)
[f 1 2 .i.g4, variation 'b' is the most accu­ 21 ... b5 22 .i.xaS 'iVxaS 23 cxbS LdS 24
r:ue: lDd2 llad8 2S llfdl �g7 with a double­
a) 12 ... .i.c8 13 .i.xc8 llxc8 l4ltif.3 c6 (or edged position, Adams-1imman, Wijk aan
14...c5 IS 1i'e2 cxd4 16 ltixd4 lle8 17 �e3 Zl.'C 2001 .
i.cS 1 8lLd1 and White has an edge) 1 5 lle1 1 3 a3
li_)xc3 1 6 bxc3lDa6 1 7 llbl llc7 1 8 c5 lld7 White has various alternatives:
19 .i.f4 and White was a bit better in Frcssi­ a) 1 3 cxdS is too simple: 1 3...ltixc3 14
net-Brodsky, Bucharest 200 1 . bxc3ltixdS 1S i.J3 cSI (but not 1 S...ltixc3??
b) 1 2... .i.xg4 1 3 ltixg4 f5 1 4 ltie3 (1 4 a3 16 'IVct with a winning double attack) 1 6 c4
lbc6 1 5 cxdS ltixd 16 bxc3 'irxJS 17 ltie3 ltib4 17 d5ltixd3 18 11'xd3 i.d61 (of course
1Wf7 is unclear) 1 4...dxc4 1 5ltixc4 c5 16 'iVb3 taking no c6 would still lose the queen) 19
�h8 1 7ltieS 1i'e8 (Black has the interesting 1Wb3 i.g4 20 h3 .i.hS 21 a4 V2-1/z Anand­
nlternativc option of 1 7...cxd41? 18 ltixe4 Adams, Dortmund 2001 .
fxe4 1 9 ltif7+ :Xf7 20 11t'xf7 'iVe8 with un­ b) 1 3 h3 is a bit too slow: 1 3. .£5 1 4 a3
.

balanced play) 18 a3 llk6 1 9ltixc6 1fxc6 20 ltic6 1S cxdS (1 5 ltixdS?I was ttied in lvan­
ll\xc4 1i'xc4 (Minakov-Moq,rado, correspon­ chuk-Ponomariov, FIDE World Champion­
dence 1 999) and here Morgado gives the ship, Moscow 2002, but after I5 ...Ld5 1 6
l'lJUalising line 21 i.d2 1fxd4 22 'iVxb7 'iVxd2 cxdS 'IVxdS 17 'iVa4 i.f6 1 8 lld1 llad8 1 9
23 1i'xe7 c4. i.e3 f4! 20 i.xf4ltixd4 21 ltixd4 .i.xd4 22
12 . �h8
. . i.e3 cS Black was already a bit better)
Stepping away from any future trouble on 1S ....txd5 16 .i.f4ltixc3 17 bxc3 .i.d6 Black
the a2-g8 diabJOnal. has comfortable equality.
12...c5?1 is premature: 1 3 i.e3 :C8 14 c) 13 lle l, however, is a reasonable alter­
dx.cS i.xcS 1 S .i.xcS D.xcS 16 1Wb3 aS 1 7 a3!? native to 13 a3. 13 ... c5 1 4 .i.e3 f5 1 5 a3 f4
(17 lladt is less clear: 1 7...'iVe7 1 8lDa4 i.f7 (Grischuk assesses 1 S...ltixc3 16 bxc3 llk6
19 11'e3 llc7 20 a3 ltia6 21 cxdS ltid6 22 17 cxdS i.xdS 1 8 dxcS f4 1 9 i.d4 as slighdy
1Wxe7 llxe7 23 .i.xa6 bxa6 is very messy but better for White) 16 i.d2ltic6 17ltixe4 dxe4
lllack seems to have compensation, Khalif­ 18 dS i.xd51 1 9 cxdS cxf3 20 dxc6 fxc2 21
man-Karpov, Denpasar 2000) 17 ...ltixc3 11'xe2 bxc6 22 i.xf4 and White's better
(17.. dxc4 allows White to play a very con­
. structure gives him a small advantage, Gris­
vincing queen 'sacrifice': 18 ltixe4 cxb3 1 9 chuk-Motylcv, FIDE World Championship,
�xeS .i.f7 20 axb4 axb4 21 i.d 1 and White Moscow 2001 .
has a healthy matc..'t'ial advantage) 1 8 1fxc3 13 ...�c3 14 bxc3 �

75
Th e Petroff Defence

11i'd7 22 .i.xf6 D.xf6 23 �cl :C6 24 :Xc6


1ixc6 25 IL!d3 l%e8 26 h3. Here the game is
level and a draw was agreed in Galkin­
Motylev, Dubai 2001.
1 7.....tg8 18 c5 lieS 1 9 l£if3 h671
This is unnecessary. The direct 19...b6!? 20
cxb6 axb6 21 .i.bS :Xet+ 22 ...xe1 �5 is
l."qUal.
20 llb1 llb8 21 ..td3 llxe1 + 22 •xe1
'ird7 23 ..tf4 lle87
This overlooks a nasty trick. 23 ...i..h7 24
.i.bS lte8 25 •dl restricts White to an edge.
24 .d2 g5
1 5ll\d2 This desperate lunge is forced. If 24...b6?
White must try to conttol c4. Instead 15 then 25 cxb6 axb6 26 .i.xh6! wins immedi­
cxd5 .i.xd5 16 .i.e3 �5 17 11a4 c6 18 ...c2 ately.
�4 gives Black easy equality. 25 llxb71
1 5...f5 Crucially, this undetmines the c6-knight.
Making f6 available to the bishop. 25 ...llc8
15...�a51? is interesting but probably not If 25 ..gxf4 White wins with 26 1i'xf4 i..g5
.

quite goo d enough to equalise: 16 cxd5 (or 26....i.d8 27 Wxh6+ i..h7 28 .i.bS) 27
.i.xd5 (or 16...1ixd5 17 .i.b2 'irg5 18 c4 .i.d6 'ifxc7 'ifxc7 28llxc7 Q..eko) .
19 .i.c3 and White has the easier game) 17 c4 26 h47
.i.f718 .i.b2 and now: This is a clever idea but there was an in­
a) If 18....i.d6 White must be a bit careful. stant win with 26 .i.bS! gxf4 27 D.xa7 as
19 ..c2 (supporting c4 is essential; not 19 Black cannot escape the pin.
.i.d3?1 .i.f4! 20 dS c6 and White's centre col­ 26 ...gxf4 27 •xf4 ..tg7 28 •xf5 'ird8?
lapses disasttously) 19....i.g6 20 .i.d3 .L:d3
21 'irxd3 and White has an edge (Belov).
b) 18...5 19 .i.c3 cS 20 dS .i.f6 21 1ic2 b6
22 .i.d3 .i.xc3 23 ..xc3 �b7 24 �f3 llXI6
25 ILleS and the protected passed pawn
makes White's poslbon preferable,
Kasimdzhanov-Adams, FIDE World Cham­
pionship, Tripoli 2004.
1 6 lle1 ..tf6 1 7 a4
Trying to gain space on the quecnside. 17
�b3 has also been tried: 17...b6 (an intrigu­
ing alternative is 17... dxc4 18 ILleS .i.c8 19
.L:c4 1Llxd4!? 20 cxd4 ..xd4 21 'irxd4 .i.xd4
22 .i.e3 .i.xa1 23 D.xa l with equality) 18 .i.f3 Now White is wmrung again. After
.i.gS 19 cxdS .i.xdS 20 .i.f4 D.c8 (or 20...1Lle7 28.....e8 29 1if4 lie7 30 .i.b5 We6 31 .i.xc6
21 i..es c6 22 .L:f6 :Xf6 23 IL!ct �g6 24 ...xc6 32lLa7 White has only a small advan­
00 h6 - Leko-Kramnik, Dortmund 2000 - tage.
and now C.alkin suggests that White gains an 29 l£ig511 hxg5 30 hxg5 Wf8
edge with 25 IL!b4 .i.xf3 26 1Wxf3) 21 .i.eS This is the only move to save the king.

16
3 liJxe5: The Msin Line with 8 c4

3 1 1th3+ -*.h6 32 .i.f5 b) 14 111'a4 has also been tried but Black
The simple 32 gxh6 l:r.e8 33 llxc7 :Ct+ 34 has no real problems: 14...f415 .i.d2 �h8 16
.tn etlc7 35 llxa7 is also winning. :ab1 l:tb8 17 l:tfc I dxc4 18 .i.xc4 .i.g4 19
32 ...J:le8 33 J:lxc7 J:le1 + 34 Wh2 liJe7 35 i..e2 a6 20 h3 .ihS 21lbg5?1 (now the tactics
g41 .i.e6 work out in Black's favour; White should
l.cko supplies the neat winning variation have settled for 21 etlh2 .i.xe2 22 l:txe2 'ifdS
'5. .l:te6 36 �g2 :a6 37 c6!? llxc6 38 l:r.xc6
. with equality) 2 l ....i.xe2 (2t....i.xg5?! is not
Q'\xc6 39 'irxh6+ 1Wxh6 40 hrxh6. the answer: 22 .i.xhS f3?1 23 .i.xgS 1i'xg5 24
36 Wxh6+ 1i'xh6+ 37 gxh6 liJxf5 38 .i.xf3! l:r.xf3 25 1i'xc61 l:txf2 26 l:r.e8+ l:lfB 27
gxf5 .i.xf5 39 J:lf7 l:r.xb8 9e3+ 28 �h 1 l:r.xb8 29 1i'xc7 and
White's pawns win easily. White should win this ending) 22 etle6 'irdS
39....i.g6 40 J:lxa7 J:lc1 41 Wg3 23 l:r.xe2 (if 2.1 etlxfB f3 Black is already much
:Xc3+ 42 �4 .i.h5 43 �5 llf3 44 c6 better because 24 gxf3 .i.xf3 25 etle6 1i'f5
.tg4 45 a5 J:lxf2 46 a6 llc2 47 llc7 1 -0 gives a winning attack) 23... f3 24l:r.ce1 fxg2!
P""'"-----. and Black was clearly better in Rohde-
Callie 5 Scirawan, Estes Park 1986.
Shirov-Gelfand 12 1Vb3
Leon 2001 Increasing the central pressure. 12 l:r.cl is
the subject of Game 6, while a worthwhile
1 e4 e5 2 liJf3 liJf6 3 lbxe5 d6 4 liJf3 option is 12 etlt:S. flor example, 12....i.f6
/li)xe4 5 d4 d5 6 .i.d3 .i.e7 7 0-0 liJc6 8 (Short assesses 12...etlxc3 13 bxc3 etlc2 14
c4 liJb4 9 .i.e2 0-0 1 0 liJc3 .i.e& 1 1 .i.e3 l:r.c I etlxc3 15 fxc3 .i.e4 as unck-ar) 13 g4
.if5 .i.e6 14f4etlxc3 15 bxc3 etlc6 16 .i.f3 .i.xeS
11 ... f5 is a h'Dod alternative. flor example, 17 dxeS d4 (White has the initiative after
12 a3 etlxc3 13 bxc3 etlc6 17...dxc418 fS .ic8 19 .i.cS 1i'xd1 20 l:r.axd1
l:r.e8 21 l:r.fcl) 18 cxd4.ixc4 19 dS etle7 20
.i.cS .ixfl 21 �ft 'ird7 22 'lfb3 and
White's bishops and impressive centre pro­
vide ade<Juatc compensation, Short-Anand,
Amsterdam 1993.
1 2...c6
Choosing to bolster the centre but
12...dxc4 is also acceptable: 13 .i.xc4 aS t 4
etlxe4(14a3 can lead to a neat draw by repe­
tition after 14...etld2! 15 .ixd2 .i.c2 16
.ixf7+ l:r.xf7 17 1fe6 .ifS 18 '1Vb3 .ic2 19
'ire6) 14....ixe4 15 a3 a4 16 'ifdl etlc2 17
.J:r.cl etlxc3 18 fxe3 .id6 19 ..tbS 'iff6 20 l:r.c3
and now: 'irh6 is unclear, Barua-Manesh, Raipur 2002.
a) 14 cxdS .ixdS 15 c4 .ixf3 16 .ixf3 f4 1 3 cxd5
17 .idS+ �h8 18 .ict etlxd4! 19 l:tb1 (19 White has various alternatives:
i.b2 does not alter the assessment: l9...c5 20 a) 13 c5 etlxc3 14 bxc3 etlc2 15 1fxb7
.ixd4 cxd4 21 .i.xb7 llb8 22 .idS .icS is etlxat 16 l:r.xa1 1fe8 17 'ira6 f6 18 etle1 'ireS
l"lJUal) 19....ic5 20 l:txb7 f3! 21 .ixf3 etlxf3+ (Kramnik states that White has compensa­
22 gxf3 112-1/z Hubncr-Yusupov, Rotterdam tion after 18 ...1Wd7 19 etld3 .ixd3 20 .ixd3)
191'18. 19 'ffxc8 .ixc8 20 .i.f4.idS 21 etlc2 l:r.eB 22

17
The Petroff Defence

.i.d3 aS is unclear, Topalov-Kramaik, Til­ 19....i.xf3!? 20 gx£3 .td6 21 9c6 llb6 22


burg 1998. 'lfe41i'b8.
b) 13ll:kS a5 14 g4 .i.e6 1S tbxe4 dxe4 16 20 'irc4 .i.d5 21 •d3 :Xf3!? 22 gxf3
a3 f6 17 axb4 fxeS 18 dS cxdS 19 cxdS .i.xdS ...d7 23 l:lfb1
20 .i.c4 .i.xc4 21 'ii'xc4+ 'ith8 22 bxaSilxaS Gelfand mentioned 23 .i.f4!?llfB 24 .i.g3
23 'l'xe4 .l:lbS and the simplifications have .l:lx£3 2S 'ii' e2 with a murky position.
led to CXjUality, Topalov-Shirov, FIDE World 23 ...1th3 24 ..i.f4 l:lf8 25 l:lxb4
Championship 2001. Or 2S .i.g3 aS 26 a3 .l:lx£3 27 11ft 'irh6!
c) The wild option is 13 g4!? .i.xg4 14 cS 28 ..e2 (28 axb4 allows a dear draw after
(the point of 13 g4!? is that Black no longer 28...llxg3+1 29 fxg3 'ire3+ 30 'l'f2 'l'e4 and
has the resource ...tbxc3, bxc3, ...tbc2) 14...aS White can only avoid the mate by aUowing a
1S a3 tba6 16 'ii'xb7 tbxc3 17 bxc3 tbc7 18 perpetual check) 28...1rhs 29 llb2 .i.f8 and
Zlab1 tllbS with an unbalanced position, Black still has dangerous play- Gelfand.
Topalov-Shirov, riDE World Championship 25 .. .l:lxf4 26 l:lb8+ �7
2001. With 27...llg4+ threatened White seems to
1 3 ...lDxc3 be lost. However...
t3...cxdS 14 l:act aS 1S tLla4llb8 16 tbcs 27 l:lb7+ !
.i.d6 17llfd1 (Shirov-Topalov, ADE World Now i f Black takes the rook, 'irb3+ wins
Championship 2001) is a safe equalising line. the bishop.
Now the simplest for 8L1ck is 17..lle8 18 27 Wg8 2Bl:lb8+ %-Y.z
•••

.i.bS.I:le7.
14 bxc3 .!bxd5 Game 6
Kotronias-Motylev
Moscow2004

1 e4 e5 2 l£lf3 lL'!f& 3 lDxe5 d& 4 lL'lf3


lL'lxe4 5 d4 d5 6 .id3 lL'lc6 7 0-0 .i.e7 8
c4 lL'lb4 9 ..i.e2 ..i.e&

1 5 1ixb7
White can play safe with 1S c4 tllxe3 16
fxe3 'ii' c7 17 .i.d3 .i.xd3 18 'lf'xd3 .llfe8 with
an equal game.
1 5 lL'lxc3 1 6 ..i.c4 llb8
••.

If 16 ....te4? then 17 ll:ks llb8 18 1i'xa7


lias 191i'b7llb8 20 11fd7 and White's queen 10 lL'lc3
escapes with a dear extra pawn. Now this game simply transposes ro the
1 7 1txc6 .i.e4 1 8 ..i.xf7+ :Xf1 1 9 1t'xc3 main line. However, the move order with
..i.b4 9....te6 gives White interesting additional
Black also has compensation after options:

18
3 lC.xe5: The Mein Lin e with 8 c4

a) 10 'i'a# is tempting but Black can de­ lle7 and ilie position is finely balanced)
fend: 10...'i'd7 11 'i'xd7+ �d7 12 ltk3 15.. ll.b8 16 .xa7 (defending ilie b-pawn
J:lhd8! (this is the way to equalise: after wiili 16 llab1 has other drawbacks: 16. ..if5
.

12...ltlxc3 13 lLk5+ �e8 14 bxc3 ltk6 15 17 llbc1 gS!? 18 .ig3 g4 19 tDes llxb2 20
cxd5 .ixd5 16 .ic4 Shirov claims an edge lDxe4 .i.xe4 21 lDxg4 .i.g7 with an unclear
for White) 13 a3 lDxc3 14 lDe5+ �e8 15 game) 16 ...1lxb2 17 lDa4 (or 17 ltlxe4 dxe4
bxc3 lDc6 16 lbxc6 bxc6 17 cxd5 .ixd5 18 18 Wa3 llc2 19 lDe5 .ixe5 20 .ixe5 f6 21
J:ld �f8 19 .if4 .id6 20 .i.xd6+ cxd6 21 .if4 .ig4 wiili a messy position) 17...1lb4 18
.id3 g6 22 f3ltdb8 231leb1 �e7 and in this a3 (White must be careful wiili his a4-knight;
dead level position a draw was agreed, Rau­ instead 18 llab1 ? 'i'bsl 19 Wxb8 llfxb8 and
sis-Bacrot, France 2003. White was losing in Al-Modiahki­
b) After 10 cS!? Black must defend 3b'llins t lordachescu, Dubai 2004) 18.. ll.c4 19 ltlb2
ideas with 'i'b3. llxd4 20 .i.e3 llc4 21 lDxc4 .ixa1 22 llxa1
b1) The natural 10 ...0-0?! runs straight into dxc4 with an 'unclear' verdict from Ior­
!rouble after 11 'ifb3! aS 12 aJ lDa6 13 dachescu.
1i'xh7. The attempt to trap the queen with 10 0-0 1 1 .te3 .tf5 1 2 llc1 dxc4 1 3
.•.

13.)tlaxc5 14 dxc5 lDxc5 15 1i'c6 llb8 is .txc4


met by 16 .ig51. White already has a pleasant
;tdvantage, but 16... f6?1 just makes things
worse: 17 .if4 .i.d6 18lDd4 .ifl19 b4 axb4
20 axb4 llxb4 21 .ixd6 cxd6 22 lDb5 and
White is winning (Nataf).
b2) tO...aS defends the knight but White
keeps the initiative: 11lDes f6 12lDd3 0-0 13
li:lf4 .ifl 14 .ig4! g6 15 lDe6 .ixc6 16
i.xe6+ �g7 17 a3lDc6 18 lDc31le8 (Nataf
points out the clever trick 18 ...1la6? 19
lLlxd51lDxd4 20 Wxd41lxe6 21lDf411le5 22
1i'xc5Q 19 lDxd5 Lc5 20 dxc5 llxe6 21
i.f4 tDes 22 llct with an cdgt:, Nataf­
Topalov, Cannc..'S 2002. 1 3 ...c6
b3) 10...ltk6 11 .ib5 .if6 (fighting for Preparing a retreat to d5 for the b4-knight.
control of d1c key e5-squarc; 11...0-0 allows White can achic..-ve an edge after 13...lDxc3:
White to seize ilic initiative in simple fashion 14 bxcJ lDc6 15 .idJ .i.xd3 16 9xd3 11'd7
with 12 .ixc6 bxc6 13lDc5 .id7 14 f3lDg5 171lb1 (instead 17 c4 J..f6 181lfd1 llfe8 19
15 lDc3) 12 Lc6+ (after 12 .if41? 0-0 13 'lfbt b6 20 h3 lDc7 was only equal in
i.xc6 bxc6 14 lDes Wbs 15 Wet .ixc5 16 G.Kuzmin-Hu7man, Rcthymnon 2003)
i.xe5 f6 17 .if4 White was a bit better in 17...b6 181lfe1 llfe8 19 .if4.
i\dams-Bacrot, Cap d'Agde [rapid] 2003) 14 lC.e5 lC.xc3 1 5 bxc3 te.d5 1 6 1ff3
12...bxc6 13 11'a4 .id7 14 .if4 0-0 15 lDcJ Also interesting is 16 1ib3 f6 17 ltlf3 b5
(trying to grab ilie c7-pawn with 15 11'a5 (Ivanchuk suggests that 17 ...1rd7!? leads to
llllows Black counterplay: 15...'1'1>8 16 b3 an unclear position after 18 l:lfet �h8 19
l:te8! 17lDe5 -if 17 Lc7 'itb7 18 .i.e5 .idS .id2 .id6 20 .in lDf4) 18 �e2 �hs 19
19 1i'a3 'i'b5 White's awkward development .id2lDb6 20 life I 9d7 21 c4 bxc4 22 .i.xc4
�vcs Black enough compensation -17...11'b7 lDxc4 23 9xc4 l:fc8 24lDh4 i.f8 and White
1 8 f3 lbg5 19 .i.xg5 .ixg5 20 f4 J..f6 21lDc3 has a very faint edge, Kir.Georgiev-lvanchuk,

19
Th e Petroff Defence

Dcbrecen 1992 27 llxc5 ltlf6 and despite White's creative


1 6 .i.e& 1 7 .i.d2
.•• efforts the position was stiU level in Topalov­
The other way to play is 17 .i.d3 when lvanchuk, Novgorod 1996.
Black has three main replies: 18 11'd7 1 9 h3lQf4
..•

b) 17 ....i.f6 18 .i.d2 .d6 19 llfet lladS Exploiting the loose bishop on c4. Black
20 c4!? is Nijboer's suggestion (20 h4 c5 21 could also play the simple 19... b5 2U .i.d3
h5 cxd4 22 cxd4 &i:k7 21 h6 lDg6 24 hxg7 llfcS 21 life1 .i.d6 22 lL!e3 ltlxe3 23 .i.xe3
.i.xg7 led to a messy position in Nijboer­ .i.d5 24 .e2 'itf7 with equality .
foressinet. Leon 2001). After 20. .�7 21
. 20 .i.b3
.i.a5 lieS 22 .i.c3Q)g6 White's space advan­ White ambitiously keeps the tension. The
rnge makes his position preferable. quiet 20 .i.xe6+ ltlxe6 21 lL!e3llfeS 22 ltlf5
b) 17...ltlxe3 1S fxe3 .i.d619lL!c4 .i.c7 20 .i.fB is level.
e4 f6 is unclear - Black's bishop pair com­
pensates for White's centre.
c) 17... .i.d6 tS .i.d2 'lh14 19llfet ltlf6 20
a4ltlg4 21 .i.f4 .i.dS 22 .i.g3 lLixeS 23 clxe5
e'h6 (Hubner pointed out the blunder
23...'itxa4? 24 .i.xh7+1 *xh7 25 'lh15+ WgS
26 exd6) 24 .i.e4 .i.xe4 25 "ifxe4 .i.c7 is
equal, Klimov-Smikovski, Toljatti 2003.

20 .i.xb3 21 axb3 lDg& 22 llfe1 .i.d&


.••

23lQe31lae8 24lQf5lQe7
A sharp try is 24....i.f4!? 25 'W'g4 .i.xd2 26
lLlh6+ gxh6 27 11'xd7 .i.xe1 28 'trxb7 lle2
with an interesting and unbalanced position.
25lQxd6 ••d6 26 c4lQg6
The greedy 26...1Vxd4?! allows a deadly
pin: 27 .i.b4llf7 28lle6.
1 7 16
••• 27 .,5 ••d4 28 .i.b4 l0114
Less accurate is 17 ... .i.gS 1S .i.xg5 'itxgS Black has to play creatively to survive. If
19 llfel llaeS 20 g3 "iff5 21 "ifdt ltlb6 2S..Jlxcl+? 29llxe1 lidS 30 ..e6+ Wh8 31
(White has a promising initiative after .eS+ .!DEB 32 "fif7 Black's weak back rank
2t..lle7
. 22 .i.d3 1i'f6 23 c4 ltlb6 24lle4- costs him the game.
Topalov) 22 .i.d3 "iff6 23 1i'c21!f> 24 llbt 29 'lrh5 g& 30 1ra5
.i.cS 25 lle2 lie7 26 llbe1 and White had Motylev points out that 30 .c5 'trxc5 31
some prcsswe in Topalov-Akopian. Linares .i.xc5 llxet+ 32lLet lids 33 .i.xa7 Wf7 is
1995. equal.
1 8� 30 Jle51
••

Or tSlLld3 1i'd7 19llfe1 .i.d6 20 h3 .i.f7 Once again 30....1Let+ leads to back rank
21 .i.b3llae8 22 "ifg4llxet+ 23llxc1 lids problems: 31 llxe1 llf7 32 .i.c5 'lrh2 33
24 Wxd7 llxd7 25 lL!es Le5 26 dxe5 fxe5 .d8+ Wg7 34 Wh2 (one clever preparatory'

20
3 �xe5: Th e Main L ine with 8 c4

move is required; instead 34 l:le8?1 allows 41 b4 llf3 42 lld7 + �f7 43 .i.e3 Jlxh3
Black to escape with a draw with 34...'iVct+ 44 :Xa7 ¢>16 45 hb6
JS 'it>h2 �f3+! 36 gxf3 'iVf4+- the king must After the alternative 4S l:la6 �S 46 l:lxb6
!(II to hl and allow a perpetual because mov­ �e6 47 bS llh4 48 gS �£5 49 bxc6 �f3+
in� to the g-file allows ...'iVgS+ and 1fxc5) Black's countcrplay arrives just in time to
_\4...�£5 35 lleH and White has a dangerous save the draw.
111tack (Motylev). 45...�5 46 .i.dB+ ¢>e6 47 ¢>g2 llb3
31 llcd 1 ! 48 lle7+ ¢>d6 49 :Xh7 I.Oxc4 50 .i.g5
Now White has t o be careful. I f 31 l:lxeS?! 1.0&5 51 llg7 ¢>d5 52 .i.f4 lbxg4 53 :Xg6
rxeS 32 �xfB 'iVe4 33 �ft 'iVxg2+ 34 �e2 lbxf2 54 lld6+ ¢>c4 55 ¢>xf2 llf3+ 56
'i'c4+ 35 Wd2 �f3+ 36 �c3 'iVd4+ 37 'iPb4 ¢>e2 llxf4 57 We3 llh4 58 llxc6+ Wxb4
'.f.>xffi Black's position is preferable: two 59 lle6 Wc4 60 :Xe4+ llxe4+ 61 Wxe4
pawns and White's exposed king provide Y.z-%
more than enough compensation for the
l'XChangc. Gamel
31 ...1rxd1 32 •xeS fxe5 Kasparov-Anand
After 32 ...1Vxb3 Black's exposed king al­ Unares2000
lows White to draw with 33 'iVd6 l:lf7 34
l:tc!t+ <l;g7 35 l:lc7 gS 36 :Xf7+ �xf7 37 1 e4 e5 2 I.Of3 �f6 3 I.Oxe5 d6 4 lbf3
'i'e7+ '.ti>g6 3M '1Fe4+. lbxe4 5 d4 d5 6 .i.d3 �c6 7 0-0 .i.e7 8
c4 � 9 .i.e2 o-o 1 o &3 b&

33 .lbd1 llf7 34 lldB+ �7 35 .i.d6 e471


This allows White's bishop too much 1 1 a3
Hcope. Instead 3S...�f6 36 llc8 .J:ld7 37 11 cxdS aUows the interesting line
i.xeS+ �f7 is equal. 1t..if5!? t21i'b3 aS with unclear play.
36 .i.c5 11 �S is certainly direct bur Black can
Now White has a tiny edge. defend: 11...�b7 12 .J:let (12 a3 was tried in
36...b6 37 .i.d4+ ¢>h6 38 .i.e3+ ¢>g7 39 Shirov-Anand, Linares 2000, but led only to
.ig5 �f6 40 g4 �h6 a draw after 12...�xc3 13 bxc3 �c6 14 �f)
40 .. h6? is a horrible blunder: after 41 �cl
. �aS 15 cS c6 16 lle l �f6 17 �g4 i..cH 18
Black must lose a piece to avoid mate. cxb6 axb6 19 a4 �4 20 �xf6+ 'iVxf6 21
41l...li:k7 is reason.-.ble but White stiU has a �a3 �xa3 22 llxa3 �e6 23 'iVb3) and now:
pull after 41 b4 cS 42 .J:le8 �6 43 bxcS bxc5 a) 12 ...dxc4 13 �xc4 �xc3 14 bxc3 .!hiS
44 �xe4. lS 'iVf3 c6 (Hracek-Yusupov, Bundesliga

21
The Petroff Defence

2000). Here Hracek suggests that 1 6 .LIJ is 23 We2 'ifh4 24 g4 hS 25 i.e3 hxg4 26 i.d5
worth an edge, the point being 1 6...ltlxc3?! (Kasparov) when White escapes the attack
1 7 ltlxf7! 'ifdS 1 8 l:r.xe7 'ifxf3 19 �h6+ gxh6 and has a good advantage.
20 .i.xh7+ Wh8 21 gxf3 .i.a6 22 .i.xh6 when 1 7 d5 lL!b8
White is dearly better.
b) 12 ... c51? (the aggressive approach is
best) 1 3 .i.f3 cxd4 1 4 �xdS f5 1 5 �xe7+
1Wxe7 16 a3 �!? 1 7 �!? hxg6 1 8 1Wxc2
'ifh4 and after some very sharp play the posi­
tion remains unclear.
1 1 ...lL!xc3 1 2 bxc3 l0c6 1 3 cxd5 1txd5
14 �1
The simple 14 .i.f4 is only level after
14....i.d6 1 5 .i.xd6 cxd6 1 6 c4 •as.

1 8 .c2
Possibly this commits the quL>en too early
Instead 1 8 ltleSI? may be the best choice
Kasparov-H.Olafsson, Kopavogur (rapid)
2000 continuing 1 8...i.f6 1 9 .i.b2 g6 (or
1 9...c6 20 d6 g6 21 cS and White's pawns are
very strong) 20 'ifd2 ti)d7 21 �xd7 i.xb2 22
�xffi .i.xal 23 ltlxg6 l:r.xel+ 24 'ifxe1 hxgt)
25 .xa1 c6 26 dxc6 i.xc6 27 1i'c3 and Kas­
parov was dearly better. After 1 8 ... i.d6 1 9
14....i.b7 1 5 .i.d3 Aae8 1 6 c4 i.b2 9£6 20 l:r.e3 White is building a strong
h is wisest to hit the queen now. If 1 6 attack, the tactical justification being
'lrc2 Black has two reasonable replies: 20...lLe5?! 21 i.xeS i.xeS 22 'ifhs :CS 23
a) 1 6...h617 i.h7+ Wh8 1 8 i.e4 (Kas­ l:r.aet gS 24 h4 when White is already dose to
parov-Kramnik, I.inares 2000) and here Kas­ victory.
parov analysed the equalising line 1 8 ... '1rh5 1 8...g6 1 9 �5 .i.f& 20 .i.b2 �7 21 f4
1 9 �5 i.d6 20 .i.f3 'ifh4 21 �xf7+ Wg8 22 If 21 lDg4 then 2t..ixb2 22 Wxb2 hS 23
ltlxh6+ Wh8. �3 �cS 24 i.c2 c6 is just unclear.
b) The dangerous 1 6...'tth5!? is a good rea­ 21 ...i.g7 22 .,2 lL!c5
son to avoid 1 6 'ifc2 here: 1 7 J:r.bt �xd4! This is fine but 22... c6!? 23 d6 ltXs 24
(Black targets the h2 square) 1 8 cxd4 {White llad 1 lle6 with a hard fight ahead wa.o; also
must avoid the back rank trick 1 8 �xd4? reasonable.
i.d6) 1 8... i.xf3 1 9 gxf3 i.d6 and Black has a 23 .i.c2 •d& 24 1le3 f6 25 � llxe3 26
threatening attack (Rogers). 1txe3 ..i.c8 27 f5 ..i.xf5 28 hf5 gxf5 29
1 6...ifd8 l0f2
If Black again tries 1 6 ...'1rh5 then White is The alternative 29 �h6+ also leads tO
better placed to meet the threats to h2: 1 7 dS equality after 29 .ixh6 30 11fxh6 ltle4.
.

i.d6!? 1 8 dxc6 l:r.xet+ 19 ..xel i.xc6 20 29 ..� 30 lrure4 % -%


.

i.e21 i.xf3 21 .i.xf3 Wxh2+ 22 Wft 'ifh1+ Here the players agreed a draw. Kasparov

22
3 liJxe5: The Main L ine with 8 c4

wves the variation 30...fxe4 31 1Wc4 f5 31 is also fine: after 1 6 dS ltleS 17 �xeS �xeS
Wc2 1Wc5+ 33 Wh 1 .i.xb2 34 ...xb2 1fxc4 3S Black has strong, active piece play) 1 6 llb1
We5. cS 17 dxcS ltle4 (the tricky 1 7...�xb2?! back­
�------., ftres after 18 1lxb2 �xc3 1 9 cxd6 �xb2 20
Game S .i.xf7+! <ifhf7 21 ...dS+ Wf8 22 ...xf5+ ...f6
Grischuk-Pavasovic 23 1i'xh7 when White is clearly better) 1 8
Istanbu/2003 �c2! �xb2 (after 1 8...lt)g3?! 1 9 fxg3 l:lxe3
-------• Black appears to be very active but a few
1 e4 e5 2 .!Of3 .!Of& 3 .!Oxe5 d& 4 .!Of3 accurate moves from White defuse the dan-
lf)xe4 5 d4 d5 6 ..id3 .!Oc& 7 0-0 i.e7 8 ger: 20 1i'd2 .i.d4 21 �xd4 1Wxd4 22 Wh2
c4 l0b4 9 ..te2 dxc4 �g6 23 ltldS! :Cs 24 �f4 1We3 25 •xe3
Immediately clarifying the siruation in the l:lxe3 26 Ld3 llxd3 27 llbe1 and White is
c.:cntre to give White an IQP. simply a pawn up) 1 9 .xd8 l:laxd8 20 llxb2
10 .bc4 0-0 1 1 l0c3 .i.xc3 21 l:lxb7 �xeS 22 .i.xcS .i.xc2 and
Also possible is the straightforward I I White still holds a slight advantage, Karpov­
li)cS. For example, 1 1 ...c6 12 �3 lbxc3 1 3 Kasparov, World Championship (Game 41),
hKc3 �5 1 4 1i'b3 (14 1id3 .td6 1 S �b3 Moscow 1 984.
WVf6 16 ...g3 .i.�.-6 is also murky) 1 4......c7 I S b2) Direct action with 1 3 ltle5 is also
.i.xd5 cxd5 1 6 .i.£4 (Sax-Pavasovic, Baden possible: 1 3...�6 (not 13...lbf5? 1 4 lbxf7!
I 'JIJ9). Now 1 6...'iraS 1 7 llfe1 .1£6 leads to llxf7 1 5 .i.xf7+ Wxf7 16 �3+, a fairly
unck-ar play. common tactical theme in this line) 1 4 .1£4
li:lf5 (Black should prefer 14...liJaS!? I S .i.e2
.i.f5 with unclear play) I S lbxc6 bxc6 16 dS
cS 17 lba4 .i.a6 18 lle1 (A.Sokolov­
Ag7.amov, Riga 198S). Now Black should try
l!L.c4 1 9 .i.c2 ltld6 20 lik5 .i.c8 21 .i.e5,
but White stiU has the initiative.
1 2 bxc3 li)d5

1 1 . . ..!0xc3
Exchanging is narural but retreating is also
pc 1ssible:
a) 1 t ...lbf6 is rather passive. After t2 ltleS
r6 1 3 .i.gS lbfdS 1 4 .i.xe7 Wxe7 1 5 lle1
.iw 1 6 .i.xdS lbxd5 1 7 lbxdS cxdS 1 8 '1Vb3
White's superior minor piece is enough for
n n edge. 1 3 'ttc2
b) More active is 1 1 ...�6 12 .i.b3 �f6 1 3 'lrb3 only wastes time. After l3. ..lbb6
and now: 1 4 .1d3 cS I S 'irc2 h6 Black has equalised
bl) 1 3 h3 (simply preventing ... .1g4) comfortably.
1 3. . ..i.f5 1 4 .tc3 l:le8 I S a3 lbd3 (1 5...�6!? 1 3 .....ta&

23
The Petroff Defence

It is best to develop this bishop immedi­


ately. Delaying with 1 3...c6 14 .td3 h6 1 5
l:b1 :C8 1 6 �5 would make i t difficult for
Black to complete his development.
1 4 �d3 h6 1 S Ab1 b6 1 6 �h7+ ¢>118 1 7
i.e4 1lc8

27 ...1WhS 28 h3 'tWaS
Threatening mate on e I as well as the
knight, but there is a solution.
29 1Wd21 1Wxa7 30 llxf6
This sacrifice is enough for a draw.
30 ...gxf6 31 1Wxh6+ Wg8 32 1Wh4
Black steps out of the pin rather than 'tWaS % -%
blocking it with 17 ... c6. After 1 8 ltle5 Wd6 Black cannot escape the checks.
19 i..d2 lLcS 20 c4 ti)f6 21 .i.£5 White
keeps a nagging initiative. Game 9
1 8 c4 lnf6 Anand-I .Sokolov
Black could have exploited the c-fLic line­ Dortmund 1999
up with 1 8...c6!?. After 19 :C1 .Li6 the posi­
tion is complex. 1 e4 eS 2 lnf3 li)f6 3 lDxeS d6 4 lDf3
1 9 i.b7 Ab8 20 i.e& i.d6 21 lld1 lnxe4 S d4 dS 6 i.d3 lDc6 7 0-0 i.e7 8
Advancing with 21 cS does not achieve c4 lnb4 9 cxdS lnxd3 10 1txd3 1txd5
any ck"U after 21 ....te7 22 :Ct bxcS 23 This is much more reliable than 10...ti)f6
l:xb8 Wxb8 24 dxcS Wb4 25 i..e3. which gives White the useful option of 1 1
21 .'W'c8 22 Ab3 .i.fS 23 'irb2 i.d7 24
•. 'ifb5+ and now:
lOes i.xc6 2S lnxc& 1Wg41 a) 1 1 ... i..d7?l I 2 'ifb3 and Black cannot
This excellent move is forced. If 2S .. ..z:r.aS?! regain the pawn. Timman-Kovacevic, Zagreb
then 26 .i.xh6! gxh6 27 dS opens the long 1985 continued 12...0-0 1 3 tlX3 bS 1 4 i..g5
diagonal with decisive effect. b4 (now Black's kingside is shattered but
26 llf3 even if 14..llb8 15 :lfet b4 1 6 i..x f6 i..x f6
Playing to win material with 26 f3 leads 1 7 ti)e4 i.e7 1 8 tiXS his position is very
only to a messy position after 26 ...Wh5 27 g4 unpleasant) 1 5 i..xf6 gxf6 16 ti)e4 �h8 1 7
ti)xg4 28 ti)xbS ltxb8 29 cS i..xh2+ 30 'itg2 l:fet a S 18 l:lact .i.bS?! (it wa s already bad
Wh4 31 fxg4 'Wxg4+ 32 'itxh2 Wxdl. but now Black is lost) 1 9 ti)h4 l:lgR 20 d6�
26 ..llbe8 27 lnxa7!?
.. i..xd6 21 .xf7 .tf4 22 l:lcS l:lg7 23 Wxf6
This pawn may seem poisoned, but White Wxm 24 ti)xf6 and Black could resign.
has seen deeply into the position. Also inter­ b) Black should play 1 t ...Wd7 but White
esting is 27 .i.xh6l? We4 28 �5 i..xeS 29 still has an edge after 12 'Wxd7+ .i.xd7 13
dxeS tl)g4 30 i..f4 with an unclear position. ti)c3 0-0-0 14 l:t:l :Ides 15 ltleS i.. b4 16

24
3 lt:Jxe5: The Main L ine with 8 c4

.td2llhf8 t 7 llac1 . play, for example t S...0-0 16 .i.d2 .i.f5 17


1 1 Jle1 .i.f5 "eb3 .xb3 1 8 axb3 f6 1 9 .i.c3 �f7 20 lbct2
llfd8 21 t.Oe4 .i.f8 22 b4 b6 23 /.Og3 .i.d3
and Black has adequate compensation, Hiib­
ner-Smyslov, Vdden 1 983) 1 S.. O-O 1 6 .i.d2
.

llfc8 17 'iVe3 h6 te .i.c3 bS (it is not clear


that this advance achieves anything; the sim­
ple 18 ...lle8!? 19 ..d2 'l'bs would have given
Black reasonable compensation) 1 9 a3 aS 20
'ii'd2 �3 21 dS! ..xdS (there is nothing
better: 21 ....i.xdS? loses neatly to 22 .i.xg7!
�xg7 23 lbct4!, threatening both the queen
and t.OfS+) 22 .i.xaS ..xd2 23 .i.xd2 llc2 24
.i.c3 .i.cS 2S I.Od4 .i.xd4 26 .i.xd4, Rogt:rs­
Volkmann, Bled Olympiad 2002. White has a
12 g4 clear extra pawn, but the presence of oppo­
This prevents Black from blocking the e­ site-coloured bishops gives Black drawing
rile with a later ....i.e6, but the weakening of chances.
White's kingside is obvious. 1 2 .i.g6 1 3 lbc3 .!Llxc3 1 4 11fxc3 f&
. . .

The solid option is 1 2 llk3 t.Oxc3 1 3 Since ....i.e6 is no longer an option, Black
1fxc3 and now: must fmd a different way to escape the pin
a) The natural 1 3 ... c6?1 allows a beautiful on the e-file, and here he prepares to unpin
tncric: 14 .i.h6! .i.e4 (the point is 14...gxh6 I S with ...�fl. An interesting alternative is
:cs 1i'd7 1 6 llaet .i.e6 1 7 dSI cxdS 1 8 :Xe6 14 ...Wd6!?, for example 1S .i.gS f6 1 6 .i.h4
fxc6 19 11fxh8+ .i.fB 20 1i'f6 and White wins) c6 I 7 .i.g3 1i'd7 with an unclear position.
1 5 .i.xg7 llg8 16 llxc4 1i'xe4 17 lle1 .xe1+
I K t.Oxe l llxg7 1 9 dS and White is much
better.
b) After 1 3... .i.e6 Black will be able to cas­
ric without any great difficulty:
b 1) 14 lieS looks aggressive but Black has
1wo acceptable replies.
b 11) 1 4...1i'c4?? is an ugly blunder: lS
W'xc4 .i.xc4 16 b3 .i.c6 1 7 dS and White will
win a piece.
bl2) 14 .....c6 1 S ..e1!? 0-0-0 1 6 .i.gS
.txgS 1 7 t.OxgS llhe8 1 8 llcl ..d7 19 •as
'it>bK 20 t.Oxe6 llxc6 21 llxe6 fxe6 22 •cs g6
2:l g3 .d6 was equal in Milos-Zarnicki. Villa 1 5 11fxc7
Martelli 1998. Taking the pawn docs not force an advan­
b 1 3) 1 4.....d7 1 5 .i.gS f6 1 6 lle2 fxgS 17 tage, but neither do the alternatives:
:acl 0-0 18 llxe6 .i.f6 is unclear. a) 1 S .i.f4 lld8 1 6 l:te3 �f7 17 ..xc7 lld7
b2) 14 ..xc7 (White wins a pawn, but 18 ..c3 l:lhd8 19 a3 .ite4 20 I.Od2 (the solid
Ulnck will have strong compensation based way out is 20 l:tdt .i.xf3 21 l:txf3 ..e4 22
on his bishop pair and light square control) l:te1 .xd4 with dead equality - Varavin)
1 4. .i.d6 1 S ..c3 (1 S .c2 leads to similar
. 20....i.h l l 21 �3 .xb3 22 I.Oxb3 .i.c6 23

25
The Petroff D e fen ce

:C3 h5 and Black has plenty of compensa­ 'ilig2 1i'c4+ with a perpetual check.
tion, Varnvin-Egin, Kaluga 1998.
b) 1 5 b3 �fl 1 6 1i'xc7 llhe8 17 "lfc4
..xc4 18 bxc4 b6! (fixing then c-pawn as a
target; Motylev points out that 18 ...:Cc8?! 1 9
c5 b6 20 .i.a3 bxc5 2 1 llact gives White a
dear advantage) 19 .i.b2 llac8 20 l:lacl .i.d6
21 l:lxe8 l:lxe8 22 c5 .i.f4 (Zhang Pengxiang­
Motylev, Shanghai 2001). Motylev now sug­
gested that White should play 23 l:let llxel+
24 tfue1 bxc5 25 dxc5 �e6 26 lDg2 .i.d2 27
lDe3 �d7, assessing the position as equal.
1 5 0-0 1 6 1lxe7
•••

'This leads to sharp play. Black has the


usual good compensation after 16 ..c3 .i.d6 23 eba+ .J:lf7 24 Ad1 'tig4+ 25 �1
1 7 �3 11'xb3 1 8 axb3 .i.fl 19 lle3 l:lfd8. Wg7 26 h3 •e4 27 1rd5 'tfc2 28 Ad2
1rc1 + 29 Wg2 lle7 30 1rd8 'tfc6+ 31
'tfd5 1rxd5 + % -%

Game 10
Nyysti-M.Rychagov
Gothenbutg 2003

1 e4 e5 2 li:lf3 li:lf6 3 lt:lxe5 d6 4 li:lf3


lt:lxe4 5 d4 d5 6 .i.d3 lLlc6 1 0-0 .i.e7 8
c4 lLlb4 9 cxd5 li:lxd3 1 0 1rxd3 'tfxd5 1 1
lle1 .i.f5 1 2 �5

1 6 •xf3 1 1 llxg7+ Wh8 18 .i.h6


•..

•xg4+
White's mating threat is demonstrated by
18....i.e4?? t9 l:lxh7+.
1 9 •g3 'tfxd4 20 .lbg6!
White very wisely heads for a draw. In­
stead 20 l:lxb7? allows Black a deadly attack
on the g-file with 20...l:lg8 21 l:le1 ..d5!
(2t....i.ffi! fails to make the most of Black's
chances: 22 llxfl l:lxg3+ 23 hxg3 llg8 is only
unclear) 22 llee7 11fdt+ 23 <i>g2 .i.e4+ 24
l:lxe4 l:lxg3+ 25 hxg3 11fd5 26 l:tbe7 f5 27 'This is cet"tainly a threatening-looking
l:t7e5 "lfxe4+ 28 llxe4 fxe4 and Black should move, but Black has convincing replies.
win this ending. 1 2 o-o-O
..•

20 hxg6 21 11h3 Wg8 22 .i.xf8 .lbf8


••• Alternatively:
Another drawing path is 22. 11Vxb2 23
.• a) Black must avoid the losing blunder
lld1 llxffi 24 Wh6 11Vc2 25 l:ld7 1fb 1+ 26 12....i.h4? 1 3 g3 lDxg3 (there is no hope in

26
3 l0xe5: The Main L ine with 8 c4

retreating: 13 ....te7 14 Wf3 g6 1 5 g4! and 1 4 lbc3 is a solid option. However, after
White wins a piece) 14 1Vf3! lbe4 15 lbc3 14...lbxc3 15 bxc3 Wx£3 16 gxf3 Black has
�xc3 16 bxc3 Wxf3 1 7 lbxf3+ i..e7 1 8 i..a3 no problemS. For example, 1 6 ...llde8!? 1 7
Ac6 1 9 .i.xe7 Wxe7 20 d5 and White wins lbxf7 llhfB 1 8 l£le5 .i.h3 with compensation.
cas�y. Zuidema-Barendregt, Amsterdam 1 4...-*.h4
1 966.
b) 12 ... f6?! is more complicated but still
bnd: 1 3 lbc3 (13 Wf3 is only enough for
Cl(uality after 1 3...g6 L4 lbc3 lbxc3 1 5 bxc3
1i'xf3 1 6 lbx£3 c;ild7) 1 3 ...lbxc3 1 4 1fxf5
�b5 1 5 ..g4 l£lxd4 (of course Black would
like to take the piece with 1 5... fxe5 but after
1 6 llxe5 Wxd4 1 7 llxc7+ Wxe7 1 8 .i.g5+
�d6 1 9 lld1 White is dearly better) 1 6 l£ld3
and the pin causes Black immense difficulty:
bl) 16 ...Wf7 1 7 l£lf41 1ff5 18 llxe7+!
�xe7 1 9 1fxg7+ �d6 20 .i.e3 llhg8 21 1ff7
cS Qf 21 ... b6 then 22 Ld4 1Vxf4 23 .i.xb6!
clears the d-file and continues the attack) 22 1 5 �3
1i'xb7 llgb8 23 Wa6+ llb6 24 Wa3 and After this White seems to be in trouble.
White's attack was obviously progressing weU FoUowing 15 lbc3 there arc many paths bu t
In Jonsson-Schandorff, Panormo 2001 . White cannot force more than equality:
b2) If 1 6...lbc2 White must find 1 7 lbb4! 1 5...lbxc3 1 6 bxc3 .i.e6 17 g5 (or 17 .xd5
�xb4 1 8 .xb4 c5 1 9 Wg4 �f7 20 i..h6! .i.xcl5 1 8 g5 f6 1 9 gxf6 .i.xf6 with a balanced
gxh6 2 1 llad1 with a massive attack, De Fir­ game) 1 7.. .llhe8 1 8 c4 .xd4 1 9 llb1 c6 20
mian-Plaskett, Copenhagen 1 985. 17 llxe 7+ .i.f4 (if 20 lbxc6 Black bails out to a draw
leads to a draw after 1 7...Wxe7 1 8 1Ve2+ Wf7 with 20 ...Wxt2+1 21 1Vxf2 .i.xf2+ 22 Wxf2
19 'irxc2 llad8 20 lbe 1 1i'd1 21 1Vxc7+ Wg6 bxc6) 20...1Wc5 21 lbxc6 lld31 22 llxe6 (after
22 1Vg3+, as the pins force White to setrle for 22 Wxd3 Black can force a draw with
n perpetual. 22..Wxf2+ 23 Wh t .i.dS+ 24 Wxd5 l'lxet+
c) A reasonable alternative is 1 2...g6 1 3 25 llxe1 .xet+ 26 Wg2 '1Ve2+ 27 Wh3 .ft+
1'13 (not 1 3 g4?1 lbxt21 1 4 W n 'eht+ 1 5 28 Wxh4 1i'xf4+; instead 22 lle3 continues
�xf2 i..h4+ when White is i n trouble) the fight: 22...llxc3 23 lbxa7+ Wxa7 24 .i.xe3
1 l .. Wxd4 1 4 lbc3 1i'xe5 1 5 .i.f4 (or 1 5 Wb8 25 We4 .i.xgS 26 i..xgS and with oppo­
lDxe4 0-0-0 1 6 .i.f4 1i'a5 1 7 llact lld7, when site bishops and both kings vulnerable, the
the game is unclear) 1 5 •a5 1 6 b4!? 1i'a3!
... position is totally unclear) 22... fxe6 23 ..xd3
(lhis is the best way to contain White's initia­ Wxt2+ 24 Wht .xf4 25 lbxa7+ �b8 26
tive; 1 6 ...1Wxb4?1 is obviously bad after 1 7 lbc6+ Wc7 27 lba5 lidS 28 1Ve2 lld2 29
ltld5, while 1 6. .ixb4 is not a clear mistake llxb7+ Wd8 30 llb8+ Wc7 lf2-1/z Wahls­
hu1 after 1 7 llxe4+ .i.e6 1 8 .te5 White's Yusupov, Germany 1 992.
initiative is very dangerous) 17 lDd5 .x£3 1 8 1 5 10xf2
.•.

!(X f3 i..dS 1 9 fxe4 .te6 20 lbxc7+ i..xc7 21 This neat trick turns the game in Black's
j_xc7 (the game has fizzled out to equality) favour. Another attractive choice is
2 l ...f6 22 a3 llc8 2l llact <t>f7 1/z- 1/z Be­ 1 5...llhg8!? 1 6 .ic3 hS 17 h3 hxg4 18 hxg4
liavsky-Smyslov, Re�o Emilia 1 986. .i.e6 when Black has a dangerous initiative.
1 3 .,3 g6 1 4 g4 1 6 'lfxd5 l0h3+ 1 7 �2 llxd5 1 8 gxf5

27
The Petroff Defence

.i.xe1 1 9 �xe1 lle8 20 �f3 llxf5 21 11tis little trick enables Black to save both
Wg3?1 rooks, so White resigned.
White should prefer 21 tik3, but after
21 ...�f4+ 22 .i.xf4 lLf4 Black is still obvi­ Game 1 1
ously on top. Shirov-Gelfand
Astana 2001

1 e4 e5 2 c!Lif3 c!Lif6 3 c!Lixe5 d6 4 c!Lif3


c!Lixe4 6 d4 d5 6 .i.d3 .i.e7 7 0-0 c!Lic& 8
c4 c!Lif6

21 �1 1 22 lbbd2
•..

The point is that if 22 �xgl then 22. ltft


.

23 l!Xi2 lLgl+ 24 Wf3 Aeel 25 tDb3 f5


leaves White hopelessly tied up.
22 �2+
•.•

Anothc:r way to win is 22...:C3!? 23 'iPf2 8....tc6 possibly commits the bishop too
.l.ex£3+ 24 Wxgt l:tg5+ 2s Wht Af2. early:
23 Wg2 g5 24 �b3 g4 25 �1 a) 9 cxdS is interesting: 9 ....i.xdS 1 0 W
There is also little hope aftc:r 2S �eS lilxc3 11 bxc3 0-0 1 2 i.f4 .td6 13 .txd6
�xd4 26 �xd4 AfxeS 27 �g3 £5. Wxd6 14 lL\g5 h6 (14. . £5 1 S c4 .i.xg2! 1 6
.

25 .. .c!wcc1 26 llxc1 lle2+ 27 Wg3 llf1 Wxg2 1Wg6 1 7 f4 h6 1 8 dS llkl8 1 9 h4 hxgS


Perhaps evc:n simpler was 27 ...hS 28 tileS 20 hxgS was a bit better for White in l..owen­
Aft 29 lilcd3 £5. thai-Morphy, London 1 8S8) 1 S 1Whs .i.xg2
28 c!Lig2 llf3+ 29 Wxg4 llxb3 30 c!Lif4 16 .i.h7+ �h8 17 �xg2 1Wd5+ 1 8 .te4
h5+1 0-1 1Wxg5+ 1 9 1WxgS hxgS 20 llab 1 llab8 21 l:l.bS
and White has good compensation.
b) 9 Act is natural and probably best
9...lilf6 to cS!? 0-0 I 1 lilc3 (preventing
....i.g4 with 1 1 h3!? may be best: 1 t ...tDb4 1 2
lilc 3 lilid3 1 3 1Wxd3 h 6 1 4 tiles i s good
enough for an edge) 1 1 ....i.g4 1 2 .te3
.i.xcSI? (now Black has counterplay) 1 3 dxc S
d4 14 .txd4 (1 4 h3 i..x f3 1S 1Wxf3 dxc3 1 6
lladl tDcS 1 7 'Wf5 ll:lxd3 1 8 Axd3 'fle7 1 9
llxc3 llad8 is equal) 1 4...tilxd4 1 S Lh7+
lilih7 1 6 .xd4 ..Lf3 17 1Wxd8 llfxd8 18
gxf3 ll:lg5 with a level position, Short
Smyslov, Hastings 1988/89.

28
3 lDxe5: Th e Main Line with 8 rc 4

Also possible is 8 ..ig4 9 ltic3 (9 cxd5 is R.xd7 23 t'fJxf7 'ttfB 24 t'iJxh8 Wxh8
loo simple an approach: 9...Wxd5 10 ltic3 (Sutovsky-Huzman, Tel Aviv 2000) . Now
lt:\xc3 1 1 bxc3 0-0 12 IZ.c l ..i.xf3 13 Wxf3 White should have continued 25 WcS :<12
Wxf3 14 gxf3 ..i.d6 1 5 :b1 lLb8 1 6 ..i.e3 b6 26 'ttc3 :d5 27 1Z.c7 with a dear advantage.
17 f4 t'iJe7 1 8 c4 c6 1 9 Wg2 :rds and Black 1 0 c5
had equalised in Kotronias-Langrock, Ham­
burg 2001) 9...t'fJxc3 (9...t'iJf6 transposes to
H ...t'fJf6 9 t'fJc3 ..i.g4) 1 0 bxc3 0-0 1 1 IZ.c1
tlxc4 1 2 ..i.xc4 Wd6 13 :b1 ltab8 14 :b5
nnd White has hi� usuaJ edge.
9 .!003
Black is okay in the variations after 9 cxd5
i;)xd5. For example, 10 t'iJc3 0-0 11 :e1 t'iJf6
1 2 a3 ..i.g4 1 3 d5 t'iJd4 14 ..i.xh7+ t'iJxh7 1 5
Wxd4 R.xf3 1 6 gxf3 R.f6 1 7 1i'd3 Wd7 18
.if4 t'iJg5 �t is important to improve the
badly placed knight immediately; instead in
Svidler-Ovetchkin, Tomsk 2001 Black played
IH..1lfe8 when 19 t'iJc4 stranded the knight Ensuring a space advantage at the expense
on h7 - White was a touch better after of taking the pressure off Black's centre.
19...b6 2.0 :e2 c5 21 llact R.d4 22 d6) 19 With the bishop already on e6, 10 cxdS
.lxg5 ..i.xg5 20 IZ.c4 WfS and White's weak­ seems less logical but it is b)• no means bad.
ened structure means that Black has no prob­ For example, lO.. .t'fJxdS 1 1 IZ.c1 0-0 12 .i.e4
lems. .i.f6 1 3 a3 h6 14 1r'd3 t'fJce7 1 5 .i.d2 c6 1 6
ltad1 llc8 17 h3 t'iJxc3 1 8 bxc3 b S 1 9 .i.f4
..i.dS 20 t'iJd21 llc8 21 1i'g3 Wh8 22 R.c21?
(dearing the way to centraJise the knight;
instead 22 Wg4 R.c6 23 'i'hs t'iJg81 24 .i..b I
rfJ 25 'irf3 .idS 26 1r'g3 .ig7 was equaJ in
Movscsian-Yusupov, Batumi 1999) 22....d7
23 t'iJe4 and White's position is mildly prefer­
able.
1 0.....tg4 1 1 ..tb5 0-0 1 2 ..txc& bxc& 1 3
h3 ..txf3 1 4 11'xf3 Ae8 1 5 ..te3 1ib8
Black can provoke many exchanges with
1S ...t'iJe4 but after 16 t'fJxe4 dxc4 17 1r'xe4
..i.xcS 1 8 •xc6 .ixd4 19 ltad1 1tf6 20 'irxf6
9 .i.e6
..• .ixf6 21 b3 White's better structure is a sig­
9...R.g4 exerts less influence on d5, so it's nificant factor.
luwcaJ for White to continue 10 cxd5. For 1 6 .:ab1 .i.f8 1 7 l:lfe1 eb4
example, 1 0...t'iJxd5 1 I ..i.e4 R.e6 1 2 Wd3 1 7 ...t'iJe4 still does not help. Gelfand
�cb4 1 3 Will c6 14 a3 t'fJa6 1 5 t'fJxdS ..i.xdS anaJyses 18 .if4 t'iJxc3 19 bxc3 llxct+ 20
J (, l'.lc1 t'iJc7 17 R.gS R.e6 1 8 ..i.xc7 1r'xe7 1 9 llxe 1 Wb2 21 lle2 as dearly berter for White.
li]c5 0-0-0 20 Wc2 l:hd4?1 (20. . 1r'd6 21 . 18 a3
Jlnd1 is better, after which White has only a 1 8 Wf4!? l:tac8 1 9 lle2 also keeps an edge.
Hmall advantage) 21 .i..xc6! 'iVd6 22 ..i.d7+! 18 11'c4 1 9 l:lbc1
..•

29
The Petroff Defence

Again there was a good alternative: 1 9 .i.d6 Ld6 31 cxtl6 :b?.


1i'g3 lie? 20 lled1 g6 21 .i.gS .i.g7 22 1i'f4 30 .i.d6 .:Xa3 31 .i.xf8 ¢»xf8 32 Ae6 Ad3
lObS 23 '1t'd2 and White has a pull. 33 .llxc6 Ad1 + 34 �2 .llxd4 35 .:CB+
1 9 .%lab8
•• �e7 36 c6 Ac4 37 Ac7+ Wf8 %-%
If 19...lle6 then 20 ..g3 reminds Black of
the weakness of c7. Game 12
Topalov-Anand
Wijk aan Zee 2003

1 e4 e5 2 lL!f3 lL!f6 3 ll\xe5 d6 4 ll'lf3


lL!xe4 5 d4 d5 6 .i.d3 lLlc6 7 0-0 .i.e7 8
c4 ll'lf6 9 h3 lLlb4
9._0-0 is also reasonable: 10 l0c3 tlxc4
(10...ltlb4 t t .i.e2 dxc4 1 2 .i.xc4 transposes
to the main game) 1 1 -*.xc4 l0a5 1 2 .i.d3
.i.e6 13 llet l0c6 (13...c5 liquidates the cen­
tre but following 14 dxc5 i.xc5 1 5 i.g5
White's superior activity is still annoying) 14
a3 a6 1 5 .i.f4 rle8 (t5...1i'd7?! exposes the
20 11ff5 queen to a series of tactics: 16 lOeS l0xe5 t 7
Now Black is able to fix b2 as a weakness. dxeS ltlds 1 8 l0xd5 .i.xd5 1 9 •c2 g6 20
20 b4 was required to keep a small advan­ :Sdt c6 21 .i.h6 :rds 22 e61 fxe6 23 .i.xg6!
tage: 20...a5 21 bxa5 :b3 22 l0b11; or and White is much better, Kasparov-Karpov
20...1i'b3 21 -*.g5 'lt'xa3 22 .txf6 gxf6 23 b5 World Championship [Game 481, Moscow
cxb5 24 1Wg4+ ..thB 25 l0xd5 and White has 1 985) 1 6 :C1 lOds 17 l0xd5 .i.xd5 and
a strong attack. Black has comfortable equality.
2o...a3 21 .t.us 10&4 1 0 .i.e2 dxc4 1 1 .bc4 0-0 1 2 lL!c3 c6
The bold 21...11'xb2 is also possible. 22
.i.xf6 gxf6 (but not 22. ltxe1+ 23 l:lxe1
.

11'xc3 24 :e3 'lfcl+ 25 Wh2 'lt'bt 26 1i'd71


gxf6 27 flg3+ .i.g7 28 'lt'g4 'lt'g6 29 'lt'f4
when Black has to surrender his queen) 23
1Wd3 .i.h6 24 llxe8+ :xeS 25 :c2 1i'xa3.
Notkin analysed this variation but under­
standably stopped now with the assessment
of 'unclear'.
22 ll)xe4 dxe4 23 .llxe4 11fd3 24 Ace1
.llxb2
It's clear that White has lost control of the
position.
25 11ff4 1 3 Ae1
Or 25 11'f3 :xe4 26 1t'xd3 rlxel+ 'Z7 Wh2 1 3 a3 is a logical alternative, for example
h5 wilh a murky, unbalanced position. t 3...ltlbd5 14 llet .i.e6 1 5 .td3 h6 1 6 .i.d2
25 .1lxe4 26 11fxe4 11fxe4 27 Axe4 f6 28
•.. rle8 Qt is best to avoid all the :Xe6 ideas
.i.f4 Wf7 29 .i.xc7 Aa2 with 1 6 ...1Wc81; after 1 7 .c2 :cs 18 �e2
29..ltb3 was another way to draw after 30 .i.d6 19 l0e5 a complicated struggle lies

30
3 l0xe5: The Main Line with 8 c4

11hcad) 17 llxe6!? (White sacrifices the ex­ 1 7 1lxe6!?


change to weaken the light squares around Boldly sacrificing the exchange is the the­
Hinck's king) 1 7 ... fxe6 1 8 �g6 l:lffi 19 1i'e2 matic continuation. The safe liner is 1 7 a3
�7 20 .Jlct .td6 21 .tbt .Jle8 (In Bologan­ ltlbdS t 8 ltla4 (18 .i.d2 led nowhere in Gcl­
llo�entalis, Belfort 1 998 Black tried fanc.I-Yusupov, Istanbul 2000: after 18....Jlc8
2l.. .•e7?1 but after 22 ltleS .i.xeS 23 dxeS 19 ltla4 .Jlc7 20 .Jle2 .i.d6 2 1 l:.ac t ltlhS 22
Q')f<.I5 24 ltlxd5 ltlxd5 25 1i'e4 g5 26 1i'g6+ .i.h7+ �h8 23 ..i.. £5 lllh f4 24 .i.xf4 ltlxf4 25
11rg7 27 1i'xe6+ *h8 28 h4 White had a dan­ .Jle3 tlld S 26 .Jl3e2 ltlf4 27 .Jle3 a draw was
Aerous attack) 22 ltle5 c5 23 llle4 cxd4 24 agreed) t 8...ll:ld7 19 .i.d2 .Jle8 (Bologan­
ltlxf(r+- ..xf6 25 ltlg4 and here Bologan Kasimdzhanov, Pamplona 2002). Here Bolo­
1tnres that White has good compensation for gao suggests 20 llc21? '1Vc7 21 .Jlae1 ltlf4 22
the material deficit. .i.xf4 'irxf4 23 ltlc3 'ire? 24 dS cxdS 25 ltld4
13 ltJbd5 14 Wb3
.•. with an edge for White. Instead, 23...�f8
'Inc other try is 14 �g5 .i.e6 15 WbJ and aUows 24 llxL-61 fxc6 25 ..tg6 .Jled8 26 .Jlxe6
nnw: with a big advantage.
a) 1 5...1i'b6 is far too obliging. White eas­ 1 7 . . .fxe6 1 8 ••2 1lrd7 1 9 ..i.d2 ..i.d8
Ily achieves a preferable structure after 1 6 If 1 9....Jlad8 20 .lle 1 .J.d6 21 lllc4 ltllx15
�xd5 cxd5 1 7 'irxb6 axb6. Following 1 8 White can claim an edge.
.ib.l h6 19 �f4 llfc8 20 ltles g5 2 1 .i.h2 20 lOe4 ltJbd5 21 l0xd8 •xd6 22 Jle1
J..b4 22 .Jle2 ltlc..-4 2.l .tZkl3 .*.fB 24 J:l.ae1 .i.g7 .llad8 23 a3 JZ.fe8 24 ..i.b 1 c5 25 lfle5
(nnt 24...lbd6?1 25 .txd6 .i.xd6 26 .llxe6! cxd4 26 1i'd3 ltJe3
fxc6 27 .Jlxe6 when White will win far too After 26...11'b6 Anand suSbrests 27 b4 li:k7
many pawns) 25 .*.e5 .*.xeS 26 ltlxc5 �g7 28 .J.a2 to keep up the pressure.
27 lbd3 White had an enduring Cllgc in 27 ltJf3 lOadS
I .cko-Adams, Dortmund 2002.
b) 1 s....JlbS 16 J:.ad 1 .Jles 11 .th4 h6 t R
i.g3 .*.<.16 1 9 .tcS lllxc3 20 bxc3 .*.xc4 (an­
olhcr convincing eyualiscr is 20...lbd5!? 21
'i'c2 b5 22 .*.d3 a6) 21 1i'xc4 llldS 22 'irbJ
hS 23 c4 bxc4 24 1i'xc4 .Jle6 25 .Jlct .llb6 26
i.xd6 .llxd6 with a balanced position, Lutz­
Dautov, Bundesliga 2004.
14 l0b6 1 5 ..id3 ..ie6 1 6 •c2 h6
..•

27...ltlf51? is a solid option, but after 28


.i.a2 ltlds 29 llc4 Topalov still claims com­
pensation.
28 ..i.xh6! ltJf4
The bishop cannot be taken: 28...gxh6??
29 9g6+ �£8 30 •xh6+ �g8 31 .Jles and
White wins.
29 ..txf4 1i'xf4 30 ..i.a2 JZ.d6 31 h4
A good alternative is 31 'it'c4!? ltlds 32

31
The Petroff Defence

/t)xd4 lZ.fB 33 l£lf3 with perfectly adequate Or 37 �3 11rxb3 38 .ixb3 lZ.b6 39


compensation. .ixe6+ llxe6 40 /t)xe6 lZ.eS 41 lZ.xd2 ltxe6
31 �6 32 .*.c4 et7 33 ..,3 d3 34
•.• and Black should win.
Ad1 37 •c6 38 g3 1tb& 39 .!Oxe& •xb4 40
•••

Topalov starts to lose his way - it was axb4 .llc8


probably time for 34 lZ.xe6. Following Since the discovered check is harmless, it
34...:dxe6 35 .t.xe6+ �h8 36 l£Je5 'l'hs the is clear that Black is winning.
position is unclear, one possible continuation 41 �1
being 37 l£lf7+ ..xf7 38 .ixf7 lZ.et+ 39 �h2 41 lDcS+ docs not help: 41 ...Wh8 42
d2 40 ..xb7 dt .. 41 ..c8+ �h7 42 ..£5+ lDxb7 lZ.b6.
with a draw by perpetual check. 41 ...Wtla 42 lOQ5 Ad4 43 lLln+ �h7 44
34 d2
.•• lOgS+ Wtl6 45 lLln+ wg6 46 li)e5+ �h7
47 l0f3 .Jlxb4 48 .!Oxd2 AdS 49 .*.c2 +
�hB 50 b3 lZbd4 51 �e1 b5

35 11'b4?1
The d-pawn had to be taken, even though
Black is a bit better afu:r 35 ltxd2 llxd2 36 52 lDb1
/t)xd2 ..xh4 37 Le6+ �h8. The queenside pawns are also decisive af­
35 .Aed8 36 l0g5
•• ter 52 /t)f3 lZ.e8+ 53 Wft ltxdl+ 54 .i.xdl aS.
Now it is too late for 36 llxd2?!. After 52. . .Ae8+ 53 �1 llxd1 + 54 .*.xd1
36 ...�1+ 37 Wh2 ltxd2 38 /t)xd2 Wet the lDe4 55 .*.c2 a5 56 lDB3 lDd2+ 57 �2
pin is deadly. llb8 58 g4 b4 59 �b1 l0xb1 60 .i.xb1
36 •c2 37 '*.b3
••• AcB 81 .*.f5 llc3 0-1

32
3 /i)xe5: The MBin Line with 8 c4

Summary
The line 6... .i.e7 7 0-0 liX6 8 c4 has been enduringly sruuied, but it is still very popular. Strate­
lo{ic understanding of the variation is the key to succL"SS - there is no real need to memorist: the
det:uls.
Although 8...lbf6 looks quite reasonable, at the moment aD the attention L� on 8 ...lbb4.
White should really retreat his bishop to e2 - Games 9-1 0 show that 9 cxdS does not lead to
�uccess against accurate black defence. After 9 .i.e2 the bishop fianchetto to b7 (Game 7) may
well not equalise (see the note to White's 1 8th move). In the case of 10 ... .i.e6 t i lDeS neither a
counterattack in the centre with 1 1 ...c5, nor ousting the centralised knight with 1 1 ...f6 shakes
off aU Black's operung problems. Mobilisation on the qut:enside with 1 1 .i.t:J seems to give
lc:ss chances for an advantage after both t t ...fS and 1 t....i.f5.
With the knight on b4 it is logical to develop the bishop to f5 as soon as possible. After 9
.le2 0-0 10 lbc3 .i.f5 1 1 lbeS, t l...cS is quite effective, while in the event of 1 1 a3 lbxc3 1 2
bxc3 llX6 the complicate<.!, strategic play has not revealed any appreciable advantage for White
�o far.

1 e4 e5 2 /i)f3 /i)f6 3 /i)xe5 d& 4 /i)f3 /i)xe4 5 d4 d5 6 ..td3 ..te7 7 0-0 Nc& 8 c4 /i)b4
8...lbf6
9 llX3 - Ga111e 1 1; 9 h3 - Gtllllt: 12
8 ..te2
9 cxdS lbxU3 10 1Wxd3 1i'xd5 t l llet .i.fS (D)
1 2 g4 Ga111e 9; 1 2 lDc5 - Ga111t 10
-

8...0-0
9...dxc4 - Ga111e 8
10 lllc3 ..tf5
10.. b6 G1111�e 7
. -

I O.. ..i.c6 (D)


t t lbes
1 1 ...cS - Ga111e 3; 1 1 ... f6 - GtJJ1Je 4
1 1 .i.e3 .i. f5
1 2 1Wh3 - Ga1ne 5; 12 llct - Gt�nJe 6
11 a3 �xc3 1 2 bxc3 t;)c& 1 3 lle1 lieS (D) 14 cxd5 - Ga111t I
14 .i.f4 Gtlll�t 2
-

1 1 . ..tf5
. . 10 ..te6
. . . t3. . . :es

33
CHAPTER TWO I
3ltJxe5: The Main Line
with 8 lle1

1 e4 e5 2 l0f3 lOt& 3 l0xe5 d& 4 .!Df3 to equality (the latter in view of 1 3 li)xc7
l0xe4 5 d4 d5 6 .i.d3 .i.e7 7 0-0 .!De& 8 �d7 1 4 .i.t4Q. Black can take on dS with his
l:le1 queen without harm if he precedes this with
After 1 e4 eS 2 li)f3 li)f6 3 lilieS d6 4 li)f3 l O... il..xO 1 1 .xf3 (Game 1 4), while in
li)xe4 5 d4 dS 6 .i.c.J3 il..e7 7 0-0 liJc6, it's true Game 1 5 Black takes the strategic decision to
that the immediate attack on the dS-pawn with occupy the square in front of the isolated
8 c4, as covered in the previous chapter, is in pawn with 1 0. .li)xd5.
.

the foreground of modem practice. Howt.-ver, While 9 c4 looks aggressive, 9 c3 appears


8 :Ct, developing the rook and attacking the distinctly restrained. However, comparing the
knight on e4, is a serious alternative. Even consequences of the lines leads to unex­
Bilguer considered the 'pawn sacrifice' 8 .il_g4
.. pcctt..od results. 9 c4 leads to the quick reduc­
to be the best reply, and he indicated that it tion of tenc;ion in the centre with approxi­
was unfavourable for White to grab and hold mately equal chances, while the conse­
on to the pawn: 9 .i.xe4 dxe4 1 0 lLe4 .i.xf3 quences after 9 c3 arc unproven. White
1 1 gxO?! (1 1 •xn li)xd4 12 1i'd3 maintains strengthens his d4-pawn and resumes his
equality) 1 1 ...fS 1 2 llf4 0-0. Instead White threat to capture on e4. Moreover, after c2
should advance his c-pawn with the aggressive c3 White's queen has a clear way to b3 and
9 c4 (Games 1 3- 1 5) or the modem 9 c3 after 9... f5 it is Jjkcly to take this road. 1 0 h3
(Games 1 6-18). has not become popular, although it looks
The intensive attack on the centre with 9 logical to force the bishop on g4 to make a
c4 virtually forces Black to defend the dS­ decision before the queen leaves for the
pawn with 9 .li)f6. If White retains pressure
.. flank. 1 0 c4?!, trying to take advantage of the
in the centre with 10 li)c3 (Game 1 3), the d4- weakening of the a2-g8 diagonal at the cost
pawn fall-; and Black equalises with either of a tempo, is unjustified: 1 o ..dxc4l 1 1 .i.xc4
.

1 0...li)xd4 or 10 ... .i.xf3 11 .xf3 li)xd4. This •d6! and .0-0-0 (Mikhalchishin). to li)bd2
..

is the reason White often releases the tension .d6!? 1 1 c4!? (Game 1 6) makes some sense
in the centre with 1 0 cxdS (Games 14-1 5). because ... d5xc4 does not work, while the
Pillsbury introduced I O...•xdS, but after 1 1 manoeuvre li)d2-ft deserves consideration
llJc3 neither the retreat 1 t...1i'h5 nor the after 1 0 li)lxl2 0-0. However, this plan is less
queen swap I t....i.xO 1 2 li)xd5 il..xdt leads appeaJjng after 1 0 li)lxl2 •d6 in view of

34
3 liJxe5: The Main L ine with 8 lie 1

Ul11ck's possibility of castling long. 1 0 cxd5 is the subject of Game 1 4.


An early 'iVb3, attacking b? and d5, is cov- 1 0 .i.xf3
..•

r!rcd in Games 1 7 1 8. A comparatively rare


· Now this capture on f3 is safer. lO...dxc4
J>lnn after tO 'Wb3 is 10 ...1i'd6, preparing is less accurate after 1 1 i.xc4 0-0 12 clS �aS
lJUccnsicle castling. Ths gives White both the 1 3 i.cl3 c6 1 4 h3 i.hS 1 5 i.gSI? cxclS 16 :es
lypical possibility of 1 1 �bd2, 1 2 1i'c2 fol- i.xf3 (not 16 ...d4? 17 i.xf6 i.xf3 1 8 'ii'x f3
lowed by the lunge with the b-pawn, and a i.xf6 19 lidS clxc3 20 :Xd8 llfxd8 21 bxc3
rnradoxical idea of 1 1 �fd2!? followed by f2- with a big advantage for White, J.Polgar-
1'3. In the main line with tO 1i'b3 0-0 1 1 Kamsky, Las Palma.-; 1 994) 1 7 'if'xO �c6 1 8
t1�bd2 Black has two main options: the king's i.xf6 i.xf6 19 llxdS, when White's pieces
removal from the dangerous a2-g8 diagonal are sufficiently active to force an advantage

n2-g8 with 1 1 ...';;,h8 (Game 1 7) offers a sacri- even in this symmetrical position.
nee of the b-pawn, while Botvinnik proposed However, Black should seriously consider
prot1.:cting b7 with 1 1 ...�5 (Game 1 8). 10 ...�xd4 here: 1 1 cxd5 i.xf3 (1 1 ...c5?! is
inaccurate: 1 2 i.bS+ WfB 1 3 .ie2 and White
Game 13 has a useful k·ad in development) 12 gxO
Shirov-Vusupov c5!? 13 dxc6 �xc6 14 i.bS 0-0 15 '5'xcl8
European Team Ch., Batumi 1999 i.xd8 16 .i.xc6 bxc6 17 i.c3 i.b6 1 8 .J:r.ad l
-------• llfd8 and the game was completely level in
1 e4 e5 2 li'Jf3 li'Jf6 3 li'Jxe5 d& 4 li'Jf3 J.Polgar-Karpov, 1-loogt:Veen 1999.
�xe4 5 d4 d5 6 .i.d3 liJc& 7 0-0 .i.e7 8 1 1 1fxf3 lt'lxd4 1 2 'W'd1
.l:le1 ..llg4 9 c4 li'Jf6 White can also try 1 2 'iVg3, when 1 2...�e6
Trying to win a pawn immediately is a is a little better for White after 1 3 .i.f5 0-0 14
mistake: 9...i.xf3?1 1 0 'if'xO �xd4 11 ..d cxd5 �xd5 1 5 �3 �f6 1 6 i.e3. However,
li:\15 12 1ih3! �fd6 1 3 cxcl5 �f6 1 4 i.g5 Black is okay after 1 2...dxc4 13 i.xc4 (Kar-
�.:'lxd5 1 5 �c3 �fB 1 6 llxe71 �xe7 1 7 �5 pov pointed out the losing blunder 13 'iVxg7?
�)xdS (this gives up material and admits that �0+1 14 �hl llg8 1 5 'ifxf6 �xc1) 1 3...0-0
It has all gone wrong; if Black tries 1 7 ... f6 1 4 .tg5 i.d6 1 5 �4 h6 1 6 i.xf6 1i'xf6 1 7
rhcn 1 8 �xf6! gxf6 1 9 i.xf6 W f7 20 •o is 1i'xf6 gxf6, Lobron-Karpov, Hanover 1 983.
fnr too strong) 1 8 i.xd8 llxd8 1 9 �4 �f6 1 2 li'Je& 1 3 .i.f5
.••

20 1t'd4 and White was much bcttet· in Lo­


hron-Handoko,Jakarta 1 983.
1 0 l0c3

After 1 3 cxdS Black defends successfuUy:


1 3...�d5 1 4 i.b5+ c6 1 5 lllxd5 cxb5 1 6
.i.f4 �xf4 1 7 llxe7+ (not 1 7 �xe7?! �6 1 H

35
The Petroff Defence

tllf5 0-0 and Black keeps a useful extra 21 llxc7+ 11'xc7 22 .rlxe7+ Wxe7 23 �+
pawn) 17...¢'£8 18 lies •d6 19 %15 lids 20 �f6 24 11'xb7 %lfc8 and White has good
lOe3 ..xdl+ 21 %lxd1 :lxd l+ 22 c!tlxd1 etle6 winning chances.
23 %lxb5 b6 24 ttlc3 q;e7 and unswprisingly 1 8 -*.e3
a draw was agreed in Kasimdzhanov­
Yusupov, Essen 2001 .
1 3 dxc4
•.•

Much riskier is t3...d4 14 .ixe6!? (this is


more dangerous than 1 4 ttle2 d3 1 5 tC!f4
ttld4 16 .ixd3 0-0 1 7 .ie3 .ib4 1 8 ltft c5
19 tlle2 ttlg4, which was unclear in J.Polgar­
Shirov, Prague 1999) 1 4... dxc3 (or 14... fxe6
15 tC!bS e5 16 %lxc5 a6 1 7 etlxd4 c5 1 8 .ih6!
and White has a vicious attack) 15 ..xdS+
:lxdS 1 6 .i.f5 cxb2 17 .ixb2. White has
good compensation: the bishop pair, a lead in
development, and the annoying pin on the e­
fale. 1 8...-*.f6
1 4 ...84+ Another option is 18 ...Wh8, for example
19 etlxd5 cxd5 20 ..d3 .i.f6 21 :ldt 1Wa5 (or
21 ....i.xb2 22 ..xd5 ..xd5 23 llxd5 b6 24
llc7 �g8 25 g3 and White's activity gives
him a smaU endgame advantage, Magem­
Lal.ic, Seville 2000) 22 �3 :ladS 23 .id2
..c7 24 .i.b4 llfe8 and White again has a tiny
edge, this time due to his better structure.
1 9 -*.c5 J:lf7 20 J:lee1
This is more conttoUed than 20 %ld6. For
example, 20.....c7 21 .rlc1 .i.xc3 22 bxc3
llc7 23 lidS+ %lxd8 24 %lxe7 b5 25 �3
(after 25 %lxc7 bxc4 26 %lxc6 c!tlb4! 27 %ld6
%lxd6 28 .ixd6 ltJxa2 Black's dangerous a­
14 c6
••. pawn gives him rhe edge) 25 ...'iff4 and Black
14.....d7?1 sees the queen stepping into had serious coumerplay in Magem-IUescas.
ttouble: 1 5 ..xc4 lids 1 6 .i.g5 and White has France 2000.
n dangerous initintive. For example, 16 ... h6 20 .. ixc3 21 bxc3 h6 22 ha711'
17 .i.xe6 fxc6 lS .ixf6 .i.xf6 1 9 %lad1 ..cS This trick gains White a small advanrage.
20 %lxe6+ 'i\lf8 2 1 ltxd8+ 11'xdS 22 ttlds and 22... llK:71 23 J:le7 •d5 24 1i'xd5 �xd5
White's attack crashes through. After 24...cxd5 25 :Xf7 q;xf7 26 .id4
1 5 -*.xe6 fxe6 1 6 •xc4 ll:ld5 %lxa2 27 :lbl c!tlb5 28 .ie5 Yusupov still
Black's simplest plan is 1 6 ...0-0 17 ..xc6+ believes White is a bit better.
:n 18 .i.g5 1l'd6 1 9 1Wxd6 .i.xd6 with (..'llSY 25 J:le8+ J:lfB 26 llxf8+ Wxf8 27 -*.c5+
equality. �8 28 c4 lDf4! 29 -*.d6
1 7 llxe6 0-0 After this Black equalises. White could
There is no time for 17...tC!xc3? because keep pressing with 29 .ie3! ttld3 30 %le2
of 1 S .i.g5 etld5 19 .i.xe7 c!tlxe7 20 :lae 1 %1£8 c!tlb4 31 f3 %lxa2 32 Axa2 c!tlxa2 33 .i.d2 Wf7

36
3 liJxe5: The Main Line with 8 :e 1

'4 Wf2 �e6 35 «li'e3 Wd6 36 Wd4 (Yusu­ bxc6 16 .i.g5 l:tfe8 with an unclear position)
pov). 13 lllc3 'iFd7 14 'iFxd7+ �d7 1 5 .i.e3 llle6
29...liJd3 30 lte2 :dB 31 i.c7 ltd4 32 1 6 l:tad1 .i.d6 17 .i.f5 We7 1 8 lllb5 Zl.hd8 1 9
g3 % -% lllxd6 cxd6 20 h 3 (White has obvious com­
pensation for the pawn) 20...b6 21 g4 h6 22
Ga1f/e 14 .i.d4 l:tac8 23 .i.c3 g6 24 .i.c2 hS 25 f3 1/z-1/z
Sax-Vusupov Kasparov-Karpov, World Championship
Thessaloniki Ofyntpiad 1984 (Game 28), Moscow 1984.
_______________. b) 12 'iFxdS should not be dangerous:
1 e4 e5 2 liJf3 lilf6 3 liJxe5 d6 4 liJf3 12 ...lllxd5 13 lllc3 0-0-0 (Black is well ad-
�xe4 6 d4 d5 6 i.d3 liJc6 7 0-0 i.e7 8 vised not to chase material with 1 3.. llldb4 14
.

J:le1 i.g4 9 c4 liJf6 1 0 cxd5 .i.e4 lllxd4 because after 15 .i.e3 lld8 1 6
..ixd4 l:r.xd4 17 a3 lllc6 18 lllb5 White will
win back the pawn and force an advantage)
14 .i.e4 i.b4! 1 5 i.d2 lC!f6 1 6 i..xc6 bxc6 17
..ie3 lCJds 18 a3 (Kamsky-Karpov, Linares
1 994 continued 1 8 %tact lllxd 1 9 fxe3 cS 20
l:tft f6 21 llf5 Zl.he8 and Black had an edge,
which grew considerably after 22 dxcS?
llxe3) 1 8...i.xc3 19 bxc3 lC!xc3 20 llact
lC!bS 21 llxc6 lC!xd4 22 Zl.c4. Despite Black's
extra pawn Karpov assessc..-s the position as
el(Ual.
1 2...11fxd4 1 3 lilc3
After 13 'iFxc7 one of BL1ck's options is to
10 i.xf3
••. force e']uality with 1 3...lllg4!? 1 4 'iff4 'ifxd3
This is much safer than 10 ...'iFxd5 1 1 lllc3 1 5 'ifxg4 0-0.
.ixf3 Qf 1 t ...'iFh5 1 2 i.b5 ..txf3 13 gxf3 �fB 1 3 0-0 14 liJb6 .g4
...

14 i.xc6 bxc6 15 i.f4 White's easier devel­ Black has a sound alternative in 14 ...1i'b41?
opment is worth a slight advantage) 1 2 lC!xd5 I S i.gS llfd8 1 6 'iFh3 '!fl 1 7 lllxc7 IL1cR 1 8
i.xd1 1 3 lllxc7+ Wd7 14 i.f41 ..tg4 (Kavalck i.xf6 .i.xf6 1 9 llld s 'iFd6 20 lllxfM 'ifxf6,
nnnlvsed 14...lC!h5 15 i.f5+ Wd8 1 6 i.c5 achieving easy equality.
li'lx�S 1 7 lllxa8 lllc6 1 8 .llaxdt i.d6 1 9 Zl.e5! 1 5 11fxg4
as winning for White) 1 5 d5 lCJd4 1 6 lllxa8 1 5 lllxc7 Zl.ad8 1 6 'ifxg4 lllxg4 1 7 i.e2
l:lxa8 1 7 i.c51 .i.f5 (or 1 7 ... ..tc5 1 8 llect lC!ge5 1 8 i.f4 ..if6 is also level.
.ib6 19 llc4 and White still has a clear ad­ 1 5...liJxg4 1 6 i.f5?1
vantage) 18 i.ft lllc2 1 9 i.bS+ Wd8 20 d6 Ths is inaccurate; 16 ..te2 lC!f6 17 lllxc7
ll'lxct 21 llxe l i.e6 22 dxe7+ Wxe7 23 .i.d4 llac8 l8 lllb5 ..ib4 is simply eCJual.
h6 24 a4 (Kavalck-Toth, Haifa 1976), when 1 6 �f6 1 7 liJxc7 lladB 1 8 i.e3 a61
•••

rhc tactics have burned out but White's A later game, Kamsky-Khalifman, FIDE
llishop pair still ensures a small advantage. World Championship, Las Vegas 1999,
1 0.. lllxd5 is studied in Game 1 5.
. continued 1 8.....ib4 1 9 lled1 g6 20 i.h3
1 1 •xf3 •xd5 12 •g3 a6 (or 20....i.d6 21 lllbS ..ie.S 22 lllxa7 lllxa7
White has two other moves: 23 .i.xa7 Zl.xdl+ 24 llxdl l:ta8 25 -'.c5
a) 1 2 .,3 lllxd4 (also playable is ..ixb2, again with equality) 21 i.h6 llxdl+ 22
12 ....xd4 13 lllc3 l:td8 14 i.b5 0-0 15 i.xc6 Zl.xd 1 l%d8 23 llxc.l8+ lllxd8 24 i.g5 i.e7

37
The Petroff Defence

with a level position. l0xd5 33 cxd5 b5.


1 9 1lac1 31 Wxe5 32 lDa8 b5 33 a4 o!bd5 34
•••

Perhaps White should try 1 9 :ed1 .i.d6 llb3 bxa4 35 llb7 l:tb1 36 �3 a3 0-1
20 .i.b6 .i.e5 21 llab1 (after 21 l:txd8 :Xd8
White has some problems since 22 l0xa6? Gan1e 1 5
fails to 22...:d5) 21 ..J:lxd t + 22 :xd1 .i.xb2, Psakhis-Cooper
when White's activity is enough to hold the Pott Erin 2003
balance.
1 9....tb41 20 llf1 lbd4! 1 e4 e5 2 o!bf3 lDf6 3 lbxe5 d6 4 o!bf3
Ths keeps up the pressure. 20..id2?! al­ o!bxe4 5 d4 d5 6 .td3 lbc6 7 0-0 .te7 8
lows White to complicate matters with 21 lle1 .tg4 9 c4 lDf6 1 0 cxd5 fud5 1 1
l:lxc61 bxc6 22 .i.c5. o!bc3 0-0 1 2 h3
21 llc4?1
Now Black's advantage is serious. After 21
.i.xd4 l:lxd4 22 l:lfdt llfd8 23 llxd4 :Xd4 24
�fl .i.d6 Black has very little.
21 ...o!bxt5 22 llxb4 l:td7! 23 :C1
Black's advantage is confmned by the fact
that White cannot grab on b7: 23 llxb7?
loses to 23...l0xe3 24 fxc3 llc8 25 llcl l0e8,
while the line 23 .i.f4? lieS 24 llxb7 li)d4! is
also hopeless.
23 . .llc8 24 llbc4 l:tcd8 25 h3 lbxe3 26
.

fxe3 �8 27 e4 �7

Twenty years earlier Ps.1khis had tried 12


.i.e4 .i.e6 (after 1 2..l0f6 Zak assessed 1 3 d5
l0b4 1 4 a3 l0xe4 15 llxc4 .i.x£3 16 'iVx£3
li)a6 17 b4 as clearly better for White) 13
'it'd3 h6 14 .i.h7+ (if 14 'lt'b5 Black draws
with 14 ...�b4!? 1 5 'lt'xb7 l:lb8 1 6 '1t'xa7
liaS) 1 4...�h8 1 5 .i.fS l0cb4 1 6 -.,1 .i.xfS
17 .xfS l0f6 18 .i.f4 l0bd5 19 .i.e5 c6 with
an equal position, Psakhis-Schussler, Tallinn
1 983.
1 2....te6
Ths is the solid option, but two others
28 l:tb4 moves are worth mentioning:
If 28 c5 then 28 ..l0d5 29 l£lxd5+ llxdS 30
. a) 1 2...i.h5? is just a blunder: 1 3 .i.xh7+!
llg4 g6 and Black has good chances of win­ �xh7 14 lbgS+ .i.xg5 1 5 'lt'xhS+ �g8 1 6
ning the rook ending. .i.xgS l0 f6 1 7 .i.xf6 'it'xf6 1 8 llkls '1t'xd4 1 9
28...l:td1 + 29 llxd1 llxd1 + 30 �2 Wd6 lt.d1 br6 20 Axd4 gxh5 2 1 llh4 and White
31 e5+? had a huge advantage in Yemelin-Ka?.akov,
Now it is all over. Also hopeless is 31 St Petersburg 1996.
l:lxb7 'ii>c6 32 lla7 �b6. but White can put b) More serious is 1 2... -*.x£3, although
up more resistance with 31 �e2 llat 32 lft!5 White should have a slight edge. For exam-

38
3 l/Jxe5: Th e Main L in e with 8 Ae 1

pic, 1 3 'i6'xf3 ll:ldb4 14 .ibl l:e8 I S dS .id6 1 6 .. .'�h8?!


I (, l:xe8+ 1fxe8 17 .ie3 lbes 1 8 'i6'e4 ll:lg6 Black could have limited his disadvantage
1 9 a3 ll:la6 20 i.c2, Dolmatov-Yusupov, with 16 ...lL\e7 1 7 lLlcS lLlc..iS 1 8 .i.d2.
1 1nm7.c 1 98 1 . 1 7 ..i.e4 ..i.d7 1 8 l/Je5
13 a3 l/Jxc3 Now Black has to make a concession.
Black has also tried the immediate 1 8. . ....i.xe5 1 9 dxe5 •ea 20 ..i.g5 •e6
l l..i. f6, for example 1 4 &4 .t.r.; 1 S Wb3 After 20 ..h6 White docs not need to re­
.

�b6 1 6 dS .ixe4 17 .i.xc4 ll:le7 1 8 .i.e3 treat: 21 .i.f6! l:tb8 22 9h5 �g8 23 lte3 and
li.)cc8 (the greedy 1 8 ...lLlexdS docs not work: the attack is devastating.
I IJ .ixb6 lLlxb6 20 .i.xb7 l:b8 21 l:tad1 lL!c.I7 21 �5 h6 22 ..i.f61 �g8 23 Ae3 l/Je7
22 'il'bs l:te8 2.1 :Xd7 l:txel+ 24 lLlxet •e8 24 Ag3 l/Jg6 25 Axg6 fxg6 26 •xg6 1 -0
25 �f3 l:txb7 26 ...xb7 1fxd7 27 ..xa7 and
White has good winning chances) 1 9 l:tacl Game 16
�16 20 .i.bt 'i6'd7 (Shirov-I.Sokolov, FIDE Leko-Yusupov
World Championo;hip, Las Vegas 1 999). Dottmund 1998
Now Ivan Sokolov suggests that 21 .i.gS!?
i.xgS 22 lLlxgS g6 2.1 ...g3 gives an edge to 1 e4 e5 2 liJf3 li.:lf6 3 l/Jxe5 d6 4 lDf3
White. lDxe4 5 d4 d5 6 ..i.d3 ..i.e7 7 0-0 lDc6 8
14 bxc3 ..i.f6 Ae1 ..i.g4 9 c3 f5
This is more aggressive than 9. .ll:\f6 lO
.

.i.gS 'i6'd7?! (Black should play 10...0-0 but


White's position is preferable after 1 1 lbbd2
.i.hS 12 Wb3 lLlas 1 3 'i6'c2 .i.g6 1 4 lLleS) 1 1
lL!bd2 (White could also choose to shatter
Black's kingside with 1 1 .i.xf6: White is bet­
ter after 1 1 ...gxf6 1 2 lLlbc12 0-0·0 13 .i.bS)
1 1 ...0-0-0 1 2 'i6'a4 h6 1 3 .i.h4 g5 1 4 .ig3
.i.xf3 (or 14 ... i.d6 1 S .i.bS and Black is still
in trouble because 1 S ...�b8? loses to 16 lL!eS
i.xeS 17 .i.xeS) 1 5 lLlxf3 g4 1 6 lbest lL!xeS
1 7 i.f51 1i'xf5 1 8 l:txeS ...d3 (l!L1i'd7 1 9
'i6'xa7 'i6'd6 20 l:te3 is also winning) 1 9 l:txe7
1 5 Ab1 :d7 20 l:te3 1i'a6 21 .xa6 bxa6 22 .i.eS 1 -0,
This is natural, but playing IS 'i6'c2 g6 and Keres-Alexander, Hastings 1 9S4/S5.
only then 1 6 l:tbl is also sensible. Now if 1 0 lDbd2
Black plays as in the game with 1 6 ... b6?!, 1 0 h3 i.xf3!? offers a promising pawn
White has 1 7 .i.xg6! hxg6 1 8 l:txe6 fxe6 1 9 sacrifice: 1 1 gxf3 (if White declines the offer
.x�:,r6+ �h8 20 llbS with a crushing attack. with 1 1 ...xf3 0-0 1 2 ..i.£4 .i.d6 the game is
Instead Vasiukov-B.Vladimirov, USSR 1981 unclear) l t ...lLlf6 1 2 i.xf5 0-0 13 Wd3 i.d6
continued 16 ....i.dS 1 7 lLld2 b6 1 8 lLlc4 lOc7 14 .igS llJc7 1 5 .i.e6+ �h8 1 6 lLld2 lL!g6 17
1 9 lLle5 .i.g7 20 c4 .i.b7 21 .i.b2 and White .i.f5?l (this lets Black take control; White had
hnd a pleasant advantage. to try 1 7 lL!b3 even though Black has com­
15 b6 1 6 ..i.f4
.•• pensation after 17 ... .i.f4 1 8 •rs .i.xgS 1 9
White could also try 16 ...c2 h6 1 7 .i.f4 'ifxgS 1i'd6) 1 7....i.f41 1 8 .i.xf6 (or 1 8 .i.xg6
.idS 1 8 .i.h7+ <t>h8 1 9 .i.e4 with a promis­ i.xg5 1 9 ..i.f5 lL!hS and Black's occupation
ing position. of f4 will be very threatening) 1 8......xf6 1 9

39
The Petroff Defence

.ixg6 .txd2 20 :ez .i.f4 21 .i.xh7 'Wg5+ 22


�h1 1ih5 0-1 Enklaar-Dvorctsky, Wijk aan
Zee 1 975.

1 1 ...0-0?1
Now castling short is a mistake - Black
has two better moves:
1 0.....d6 a) 1 t ...ltlxd2 12 .i.xd2 dxc4 (not
Possibly preparing to castle long. Instead 12...ltlxd4? 13 lihd4! .i.xdt 1 4 ltlxf5 1fc5 1 5
1 0... 0-0 is also reasonable, with the foUowing: b4, when White wiU have a decisive material
a) t 1 '1Fb3 transposes to Game 17. advantage) 1 3 i.xc4 0-0-0 14 .i.c3 .i.f6 with
b) 1 1 lbft with a further split: an unclear position.
bt) t t ....i.d6 is tricky but probably inaccu­ b) 1 1 ...0-0-0!? leads ro sharp play: 12 cxd5
rate after 1 2 ll)g3 (not 12 lbe3? .i.xh2+1, ltlxd2! (once again 12 ...ltlxd4? is a mistake:
when 13 �? is imposs.ible because of 1 3 lihd41 .txd1 14 ltlxfS ..eS 1 5 ltlxe7+
13...l£lxf2 1 4 ..c2 ltlxd3 1 5 1Wxd3 .i.xt3 ..xc7 1 6 .i.xe4 and White's extra material
winning for Black - the problem with 1 6 ensures a clear advantage) 1 3 dxc6 ltlxf3+ 14
gx f3 is 16...1ih4+, winning the rook) 12.....f6 gxf3 i.h3 1 5 cxb7+ Wb8 16 .tn i.xfl 17
1 3 '1Fb3 and White is a bit better. D.xft i.f6 with a very mes.c;y position (Leko).
b2) t t .. .th4 12 g3 .tgs 13 .txg5 l£lxg5
. 1 2 cxd5 c!Dxd4
14 .i.e2 l£lh3+ 1 5 �g2 f4 16 ltl3d2 .i.xe2 17 White a bit better after 12...l£lxd2 1 3 dxc6
1fxe2 ltlg5 1 8 1Wg4 fxg3 19 hxg3 ..f6 with l£!xf3+ 1 4 gxf3 i.hS 1 5 cxb7 llab8 1 6 i.c4+
an unclear pos.ition, Ljubojevic-Hjartarsson, ll.f7 17 'irb3.
Belgrade 1989. 1 3 .*.xe4l fxe4 1 4 1llxe4
1 1 c41?
The other option is 1 1 ltlft. For example,
1 1 ...0-0-01? 1 2 ltk3 h5!? 1 3 h3 g6! 14 hxg4
(mstead 1 4 i.c2 i.xt3 1 5 i.xf3 .th4 1 6 1le2
lbe7 1 7 l£lf1 i.f6 1 8 1lc2 'itb8 19 a4 1lh7 20
aS g5 was unclear in Adams-Makarychev,
Oviedo 1992) 14...hxg4 15 .txc4 (White
must avoid 1 5 ltld2? because of 1 5...11fh2+
1 6 'itft ltlxf21 17 'iPxf2 .th4+ 18 Wc2 1ldc8
with an immense attack) 1 5...dxc4 1 6 ltlc4
..f6 and Black has good compensation - 1 7
.i.gS? loses to 1 7...1fg7 1 8 Le7 1ih6 19
l£lh4 l£lxc7 20 g3 g5.

40
3 '!Jxe5: The Main L ine with 8 .l:l e 1

14 .i.xf3
..• 49 'it>e3 .l:lb3+ 50 �4 .l:lb4+ 51 'it>g3
Leko assessed 14...�xf3+ 1 5 gxf3 .i.xf3 White prefers to avoid the pawn ending
H i �xd6 .i.xdl 17 �xb7 l:labB 18 l:he7 51 WgS ltbS+ 52 �h6 1lxh5+ 53 WxhS even
:xb7 1 9 d6 as clearly better for White. !hough he wins neatly after 53 ... b5 54 f4 b4
1 5 1i'xd4 1fb4 1 6 1t'xb4 ..txb4 1 7 .l:le3 ss 5 b3 56 f6 b2 57 fl b1W ss f89+ Wa2
.i.h5 1 8 .l:lb3 .i.a5 59 Wa8+.
White also has a clear plus after 1 8...a5 19 51 b5 52 h4 .l:lb1 53 .l:lxh7 b4 54
.••

a."\ .i.g6 20 f3 .i.xc4 21 fxe4 .i.d6 22 .i.d2. .l:la7 + �b2 55 �4 .l:lh1 56 'it>g5 b3 57
19 ..te3 ..tf7 20 .!Lig5 h5 'it>c3 58 .l:lc7 + �d4 59 .l:lb7 Wc3 60
Or 20 llxb7 .i.xdS 21 :lbS .i.xc4 22 l:lxaS h6 b2 61 g4 b11i' 62 .l:lxb1 .l:lxb1 63 h7
n6 23 :lc1 and, despite the opposite-coloured .l:lb8 64 �6 .l:lb6+ 65 �g7 .l:lb7+ 66
bishops, �'hitc has an obvious advanta�:,>c. ¢>h6 .l:lb6+ 67 Wh5 .l:lb8 68 g5 1 -0
20 ..tb6 21 .!Lixf7 .l:lxf7 22 ..txb6 axb6
.••

23 d6! cxd6 24 .l:lxb6 .l:lc7 25 �1 .l:la6 Game 1 7


26 .l:lb4 lvanchuk-Shlrov
The direct 26 :Xa6 should also win. For Dortmund 1998
example, 26 ... bxa6 27 b4 :lc2 28 a4 :lb2 29
bS axbS 30 aS l:lc2 31 a6 :lc8 32 a7 :laB 33 1 e4 e5 2 ltlf3 .!Lif6 3 d4 ltlxe4 4 .!Lixe5
.Ji!c2 etc. d6 5 .!Lif3 d5 6 ..td3 &6 7 0-0 .i.e7 8
26 �7 27 a3 d6 28 .l:ld1 �6 29 .l:lb5
•.. .l:le1 .i.g4 9 c3 f5 1 0 'iJbd2 0-0 1 1 'irb3
lld6 30 'it>e2 'it>e5 31 'it>d3 .l:ldc6

11 •.• Wh8
32 �3 Sacrificing a pawn. 1 1 ...�5 will be stud­
Or 32 :lel+ �d6 33 l:le8 :lcS 34 l:lxcS ied in the next game.
l:.xcS 35 l:lb8 and White's advantage is t:vi­ 12 ..xb7
tlcnt. White can opt to include 12 h3 .i.hS be­
32 .l:lc5 33 .l:lxc5 .l:lxc5 34 'it>d3 .l:lc4 35
..• fore playing 13 Wxb7. For example, 13 ...l:lf6
lle1 + Wd6 36 .l:le8 .l:lf4 37 f3 .l:lh4 38 h3 1 4 9b3 l:lg6 (or 14 ...g51? 1 5 .!?lfl l:lbB 1 6
b6 39 .l:lb8 'it>c5 40 .l:lc8+ 'it>d6 41 .l:lf8 g5 Wc2 with a murky position) 15 .i.e2 .i.d6
42 .l:lf6+ 'it>c5 43 b4+ 'it>b5 44 .l:lf5 �a4 (1 5 ... .i.h4?! is the wrong way to build up the
45 .l:lxg5 'it>xa3 46 .l:lxd5 attack: 1 6 l:ln .i.xf3 1 7 o!?lxf3 and White is a
46 bS �b4 47 l:lxdS is even more straight­ safe pawn up because 17....i.xf2+?! fails to 1 8
forward. llxf2 .!?lxf2 1 9 � f2 Wd6 20 o!?lgS! l:lf8 21
46 .Axb4 47 .l:lh5 llb3+ 48 �e4 .l:lb4+
•• 9a3 Wd8 22 .i.f4 h6 23 o!?lf3 lle8 24 .i.d3,

41
Th e Petroff Defence

Katpov-Korchnoi, 6th matchgame, Moscow Bayramov, Batumi 2001) 1 7 .ixc6 tlg6 1 8 g3


1 974) 1 6 ltk5 I!Dxe5 17 .i.xh5 ltxg2+! 1 8 .i.e4 and Black's kingside play provides ob -
�xg2 .gS+ 19 �ft 1Wh4 20 I!Dxe4 (not 20 vious play for the pawn.
ltxe4? 11'xh.J+ 21 Wg1 �! and Black wins) 1 5....th5 1 6 .i.e2 llg6
20...1i'xh}t 21 �g1 I!Df3+ 22 Lf3 .i.h.2+
and Black has a perpetual check (O'Kelly).
1 2...1lf6

1 7 Wf 1 17
Trying to escape the pressure along the g­
file. In this complex position White has two
1 3 ...,5 other main options:
The other retreat 1 3 'tl'b3 also allows a) 1 7 I!Dxe4 leads to very unclear play after
Black counterplay: 1 3 ...ltg6 (1 3...ltb8 14 .a4 1 7 ... fxe4 1 8 li)gs :Xgs 1 9 .ixgs •xgs 20
simply transposes to the game) and now: .i.xhS lik71 21 .d 1 g6 22 .i.g4 h5 23 .i.d7
a) 1 4 .ib5 I!Dxd2 1 5 I!Dxd2 .i.d6 1 6 g3 (or lids 24 .i.b5 I!D5.
1 6 .ixc6 Lh.2+1 17 Wxh.2 1Wh4+ 1 8 �gl b) Curiously, 1 7 lbcs allows Black a
lth6 1 9 f3 •xel+ 20 .!Dn lthl+l 21 �ht choice of forced draws: 17 ...�e5 1 8 dxe5
.xft+ 22 Wh2 .i.xf3 and now White should .ixe2 19 exd6 ltxg2+! 20 Wxg2 .gS+ 21
accept that Black has a perpetual check be­ �h 1 I!Dxf2+ 22 Wh2 f4! 23 lOft � 24
cause if 23 'ii'c2? .ie4 24 .d2 :ciS Black has hxg4 1Wh4+ 25 Wg2 1i'xg4+ with a perpetual
a deadly attack) 1 6 ...ltk7 1 7 .i.d3 h5 1 8 .!Dfl check, or 1 8...•g5 1 9 .if1 I!Dxf2! 20 exd6
h4 and Black has a dangerous initiative, .!l)xh.}t 21 Wh1 I!Df2+ (lvanchuk).
Peshkov-Raetsky, correspondence 1985. 1 7 -*.14 1 8 �3 .i.xc1 1 9 llaxc1 •d6
.•.

b) 14 g3!? 1:lb8?! 1 5 .c2 .i.d6 1 6 b41 20 1lc2


(Movsesian pointed out that the imprecise 16 The ambitious 20 .ib5?1 backfires after
.ift ..f6 17 .i.g2 ltfB would allow Black un­ 20.. .:Xg2! 21 Wxg2 ..g6+ 22 Wh2 .ixf3 23
deserved compensation) 16.JWf6 17 b5 ltk7 ltgt 1i'd6+ 24 .l:[g3 �7 25 :n hS, after
18 � .ixeS 19 dxeS and White was clearly which Black's attack is very strong.
better in Movsesian-Kroeze, Nethedands lvanchuk analysed 20 I!Dh4 as leading to a
2001 . More accurate is 1 4...•d6 1 5 .ib5 ltb8, draw: 20...•f4 21 I!Dxg6+ hxg6 22 f3 l0c5! 23
when Black has some compensation. 1i'xa7! I!Dxf3! 24 1i'xb8+ Wh7 25 .ixf3 .ixf3
1 3 1lb8 1 4 1ra4 i.d6 1 5 h3
•.. 26 :Xe4 .ixe4+ 27 �e2 .g3 28 .l:[g1 .d3+-
Or 1 5 .ibS I!Dxd2 1 6 .i.xd2 .ixf3 29 Wf2 f4 when to avoid mate White must
(1 6...ltk7?! is too passive: 17 �5 a6 18 .if1 play 30 .!DeS and allow perpetual check.
:xb2 1 9 .tgs .i.h5 20 .ixf6 gxf6 21 •d7 is 20 11ff4 21 ,j_d1 1le8 22 libl2
.••

much better for White, Gagunashvili- It is vital to support f3. If 22 �5?! then

42
3 liJxe5: The Main L ine with 8 lle 1

22... ..i.xf3 23 ..i.xf3 'Wh2 24 g3 .1Lg31 25 41 'it'e8+ �h7 42 'it'xc6 'ti'd2+ 43 'iVg2 e2 44
.lxe4 'ti'xh3+ 26 We2 .J:g6 and Black's attack liJgS+ 'it'xg5 45 'it'xg5 e11i' 46 'it'hS+, with
i� venomous. White having reasonable winning chances.
33 hxg4
After 33 fxg4? Black can draw with
33...Wfe4+ 34 tll f3 l:txg4+! 35 hxg4 'ifxg4+ 36
'ith l 1i'xf3+ 37 .llg2 c2 38 �gl .e3+ 39
'itht •n.
33 ...llh6 34 1txd5 /:i)e7 35 •e5
White had a faster win with 35 'ild7! l:tb8
36 'ifxe7 llxb2 37 l:lxb2 llh2+ 38 Wfl l:txb2
39 l:lxe3.
35 ....-xe5 36 dxe5 /:i)d5 37 �g3 lie&
38 /:i)h3 llxe5 39 /:i)f4 /:i)xf4 40 �xf4 lla5
41 a3 llb5 42 �xe3 llcb6 43 b4 liaS 44
Jla1 c5 45 llb2 cxb4 46 llxb4 llba5 47
22 1lge6 23 liJxe4 fxe4 24 g3! 1tf5 25
•.. a4 llc5 48 Jla3 1-0
t;)g1 .i.xd1 26 llxd1 e3?1
This is the wrong approach. 26...h5 leads Game 18
In an unclear position after 27 c4 e3 28 cxd5 Anand-Gelfand
1i'xd5 29 'it'c4 liJe7. However, the best line is Moscmv 2004
1hc simple 26...l:tf6 27 l:te2 l:tcf8 when Black
has a definite edge. 1 e4 e5 2 .!Llf3 .!tlf& 3 c!Dxe5 d6 4 .!tlf3
27 �2 1We4+?! o!Llxe4 5 d4 d5 6 .i.d3 .i.e7 7 0-0 lbc& 8
One slip follows another... After 27 ...l:tffi l:te 1 .i.g4 9 c3 f5 1 0 'lrb3 0-0
2flf3 fi:k7 29 l:te2 White would have b�o.-cn A major alternative here is lO...'ifd6 with
only slightly better. the idea of ca!ltling lJUeenside. White has
28 f3 1Wf5 29 h2 1lg6 30 llde1 1Wf4 31 three main answers:
g4 h5 32 'lrb51 hxg4

a) Taking the pawn with 1 1 'lrxb7 should


A better try is 32 ... l:tcd)!? when lvanchuk certainly be analysed, but Black has good play
analysed 33 'lrxd5 hxg4 34 hxg4 l:tx�:,t4+ 35 after 1 1 ...l:tb8 1 2 'ifa6 ..i.xf3 1 3 gxf3 0-01? 1 4
fKg4 'ifxg4+ 36 �h 1 l:th6+ 37 l:th2 llxh2+ 38 fxe4 (not 1 4 ..i.bS?! l:tb6 1 5 'lra4 .i.h4 with a
Wxh2 'irh4+ 39 liJh3 'it'xe1 40 'it'hS+! Wg8 dan�rous initiative; for example, 1 6 fxe4?

43
The Petroff Defence

allows a winning attack after 1 6....txf2+! 17 .!DeS i.xf2+ (White is also clearly better after
�£2 9xh2+) t 4...fxe4 1 s .tn .th4. the alternatives 17...f4 1 8 :e6 9d8 19 l0d3
b) 1 1 .!Dfd2!? is interesting but also poten­ f3 20 g3 and 17 ...lbd8 18 .tn b6 19 .!Ddl
tiaUy risky: 1 1 ...0-0-0 1 2 f3 i.h4 (the tricky .!De6 20 1tb4! - Anand) 1 8 Wxf2 ....2 19
12 ....!De5?! fails to 13 i.xe4 dxe4 1 4 fxg4 i.xc6 bxc6 20 11'xc6 f4 21 ..xdS+ Wh8 22
.!Dxg4 1 5 .!Df1 and White is clearly better - 1WxhS f3 23 11'xf3! :Xf3+ 24 Wxf3 and White.
Timman) I 3 :n and now: has cleverly consolidated his advantage, An·
bl) The creative 1 3...i.h3?! is refuted by and-Kramnik, Tilburg 1998.
accurate defence: 1 4 ..c2 11'g6 1 5 .!Db3 :h£8 b) 1 3 ...i.h4! with a further branch:
16 .!Da3 llde8 17 i.f4! {but not 17 Wht??
.!D£2+! t 8 llxf2 i.xg2+! 0-1 Ivanchuk-Anand,
Reggio Emilia 1 989) 17 . .:e6 18 .!DbS and
.

White has a clear advantage.


b2) 1 3 ...i.f2+!? 1 4 :xf2 lOx£2 1 5 Wx£2
..xh2 1 6 .!Dn 11'h1 1? 1 7 fxg4 fxg4 and the
situation is impossible to assess with any
confidence: White has a large material advan­
tage but a dangerously exposed king.
c) 1 1 .!Dbd2 is the safest option. 1 1 ... 0-0-0
12 ..c2 (12 .lb5is less clear, for example
12..if6 13 1i'a4 :he81? I 4 i.xc6 bxc6 I 5
l0xe4 :xe4 16 :xe4 fxe4 17 .!Des i.xeS 1 8 .I
dxeS ..cS 1 9 i.e3 'irbs with an unbalanced bl) 14 g3 (this is usual) 14 ... i.f6 t S Let
position) 12..Wb8 13 b4 i.f6 1 4 bS .!De7 1 S bxc6 1 6 1i'xc6 :es 17 .!Des i.xeS 18 dxe5':
.!Des i.xeS 1 6 dxeS 1Wg6 1 7 .!Db3 lbgs 1 8 f4 lbgs 19 f4 (Kramnik mentions 1 9 c4!? d41�
.!De4 19 i.e3 and White had an edge in S7.e­ 20 WdS+ 1WxdS 21 cxd5 llad8 as still giving
lag-Naumann, Rostock 2002. compensation) 1 9....!Dh3+! (White has an;
1 1 lbbd2 lba5 edge after 1 9...:C6?!, for example 20....!De4?1;
21 l0xe4 dxe4 22 .i.e.� a6 - Shirov-Kramnik.j
Belgrade 1 999 - and now 23 :OCt i.f3 24J
:c2 would have clarified White's advantage).;
20 Wg2 :b8 21 c4 (21 .!Db3!? :b6 22 -�
d4! 23 c41 •c8 is messy) 2t ...dxc4 22 lbxc4
1t"d3 23 tbe3 (White has to watch out fen'
mating trick.o;: 23 i.d2? :eds 24 .tc3 •cZ+
25 Whl ..a4!) 23 .. 1E.xb2+! 24 .i.xb2 ..dz..+;
25 Wht l0f2+ 26 Wg1 .!Dh3+ 112-lfz Anand�
Kramnik, Wijk aan Zec 1999.
1
b2) 1 4 :n !? is also worth considering;
14....!Dxd2 1 S .!Dxd2 f4 1 6 i.xc6 bxc6 11,
.xc6 .gS 18 c4 (18 g3 allows Black to fore �
1 2 Wc2 a draw with 18...fxg3 19 hxg3 .txg3! 20 fxg�
12 1Wa4 is a very important alternative. 1i'e3+ 21 Wg2 9e2+) 18...1lad8 19 cxdS :d�
Play continues 12....!Dc6 1 3 i.bS and now: 20 1Wc3 llg6 21 .!Df3 Whs 22 .!Des f3 �
a) White has an easy time after 13 ....!Dxd2?! lbxg6 hxg6 24 .i.e3 .i.c8! is totally unclear�

j
14 .!Dxd2 '1Fd6 I S h3 i.hS 16 .!Db3 i.h4 17 Koziak-Motylcv, Lvov 1999.

44
3 lbxe5: The Main Line with 8 lle 1

keeps an edge after 15 g3 �g3 1 6 hxg3


i.xg3 1 7 lle2, whereas 1 5 fxe4 i.xel 1 6 cxfS
c4 17 i.fl .i.xd2 18 .i.xd2 �c6, Moro­
zevich-Adams, Dorttnund 2002, is rather
unclear.
1 3 b4
This is reasonable but perhaps White
should prefer 1 3 �5!? .i.xeS 1 4 dxe5 11Vh4
1 5 llfl with a pleasant edge.
1 3... ..i.xf31 14 l'Oxf3 l'Oc4 1 5 ..i.xc4
Or 1 5 a4 c6 1 6 11Vb3 b5 17 axb5 cxbS 1 8
lla6 llf6 with unclear play.
1 5 dxc4 16 ..i.e3 J:leB 1 7 J:lad1 •te 1 8
•••

1 2....i.d6 lbe5 ..i.xe5


This is sttongcr than 1 2...c5 1 3 llk:S .i.hS This is the obvious choice but 1 8.. .Axe51?
(not 1 3 ...cxd4?1 14 �xg4 dxc3 1 5 bxc3 fxg4 1 6 is also interesting, 1 9 dxe5 '1Vxe5 20 g3 1ixc3
ti)xe4 dxe4 17 .i.xe4 i.c5 1 8 i.xh7+ �h8 1 9 leading to a messy position.
.ic3 when White defends easily) 1 4 t3 cxd4 1 9 dxe5 •xe5 20 .i.d4 'MI5 21 a4 •d7
1 5 fxc4 fx<:4 (1 5...dxc3 is tricky but White White also has compensation for the
keeps the initiative with 16 cxd5!? cxd2 17 pawn after Black's two other options:
.ixd2 11Vxd5 1 8 i.xaS 11Vxa5 t9 li)d7 1lfe8 20 21 ...1i'a6 22 f3 �6 23 .i.c5 and 21 ...11Vc6 22
lieS) 1 6 �e4 dxe4 17 .i.xe4 h6 (Anand­ f3 �6 23 b5 'W'd7 24 .i.eS llad8 25 11Vf2.
Uoi<Jb>aO. Wijk aan Zee 2004) and here White 22 ..i.xa7 rod& 23 J:lxe8+ •xeS 24 .i.c5
can gain a large advantage with 18 b41 dxc3 19 •xa4 25 •xa4 J:lxa4 26 .i.xd& cxd6 27
!Jxa5 1id4+ 20 �hl 11Vxe5 21 .i.a.�!. Wf1 .IZa& 28 J:ld4 b5 29 J:ld5 J:la3 30
Black has also tried 14....i.h4 but White J:lxd& J:lxc3 Y.z Y.z
-

45
Th e Petroff Defence

Summary
Nowadays the active continuation 9 c4 docs not seem to offer White an advantage in the
opening. It can be recommended to fans of the isolated quecn's pawn, which arises after
9 ... l0f6 10 cxd5 l0xd5. However, we should not forget that 1 0... .A.xf3 1 1 1i'xf3 l0xd4 (Game
1 4) has proven to be a reliable way of equalising.
Despite initial appearances, 9 c3 is one of the sharpest system� in the Petroff Defence. Of
course, opposite side castling after 9. . . f5 1 0 'irb3 1Vd6 and ...0-0-0 leads to many complica­
tions. After 1 0 ...0-0 11 l0bd2 l0a5 Black maintains material equality but loses control over the
important eS-square, and following 1 2 "it'c2 .i.d6 (or 1 2. ..c5) 1 3 lOeS White's chances are pref­
erable. The line with 1 1 ...'�h8 (Game 17) is extremely intricate and every move must be care­
fully analysed. After 1 2 1Wxb7 Black's queenside becomes (.."VCO more vulnerable but the offside
position of White's queen exposes his king to danger. Black has .. .llf8-f6-g6 at his disposal
increasing his initiative on the kingside, and despite many games it is still impossible to give an
exact assessment of this position.

1 e4 e5 2 lDf3 lDf6 3 lDxe5 d6 4 .IL!f3 lllxe4 5 d4 d6 6 i.d3 i.e7 7 0-0 .IL!c6 8 .lle 1
i.g4 (D) 9 c4
9 c3 f5
10 l0bd2 - Game 16
10 1Wb3 0-0 1 1 li.:lbd2 (D)
1 t ...lOas - G0111e 18
1 1...Wh8 - Game 1 7
9••. lllf6 (D) 10 cxd5
10 l0c3 - Go1ne 1 J
1 0...i.xf3 - G0111e 14
10...l0xd5 - Come 15

B. . . i.g4 1 1 lllbd2 9. . .li:Jf6

46
CHAPTER THREE I
3ltJxe5: Black Plays 6 . . . J..d 6

1 e4 eS 2 c!Llf3 I.Llf& 3 c!LlxeS d6 4 I.Llf3 IU ....i.g4(Game 20) or to... fS (Game 21).


c!Llxe4 5 d4 dS 6 ..Q.d3 ..Q.d& Attacking the e4-knight with 9 lle I was
The line I e4 eS 2 lDf3 lDf6 3 lDxeS d6 4 unpopular for many years, but it has become
lDf3 lDxe4 5 d4 ciS 6 i.d3 .i.d6 is sometimes quite fashionable rccendy. Defending the
called the symmcttical variation. This is a knight with 9....i.f5 and the !,>ambit continua­
pretty formal name as the symmetry is ruined tion 9....i.�4 arc discussed in Game 22, while
lJuite quickly, and it's actually more accurate to 9. .:es is the subject of Game 23.
.

call it the Marshall Variation. At the lx.-ginnin� Despite the obvious transposition after 9
of d1c 20th century Marshall often playt:d the lDc3 lDxc3 10 bxc3 .i.g4 1 1 cxdS cxdS and 9
J..,'lUllbit 7 0-0 .i.g4?! 8 c4 0-0 9 cxdS f5, a line cxdS cxdS I0 lDc3 lDxc3 11 bxc3 .i.g4, this
that is nut quite correct, and eventually White docs not mean that there arc no signitlcant
players learned how to dampen Black's dan­ differences berween 9 lDc3 and 9 cxdS. After
�-,rcrous initiative (sec Game 1 9). 9 lDcJ (Games 24-25) Black can proceed
·n,e tabiya of the system arises after the with 9...lDxc3 10 bxc3 dxc4 1 1 .i.xc4 .i.J..,r4,
moves 7 0-0 0-0 8 c4 c6. Black has strcllbrth­ when the sharpest line is 12 'ii'd3 lDd7 13
ened the <.IS-pawn. which lx.>Came vulnerable lbgs lDf6 1 4 h3 i.hS 1 5 f4 h6 1 6 g4 as
;�ftcr ....i.d6. 9 'iVc2, attacking the kni1,rl1t on played by Capablanca (Game 25).
e4, is covered in Games 1 9-2 1 . Defending the The theory is very far advanced in the line
knight with 9...f5, as played in the historic of 9 cxdS cxdS 10 lDc3 lDxc3 11 bxc3 .i.g4,
J..,>amc Williams-Staunton, London 1851 is and here White manages to develop a strong
certainly possible (see the notes to Grune 19), initiative on the kingsidc. After 12 llb1 the
but these days most Black players prcfer game branches: 12 ...b6 is studied in Game
Krau7.e's su&�,rcstcd pawn sacrifice 9 ...lDa6!?. 26, while 12 ...c!'Dd7 is the subject of Games
'I hen to .i.xe4 dxe4 11 'iVxe4 (Game 19) gives 27-28.
Black active play as compcm;ation for the
pawn after either 1 1.. .lle8 or 11...lDb4!? (Raet­ Gamc 19
sky's b'llmes are important here). Nowadays, Burkov-Raetsky
j.,>randmasters, including the elite, prefer to Correspondence 1985
prcvent the possibility of ...lDa6-b4 with 10 a3,
;�ftcr which Black continues with 1O...lle8, 1 e4 eS 2 c!Llf3 I.Llf& 3 c!Llxe5 d6 4 c!Llf3

47
The Petroff Defence

l'Dxe4 5 d4 d5 6 .i.d3 .*.d6 7 0-0 li)xd5 17 'tth5 li)f6 18 1Wh4 White is much
White has a couple of minor alternatives: better - Cordel) 1 6 .tgS .te7 17 .i.xe7 li)xe7
a) 7 lbbd2 f5 8 �5 0-0 9 0-0 c5 to c4!? 1 8 .i.e4 .i.xf3 19 .1xf3 li)f5 20 11fxb7 l:lb8
(the passive 1 0 c3 allows Black to create 21 ..xa7 lbxd4 22 .lg4 l:lxb2 23 l:lad1 l:lb6
strong counterplay after 1 o...cxd4 1 1 cxd4 24 1Va4 and White has small advant!lh>c,
..i.xeS!? 1 2 dxeS llk6 1 3 lbf3 f4, for example Bernstein-Marshall, San Sebastian 1 91 1 .
14 .txe4?! dxe4 1 5 ..xd8 l:lxd8 16 tbgs ..i.f5 b) 1 5 g41 (this is even stronger) 1 5...lbxd5
17 ..i.xf4 h6 18 lbh3 - Shafranska-Raetsky, 1 6 .1e61 .tfl 1 7 lbg5 .txL-6 18 lbxe6 ..f6 1 9
Budapest 1 991 - and here 1 8...g5 gives Black li)xfB ltxf8 20 11fd3 and White is much bet­
an edge) to...cxd4 1 1 tben (not 1 1 tbdf3? ter, Gipslis-Christiansen, Gausdal 1 992
1fc7 1 2 .if4 g5!, when Black is a1ready al­ 8 c4
most winning) 1 1...dxc4 1 2 lbxe4 fxe4 1 3 This is the critical try, but 8 l:lc1 is also
.1xe4 d 3 and the position i s complex. reasonable. Now 8... .tf5 9 c4 c6 transposes
b) Playing 7 c4 before castling allows to 8 c4 c6 9 l:le 1 .if5, while 8...l:lc8 9 c4 c6
Black an additional option: 7 ....tb4+ 8 lbbd2 transposes to 8 c4 c6 9 :C l l:le8. This leaves
0-0 9 0-0 .txd2 1 0 tbxd2 (also possible is 1 0 8....tg41? as the independent try: 9 .i.xc4
.i.xd2 .1g4 1 1 .te3 li)c6 1 2 h3 .thS t 3 llc1 dxe4 10 l:lxc4 f5 1 1 l:le1 llk6 1 2 c.1 'iii'hH 1 3
l:le8 14 a3 dxc4 15 .txc4 'irf6 16 .1e2 - lbbd2 11'£6 14 h3 .ths 1 5 ltlc4 l:lae8 1 6 :n
Short-Adam.-;, Wijk aan Zee 2000 - and here (Black also has compensation after 16 ltxe8
Adams suggests 1 6 ...lbg3! 17 fxg3 l:lxc3 1 8 l:lxe8 1 7 ltlxd6 ..xd6 18 .td2 f4) 1 6... £4 1 7
g4 .tg6 1 9 ..d2 lte4! leading to an unclear lbxd6 cxd6 (White was a bit better after
position) to._li)xd2 1 1 .i.xd2 dxc4 12 .txc4 17 .....xd6 1 8 b3 b5 1 9 .tb2 •ds 20 a4 a6 21
.i.e6 (1 2...Wxd4 t 3 .tb4 ..xdt 14 ltaxd 1 axb5 axb5 22 .ta3 l:l£6 23 .1c5 in Kulaots­
l:le8 15 l:lft:1 l:lxet+ 16 l:lxel .te6 17 .i.xe6 Rozentalis, Cappelle Ia Grande 2004) 18 b3
fxe6 18 l:lxe6 li)c6 is also equal) t3 .1xe6 g5! 1 9 .1b2 l:lg8 20 g4 (not 20 d5?! .ixf3 21
fxe6 1 4 .g4 1Wd7 1 5 l:lfe 1 l:lf6 1 6 .tg5 l:lf5 'lrxO g4! 22 hxg4 ltle5 23 "irh3 ltxg4, when
17 d5! l:lxd5 1 8 llad1 tbc6 is level. Tiulin­ White is in serious trouble) 20 ... fxg3 21 fxg3
Raetsky, correspondence 1985. l:le3 22 g4 ..£41 23 lbe 1 l:lxh31? 24 l:lx£4 gxf4
7 0-0
••• 25 Wg2 l:lh4 26 �f2 l:lgxg4 27 ltlo l:lh3
with an unclear position (Kulaots).
8 c6 9 1tc2 tl)a&
..•

9 ... f5 is a major alternative here. Play


continues 1 0 lbc3

7....tg4?! is premature after 8 c4 0-0 9


cxd5 f5 10 h3 .th5 1 1 tbc3 tbd7 12 lbxe4
fxe4 1 3 .i.xe4 li)f6 1 4 .1f5 Wh8 and now:
a) 1 5 Wb3 tbxd5 Qf 1 5....txf3 1 6 ..x£3

48
3 ltl xe5: Black Pla ys 6 . . . .i.d6

and now: 1 5... £5 1 6 ltlb3 .i.xf3 1 7 gxf3 1Wf7 18 f4 ltlb4


a) lO. ..lba6 1 1 -.,3 �h8 12 cxd5 cxd5 1 3 19 'ilc3 Whs with a dangerous attack,
lbb5 (13 .i.xa6 can lead to a dmw after Kruppa-Rozentalis, Lvov 1985.
1 3...bxa6 14 1i'xd5 l:lb8 1 5 �5 .i.b7!? 1 6 1 2 /l\g5
lbt7+ l:lxf7 1 7 1i'xf7 lbxc3 1 8 bxc3 .i.xh2+! White's two other possible options also al­
1 9 �2 Wh4+ 20 Wgt ..txg2! 21 Wxg2 low Black good counte�play: 12 ltlc3 l:le8 1 3
1Wg4+) 13....i.b8 14 lle1 ..te6 15 a3 .i.gs 1 6 'libt 'lfJ 1 4 tlk4 .i.g4, or 12 .lle1 l!fl 13 .ih6
lbc3 ..Ll6 17 ..t ft g5!? 18 1Wxb7 g4 is un­ .if5 14 ..e2 l:le8 1 5 1Wxe8+ 1fxe8 16 llxe8+
ck."llr, Volchok-Borisov, correspondence llxe8 1 7 lbc3 ltld3.
1 984. 1 2 ... f5 1 3 ••2 f4 14 lD84
b) JO ...Wh8 1 1 cxd5 cxd5 12 ltlxd5 White wants to avoid the nasty pin that
.ixh2+ 13 �xh2 1Wxd5 1 4 .i.c4 ..d6+ 1 5 arises after 14 ltlf3 ..tg4. For example, 15 a3
tbcs lbc6 16 .i.f4 1Wf6 (not 16...lbxd4? 1 7 Lf3 16 gxf3 Wh4 17 'A'eM �h8 18 1Wg4
1Wd3 and White i s winning, but perhaps (other tries lose by force: 1 8 ..xd6 lbc2 19
Black shoukJ tty 16 ...ltlxe5 17 .i.xeS Wh6+ lla2 'irh31 20 .ixf4 llad8, and 18 axb4 l:lf6
1 8 Wg1 .i.e6 1 9 dS i.g8 with unclear play) 19 1Wd7 llg6+ 20 Wh1 :cis 21 1Wf5 llg5 22
17 ltlxc6 bxc6 1 8 i.eS Wh6+ 19 Wg1 ltld2 1ic6 �1.51 23 lla3 llffJ 24 1We4 .g5)
(19...lbg5?! allows White a strong advantage 18...1Wxg4+ 1 9 fxg4 ltlc2 20 lla2 lbxd4 and
after 20 f4 ltle4 21 i.d3, Macieja-Meijers, Black was slightly better in Kuznetsov­
Istanbul Olympiad 2000) 20 l:lfel (Macieja Raetsky, correspondence 1 985; or 1 5 ltla3
pointed out 20 1Wc1 ?! ltlf3+1 21 gxf3 f4 22 1if6 16 c5 .i.c7 17 1ic4+ ltlds and Black's
ltc1 .i.h3, when White is in trouble) position is preferable.
20...ltlxc4 21 'ilxc4 and White has an edge. 14 f31 1 5 gxf3 J.h3 1 6 lle1 •c7
..•

1 0 J.xe4
1 0 a3 is studied in Game 20.
10 dxe4 1 1 •xe4 lilli4
..•

1 7 a3
Black is also much better after other tries:
a) 17 lbxd6 1fxd6 18 �hl 1Wxd4.
This is strong but Black has a good alter­ b) 17 ltlg3 llae8 1 8 .ie3 (Black has as
native in 1 t ...l:le8 12 1Wd3 (or 1 2 1Wc2 lbb4 strong attack after 1 8 1Wxe8 llxe8 19 llxe8+
1 3 -.o3 .i.f5 14 lba3 aS and Black has excel­ ..tf8 20 lba3 1Wf7 21 lle3 hS) 1 8...1Wt71 19
lent compensation) 12....i.g4 1 3 .i.gS 1Wd7 14 l0a3 lbd31 20 lleb1 lbf4 21 ...<12 �! and
ltlbd2 h6 1 5 i.e3 (White should probably White is in terrible trouble.
prefer 15 .i.h4, but Black stiU has good play 1 7 ... ..bh2+ 1 8 �1 lOa& 1 9 ll\g5 .if5
after 1 5...lbc5 1 6 1Wc2 ..tf5 1 7 'l'c3 ltld3) 20 ll\c3

49
The Petroff Defenc e

20 tik6? leads t o a quick loss foUowing 31 ..g2 wouW have aUowed a neat finish:
20...W'f7 21 cJilxh2 D.aeS 22 d5 'WbS+ 23 cJilgt 3t...llh3+ 32 Wgt .i.f3!.
11fg6+ 24 cJilh2 .i.xe6 25 dxe6 llf5. 31 ...•f5 32 1t'e5 llh3+ 33 �g2 1Vf3+
20. . h6. 34 Wf1 llh1 35 .g3 .th3+ 36 �.1
Black can also tty 20... .i.f4!? 21 tlk6 .i.xe6 1t'xg3 0-1
22 ..xe6+ Wh8, when his attack continues.
2 1 lilge4 Game 20
Kasparov-Shirov
Wijk aan Zee 2001

1 e4 e5 2 lilf3 �f6 3 li)xe5 d6 4 lilf3


li)xe4 5 d4 d5 6 i.d3 i.d6 7 0-0 0-0 8
c4 c6 9 1t'c2 lila6 10 a3

Playing to block the e-file. After 21 o!L!e6


1i'f7 22 o!L!xg7 Wxg7 23 Wxh2 l:lae8 Black
has exceUent play.
21 .Dae8
..

Also promising is 2t....i.g61? 22 �e3 llf5


23 Wg2 l:a£8, when Black has a powerful
attack. For example, 24 o!L!d2? l:lh51 (instead 1 0 ....tg4
24 ....i.h5 25 llh1 llxf3?? 26 1l'e1! was sud­ This and 10. .£5 (Game 21) are the usual
.

denly winning for White in Khramov­ moves but 10..lle8 is also reasonable. Play
Raetsky, correspondence 1985) 25 l:lht .i.f5 continues 11 o!L!c3 and now:
and White has no hope of defending. a) 1t....i.f5 is a solid line: 12l:le1 h6 13 cS
22 .f1 i.e? 14 .i.d2 .i.a5?1 Onstead Anand assesses
White could try 22 .i.xh6!? but Black still 14...o!L!xd2 15 l:lxc8+ Wxe8 16 'it'xd2 .i.xd3
has compensation after 22..lle6 23 .i.e3 17 11'xd3 as cquaQ 15 .i.f4 .i.xc3 16 bxc3
.i.f4. �c7 17 h3 l£le6 18 .i.h2 (Anand-Kasimdz­
22 ...1t'f7 23 1t'g2? hanov, Hyderabad 2002) and here White is
Now White is completely losing. After 23 slightly better after 18....i.h71 19 llab1 b6 20
Wxh2 W'hS+ 24 Wg1 .t.h3 25 1l'e2 Wg6+ 26 lDes.
.i.g5 o!L!c71 27 ..e3 lbe6 at least White has b) tt....i.g4 121£lxc4 dxc4 13 .i.xe4 .i.xf3
chances to fend off the attack. 14 .i.xf3 'iVh4 15 g3 11'xd4 16 .i.g5 1i'e5 17
23...i.b8 24 i.d2 ..d2 11'f5 18 �2 o!L!cs 19 l:ad1 (According
Or 24 f4 :C6 25 dS cxd5 26 cxd5 llg6 to Nataf, 19 'irxd6 is also equal after
and the attack is too strong. 19...1i'xg5 20 £4 ..f5 21 b4 l£lc6 22 1i'd7
24...•5+ 25 �g1 i.h3 26 lilg3 i.xg3 l:ab8 23 l:ladt lled8 24 11'e7 :C8 25 1Vh4
27 •xg3 :Xe1 + 28 :Xe1 :Xf3 29 •2 lLif8) 19 . 1£le6 20 .i.e3 l:ad8 21 ..c3 .i.cS 22
. .

.g6+ 30 �1 i.g4 31 Jlg 1 b4 llxd1 23 llxd1 .i.xe3 24 .xe3 aS! and

50
3 f0xe5: Bisek Plays 6 ... �d6

Black had equalised in Nataf-Vo17.hin, Stock­ b2) 13 f3 �xd3 14 1i'xc.I3 �cS ! 15 1i'd4
holm 2001. �b3 16 Wxg4 �xal 17 .i.h6 g6 (if Black
1 1 �5 inserts the moves 17 ...'1Fb6+ 18 Whl before
11 cS .i.b8 12 l£les gives Black more playing 18...g6 then 19 Wf4f6 20 .i.xf8 llxf8
chances: 12...Le5 13 dxeS �exeS (but not 21 cxdS cxdS 22 Wd2 promises White a
l3...lt)axc5?! 14f3 �xd3 15•xd3, winning a strong initiative) 18 llk3 (an important
piece) 14 .i.xh7+ 'iPh8 15 b4 1Wh4 16 .i.d3 choice) 18-.'1Fb6+ t91lf21lfe8 20 Wf41i'c7
�xd3 17 •xd3 Whs 18 lle1 .i.fS 19 Wg3 (in Shirov-Leko, Linares 2000 White was a
ftk7 20 �2lbe6 21 �b2 b6 and Black has bit better after 20...f5 21 cxdS lDb3 22 e6
sttung counterplay, Jobava-Mamedyarov, cxdS 23 �xdS 1i'xe6 24 �c7 1i'c6 25 lDxe8
Plovdiv 2003. llxe8 26 g4Q 21 lle2 lle6, which Shirov as­
1 1 .....th& sesses as unclear. Instead of 18 ru, 18 cxdS
This is a major decision for Black as the is very messy after 18...:C8 19 �2 Wb6+ 20
alternatives are very interesting: Wh1 llxeSI 21 llxa1 1i'xb2 22 lld1 .ll xdS.
a) 11...�f5 12 �e3 �acS 13 c.lxcS �xeS Perhaps the best move is 18 �xf8 as White
14 cxdS 'ii'xdS 15 llk3 �xc3 16 bxc31lad8 has an edge after 18...9xf8 19 cxdS 'ii' cS+ 20
17 .i.d4 �c3! lH �xfS (after 18 .xc3? Wh l cxdS 21 e6 f5 22 1i'a4.
Motylev assesses 18...�xd3 19 �xg7 �xft 1 2 cxd& cxd& 13 �c3
20 llxft llfe8 21 .i.h8 f6 22 'ti'xf6 lld7 as A good, simple developing move. Taking
clearly advantageous for Black) 18. ..Wxd4 19 the pawn with 13.i.xc4dxc414Wxe4aUows
lLct gp 20 1i'xc31i'xc321 llxc3gxfS is very Black sufficient countcrplay: 14..lte8 15.i.e3
llrawish, Smimov-Motylev, Russia 2004. (or 15 �f4�xeS 16 dxeS llks 171We3l0d3
b) 1t....i.xe5 leads to heavily analysed and Black has excellent compensation)
complications after 12 dxeSlbacS 15...lDcs 16 1i'c2 t:&6 17 'ii'rs Wh4!? 18
�c.l2 �g6 19 1i'g4 1i'xg4 20 �xg4f5 21 �c4
�f8 22 dS fxbr4 23 dxe6 .llxe6 with an I.:<Jual
position, Potkin-Rozentalis, Bad Wiessee
2003.

and now:
b1) Black has an easy position after 13
cxdS ?! 'ii'xdS Qess exact is 13... cxd5 14 f3
llc8 15�e3- not 15 fxe4? dxe4!with a clear
advantage - 15... d4 16 �xe4 .i.e6!? 17 1 3 ...�xc3
.i.xh7+ ¢'h8 with an unclear position) 14 This may seem natural, but Black should
.i.xe4 �xe4 15 llk3 (if 15 f3 �cS 16 lldl seriously consider 13...Le5. For example,
1i'b3! Black has tremendous play) 15...lthc3 14dxcS lDacS 15�e4lDxd3 16 'iixd3�gp
16 Wxc3llfdK. 17 'ilb3.i.xe4 18 Wxb71Wh4 19 1ib3 d4 20

51
Th e Petroff Defence

1i'g3 1i'xg3 21 hxg3 l:lfc8 and Black had suf­ 24 .c2


ficient compensation in Leko-Shirov, Linares 24 llf7 is also winning for White: 24...1i'c6
2004. 25 "ifc2ll:le6 26 Wxg6 11f'xc3 27 .i.e7! 'lfxd4+
14 bxc3 �h8 28 Wh1 1i'xe5 29 ..if6 1i'c4 30 ..ixg7+ �g8
Black is clearly worse after 1 4 ... ..ixe5?! 1 5 31 1i'g3 h5 32 .i.eS+ 1i'g4 33 1i'f2 (Dok­
..ixh7+ �h8 1 6 dxc5 g6 1 7 'itd2! �h7 1 8 hoian).
'ii'h6+ 'iPg8 1 9 ..igS f6 20 ..ixf6 l:lxf6 21 exf6 24•..•g4
1i'xf6 22 £31 (Dokhoian). There is no way to save the pawn on g6,
1 5 f4 for example 24....z:r.e6 25 l:lf7 1i'c6 26 .: 1 f6!.
A natural move but White has a more 25 Ag3 'trh5
threatening choice in 1 5 .ixh7!?. After The rook ending after 25...•e4 26 1i'xe4
1 5...g6 the bishop is trapped but 1 6 1i'd2 dxe4 27 .i.xc7 ltxc7 28 :Ct is easily winning
�h7 1 7 9'h6+ �g8 18 ..tg5 ..ie7 1 9 ..txe7 for White.
1i'xe7 20 llae1 1i'c7 21 l:le3 creates a dan­ 26 JDI3 ...g5 27 llg3 'trh5 28 .ixc7
gerous attack. llxc7 29 llxg6 1th4 30 h3 •xd4+ 31
1 5 ....txe571 cxd4 llxc2 32 llf7 J:lg8 33 J:ld6 J:lc4 34
Now White's rooks arc too strong. In­ J:lxd5 J:lxa4 35 J:ldd7 lla1 + 36 �2 1la2+
stead 15 ...f6!? 16 li:l£3 1i'd7 17 llb1 l:lab8 is 37 Wf3 Wh7 38 e6 �g6 39 d5 J:lc8 40
unclear. llc7 J:le8 41 g4 a5 42 llxg7+ Wf6 43
1 6 fxe5 J:lgf7+ �5 44 1lf5+ �d4 45 e7 1 -0
White would even have a slight advantage
after 16 dxe5 lLic5 1 7 .i.c3 li:lxd3 1 8 .xd3 Game 21
b6 1 9 h3. Grischuk-Gelfand
1 6 .tg6 1 7 a4 ...d7
•.•
Wijk aan Zee 2002
Or 17 .....txd3 1 8 Wxd3 f6 19 .i.a3 l:lf7 20
l:lf5 1i'd7 21 l:laft Wxa4 22 .i.d6 and the 1 e4 e5 2 lL!f3 lL!f6 3 li)xe5 d6 4 lL!f3
attack continues. lL!xe4 5 d4 d5 6 .id3 .id6 7 0-0 0-0 8
1 8 .ia3 llfe8 1 9 .ixg6 fxg6 20 11fb3 c4 c6 9 •c2lLla6 10 a3 f5 1 1 &3
White has a sound alternative in 11 c5
.i.b8!? (1 t....i.e7 12 ll:lc3 lLic7 transposes to
11 lLic3 ll:lc7 12 c5 ..ie7) 12 lLic3 lLic7 1 3
ll:le2 (White should try 1 3 ..tf4!? ll:le6 1 4
.i.xb8 l:lxb8 1 5 lLic2 1i'f6 16 b4 g 5 1 7 'iVb2
with a slight edge) 1 3...lLic6 1 4 b4 1i'f6 1 5
..tb2 ..td7 16 llae1 ..tes 1 7 'lfcl .i.h5 1 8
ll:le5 ..ixe2 (also promising is 1 8...li:lxd4!? 19
.i.xd4 ..txe2 20 li:l£3 ..ixd3 21 .i.xf6 l:r.xf6
with excellent compensation - Piket) 19
.i.xe2 o!Dxd4 20 l2Jct7 o!Dxe2+ (not 20...9'h4?!
21 o!DxbB o!Dxe2+ 22 .llxe2 Lxb8 23 £3 .!Df6
24 l:lfe1, when White had plenty of compen­
20 b6
•.. sation, Moro7.evich-Piket, Wijk aan Zee
Against 20...ll:lc7, Dokhoian supplies the 2002) 21 .D.xe2 'lff7 22 o!Dxb8 l:r.axb8 23 £3
neat winning line 21 1i'xb7 :CbS 22 e6! 1i'd8 .!Df6 24 1i'g5 ltbe8 and Black is slightly better
23 1i'xa8! l:lxa8 24 e7 Wg8 25 l:r.f8. (Morozcvich).
21 .id6 lfx;7 22 llf3 llac8 23 llaf1 h6 1 1 &7
.•.

52
3 liJ.xe5: Blac k Plays 6 . .. �d6

llde8 26 llac1 when White had a definite


edge in Gipslis-Raetsky, Berlin 1993) 17 .i.e5
1fh5 18 life I llae8 and Black has managed
to complicate the game.
1 2 ...�8 1 3 b4 .td7 14 .tb2liJ.e6
This is an ambitious move. Black could
equalise easily with 14...dxc4 15 .ixc4 �xc3
16.i.xc3�5
1 5 cxd5
This is probably best White is in trouble
after 15ltle2 �6g5 16 �5.i.xcS 17 dxc5 f4
18 ltld4 (or 18 f3 �x£3+1 19 brxf3 1ig5+ 20
ltlg.'i fxg3 and Black's attack crashes through)
12 .lle 1 18...£3, when Black has a dangerous initiative
White has plenty of alternatives: (Gelfand).
a) 12 b4 a5 13 b5 .i.d7 14 llb1 (or 14 b6 1 5 ...cxd5 16liJ.xd5 llc8 1 7 •d1 liJ.6g61
c!frc, 15 cxd5 cxd5 16 �d5 llc8 17 'W'a2
lt.?h8 with compensation - Svidler) 14...�h8
15 c5 .i.e? 16 bxc6 bxc6 17 .if4 lbe6 18
.ic5.i.e8 19 llb6 'W'c8 20 �2.i.g5! (White
has excellent compensation after 20.....td8 21
11 .Lb6 22 cxb6) 21 �b3.i.hS 22 a4 'ireS is
unclear, Svidler-Morozevich, Krasnoyarsk
2003.
b) 12llk2 � 13 b4 .id7 14 cxd5 cxd5
IS �3 .ie8 16 llk5 .i.e? (if 16...�xd4,
A.Sokolov claims an edge for White after 17
..xdS+ .if7 18 'W'xd4 .i.xe5 19 'ii'xe5 'W'xd3
20 .i.b2 �f6 21 �f4) 17 .i.e3 .ih5!?
(17 ....if7 is inferior. 18 f3 .i.xe5 19 dxe5 d4 ·n1is keeps up the initiative. After 17....i.c6
20 fxc4 dxe3 21 .ic4 �4 22 'W'a2! �xe2+ 18 �3 �3!? 19 .i.xc3 L£3 20 1txf3 :Xc3
2.1 'W'xe2 'W'e7 24 :Xf5 Lc4 25 'ii'xc4+ �h8 21 ltlxf51 �xd4 22 11fg4 �f5 23 .ix£5
26 'W'c3 and White was clearly better in White escapes with an edge.
A.Sokolov-Frcssinet, Val d'lscre 2002) 18 f3 1 8 liJ.e5
lL\4g5 19 llfd 1 f4 20.i.£2 and White has only After t8lbc3ltlxf3+ 19 'W'xf31th4120 g3
a small advantage. 1th6 the attack continues (Gdf.1nd).
c) 12 c5 .ie7 J3lbc2 with a branch: 1 8 ...lLlh3+ 1 9 gxh3 •us+ 20 �1
cl) 13....i.f61? 14 b4 &1> 15 .ib2 'ii'c7 16 Tlus is forced. 20 l0g4? loses immediately
l:lad1 rffi 17 'ii'b1 'W'g7 18 .i.at .id7 was to 20...ltlxf2!.
unclear in Tseshkovsky-Motylev, Ekaterin­ 20 ...�xe5
hurg2002. In such a position it is not surprising that
c2) 13...�c6 is a viable alternative: 14ltlf4 Black has more than one possibility:
.if6 15 �xe6 .i.xe6 16 .if4 'W'e8 Oess 20...'ilh4!? 21 lle2 (after 21 'it'c2 .ixe5 22
promising is 16....i.f7 17 .icS .ih5 18 .ixf6 dxe5 .ic6 23 .ic4 1ixh3+ 24 �g1 llKl2! 25
1i'xf6 19 lDc5 llad8 20 f3l0g5 21 .llfellbc6 1txd2 'it'g4+ 26 �hl 1i'xc4 Black has a pleas­
22 '1Vf2 f4 23 b4 .llfe8 24 11fh2 .lle7 25 l:le2 ant advantage) 2t...1fxh3+ 22 �g1 .ixe5 23

53
Th e Petroff Defence

�xeS 23 dxe5 Wg4+ 24 �f1 11t'h3+ when 1i'xh3+ 26 �gl e3, while 24 ltaz?! e3! 25
White should accept a draw (Gelfand). :xe3 1fc4+ is not much better.
21 dxe5 .*.c6 24...Wxh3+
Now the game must end in a draw. Black
could play on with 24...:0!? but after 25 b51
�xbS + 26 �gl .llxh3 27 �f4 :d3 28 �g3
Wd8 29 1fg4 l:txd5 30 1fxe4 the position is '
simply unclear.
25 �1 hd5 26 'irxd5 'irg4+ 27 Wf1
1ih3+ 28 Wg1 1tg4+ 29 wn % - %

Game 22
Shirov-Piket
Wijk aan Zee 2001

1 e4 e5 2 l0f3 l0f6 3 l0xe5 d6 4 l0f3


This is promising, but so is 21 ...�e6!? 22 l0xe4 5 d4 d5 6 .*.d3 .*.d6 7 0-0 0-0 8
c!De3 Wf4! 23 :e2 1fxh2 24 Wet Whl+ 25 c4 c6 9 1le1
c!Df1 c!DgS 26 .llc2 :cd8 27 f4 ..xh3 (if Black
wishes he can force a draw with 27 ...c!Df3+ 28
W£2 �d5 29 �g3 lDel! 30 Wxel 1ff3+ 31
Wh2 1lt'h 1 +) 28 fxg5 :xd3 29 1fe2 1lfd8
with excellent compensation (Gelfand).
22 �c1 1
A superb defensive move. After 22 c!De3
:Cd8 23 1fc2 :Xd3! 24 9xd3 1ff4 Black has
a terrific attack. Por example, 25 .!&II l.lkl2+
26 Wg1 c!Df3+ 27 Wf1 1fxh2 28 We2 1fxh3
and Black wins (Gelfand).
22 ...11114 23 �xe4 fxe4

9 -*.fS
••.

9. .:es is examined in Game 23. Black's


.

two other options make life easier for White:


a) 9... f5 1 0 lDc3 Wh8 1 1 1Vb3 lDa6 1 2
cx.d5 cxd5 13 c!Db5 �b8 1 4 �d2 .t.c6 1 5
llact :f6 (or 15...c!Dxd2 t 6 c!Dxd2 :e8 1 7
c!Df1 �f4 1 8 lDe3 and White has a pull) 1 6
�g5 lDx.gS 17 c!Dxg5 .tgS 18 c!Dt3 .tf4 1 9
:c2 1lb6 20 1ia' 3 and White was a bit better
in Naiditsch-Meijers, Senden 1999.
b) 9....ig4 and now:
b1) tO c5? leads only to trouble:
24 �e3 1 0....ixh2+! 1 1 Wft (or 1 1 Wxh2 c!Dxf2 and
White continues to defend stoutly. 24 Black is better because 1 2 ..e2? loses to
c!Df4? loses rapidly to 24.. Jixf4 25 �xf4 12...lfud3 1 3 ..xd3 .ixt3) 1 1 ...5 12 lDc3

54
3 li)xe5: Bisek Plsys 6... i.d6

.!l:Xt7 13 �xe4 fxe4 14 .i.xe4 J.c7 15 .i.g5 b1) 12...J.c6 l3 a3 lilc7 14 ...c2 f5 15
lilf6 16 .tc2 11'd7 17 1l'd3 �4 and Black liJc5 ltc8 16 .if4 g51 17 .td2 (also worth
was obviously better in Yemelin-Ractsky, considering is 17 J.e3 lila6 18 f.3 lL!xc3 19
Rostov on Don 1993. bxc3 .txe5 20 dxe5 �5 21 .td4 with coun­
b2) The simple 10 .i.xe4 dxt.-4 11 Le4 is terplay - Gelfand) 17...�6 18 1i'd1 .i.xeS 19
best. Following 11... f5 White has two op­ dxeS �acS 20 .ic2 �xd2 21 9xd2 d4 22
tions: �4lilb3 23 .txb3 .txb3 24 c6! was unclear
b21) 12 l:e6 gives Black chances after in Frcssinet-Gclfand, Cannes 2002
12...11'd7 13 :Ct lL!a6 14 11'b3 (or 14 lilc3 b2) Also interesting is 12...�b4 13 .i.b11?
l:ae8 15 l:xe8 l:xe8 16 .i.d2 J.b8 with b>t>Od (Black has enough compensation after 13
attacking possibilities) 14...J.xf.3 15 11fxf.3 .i.xe4 dxe4 14 .ig5 1ih6 15 �xe4 .i.c6 16
ltae8 16 J.e3? {White can hang on to equal­ 'ild l �LacS) 13....ie6 14 a3 �6 15 1Wxb7
ity after 16 J.d2 J.b4) 16...�b4 17 c5? (after �5 16 ...a6lL!xc31 (not 16...lilb3?117lilxe4
17 lLla.3 lDd3 18 :n J.xa3 19 bxa3 f4 20 dxe4 18 .ixe4 �xa1 19 .i.xa8 �c2 20 l:te2
'ii'd1 •f5 21 J.d211'e4 Black is much better, �xd4 21 �d4 J.cS 22 .ie3 .txd4 23 lld2
but there is no forced win) 17.. .ru 18 cxd6 and White was comfortably better in Palac­
f4 19 J.d2lilxe1 20 11'b3+ *h8 21 � f.3 M.Ivanov, Cannes 2004) 17 bxc3 �b3 18
0-1 Kosteniuk-Raetsky, Bicl 2004. .ig5! J.xh2+ 19 '1Ph1 ...c7 20ltxe6!? fxd) 21
b22) 12 l:e 1 (again the simple move is the 11fxe6+11'f71 22 9xf7+ :Xf7 23 .i.a2!?lilxa1
best) 12...J.xf.3 13 11'xf.3 11'h4 14 h3 ..xcl4 24 .ixdS and after all the complications the
1511'b31ib6 16 11'xb6 axb6 17lL!c3lbd7 18 position remains unclear.
ltdt l:f6 and White had a tiny edge in c) 10 �3 �c3 11 bxc3 .txd3 12 1l'xd3
Ehlvest-Mamedyarov, Moscow 2002 dxc4 13 'lfxc4lLld7 with another branch:
1 0 1rc2 ct) 14 .igSI? 11'c7 15 .te7!? (15 11'b3 is
This is probably the best of a wide choice: less enterprising: 15.. 1lfe8 16 .lab1 h6 17
a) 10 cS .tc7 11 � l:eR 12 1l'c2lbd7 .i.e3 b6 with a dull equality) 15...�b6 16
13lL!xe4 dxe4 14 .i.xe4 .e7 15 �5lL!f6 16 .i.xd6 9xd6 17 'iib3 l:tacS 18 c4 'flc7 19 a4
f3 (Socko-Skato;hkov, Cappelle Ia Grande lLld7 20 l:tab1 b6 21 g3 1l'd8 22 aS 11'f6 23
2004) and hert: Skatshkuv suggests that axb6 axb6 24 *g2 and White had a marginal
16 ...lilxe4 17 fxe4 h6 18 ...f2 .tg6 is slightly edge in lvanchuk-Gelfand, Lvov 2000.
bcrtcr for Black. c2) 14 1tb3 has also bet."' tried: 14...1i'c7
b) t O eb3 �6 t1 cxd5 cxd5 12 �3 15 c4 :rea 16 .ib2 h6 17 g3l:tad8 18 �4
.i£8 19 11'c2 (perhaps a bertcr try is t9l:tad1
1Wb6 20 .ic3 with unclear play) 19...1fa5! 20
lL!g2ltxe1+ 21 l:txe1 bS 22 .ic3 b4 23 .i.a1
�b6 24 91>3 cS 25 dS Wa6 26 �e3 lte8 27
l:tct 11fc8 and Black was a bit better in Ad­
ams-Morozevich, Dortmund 2002.
1 0 ...i.g6
Solidly defending the bishop. 1o.. .lLxl7 is
also playable: 11 �3 lLldf6 12 cS (12 cxdS
cxdS 13 �3 :Cs 14 "xb7 ltb8 15 'W'xa7
l:e7 16 1i'a6 lte6 17 1l'a7 l:e7 is simply a
draw- Piket) 12....ic7 13 b4 J.g6 14 .i.b2
�c3 15 11fxc3 .ixd3 16 1i'xd3 �e4 17 l:te2
and now: l:e8 18 l:taet l:e6 19 �cS 11'h4 20 J.c1 (20

55
The Petroff D e fence

f3? loses after 20..llh6 21 fxe4 'lfxh2+ 22 lies 22 �d6 tnxc3 23 llc3 lDe4 24 :th3 h6
wn Wht+ 23 Wf2 llf6+ 24 tnn �g3+! 25 25 :tft with reasonable play for the pawn.
Wxg3 �>6+ - 1-Jaba) 20...f6 21 ti)f3 'iVh5 22 20. . .�xg5 21 •xg5 h6
g3 llaeS 23 Wg2 g5 was unclear in Lanka­ Trying to win a piece with 2t...f6? loses to
Haba, Hamburg 2003. 22lDg6+! hxg6 23 1Wxg6lDxc5 24 :te7.
1 1 c5 J..c7 1 2 l0c3 22 -.,5 �f6 23 �xf7+ �h7 24 �g5+
Wg8 25 •s6 hxg5 26 l:le7!

1 2 l0f6
•••

Alternatives are not so good: Black defends easily after 26 fxg5? lbe4 27
a) 12...tnxc3 13 bxc3 tnd7 14 �xg6 hxhr6 :th3.f5.
15 �g5 lDf6 16 &5 and White has a pleas­ 26 ...�8 27 l:lae1 1tg4
ant initiative - Pikct. Black would be mated after 27.-gxf4? 28
b) 12...f5 13 'lfb3!? and now 13...:tt7 al­ llxg7+! tnxg7 29 l:te7.
lows a trick: 14 ti)g5! b6 (the point is 28 fxg51th4
14...tnxg5? 15 �xg5 Wxg5 16 :teS+ J:tf8 17 The safest line is 28. .1Wf4! 29 h3 'ifg3 30
.

l:lxfB+ �xf8 18 'ifxb7 and White wins) 15 �h1 'ifxc3 31 :XeS Lc8 321lxc8 'ifct+
�xe4 fxe4 16 lDxt7 �xt7 17 cxb6 axb6 18 33 Wh2 1Wf4+ with a draw (Piket).
lDe2 .f6 19 �e3 tnd7 20 llac I with a clear 29 g3 .3 30 llxb7 .,5 31 ••6+ 1
advanmge, Anand-Morozevich, Monte Carlo 11rxe6 32 l:lxe6 l:ln?!
(blindfold) 2003. Instead, 13... b6 14 cxb6 32...J:tf31? 33 llxc6 J:txc3 is still unclear.
axb6 15 tnxe4 fxe4 16 �xe4 �xc4 17 :Xe4 Now Black is in trouble.
gives White a small advantage. 33 llxn �xn 34 llxc6 l:lb8 35 l:la6 l:lb7
1 3 J..g5 lbbd7 1 4 r!Lle5 .bd3 15 1txd3 36 h4 fi:jc7 37 l:ld6 �5 38 g6+ h7 39
1tc8 1 6 f4 llxd5 �c3 40l:le5+ � 41 c61 l:lb1 +
If 16 b4 lDxe5 17 dxe5 tnd7 18 f4 f6 19 There is no way to save the game: 41.. llc7
exf6 tnxf6 20 g3 �dB Black's stands well. 42 h5 l:lxc6 43 g4 llc6 44 gS+! �c7 45 h6
1 6 .. .i.a5! 1 7 1tg3 �8 18 e.4 .i.xc3 gxh6 (45...lbe2+ 46 :txe2 :txe2 47 hxg7 is a
Preparing ...tne4
. The immediate beautiful finish) 46 g7 l:lxc5 47 dxe5 �t7 48
18...lDe4!? is also possible: 19 lle3 tnxe5 20 e6+ Wxg7 49 !,rxh6+ and a pawn will queen
fxe5 ..i.xc3 21 bxc3 f6 22 exf6 gxf6 23 �h6 (Piket).
llt7 24 c4 1Ve6 25 cxd5 cxd5 26 :lfl l:g8 27 42 �g2 l:lb2+
h3 with an unbalanced position - Piker. 42...tnb5 doesn't help: 43 J:ld5 :tb4 44 a3
1 9 bxc3 �e4 20 l:le3 :tb2+ 45 �h3 <iii'e6 46 lld7 and White wins.
White couJd try 20 �c7 tnxc5 21 fxc5 43 Wf3 1-0

56
3 l'Dxe5: Bla ck Plays 6 ... �d6

bxc3 ..i.g4 14 ..i.e3 .dB?! (Black should play


Game 23 to exchange White's better bishop; after
Kovalev-Rozentalis 14.....d7 15 h3 ..i.fS 16 i..xfS WxfS the game
Glogow 2001 is leveQ IS llb1 ll\d7 16 h3 ..i.x£3 {after
16 .....i.h5 White can simply win a pawn with
1 e4 e6 2 l'Of3 ll)f6 3 l'Oxe5 d6 4 l'Of3 17 llxb7 ll\b6 18 g4 i..g6 19 i..xg6 hxg6 20
ll)xe4 5 d4 d5 6 �d3 �d6 7 0-0 0-0 8 ll\g5 ..i.c7 21 'iWD) 17 ..xf3 (White's bishop
c4 c6 9 .lie 1 .lieS pair ensure an edge) 17...ll\b6 1 8 a41lb8 19
a5ll\a4 20 i..d 2 ..xa5 21 l:lb5 •c7 221lxd5
.1Zd8 23 llb5 ll\b6 24 h4 a6 25 l:lg5 lieS?
{White had a promising attack, but this al­
lows a quick kiU) 26 .J:lxg7+1 1-0 Timofeev·
Raetsky, Abu Dhabi 2004 {26.. .'�xg7 27
..i.h6+! leads to mate).
1 0 l'Dxc3 1 1 bxc3
...

Also playable is 1 1 l:lxe8+ ifxe8 12 bxc3


..i.g4 t 3 ..i.d2 (with the idea of Wb t and a
double attack) 1 3...h6 (t3....d8?! mio;ses the
threat: 14 cS ..i.c7 15 Wb1 i..xf3 16 Wxb7
iDd7 17 gxf3 .J:lc8 18 llet g6 19 h4 and
White was much better in Raetsky-Varlamov,
10 ll:lc3 correspondence 1985) 14 Wb3 dxc4 15
This is the most logical of a range of pos­ i..xc4 (Black can draw after 15 .xb7 cxd3
sibilities: 16 IZet 'ilfd7 17 'ilxa8 ..Lf3 18 gxf3 1Wh3 as
a) 10 ..c2 ..i.g4 11 lLic5 ..i.xeS 12 dxe5 there is no good way to avoid perpetual
lfr.t6 13 ..i.xe4 dxe4 14 'ilxe4 f5 15 'ilc2lLib4 check) 15...b5 16 l:le1 1Vf8 17 i..x£7+1 ..x£7
16 ..c3 lLid3 and Black's huge lead in dcvcl­ 181lc8+ ..i.f8 19ll\e5 Wxb3 20 axb3 ..i.hS 21
' 1pmcnt provides ample compensation for .J:ld8 and White has sufficient compensation
the pawn. for me piece: the pins are nlmost impossible
b) 10 lL!bd2 ..i.t'!i 11 c5 (Black equalises to escape.
easily after t 1 'ifc2 lLixd2 t2 ..i.xd2 i..xd3 13 11 �g4 12 .i.g5 Jlxe1 + 13 1rxe1
.••

1rxd31lxc1+ 14 l:lxe1 dxc4 15 ..xc4lLia6 16


.!DgS ..d7) 11...i..f8 12 Wb3 b61 13ll\xe4 (or
13 cxb6 axb6 14ll\fl ..i.d6 with an unclear
position) 13 .....i.xc4 14 ..i.xe4 dxe4 15 ll:x:S
16'd5! 16 cxb6 axb6 and Black has equalised.
·n1e point is that 17 Wxb6?! is a mistake after
17... £6 18 Wb7 (Ulibin-Raetsky, Makhachkala
1993). Black should continue t8...lLia61 19
�d7 ll\b4t 20ll\xf8 (or 20ll\b6 Wbs 21 a4
'ii'a6 and Black is clearly better) 20...ll\c21 21
iLh61? ..f7 (of colU'SC not 21 ...gxh6?? 22
lZ'Ixh7 and White wins) 22 lik!7 ll\xct 23
Axel llcd8 with a smaU advantage.
c) I 0 CXliS should not be dangerous: 1 3 1rc8
•••

IU ..cxdS II �3 ll\xc3 121lxe8+ Wxe8 1 3


. 1 3 ...1i'd7 is another option: 14 c5! ..i.c7 15

57
The Petroff D efence

�h4!? h6 16 .i.d2 'lidS? (now Black is in 1 7 h6?!


•.•

desperate trouble; instead t6...liJa6 17 'lfbt This is too weakening. Black should play
Zle8 18 g3 Wc8 19 f3 .i.h3 20 IllES is only 17...f6, for example 18 fS fxgS 19 fxc6 �xeS
slightly better for White) 17 f4l .i.c8 18 Wg3 20 dxcS .i.xc5+ 21 Whl gxh4 22 1i'xh4 g6 23
(surprisingly White is already winning by %let and White obviously has compensation
force) 18...b6 (after 18...1i'f8 Anand analyses for the pawn but Black can hope to defend.
a beautiful winning line: 19 %let .i.d8 20 1 8 f5lt:lxc5
lt\g6! fxg6 21 1i'xg6 .i.e? 22 �7+ Wf7 23 Black prefers to sacrifice a piece rather
.i.g6+ Wf6 24 .i.hs .i.fS 25 g4l .i.xh7 26 g5+ than take one: 18...hxg5 19 fxe6 Wxc6 (the
hxg5 27 fxg5+ WfS 28 Zlc:S mate) 19 Zlc:t problem is that 19...gxh4? loses to 20 cxf7+
bxc5 20 dxc5 'iff8 21 .i.c:31lla6 22 .i.d4 g5 'ii?xf7 21 Wxh4, when the king is doomed) 20
23 ...f2 and Black is defenceless and resigned 'lfxe6 fxe6 21 �f3 .i.e7 22 Zlb1 and White
in Anand-Piket, Wijk aan Zce 2001. has a clear advantage - Rozentalis.
1 4 lDh4 1 9 dxc5 hxg5 20 fxe6 .i.xc5+ 21 �1
This prepares a kingside advance. Whi.te 1be6
has tried two other possibilities: If Black tries to recapture the piece Whi.te
a) 14 .i.e7 .i.f4! 15 .i.h4 �7 16 .i.g3 gains a d<.-adly attack: 21...gxh4 22 Wxh4 g6
.A.xg3 17 hxg3 .i.xf3 18 gxf31£lf6 (Black has 23 exf7+ Wg7 24 %let 'ifd7 25 J:.xg6! Wxg6
equalised) 19 WeS dxc4 20 .i.xc4 Wd7 21 26lle5.
Wf4 ZleB 22 11fd2 hS 23 %let :Xet+ 24 'lfxel
bS Vz-112 Movsesian-Rozentalis, Neum 2000.
b) Perhaps White's strongest line is 14 cS,
for example 1 4....i.f8 15 'ltb 1! h6 (after
15 ..g6 16 �eS .i.e6 17 f4 �7 18 �f3
.

White has a promising attack - Rw.entalis)


16 �eS .i.e6 17 .i.f41£ld7 18 a41? .i.e? 19 aS
.i.d8 20 'i'b4 and White's easier devclop­
ment promises an edge, Degraeve-Rozentalis,
Montreal 2002.
1 4...lt:la6
If 14... h6 then 15 J:.e7 .i.e? 16 .i.a3 and
White still has pressure.
1 5 c5 .i.fB 1 6 h3 .i.e& 1 7 f4! 22 11xe6 fxe6 23 li:lf3 g4 24 hxg4 �7
25lt:le5+?!
Now Black's central pawns have a chance
to roll. White can keep control with 25 g3
ri;e? (25. id6 26 Zlfl Wc7 27 �5 is also!
.

unpleasant) 26 Wg2 'ii?d6 27 Zlcl with good


winning chances.
25 .. .'�87 26 g3 �d6 27 lt:lf7+
White should probably give up a pawn to
blockade Black's centre: 27 Zlc:t .A.f2 28
�f7+ We7 29 Zlfl .i.xg3 30 'ii?g 2 i.e? 31
ll!gS and Black's pawns are not a threat, so.
White is a bit better.
27 ...h7 28lt:lg5 e5

58
3 fi)xe5: Black Plays 6 ... .i.d6

Now Black has significant counterplay. and now:


29 :n e4 30 J:lf7+ Wei& 31 J.e2 a) 1 1 c5 (stabilising the space advantage)
After 31 .i.xe4?! .i.e3! 32 llxg7 .i.xgS 1 t....i.c7 1 2 :C1 �7 1 3 .i.g5 li:)f6 (or
Black escapes with a draw. 13 ... f6 1 4 .*.h4 llfl 1 5 1fc2 g5 1 6 .i.g3 .i.xg3
31 b5 32 :Xg7 J.e3 33 J:lg6+ Wd7 34
..• 1 7 hxg3 - Stellwagen-S.Ernst, Groningen
liJf7 J:le8 35 g5 J.d2 36 J:lg7 J.xc3 37 2002 - and here White can gain an edge with
.:117 d4 38 .i.g4+ �7 39 g& d3 17 ...li:)f8 1 8li:)h2 .i.hS 1 9li:)f1) 1 4 111d2!? (14
h3 leads to an unclear position after 1 4...i.h5
15 g4 i.g6 1 6 �5 h6 17 .i.h4 .i.xd3 1 8
'iVxd3 g5 1 9 .i.g3 �4) 14.....ixf3 1 5 gxf3 b6
1 6 lle2 bxc5 17 dxc5 h6 1 8 .i.xh6! �7 (not
1 8...gxh6?? 1 9 1Wxh6 l:tb8 20 �h1 and .l%gt +
will lead to mate) 1 9 .*.gS f6 20 i.f4 is messy,
Glauser-Ractsky, Lugano 2000.
b) 1 1 lib 1 allows 1 1 ...dxc4 1 2 .i.xc4 bS.
For example, 1 3 .*.e2 �7 14 a4 a6 1 5 c4
bxc4 1 6 11fc2 .i.xf3 17 .Lf3 Wh4 18 g3
1Wxd4 1 9 lld1 (in Ivanovic-Khalifman,
Plovdiv 1986 White now lost the thread with
19 .i.xc6?! llad8 20 .i.gS?! WcS 21 .i.xd8?!
40g77 'ifxc6 22 .i.a5 �5 23 llfd1 li:)f3+ 24 Wf1
In time trouble White slips up. Instead 40 li:)xh2+ 25 Wg1 li:)f3+ 26 Wft :CB! and
'ifi!g2 cS 41 li:)d8+! Wb6 42 llb7+ Wa6 43 Khalifman had a winning attack) 1 9. .'iff6 20
.

:c17 gives good chances. .i.xc6 llad8 21 .i.xd7 l:lxd7 22 .i.f4 l:lfd8
40...J:lg8 41 J.f5 d2 42 J.g4 e37! with a level position (Khalifman).
Now it is only a draw. 42...1lxg7? 43 :Xg7 1 1 J.xc4 �g4 1 2 h3
.i.xg7 44li:)gS, leaving White clearly better, is 12 Wd3 is the subject of the next game.
even worse, but Black could have kept de­ 1 2....i..h5 1 3 J:le1 l0d7
cent winning chances after 42....i.f6! 43 ¢'g2

cS.
43 lt)h& lbg7 44 :Xg7+ J.xg7 45 lOts
.i.e5 46 �2 c5 47 lrule3 c4 48 J.e2
'>Pd& 49 g4 .i.f4 50 Wt3 WaS
Of course not 50....i.xc3? 51 Wxe3 c3 52
'1Pd3 b4 53 .i.dt when White wins as all
Black's pawns arc worthless.
51 �c2 J.g5 52 J.d1 % - %

Game 24
Firman-Bick
Stratton Mountain 2003
1 4 g4
1 e4 e5 2 l0f3 �f& 3 lL!xe5 d& 4 liJf3 White can also play Jess aggressive])' with
lllxe4 5 d4 d5 6 J.d3 J.d& 7 0-0 0-0 8 14 .i.ft. Now 1 4...l:lc8 1 5 llxe8+ Wxe8 1 6
c4 c& 9 liJc3l0xc3 10 bxc3 dxc4 i.e3 b5 1 7 a4 a6 (17 ...b4 1 8 cxb4 .*.xb4 1 9
Black can also try the immediate 10....i.g4 J:lct li:)b6 is also equal) 18 axb5 axb5 1 9

59
Th e Petroff Defence

llxa8 11fxa8 20 dS cxdS 21 g4 i.g6 22 i.xbS c4 c6 9 lDc3 clDxc3 10 bxc3 dxc4 1 1


�7 was completely level in Ehlvest-Anand, .i.xc4 .i.g4 1 2 'il'd3 �d7
Unares 1 991 .
1 4 .. ..ig6 1 5 .*.g5 ••6 16 .*.e7! .*.xe7 1 7
llxe7clDb6
Or 1 7 ... :ad8 1 8 11t'e1 cS 19 .!DeS c!DxeS 20
:xeS lk8 21 f4 and White's initiative con­
tinues.
1 8 .i.b3 llab8 1 9 1i'e1
Taking control of the open file. lnst(!ad 1 9
c4 l(fe8 20 11'e1 'ifxet+ 2 1 ltaxe1 :xe7 22
:xe7 �fB 23 :C3 l(d8 is only equal.
1 9...•a3 20 1fe5 lDd6
If 20...'it'b2 White plays 21 l(d1 11fxc3 22
c!Dgs, when 23 c!De6 is a nasty threat.
21 .i.xd5 cxd5 22clDg51 1fxc3 23 �61 Sensible development. Black has a solid al­
ternative in 12 ... i.hS 13 c!Dcs bS 1 4 'il'h3
bxc4 1 S 'il'xhS 'WaS 16 i.d2 1t'dS (Black
doesn't win a piece with 16...f6?! because of
17 1i'g4 heading for c6; for example,
17 ...i.xeS 1 8 dxeS 'ifxcS 19 ltfel and
White's lead in development is overwhelm­
ing) 1 7 11'e2 :C8 18 i.e3 i.xeS 1 9 dxcS
"iVxeS 20 'ii'xc4 'ife4 21 1i'xe4 1lxe4 22 llfcll
c!Da6 23 l:ld7 and White had a solid advan­
tage in Abropov-Solozhenkin, Helsinki 2000.
Instead of 13 ...b5, Black should play the sim­
ple 1 3...i.xeS 1 4 dxeS 'iVxd3 1 S i.xd3 Zld8
1 6 i.c2 i.g6 17 i.xg6 hxg6 1 8 ltb1 b6 with
23...1i'xa1 +1 c:quality.
Taking a rook with check is always tempt­ 1 3 li)g5
ing, but now Black is lost. Black has to tty Starting a kingside attack. 1 3 :C:1 allows a
23... £6, although White is clearly better after clever trick with 13...bS 14 i.b3 .!DeS! 1S.
24 :xg7+ Wh8 2S 'ifc7 'ifxc7 26 :xc7 l(fe8 jfc2 (the idea is 15 dxcS?! i.xh2+, now 16
27 .!DeS. .!Dxh2 'iVxd3 1 7 c!Dxg4 :ac8 1 8 .i.e3 'il'xc3 is
24 Wh2 f6 25 :Xg7+ WhB 26 'il'c7 1 -0 clearly in Black's favour, Klingcr-Fuchs,
White threatens 27 :xh7+ followed by 28 Obcrwart 1 993) 1 S...c!Dxb3 1 6 c!DgS g6 1 7
11t'g7 mate, and 26...)%g8 is met by 27 fle7 axb3 .i. f5 and Black has equnlised. I
1 3 ...�f6 1 4 h3 .i.h5 I
after which f6 cannot be defended.
This leads to immense complications. AI
Game 2J 1
safe, reliable option is 14 ....i.d71? 1 5 i.b3 h6 '
Hamdouchi-Le Roux 1 6 c!Df3 jfas with equality.
B e!fort 20()3 1 5 f4 h6 1 6 g4
This is the critical try. Black has no prob- i
1 e4 e5 2 lDf3 clDf6 3 �xe5 d6 4 .lDf3 lcms after 1 6 c!Df3 .i.xf3 17 ltxf3 (or 17 11'xf31
clDxe4 5 d4 d6 6 .*.d3 .i.d6 7 0-0 0-0 8 'flc7 18 i.d2 cS with good counterplay)l
I

60
3 li:Jxe5: Black Plays 6 ... .i.d6

17...:e8 18 .i.e3 .i.c7 19 llafl 1Wd6 20 .i.f2 tnkes over the attack after 21 gxf7+ .i.xf7 22
:C4 21 .i.h4 ltae8 22 �xf6 11'xf6 23 �ht :Xf7 ltxf7 23 1Wg6 .f6 24 .i.xf7+ 1Wxf7 25
Act 24 a4 hS 25 g3 a6, Kamsky-Bareev, 'iPxd6 ltf8) 2t....i.xb3 (21....e8? allows a
Linares 1993. neat forced win: 22 'trh3! :Xf7- the trick is
22_..i.xh3? 23 ltfS+I- 23 Wh7+ Wf8 24 .i.g5
l1f6 25 ..ixf6 gxf6 26 g7+ and Black wiU be
mated -Nunn) 22 'iPh3 l:lxf7 231Wh7+ �£8
24 WhB+ We7 25 .i.g5+ J:l£6 26 .xg7+ We6
27 .i.xf6 'iPg8 28 axb3 .xg7 29 .i.xg7 WfS.
After all the tactics White has a slight edge in
the ending.

1 6...b5
It is also posl>ible to take the piece imme­
diately: 16...hxg5 17 fxgS lLlxg4 18 hxg4 1Wd7
(18...�xg4? is a mistake: 19 1i'e4 1i'd7 20 g6
..ie6 21 .i.xc6 fxe6- Capablanca-Northrop,
New York 1909 - and here the simple 22
Wh4 is winning) 19 �3!? (the independent
try; 19 gxhS 9g4+ 20 �f2 ltaeB 21 ltg1 20 ..1671
'1Vh4+ 22 �g2 b5 23 .i.b3 trans�cs to Now Black has an edge. The correct way
16...b5) and now: to hold equality is 20 gxh5 .g4+ 21 Wf2
a) 19 ...'.xg4+ is safest: 20 1i'xg4 �xg4 21 l:aes 221tg1 'trh4+ 23 wg2 "W'h2+1 24 wn
llxf7 b5 22 .i.b3 �h7 with equality (Byrne, .i.f4 25 1i'f3 (not 25 .i.x£4? .xf4+ 26 ..t.>g2
Mednis). llc3 and White must give up his queen)
b) 19 ....i.xg4 is interesting but also risky: 25..llc1+ 26 �xel 'ii'xgl+ 27 We2 .i.x.cl 28
20 Wh4 .i.fS (or 20....i.e6 21 .i.d3 g6 22 :Xcl .xcl 29 g6 lle8+ 30 Wd3 'iWb 1 +
l:t£6!? c5 23 .i.e3 and White has a dangerous (30 .lle7 31 gxf7+ wm 32 �3 'ilrbt+ 33
.

attack) 21 .i.£4 .i.xf4 22 llxf4 bS 23 .i.b3 and .i.c2 .xa2 34 1i'c8+ on - also draws) 31
-

White certainly has compensation. Now in Wd2 .el+ 32 Wd3 'ilrbt+ 1/z-'12 A.Sokolov­
the game Moro7.evich-Ippolito, New York OU, Odessa 1989.
1997 Dlack erred with 23...ltae8?, after which 20 ...hg4 21 •xd7 .i.xd7 22 lbf7 lbf7
White could have won with 24 Wf2! 11fe7 25 23 g& .i.e8 24 .i.g5
l:txfS 1t'e2+ 26 Wg1 :C4 2711fh3 (Ippolito). White is also slightly worse after 24 �e3
1 7 .i.b3 hxg5 1 8 fxg5 li:Jxg4 1 9 hxg4 aS 25 .i.e6 Wf8 26 gxf7 .i.xf7 27 llft lla7,
..d7 Nunn-Salov, Brussels 1988
Another hugely complicated line is 24 85 25 �g2
•..

19....i.xg4 20 g6 (or 20 1We4 .i.h3 21 g6 Black also has an edge after 25 l:lfl llaa7
.i.h2+!- not 21....i.xft? 22 1Wht! and Black 26 llf3 a4 27 .i.e6 llac7 28 ..i.d8 l:lb7 29
l"lln resign - 22 Wxh2 .i.xfl with a very .i.gS b4 30 cxb4 �xb4.
complex position) 20....i.e6 21 llxf7!? (Black 25 84 26 .i.e& �8 27 l:lh1 l:lf& 28
...

61
Th e Petroff Defence

hf6 gxf6 29 .*.f5 Wg8?1 rer but Black still has an edge) 18 . .1lxc3 19
.

Black fails to make the most of his llxb6 Axd3 20 .i.e3 1i'g6+ 21 �h1 'WfSI (in
chances. Better is 29.. 1[b8 30 'itf3 (White Capablanca-Marshall, New York 1909 Black
achieves nothing after 30 llh8+ 'itc7 31 missed his chance with 21...'1Fe6 22 Wg2
llh7+ Wd8) 30 ..b4 31 cxb4 llxb4 with a
. 1i'g6+ 23 �h1 1i'e6 when a draw was a&rreed)
definite edge for Black. 22 Wg2 Qf 22 llxd6 Black's idea is
30 l:.h7 22.. '1Fxh3+ 21 �g1 ltxe3! 24 fxe3 'Wg3+ 25
.

Now the threat of 31 .i.e6+ forces Black Whl 1i'xd6 with a clear advantage because of
to bail out to a drawn ending. White's exposed king) 22. ..1la3! 23 1rb7
30 hg6 31 ..i.xg6 .*.f4 32 .llb7 .*.d2
•.. 1i'g6+ 24 Ciilh1 llxe3! 25 :Xd6 Wh5 26 fxe3
33 ..i.e4 l:.c8 34 llb6 ..i.xc3 35 :.Xc6 'irxh3+ 27 Wg1 1i'g3+ 28 Wh1 1i'xd6 and
:.Xc6 36 ..i.xc6 % - % Black has an obvious advantage.
1 2...b6
Game 26 12...�7 is studied in Games 27-28.
V .Gurevich-Meijers 1 3 .llb5 ..i.c7
Germaf!Y 1999 The d-pawn was genuinely threatened. For
example, 13...a6? 14 %hd5! wins a clear pawn
1 e4 e5 2 lbf3 lbf6 3 lbxe5 d6 4 l/)f3 because 14...i.xh2+? 15 �xh2 i.xdt 16
�e4 5 d4 d5 6 ..i.d3 ..i.d6 7 0-0 0-0 8 llxd8 ltxd8 17 llxdl is winning for White.
c4 c6 9 cxd5 cxd5 1 0 lt!c3 ltlxc3
This is the automatic answer to 10 lbc3,
but the rarely played 10...1le8 is also a con­
sideration. Now I I lle1 transposes to note
'c' to White's 10th move in Game 23, 11
.ixc4 dxe4 12 �g5 .ifS 13 f3 c3 14 �ge4
.if4 was unclear in Morozevich-Gelfand,
Cannes 2002, while 11 h3 and 11 ..c2 are
also serious possibilities for White.
1 1 bxc3 ..i.g4 1 2 llb1

1 4 h3
14 c4 dxc4! leads to very sharp play after
15 i.e4 lbc6 16 llg5 (if 16 .i.xc6?! Black's
idea is 16...'1Fd6 17 ..LaB i.xf3 18 i.f4 1i'xf4
19 g3 i.xd1 20 gxf4 i.c2, when the queen­
side pawns will be very dangerous) 16...i.xt1
17 1i'xt1 '1Fd6 18 llg3 (18 g3 eventually pe­
ters out to equality after 18... �d4 19 1i'e3
llfe8 20 :d5 1i'xd5 21 .i.xd5 llxe3 22 fxe3
tbe2+ 23 �g2 :ds 24 i.xt7+ Wh8 25 i.xc4
Instead 12 h3 .i.h5 13 1rb3 is too ambi­ �xcl 26 llxct) 18...�xd4 19 1i'g4 g6 20
tious: 13....i.xf3 14 ..xb7 �7 15 gxt1 �b6 .i.xa8 f5 21 'Wh4 l:lxa8 (the tempting
16 ltbt 'it'f6 17 �g2 llac8 18 1i'xa7? (18 2L.l&2+? fails to 22 Wht �xg3+ 23 hxg3-
llb31lc7 19 Wa6 .i.f4 20 .i.xf4 Wxf4 is bet- White is clearly better because 23..1l . xa8?

62
3 liJxe5: Black Plays 6. .. �d6

loses to 24 ...xc4+ rtlg7 25 .i.f4) 22l:r.h3 hS, 21 .c6 liJf671


OU-Akopian, Manila 1992. Black overlooks the coming tactic. Afrcr
1 4 a6 1 5 hxg4
.•. 21...lLlffi 22l:r.e l .i.f4 23lLle5l:r.h4 24 i.xf4
White has a promising alternative in 15 :xf4 the position is unclear.
%lxd5!?. For example, 15...'tWxd5 16 hxg4 22 o!tle5 l:lxd47
lbc6 17 lLlgS h6 18 .i.h 7+!? Wh8 19 .i.e4 Now Black is lost. Instead 22...%lc8 23
1i'd6 20 lLlf3 l:.fe8 21 .i.c2 'ii'd7 (White has cli)xt71? .xd5 24 1Wxd5 cli)xd5 25 i.d7l:r.xd4
an edge after 2t...b5 22 'ird3 'iVg6 23 'ifxg6 26 i.xc8 Wxf7 27 lie I is only a little better
fxg6 24 .i.xg6) 22 g5 Wg4 (Benjamin­ for White.
Zamora, Philadelphia 1999). Now Benjamin 23 liJxf7! llh4
assesses 23 'ii'd3 g6 24 gxh6 .i.f4 25 i.b3 The obvious point is 23 Wxf7 24 11fe6
...

.i.xc1 26 llxct as clearly better for White. mate!


1 5...axb5 1 6 •c2 g6 24 g3 1Vxd5 25 gxh4 1Vxc6 26 �xc6
lba2 27 o!tlg5 �f4 28 l:ld1 ..i.d2 29
liJxh71 .i.xh6 30 liJxf6+ �7 31 liJg4
�d2 32 Wg2 l:lc2 33 �e4 llb2 34 lla1
1 -0

Game 27
Pavlovic-Raetsky
Bie/ 1999
1 e4 e5 2 liJf3 liJf6 3 liJxe5 d6 4 liJf3
liJxe4 5 d4 d5 6 �d3 �d6 7 0-0 0-0 8
c4 c6 9 cxd5 cxd5 10 o!tlc3 liJxc3 1 1
bxc3 �g4 1 2 lib 1 liJd7
1 7 �h6
This seems the natural try. White has also
played 17 .i.xbS 'ird6 18 g3lLlc6 19 Wg2 f5
20 .if4 'ird7 21 gS i.xf4 22 gxf4 'irc7 23
.ixc6 'irxc6 24 �5 'ird6 (Black should
avoid chasing material with 24...'1'a4 because
;tfter 25 Wd3 Wxa2 26 l:r.ht White has a
promising attack - Nikcevic) 25 a4 %laS 26
l:tbt JibS with an unclear position,
i\.Sokolov-Nikcevic, Vmjacka Banja 1998.
17 ...Ilea 1 8 �xb5 Jle4 1 9 c4 lbg4 20
.

cxd5 liJd71?
Completing development is logical but
20....i.f4 is also worth considering: 21 lLleS 1 3 h3
lth4?! (Black has to play 2t ...l%xg2+1 22 Wxg2 Forcing the bishop to the kingside. In­
..ixh6 with an unclear position) 22 .i.xf4 stead 13 l:r.b5 lLlb6 14 h3 gives Black the
llxf4 23 'ird2l:[f6 24 cli)g4l:[d6 (Kotronias­ option of 14.. ..i.d71?. For example, 15 l:r.bt
Roi'.entali�, Debrecen 1992) and here White i.a4 16 'ire2 lieS 17 .i.e3 %lc8 18 cli)gs (the
could have gained a dear advantag.: with 25 game is level after 18 �5 .i.xeS 19 dxe5
llc I lLld7 26 1Wf4 cli)f6 27l:r.c61 (Lepeshkin). l:.xe5 20 .i.xb6 :Xc2 21 Ld8 Zld2) 18...g6

63
Th e Petroff Defence

(18 ...h6 is very risky: 19 �xt7!? Wxt7 20 1 4...�6 1 5 c4


Wh5+ �g8 21 .i.xh61? and White has a This is the normal move but 15 a4 is also
strong attack) 19 �xh7! (mstead 19 11'f3?! worth considering: 15...a6 (after 15....i.g6 16
l:lc7 was a bit better for Black in Kruszynski­ .txg6 hxg6 17 1i'b3 Ilea 18 .tgs 11fd7 19 aS
Raetsky, Katowice 1990) 19..1lxc3 (an ambi­ �4 20 llxb7 .1c7 21 a6 11'c6 221Vb5 White
tious attempt to play for a win; after has a pleasant advantage) 161lxb6!? (the less
19...�xh7 20 'ifhS+ �gB 21 Lg6 fxg6 22 ambitious 16 lib1 leads to an unclear posi­
.xg6+ Wh8 23 .i.gS .i.e7 White has a per­ tion after 16...1lb8 17 .i.gS Wc7 18 aS �4
petual check) 20 �gS 11'f6 with a complex 19 .i.xc4 Wxc4) 16...1l'xb6 17 .i.xh7+ �h8
position. (if 17...Wxh7 18lbg5+ Wg6 19 g4 White has
1 3 ih6
.. a strong attack) 18 .i.d3 with an unclear posi­
tion (Miralles).
1 5 .txf3
.••

15...�xc4 is less accurate because of 16


llxdS! (16 .txc4 .i.xB '17 1i'xR dxc4 18
l:lxb7 1i'e8 19 .i.e3 .e6 20 llct c3 21 dS
•es 22 g3 llac8, as in Kulaots-Lauk, Tallinn
2000, is only equal) 16....th2+ 17 �xh2
11'xd5 18 i..xc4 1i'xc4 19 •xhS 'iixd4
(19....xa2 is even worse: 20 dS llad8 21 l%d1
lld7 22 d6 and White is much better) 20 �0
.c4 21 a3 b6 22 .i.e3 llad8 23 l%ct 1l'd3 24
a4 and White has an edge, Dvoirys­
Vladimirov, Bamaul 1988.
1 4 1lb5 1 6 •xf3 dxc4 1 7 .tc2
This is probably the most promising of
White's options. Alternatives include:
a) 14l:lxb7 �b6 15 .i.d2 .i.g6!? 16 .i.xg6
hxg6 171Ve2 (or 171Vb3 .c8 18l:lxb6 axb6
19 .xdS 11'c7 with an unbalanced position)
17 ...•c8 18 11'a6 l:ld8 is again unclear. Oliv­
ier·Raetsky, Geneva 1999 continued 19 llks
.i.xeS 20 dxeS l:ld7 21 l:lxb6 axb6 22 .xb6
l:lb7 23 11'd4 1Vc4 24 1fxc4 dxc4 and here
Black is a little better.
b) 14 a4 b6 15l:lb5 �£6 16 .tg5 .i.c7 17
g4 a61? 18 l:lb3 .i.g6 19 �eS .id6 20 �xg6
(an aggressive choice is 20 f4 .1xd3 21 1i'xd3
1i'c7 22 .i.x£6 gx£6 23 �B .1xf4 24 �4 1 7 •..•d7
l:lfe8 25 �fS. when White has reasonable Black's two other options are probably in,..
compensation) 20...fxg6 21 .i.h4 .1c7 22l:le1 ferior:
gS!? 23 .i.xgS 1l'd6 24 l:les (White must a) 17...a6 allows a dangerous attack: 18
avoid the blunder 24 f4? h6 25 .tx£6 11'xf41 .i.gS! 1i'c7 (the queen sacrifice 18...axb5 19
26 llb21lxf6) 24...1i'c6 and in this level posi­ .i.xd8 llfxd8 is not sound: 20 11Fhs g6 21
tion the players agreed a draw in Semenova­ 1Wxb5 .i.c7 22 a4 l%xd4 23 •cs lld7 24 g3
Chetverik, Balatonbereny 1994. and White was clearly better in Beliavsky-

64
3 /i)xe5: Bla ck Plays 6 . • . �d6

Petursson, Reykjavik 1988; or 18...f6 19 'ii'hS with unclear play) 25 9f3 � 26 L3 �b4
h6 20 .i.xh6! axbS 21 Lg71 f5 22 .i.xfB 27 h5 �3 28 hxg6 hxg6 (Ponomariov-Safin,
.i.xfB 23 i.xf5 and Black's king is in trouble) Yerevan 2001) and here White should force a
19 .i.f61 g6 (19...axbS? nms into a mating draw with 29 'Wf6! .i.xa3 30 llhSI gxhS 31
attack after 20 .i.xh7+! Wxh7 21 'ifh5+ Wg8 'Wg5+.
22 9gS g6 23 1th6) 20 llb2 and White stiU 1 9 �e3
has a strong initiative. Once again there is a choice:
b) 17...1i'c7 18 a4 a6 19 llb2 llab8 20 a) 19 .i.b2 should be equal after 19..ic7
.i.h6! 9c6 (the point is if 20... gxh6 then 21 20 J:lcS l:fd8? (this is a fatal slip; 20....i.d6 2 1
9f5 Wg7 22 'lfxh7+ Wf6 23 .xh6+ Wc7 24 l:b5 .i.c7 is level) 21 llh5 g6 22 d511fd6 23
lie 1+ Wd7 2S .i.f5+ Wc6 26 .i.t->4+ Wd7 27 g3 gxh5 (If 23...f6 then 24 llxh71 and the
llcbl and the attack breaks through) 21 9f5 attack crashes through) 24 9xh5 and White
g6 22 9a5 �d7 23 9c3 1fdS 24 .LfB �xfB was winning in Zagrebelny-Chctverik, Gyula
2S llet and \Vh.ite had an edge in Yandar­ 1992.
biev-Skatchkov, St Petersburg 2001. b) 19 .i.g5!? is a promising alternative to
18 a4 19 i.e3. For example, 19.. .1lfc8 20 l:lfbt g6
21 .i.e3 Wc6 22 a5 'lfxf3 23 gxf3 �8 24
l:xb7 l:lxb7 25 l:xb7 lle7 26 l:b1 with an
edge for \Vh.ite, De Firmian-Kosebay, Co­
penhagen 1996.
1 9 /i)cS 20 llfb1 b6
...

18 llab8
•..

Supporting the b-pawn, but this is not the


only move:
a) 18...g6 is an important option - sec the
next game.
b) Black should avoid 1B ...i.c7?! 19 g3 a6
. 21 llh51?
20llhS g6 21 lies .i.ds 22 aS 9xd4, as in D. nus is not \Vh.ite's only attacking method.
Kaiumov-Chetvcrik, Harkany 1992. Now 21 h4 liJc7 22 h5 h6 23 g4 .i.c7 24 gS a6 (if
White could have built a powerful attack with 24...hxg5 25 l:xg5 f6 26llg2 Wh8 27 1fe4 f5
23 .i.e3! 1ic3 241lk l llcB (not 24...�7? 2S 28 1fh4 White has a promising attack) 25
.ixg6 1i'xcl + 26 .i.xcl �xeS 27 .i.f5 and l:l5b4 bS 26 gxh6 f5 27 axbS axb5 28 1fg2
White wins) 2S llhS. c) Worth considering, llf7 with a complex position, Shilov-Ractsky,
however, is 18...llfe8!?, for example 19 .i.fS Lugansk 1989.
ti'c7! 20 i.e4 a6 21 llg5 llxe41 (the careless 21 ...g6 22 .i.h6 lld8 23 .f6 �fS 24
21 h6?? is refuted by 22 llxg7+! Wxg7 23
... .i.xfB •xd4
.i.xh6+!) 22 'lfxe4 �a4 23 lle l g6 24 h4 The only move. 24...llxfB? loses to the
li'lc3 (also possible is 24...lld81? 25 h5 b5 clever 2Sllxh71 Wxh7 26llb5. For example,

65
The Petroff Defen ce

26...Wh6 27 llhS+I Wxh5 28 'iff41 and mate which gave White an edge in Svidler­
is unavoidable. Akopian, Yerevan 1996) 23 axb6 .i.e7! 24
25 •xd4 bxa7 b6 25 i.e4! 11'xa7 26 i.dS i.d6 27
After 25 Jl..g7 'ifxf6 26 .i.xf6 ltd6 27 ::lh4 :Xb6 :bSI 28 :Xb8 llxb8 29 llxb8+ i.xb8
llxf6 28 ltxc4 �6 Black has won back the 30 Jl..xc4 tli'at+ 31 .tn tli'xd4 with a drawn
piece and keeps a clear extta pawn. position (Svidler).
25 ...:txd4 26 .i.c5 ltd2 27 :te5 :txc2 28 b) 19 i.d2 c3 20 i.xc3 :ac8 21 i.e4 :c4
:tea+ �7 29 .i.d4+ 22 :bb1 :xa4 (Black has a promising alter­
29 .i.fB+ is le::;s clear: 29. ..'itf6 30 :dt c3 native in 22...:fc81? 2'\ .tat :xa4 24 i.xb7
3t:dd8 with an ob::;cure position. :c7 25 i.e4 :a3 with counterplay - Barlov)
29 f6 30 .i.e37
••. 23 i.xb7 :V 24 :ret (or 24 i.c6 11fc7 25
This spoils White's advantage. 30 ltdt c3 :at :b3 26 :fb1 :xbt+ 27 :xb1 rlcs 28
3t Jl..e3 Wf7 32 ::ldd8 lta2 33 ltxc8:Xc8 34 dS liJxd5 29 :b7 1Wxb7 30 i.xb7 :Xc3 31
llxcS :Xa4 35 llxc3 promises some winning 'IVdt i.f4! heading for a drawn opposite­
chances. coloured bishop ending - Kaspamv)
30 . :tc3 31 ltd1 ltd3 32 :C1 ltd7 33
.. 24...tli'c7! 25 ltat %lb8 26 i.e4 ltb3 was
¢1 1 7 equal in Kasparov-Shirov, .Iinares 2000.
Instead 3 3 Jl..f4 :laS 34 llxc4 tbd6 35 1 9...:ac8
:c7! holds on to equality.
33...g5 34 �e2 ¢17 35 llh8 Wg7 36
lte8 lte7 37 :td8 :C7 38 h4 h6 39 f47
The final mistake. 39 :d4 :bb7 40 hxgS
hxgS 41 ltcxc4 liJe7 still leaves White some
hopes of saving the draw.
39 ...gxh4 40 f5 :tbb7 41 .i.f4 ltc6 42
ltcd1 a6 43 lt1d4 h5 44 .i.c1 fi::Je7 45
:t8d7 ltcc7 46 :Xc7 :Xc7 47 .i.f4il)xf5
48 .i.xc7 il)xd4+ 49 �e3 lLlf5+ 50 ¢14
il)e7 5 1 We4 b5 52 a5 b4 53 � b3
0-1

Game 28 20 ltfb1
Anand-Shirov After 20 i.h6 :res White has an unusual,
Unarr:s 2000 way to force a draw: 21 i.fS! gxfS 22 '1Vg4+11
fxg4 23 ltgS+ Wh8 Va-lfz Zaw Win Lay­
'

1 e4 e5 2 il)f3 il)f6 3 lLlxe5 d6 4 il)f3 Pacicncia, Surabaya 2002.


il)xe4 5 d4 d5 6 .i.d3 .i.d6 7 0.0 0-0 8 20 c3 21 a5il)c4 22 :txb7
.••

c4 c6 9 cxd5 cxd5 1 0 il)c3 fi::Jxc3 1 1 This is promising but 22 .i.h6!? is also in­
bxc3 .i.g4 1 2 :tb1 il)d7 1 3 h3 .i.h5 1 4 teresting: 22.. %lfe8 23 :xb7 'ifc6 (not
.

ltb5 fi::Jb6 1 5 c4 .i.xf3 1 6 •xf3 dxc4 1 7 23 ..11t'xb7? 24 '1Vxb7 %lb8 25 a6! and White
.

.i.c2 •d7 1 8 a4 g 6 1 9 .i.e3 has a clear advantage) 24 :lfl! lte7 25 ::lxe7


White's two other options are less threat­ i.xc7 26 ltdt (Greenfeld mentioned the·
ening: unclear line 26 d51? 'ife5 27 .i.f4 'iff6 28
a) 19 Jl..h6 :res 20 :fbt :ladS 21 i.gS 1llg4:cs 29 .i.h6) 26...�2 27 'ild3 .i.d6 28
Jl..e 7 22 aS Jl..xgS! (this i::; more accurate than a6! 'ifc4! and White was a bit better in
terman-Greenfcld, Haifa 2002.
AI-11
22...�5 23:Xb7 'ife6 24 Jl..d2 .i.f6 25 ltet,

66
3 f.i)xe5: Bla ck Plays 6... J.. d6

22.......6 30 lla2?
More solid than 22...:C7 23 .ih6lle8 24 Now White loses control of the position.
Axc7 'flxc7 25 'flxc3 when White has a dan­ Even worse is 30 'ifxa7? llxe6 31 dxe6 .i.d4
gerous initiative. For example, 25...loa3? 26 32 l:xh7+ WgS 33 'flb7 Wxf2+ 34 Wht c2!
llet! llb8 27 'flxc7 .i.xc7 281le7 ltlxc2 29 35 :t? .i.xal 361lxf2 ct'fl+ 37 Wg2 :xl2+
:Xc7 ltlxd4 30 llxa7 gave White a winning 38 Wxl2 .td4+ when Black has decem win­
ending in Motylev-Ristic, Novi Sad 2000. ning chances. White could have maintained
23 1la1 an edge with 30 :b4!? llb8 31 :c4 c2 32
Less threatening is 23 .ib3 'fit'S 24 11fxf5 llct llbl 33 ll4xc2 .td4 34 Wf41lxc1+ 35
gxfS 25 llal (Grischuk-Shirov, FIDE World llxct .i.xf2+ 36 Wg2 (Rogers). However, the
Championship, New Delhi 2000 ) and here best option is Shirov's later suggestion: 30
Grischuk suggested the equalising 25...ltld2!? lla41 c2 31 llc4 .td4 32 1i'f4 Wxf4 33 gxf4
26 g3ltlxb3 27 :Xb3 c2 281lct llc4. .i.xf2+ 34 Wg2 with good winning chances.
23...J..b8?1 30....id4 31 ...,
Now Black is in real trouble. 23...Ac7 24 31 1td3!? .txf2+ 32 Wg2 .i.d4 33 llc2
a6 lDxe3 25 fxe3 'fle7 26 l:.abt is only mar­ with an unclear position is better.
ginally better for White. 31 .....f3 32 Wh2 •xd5 33 .ixd5 llxe1
24 i.b3 .d6
If 24 ... c2 25 Act Wd6 26 :Xc2 9hz+ 27
Wft ltlxe3+ 28 fxe3 lLc2 29 .i.xc2 White is
clearly better (Shirov).
25 g3 f.i)xe3 26 J..xf7 + �h8 27 •xe3
.,6
Pushing the pawn again achieves nothing:
27...c2 28 llct llc7 29 llxc7 .ixc7 30 .ic6
i.xa5 31 d5 (Shirov).
28 .i.e& llceB 29 d5 .ie5

34 �2?
Now White is lost. 34 f4 :res 35 .to
llct still leaves White with good drawing
chances.
34 ...i.xf21 35 llf7
The point is 35 llxf2 llxf2+ 36 Wxf2 c2
and the pawn queens.
35 1lxf7 36 .ixf7 .ic5 37 .ib3 ¢>g7 38
•..

:C2 .id4 39 a6 ¢>16 40 lla2 'ite5 41 h4


�e4 0-1

67
The Petroff Defence

Summary
The variation 6...J.d6 7 0-0 0-0 8 c4 c6 has become fashionable and is likely to remain popular
because of the huge number of complicated, untypical and insufficiently studied positions.
After 9 .c2, defending the knight with 9... f5, as in the historic Wtll.iams-Staunton game
(London 1 851) isn't had because it's not that easy for White to take advantage of the weakened
a2-g8 diagonal. After 9 .c2 �a6 10 a3 Black has a few effective responses: 1 O J:e8 1 1 llk3
•.

J.f5; t t ...J.g4; or the immediate 1 0....i.g4, responding to 1 1 lDe5 with 1 1 .....tf5, 1 t ..J.hS or
1 1 .....txe5 12 dxe5 �c5. The fashionable 1 o...f5 appc..-ars to give White the better chances after
an accurate attack on the queenside with 1 1 �c3 �c7 1 2 lDe2 � 1 3 b4.
In the event of 9 1let. the pawn sacrifice 9.....tg41? deserves careful consideration. After 10
Le4 dxe4 1 1 llxe4 f5 1 2 lle6 1id7 Black has full compensation, while he has prospects of
equalising after 12 1let ..txf3 13 Lf3 1t'h4. In the case of 9 ... J.f5 White can claim an advan­
tage with 1 0 llk3 �xc3 1 1 bxc3 ..txd3 1 2 Wxd3 dxc4 1 3 11'xc4 �7 14 ..tgS!?, while after
9...1le8 10 ru �xc3 1 1 bxc3 the position appears to be favourable for White.
To us, the move 9 �c3 seems less accurate than 9 cxd5 cxd5 1 0 llk3 because Black can
proceed with 9...�xc3 10 bxc3 dxc4 1 1 Lc4 ..tg4. We can't find any white's advantage in
either the calm variations or the sharp 1 2 1id3 �7 1 3 clOgs �f6 1 4 h3 .i.h5 1 5 f4 h6 1 6 g4.
Recently 9 cxd5 cxd5 1 0 ru �c3 1 1 bxc3 .i.g4 has not been so popular. White's position
seems to be more promising after 1 2 1lb1 b6 but. if Black defends accurately. White will not
gain an advantage after 12 ..�7. Nowadays there are fewer fans of this complex line, which
requires knowledge of variations developed earlier and where t.-very moment is critical.

1 e4 e5 2 �f3 �f6 3 lnxe5 d6 4 �f3 lnxe4 5 d4 d5 6 ..i.d3 .i.d6 7 0-0 0-0 8 c4 c6


(D) 9 cxd5
9 1fc2 �a6
10 .ixe4 Ga111t 19; 1 0 a3 .i.g4 - GtUJJe 20; 1 O... f5 - Game 21
-

9 1le l
9....if5 - Game 22; 9...1le8 - Gan1e 23
9 llk3 �xc3 1 0 bxc3 dxc4 1 1 .lxc4 .i.g4 (D): 12 h3 - GttnJe 24; 1 2 .d3 - G011Je 2;
9 cxd5 10 �c3 lnxc3 1 1 bxc3 .i.g4 1 2 .Rb1 (D) �7
••.

12 ...b6 - Gm11e 26
1 3 h3 .i.h5 14 .Rb5 �6 1 5 c4 .i.xf3 1 6 11xf3 dxc4 1 7 .i.c2 ...d7 1 8 a4
1 8. 1lab8 Game 27; 18...g6 - Gm11e 28
.. -

B. . . c6 1 1 ... i.g4 12 fl.b 1

68
CHAPTER FOUR I
3 ltJxe5: Deviations
from the Main Line

1 e4 e5 2 t0f3 �f& 3 o!bxe5 d& 4 �f3 After S ... dS 6 ..id3 i.e7 7 0-0 the idea
o!bxe4 5 d4 7 ...ll:)d6 is played very seldom, with Bilguer
A fter 1 e4 cS 2 lLlf3 lL!f6 3 lLlxeS d6 4 declaring this move a 'motiveless retreat'. In
ll:)f3 lLlxe4 5 d4 we've already identified two fact, there arc some motives, i.e. a counter­
principal trends: S...dS 6 �d3 �e7 7 0-0 c!bc6 measure to c2-c4 and a preparation of .....i£5.
(< :hapters 1 -2) and 6 ....i.d6 (Chapter 3). It is However, a fter 8 i. f4! Black fails to develop
reasonable to devote a separate chapter to his bishop to f5 and is forced into passive
deviations from these �nes. Black rarely dc­ defence.
c�nes to play the main continuation S...dS in If Blacks wishes to move his bishop to f5,
favour of the passive s...i.e7 (Game 32). he should do so immediately after 5 ... d5 6
After 6 �d3 the knight retreats to f6 rather �d3 i.e7 7 0-0 �f5. Then 8 c4 can be met
than to gS (in view of 7 li:)xgS ..ixg5 8 by 8...dxc4 since the rook is not yet attacking
'ii'c2+). l11en White normaUy proceeds with the knight on c4, and the arising position
7 h3 in order to restrict the c8-bishop. with an isolated pawn gives approximately
After 5 d4 dS 6 ..td3 Marshall suggested equal chances. 8 :Ct lL!c6 (Games 29-30) is
6 .....tg4, which failed the test of time in view more common, after which the position
of 7 "iVe2+1 'fle7 8 0-0 li:)c6 9 �bS, when it is branches. Game 29 reviews the fashionable 9
difficult for Black to complete his develop­ lbbd2 as weU as 9 lbc3 and the pin 9 i.bS,
ment. In Game 31 we discuss MarshaU's idea taking control of the eS-squarc; Game 30 is
of an early .....tg4 improved by the insertion devoted to the more fundamcnral 9 c4.
of the moves 6 ...li:)c6 7 0-0. The two critical
lines arc 6 ...c!bc6 7 IJ-0 ..ig4 8 c4 lLlf6 9 cxdS Game 29
..txf3 10 'tixf3 'it'xdS 1 1 'it'xdS (or 11 "iVe2+; Kasparov-Karpov
I I l:cl+ �c7 reaches a tabiya from Chapter Mo.rco»' 1981
2) and 9 li:)c3 i.xf3 10 "iVxf3 li:)xd4 1 1 'it'c3+
(or 1 1 'it'h3; again 1 1 l:e l + ..ic7 rt-aches 1 e4 e5 2 t0f3 �f& 3 lDxe5 d& 4 �f3
Chapter 2). l r should be said, however, that �xe4 5 d4 JJ..e7 6 .i.d3 d5 7 0-0 JJ..f5
in modem practice Black usually avoids giv­ Retreating voluntarily with 7...li:)d6 makes
ing his opponent a variety of possibilities and White's task easier: 8 i. f4 0-0 9 :let ..ie6 (or
often develops his bishop to e7 early on. 9.....ig4 10 h3 ..ihS I I l:leS!? i.xf3 12 1i'xf3

69
Th e Petroff Defence

c6 1 3 c3 �7 14 Ae2 ltlf6 1 5 � with a edge, Zhang Pengxiang-Delchev, Unares


clear advantage, Lau-Mathe, Munich 1992) 2002).
t o c3 �7 1 1 1Vc2 h6 (this is forced as b2) 1 1 lle1 �6 1 2 i.d3 i.f6 1 3 dS
1 1 ...g6?! allows the standard tactic 1 2 i.h6 i.xd3 14 1Wxd3 ltlb4 1 5 1i'd1 aS (a sharper
lle8 1 3 lLe6! fxe6 1 4 i.xg6, when White is option is 1 5...ltlf5 1 6 g4 �6 1 7 a� ltla6 1 8
much better) 12 ltlbd2 :e8 1 3 ltlft ltl£8 14 i.f4 1Vd7 19 h3 with an unclear position
ltlg3 i.g5 15 1Vd2 f6 16 h4!? Lf4 1 7 1fxf4 according to Tal) 16 a3 t'ba6 17 .i:e3 :Cs 1 8
'ird7 1 8 ltlh2!? 1ff7 1 9 ltlg4 and White had a 'iVa4 �b8!? 1 9 i.d4 (Komeev-Ro.Perez,
promising initiative in Timman-Skembris, Collado Villalba 2000) and here Perc?. now
Corfu 1 993. suggests the correct route to equality is
1 9...ltld7!? 20 :xeS+ �xeS.
8 &6 9 �bd2
...

This is the best of White's alternatives (the


main move 9 c4 is discussed in the next
game). Other options include:
a) 9 i.bS i.f6 t O �bd2 0-0 1 1 �f1 �e7
1 2 c3 ltlg6 1 3 i.d3 �6 14 i.xf5 ltlxf5 1 5
'iVb3 b6 16 1rbs a6 1 7 'iVd3 1Wd7 1 8 ltlg3
�xg3 1 9 hxg.� aS 20 i.g5 i.xg5 21 ltlxg5
:reB and Black has easy equality, Karpov­
Korchnoi, World Championship (Game 4),
Merano 1 981 .
b) 9 ltlc3 ltlxc3 10 bxc3 i.xd3 1 1 cxd3
8 �1 (1 1 'irxd3 also gives nothlng after 1 1 ...0-0 1 2
This is more promising than the alterna­ c4 dxc4 13 1rxc4 i.f6 1 4 c3 :eS) 1 1 ...0-0 1 2
tives: 1i'b3 llbs 1 3 :e2 bs 1 4 i.d2 1Wd7 1 s :net
a) 8 ltles ltld7 9 1ff3 ltlxe5 10 dxe5 'ird7 :res and Black had equalised in A.lvanov­
1 1 ltlc3 i.g4 1 2 e61 (a neat trick but Black Pierrot, Buenos Aires 2003.
has a defence) 1 2...i.xe6 (not 12 ...'irxe6?! 1 3 9 .c!Dxd2 1 0 'irxd2 .i.xd3 1 1 'irxd3 0-0
••

ltlxd5! 1Wxd5 14 1Vxg4 ltlf6 1 5 Wa4+ 1Wd7


16 1Vxd7+ �xd7 1 7 :e1 - Ro. Perez - when
the bishop pair and especially the nasty pin
give White a clear advantage) 1 3 ltlxe4 dxe4
14 i.xe4 c6 1 5 :d1 1Wc8 1 6 :ct 0-0 1 7 Whs
f5! 1 8 i.d3 i.f6 19 c3 :Cs and Black had
equalised in Sutovsky-Ro.Pcrcz, Istanbul
Olympiad 2000.
b) 8 c4 has the obvious drawback of los­
ing time with White's Icing's bishop. 8...dxc4
9 i.xc4 0-0 10 �c3 lL!c6 and now:
b1) I f t 1 i.d3 rhe correct answer is
1 1 ...�c3 12 bxc3 'iVd7 13 :e1 i.d6 14 :b1
b6 1 5 :bs i.xd3 1 6 'irxd3 :res, when Black 1 2 c3
has no problems (but not 1 1 ...�d6 12 i.xfS Kasparov quietly defends d4 before start­
ltlxf5 13 d5 ltlb4 1 4 a3 lba6 1 5 1rc2 �h4 1 6 ing active play. 12 i.f4 is also possible:
�xh4 i.xh4 17 :ld 1 , when White has an 12 ... i.d6 13 ll)g5 g6 1 4 Wh3 h5 1 5 i.xd6 (1 51

70
3 li:lxe5: D e viations from the Msin L ine

i.e3 illustrates the potential weakness of d4: Playing directly for exchanges with
1 5...lnxd41 1 6 ttlxf7 ttlf3+! 1 7 Wxf3 lLf7 1 8 1 9...l:te7 leads to some difficulty: 20 tL\b3
'lfh3 llf5 and BL1ck has equalised) 1 5...cxd6 llxe2 21 Wxe2 i.e? 22 'l'g41 Wxg4 23 hxg4
1 6 ttlf3 1Vb6 1 7 l:tab 1 ttlxd4 1 8 ttlxd4 Wxd4 i..d6 24 i..xd6 cxd6 25 f4 and Black's dou­
1 9 Wd7 and White's activity certainly com­ bled d-pawns ensure White of a slight edge in
pensates for the missing pawn but is proba­ the ending.
bly not enough to force an advantage. 20 ltlf1 .i.f8 21 1t'f3 J:le7 22 ltle3
1 2...1t'd7 1 3 .i.f4 a& 1 4 J:le3 J:lae8 1 5 The assessment is typical of this variation:
J:lae1 White has a nagging pull.
White is more active but Black is very 22...�8 23 .i.xc71?
solid and simplifying exchanges seem likely. An interesting way to unbalance the play.
1 5.....td8 1 6 h3 23 ...1t'xc7 24 ltlxd5
Safeguanling the back rank and keeping
the tension. 1 6 �5 eases Black's task after
1 6...ttlxe5 17 dxe5 'Wbs 1 8 b3 Wxd3 1 9
l:lxd3 c6 20 c4 i.a5 2 1 llbt dxc4 22 bxc4
lld8 23 lla3 (Kavalck-Smyslov, Amsterdam
1 981) and here the simplest path to equality
is 23. ..i.d2 24 i.xd2 llxd2.
.

1 6 J:lxe3 1 7 llxe3
••.

24 1t'd6?1
•.•

Black should prefer 24...:Xe21? 25 ttlxc7


l:tet+ 26 ..t>h2 i.d6+ 27 g3 .ixc7 28 ...d5+
�f8 29 Wg2 with unclear play: Black has
plenty of material for the queen but regaining
coonlination without losing more pawns will
be very difficult.
25 li:lxe7+ .be7 26 1t'e4 .i.f8 27 1t'e8
1 7 f6
.•. Kasparov su�>ested that 27 c4 b6 28 d5
Avoiding a nasty back rank trick: gains a clear advantage.
1 7...lte8?! 18 Wf5! l:te6 1 9 h4 g6 20 'irh3 27 g6 28 a4
•..

1i'c8 21 ltxe6 'l'xe6 22 'irxd) fxe6 23 tl\g5! Now Black manages to untangle. 28 h4!?
and White has a clear advantage (Kasparov). Wg7 29 h5 keeps an edge.
18 J:le2 28 ...Wg7 29 b4 1t'c7 30 J:le3 li:lf7 3 1
This simply improves coordination while 1t'e& •d8 32 a5 h5 33 ee4 ed7 34 ••6
awaiting developments. In a later game 1 8 1t'd8 35 �1 li:lh6 36 g4 hxg4 37 hxg4
ll:\12 was tried and following 1 8...ttle7 1 9 li:lf7 38 We2
tL'lbJ 'irf5! 20 'l'xf5 ttlxf5 2 1 l:tc2 b6 22 tl\c1 Or 38 f4 'ire? 39 llf3 i..d6 40 1i'e4 Wc8
<tlf7 23 ttld3 aS 24 J,>4 ttle7 a draw was and Black still holds on to CI.Juality.
a�-,rreed in Leko-Kramnik, Dorunund 1 999. a8 ltlg5 39 as •d7 40 Wc�3 ..td& 41
.••

18 Af7 1 9 li:ld2 .i.e7


•.. �2 % -%

71
The Petroff Defence

17 .tf4 l:ad8 18 llbd1 and White has a slight


Came JO edge. Now the game Bruzon-Ro.Perez, Santa
Herrera-Ro .Perez Clara 2000 continued 18 ... .i.f6 19 h3 :Xe1+
Varadero 2000 20 llxet b5?! 21 .i.g5 llk4 (Nogueiras gives
2t ....i.xg5 22 li)xg5 5 23 d61 1Wxd6 24 Wx5
1 e4 e5 2 �f3 �f6 3 �es d6 4 �f3 with a winning advantage for White) 22
�e4 5 d4 d5 6 .id3 .te7 7 0-0 .tf5 8 .i.xf6 gxf6 23 llc..--41 and White has a clear
lle1 �c6 9 c4 plus.
b2) 1 3....td6 14 �5 ..f6 15 llbt (1 5
..xb7?1 grabs a poisoned pawn and Black has
tricks based on a later ....txh2+, i.e. 1 5.. .1lab8
1 6 ..a6 li)xd41 17 ..d3 li)xf3+ 1 8 WxO
1rxt3 19 gxO a6 and Black's stronger pawn
structure gives him an edge) 15 ... a6 16 1Wd3
b5 17 a41? ...g6 (or 17 ...bxa4 1 8 1la1 l:lfe8 19
l:lxe8+ llxe8 20 llxa4 aS 21 .i.d2 llb8 22
l%a 1 and White is a bit better) 1 8 1Wxg6 hxg6
19 axb5 llfb8 20 c4 axb5 21 c5 .i.f8 22 .i.f4
lla4! and despite Black's enterprising play
White still has an edge, Abreu-Ro.Percz,
Cuba 2003.
9 o!Ob4
••• 1 0 cxd517
Provoking a tactical sequence while fight­ Now the play becomes very complicated.
ing for control of d5. The simple 1 0 .i.ft also allows White to try
Instead 9 ...0-0 makes it much easier for for an edge, for example:
White to fight fur the advantage: a) 1 o... dxc4 leads to very sharp play after
a) 1 0 cxd5 ..xd5 1 1 llk3 (the tactics fa­ 1 1 li)c3 (or 1 1 .1xc4?1 0-0 12 a3 ll)d6! 13
vour Black, who is clearly better after 1 1 .txf7+!? :Xf7 14 axb4 .i.g4 1 5 li)bd2 .i.g51?
..c2? li)b4 12 .Le4 li)xc2 13 .i.xd5 .llaeB! and Black has superb compensation for the
14 .i.xb7 .i.b4 15 :XeS llxe8 16 .i.c6 l:b8, pawn) 1 1 ...lbf6 1 2 .i.xc4 0-0 1 3 a3 llk6 (try­
Shirov-Ivanchuk. Monaco (rapid) 2002) ing to win material with 13 ...lbc2?! fails to 14
1 t ...li)xc3 12 bxc3 llfe8 (Black can try to lilh4 .i.e4 1 5 lllxe4 lllxa1 16 lilts, when
blockade with 1 2... b5 but after 13 L5 White has a large advantage - the knight on
•x5 t4 li)e5 li)a5 1 5 a4 b4 16 .td2 White at has no chance of escaping) 14 d5 ll)a5 15
has a strong initiative) 13 .i.f4 .i.xd3 1 4 .i.a2 c5 16 .i.g5 :Cs 17 1Wa41 .i.d7 18 1Wc2
..xd3 ..d7 1 5 :C3 .i.f6 16 llae1 lle7 1 7 h6 19 .i.h4 and White had a strong initiative
.i.g5 :Xe3 1 8 fxe3 .i.xg5 1 9 li)xg5 J!f1 20 e4 in Karpov-Portisch, Tilburg 1982.
and White's impres.'iive centre gave him good b) 10 ...0-0 1 1 a3 llk6 12 cxd5 (1 2 ltlc:J
chances in Yagupuv-Sorokin, St Petersburg ltlxc3 13 bxc3 dxc4 14 .i.xc4 rransposes to
2001 . 7...liJc6 8 c4 ltlb4 9 .i.e2 0-0 10 li)c3 .i-5 I 1
b) Also good is 1 0 llk3 lbxc3 1 1 bxc3 a3 lilxc3 1 2 bxc3 liJc6 13 lle1 dxc4 14 .i.xc4
.i.xd3 1 2 ..xd3 dxc4 1 3 ..xc4 with a further - see Game 2) 12...1l'xd5 13 ltlcJ lllxc3 14

branch: bxc3 .i.f6 1 5 .i.f4 ..d7 16 lla21? llac8 17


b1) 13.....d7 1 4 llb1 b6 15 d5 � 1 6 l:ae2 :Xe2 18 1fxe2 li)a5 19 lbe5 1Wa4 20
1Wd3 llfe8 (or 1 6...c 6 1 7 c4 cxd5 1 8 cxd5 g4!? (White's advantage is his activity rather
llad8 19 .i.gS! and White keeps the initiative) than his structure, so it is vital to continue

72
3 li:Jxe5: Deviations from the Main L ine

aggressively) 20....i.e6 (or 20....i.xe5 21 .i.xe5 the wayward knight wiU soon be coUected.
.i.d7 22 .i.xc7 lieS 23 'ifd2 'ifxa3 24 .:.xeS+ 14 i.xd7 /l)xc1 1 5 �c3
.i.xe8 25 •g5! and White keeps an edge) 21
g5 .i.d8 and White was a bit better in Korn­
ec.'V-Ro.Percz, Albacete 2000.
1 0 ...lnxf217
This is the critical try. 10...�xd3 1 1 '1Vxd3
'ifxd5 transposes to 7...�6 8 c4 �b4 9 cxdS
�d3 10 9xd3 'IVxdS 1 1 .:.e1 .i.t5 (sec
Games 9-1 0).
1 1 .84+
The only good move. If 1 1 .tb5+?! c6 1 2
.a4 0-0 1 3 dxc6 bxc6 14 Lc6 �fd3 Black
has a clear advantage. 1 2 dxc6? is even worse:
1 2...�xd1 13 c7+ '1Vd7 14 .i.xd7+ .i.xd7 1 5
�a3 �f2! and Black i s winning. 1 5... lDcd3
1 1 ....id7 Black can also try 1 S...ltkxa2, after which
White's best is 1 6 .:.xa2!? �xa2 1 7 �a2 a6
1 8 .i.t5 with an edge. Instead after 16 �xa2
�xd5 Black hits c3 and thn:atens ...�b6. 17
:act �xe3 1 8 .:.xe3 .i.d6 leading to an un­
clear position.
1 6 a3 lDc5
Ro.Percz as.'ICsses 16.Jnxb2 17 9b3 lilc4
1 8 llxe7 'flxe7 19 axb4 as a bir better for
White.
1 7 dxc5
This aiJows White to keep some control
lnstc.-ad 17 '1Vxb4 aS 18 'it'bs �xd7 1 9 :act
.i.d6 20 1Wxb7 .:.b8 21 Wa7 .:.xb2 22 '1Vxa5
1 2 i.b5 �b6 is very messy.
Again White has no choice. After 12 1 7 ....hc5 1 8 axb4 i.xe3+ 1 9 �1
1i'xb4? �xd3 1 3 .:.xc7+ '1Vxe7 14 '1Vxb7 0-0 White's two minor pieces are preferable to
1 5 .i.d2 Wc2 Black wins easily. Black's rook and pawn.
1 2.../l)fd3 1 9...1tf6
Advancing further into White's position t 9... a6 20 .:.et .i.b6 21 .i.£5 is also mar­
with 1 2...�bd3?! leads only to trouble: 1 3 ginaUy in White's favour.
.:.e2 a6 (or 1 3...�xc1 14 .:.x.£2 �3 1 5 .:.d2 20 i.b5 1lad8 21 11rc2 •f4 22 .i.d3 g6
�f4 16 �5 .i.xb5 1 7 9xb5+ <i'fB 1 8 .:.f2 The other way to cover h7 is 22...h6, but
and White is much better) 1 4 .i.xd7+ '1Vxd7 White keeps the attack going with 23 bS
1 5 'ti'xd7+ <i'xd7 1 6 .:.xf2 �xf2 17 <i'xf2 .:.res 24 .:.n .
.i.d6 1 8 � and White's material advantage 23 b5 lltea 24 :n ts 25 a3 •de 26
is almost decisive. 1i'c4 26 ...1tc57
1 3 1le3 0-0 Overlooking a nasty trick. 26.. Wh8 27
.

1 3....i.xb5 is no improvement 14 'IVxbS+ .i.c2 :d7 28 �4 limits White to ju<�t an


�f8 15 .:.xd3 �2 16 9c4 �xa1 17 b3 and edge:.

73
The Petroff Defence

10 cxdS .i.xf3 1 1 1ixf3 l0xd4 12 'lre3 (this is


simplest but 12 1id1 1? also works:
1 2 ....i.xh2+ 1 3 �fl 'ifxdS 1 4 ltic3 and White
is much better) 1 2......f6 1 3 .i.xe4 fxe4 14
l0d2! (less ambitious is 1 4 'lrxe4+ Wf7 15
.i.gS 'iVxgS 16 'ifxd4 llhe8 1 7 l0c3 when
White was only a litde better in Capablanca­
Marshall, New York 1910) 1 4...l0f5 1 5
9xe4+ liJ.c7 1 6 l0 f3 h6 17 9a4+ Wd8 1 8
'iVbs Wc8 19 .i.e3 when White's extra pawn
and safer king ensures a huge advantage.

27 1i'xc5?1
White misses it as weU: 27 d6+! q;g7 28
dxc7 :d7 29 b6 is effectively winning.
27 i.xc5 28 i.c4 h6?1
••.

This aUows White a second chance for the


same tactic. Instead 28 ....i.d6 29 l0g5 lieS 30
l0e6 lieS 31 g3 is very unpleasant, but at
least Black can play on.
29 d6+ Wg7 30 dxc7 .ZZ.d7 31 b61 a6 32
i.e& 1 -0

Game 3 1 8 ...lbf6
Kupreichik-Vusupov Trying to grab material with S. ...i.x£3 9
.

Minsk 1987 'lrxf3 l0xd4?1 backfrres after 10 'lre3 l0f5 1 1


______________.. 'lrf4 'lrd7 1 2 .i.xc4 dxe4 13 'ifxe4+ .i.e7 1 4
1 e4 e5 2 lbf3 lbf6 3 lbxe5 d6 4 li)f3 'lrxb7 0-0 1 S l0c3 when White is clearly bet-
lbxe4 5 d4 d5 6 i.d3 i.g4 ter.
Or 6...l0c6 7 0-0 .i.g4. 9 l003
7 0-0 Developing and increasing the pressun: is
Also interesting is the simple 7 'lre2 'fie? the correct approach. The simple 9 cxdS
(instead 7 ... f5?! allows an instructive trick: 8 leads nowhere after 9 ... .i.xf3 (9...l0xd5 10
h3 .i.hS 9 g4! fxg4 to ltics .i.f7 1 1 hxg4 and ltic3 .i.e? transposes to 6....i.e7 7 0-0 ltic6 8
White is clearly better) 8 0-0 l0c6 9 .i.bS and c4 l0f6 9 ltic3 .i.g4 10 cxdS lClxdS - see
White has an edge, which will grow consid­ Game 1 1) 10 'iVxO 'iVxdS 1 1 'iVe2+ (1 1
erably if Black plays too direcdy with 9...a6 'lrxdS ltixdS 12 �3 0-0-0 1 3 .i.c4 ltice7 1 4
10 .i.xc6+ bxc6 I 1 lle1 'iie6 1 2 l0bd2. .i.d2 't!f> 1 5 D.fe1 .i.g7 1 6 .i.g5 1ld7 17 .i.xe7
7 lbc6 8 c4
.•. ltixe7 1 8 .i.xf7 Wb8 1 9 lle2 lOeB 20 .i.b3
Immediately undermining the e4-knight llxd4 21 g3 aS, as in Adams-Piker, Dort­
but 8 :e1 is also possible. Now 8 ....i.e7 mund 2000, is also a1ual) 1 t ..ie7 12 .i.bS
transposes to Chapter 2, while 8... f5 is an­ 'ii'd6 1 3 ltic3 0-0 1 4 .i.xc6 bxc6 1 5 1ldt llfe8
other possibility: 9 c4!? .i.d6?! (Black must 1 6 1WO lOdS t 7 l0xd5 'lrxdS 1 8 'i'xdS cxdS
play 9 ....i.b4, which leads to very unclear play 19 .i.f4 c6 20 JLct llac8 and White had no
after 10 lle3 0-0 1 1 a3 .i.aS 1 2 cxdS 'lrxdS) more than ei.Juality in Short-1-I.Oiafsson,

74
3 fi:lxe5: De viations from th e Main L ine

Reykjavik 1987. This is certainly the natural move but


9 .i.xf3
•.• Black should seriously consider 1 3 ...1i'c8. For
9..ie7 ttansposes to 6...i.e7 7 0-0 ltk6 8 example, 1 4 'ire3!? �6 (14...ltk2 is too
·

c4 lt.)f6 9 lt.)c3 ..tg4 (sec Game 1 1). greedy: 1 5 We2 'lrg4 - 1 S...lt.)xa1? ill refuted
10 1i'xf3 li:lxd4 1 1 Wh3 by 1 6 ..ixf6 gxf6 1 7 l0ds - 1 6 ..xg4 �g4
1 7 Jl.xe7 Wxe7 1 8 lfitS+ Wf8 1 9 :act l0d4
20 lt.)xc7 l:d8 21 J:lfe1 and Black's awkward
king delays his development, giving White a
pleasant edge) 1 5 Jl.xf6 .i.xf6 1 6 lDd5 1Wd8
1 7 1tb3 i.d4 18 •xb7 0-0 1 9 llab1 ltb8 20
1fa6 c6 21 1fxc6 i.xb2 and Black has es­
caped to �::quality, Bologan-Koch, Belfon
2002.
1 4 llad1 c5
If 14 ... h6 White should play calmly with
IS ..tf4 cS 16 lt.)bS when his initiative is very
threatening. 1 5 .i.xh6?! ill premature:
I S...gxh6 1 6 J:ld3 Wh7 17 llg3 �8! and
This is probably the most dangerous op­ Black defendll successfully.
r:ion but 1 1 'ire3+ is also worth considering: 1 5 llfe1
t l ...c!De6 12 cxdS lOxdS 1 3 lt.)xdS ..xdS 1 4 White's lead in development and the vari­
..ic4 'irbs I S a4 Wa6 1 6 llcJt Jl.c7 (the seem­ ous potential pins constitute good compensa­
ingly more active 1 6....i.c5 simply helps tion for the pawn.
White to open lines: 1 7 1i'f3 c6 1 8 b4!? i.xb4 1 5 h6 1 6 hh&l
•••

19 lib I Jl.cS 20 :Xb7 and White has the This is even stronger than the tempting 1 6
initiative) 17 b4! 0-0 (1 7 ...Jl.xb4 is too greedy: llxe7!? hxgS (Black can give up the queen
1 8 .i.b2 0-0 1 9 11h3 �5 20 Wg4 .i.e7 21 h4 with 1 6...1fxe7 1 7 lOds lt.)xdS 1 8 .i.xe7
and White has a strong attack) 1 8 1Vh3 br6 1 9 lt.)xc7, but after 1 9 b4 b6 20 bxcS bxc5 21
'ifc3 (keeping Black'll queen out o f the game; 'ifhs White still has an edge) 17 llxb7 llb8
instead 19 Jl.b2 allows 19 ...Wc41 20 l:d7 1 8 llxb8 ..xb8 1 9 b3 'ifeS when the position
J:lae8 21 i.dS 'ifxb4 22 Jl.c3 lt.)f4 23 .i.xb4 is unclear according to Yusupov.
lt.)xh3+ 24 gxh3 i.xb4 25 :Xc7 b6 26 llxa7 1 6...gxh6 1 7 1i'xh6 fi:lh7
�g7 when Black has cleverly equalised, Kas­
parov-Karpov, World Championship (Game
6], London/Leningrad 1 986) 19 ...lt.)g5 20
..ixgS (20 i.b2 may seem more natural but
20 ...i.f6 21 1i'xf6 1fxf6 22 ..txf6 lt.)xe4 23
.

.i.e7 J:lfe8 24 l:ld7 lt.)c3! equalises according


to Kasparov) 20...i.xg5 21 Wxc7 llad8 22
ltxd8 llxd8 23 ..txb7 We2 24 i.o -.,2 25
lift Wxb4 26 'irxa7 (Bauer-Koch, Belfort
2002). Clearly White's extra pawn makes his
position preferable, but the oprosite­
coloured bishops will make any winning at­
tempt very difficult
1 1 dxc4 1 2 hc4 .i.e7 1 3 .i.g5 0-0
.•. 1 B lld3

75
The Petroff D e fence

This is the right rook. After 18 lle3?! �g5 22 h41


19 D.g3 cRh8 20 �5 f5 Yusupov again gives By freeing his back rank, White threatens
the verdict 'unclear'. the queen. Now White has a large advantage.
1 8...J..g5 1 9 Wh5 22 .1Z.xe4
••.

1 9 D.g3 allowsBlack some additional de­ If 22...1i'h6 Kupreichik suggests 23 llxgS!


fensive resources: 19 ./i�fS 20 1Wg6+ li)g7 21
. W'xhS 24 J:l'th5 f5 25 llxh7+! �7 26 li)f6+,
/&4 (not 21 h4? �f6 22 ll:)ds Wh8 and when White is much better.
Black refutes the attack) 21 ...Wh8 22 .ixf7 23 .IZ.xe4 J..f4 24 Jlg4 J..h6
We? with a position that is very difficult to Or 24. ..td6 25 :Xd4! cxd4 26 .i.d
. . 3 and
assess accurately. For example, 23 ..xh7+ Black must resign.
Wxh7 24 li)xg5+ 'irxg5 2S :xgs l:txf7 26 25 1rxc5 �6 26 Wh5 1rd& 27 ,j,xn
l:txcS :ld8 and Black's extra knight wiU have 1t'd1 + 28 *h2 •d6+ 29 f4 •c7 30 J..b3
an interesting battle against the three pawns. 1rd6 31 .IZ.e8 hf4+ 32 g3 .txg3+ 33
1 9...1rf6 20 .1Z.g3 �h3 1 .()
A strong alternative is 20 .rlh31? 1i'g7 21 f4 Mate is inevitable.
.L£4 22 ttxis .tgS 2.1 li)e7+ i..xe7 24 l:txe7
li)c6 25 1lee3 with a dangerous attack. Game 32
20. . ..1Z.ae8 Tukmakov-Bronstein
Al'lo possible is 20 ll:)f5 21 llh3 1i'g7 22
...
Moscow 1971
f4 llaeB 2.1 lle2 when White's attack contin­
ues although there is no clear win. However, 1 e4 e5 2 l0f3 tnt& 3 tnxe5 d6 4 l0f3
after 22 lieS? (I! - Kupreichik) 22 li)f6 23
... tnxe4 5 d4 J..e7
1Wf3 li)h4 24 'ti'g3 .tf4! Black successfully Of course S...dS is the more popular and
defends. reliable move.
21 l0e4 6 J..d3 l0f6

21 Wh8?
..• Trying to equalise by seeking exchanges
Black cracks under the pressure. 21 ...q.,g7 with 6 ...�g5 just falls short: 7 ��5 i.. xgS 8
is essential, when the fight continues. Follow­ 1i'c2+ Wffi (or 8....te7 9 0-0 0-11 l O �c3 /0c6
ing 22 �xf6 (Black has counterplay after 22 1 l �5 lieS 1 2 .ie3 and White also has a
h4 'ti'g(i 23 11fxg6+ fxg6 24 hxgS llf4 - Yu­ slight edge, while after 8. .i..e6 9 f4 i.. h4+ t O
.

supov) 22. .:Xet+ 23 .tft �f6 24 f41 .txf4


. g3 .ie7 1 1 f5 .idS 12 0-0 0-0 13 �3 .ic6
2S llf3 �xf3+ 26 ..xf3 the position is still 1 4 dS .id7 1 5 llk4 White has a dangerous
unclear. initiative) 9 0-0 .i.xct 1 0 llxct �c6 1 1 c3 g6

76
3 l:iJxe5: De viations from th e Main L ine

12 �2 '*g7 13 �1 .i.d7 14 f4 �8 15 'iff3 a) to. .ll)fB 11 :C1 ll)g6 (White has a clear
.

11fh4 16 g3 'ifg4 17 11ff2 and White was a bit plus after 11...d5 1 2 �5 c6 13 cxdS li)xdS
better in Palac-Mun.1', Pula 2002 14 li)xd5 cxdS 15 .i.b5) 12 11fc2 .i.d7 13
7 h3 .i.g5 .i.c6 14 d5 i.d7 15 l:lad1 a6 16 �2
White plays to dominate his opponent's li)hS 17 .tel •cs 18 �d4 cS 19 dxc6 bxc6
queen bishop. 7 0-0 is a sensible alternative, 20 li)5 .i.x£5 21 .i.x5 11fc7 22 1Wa4 li)f6 23
t(,r example 7.....i.g4 8 �bd2 0-0 9 �1 c5 1 0 l2XI4 c5 24 li)c6 and White was much better
h3 ..i.hS 11 ll)ft llk6 and now: in Luther-Bellin, Catalan Bay 2003.
a) Breaking the pin with 12 g4 is weaken­ b) 10. c6 11 J:le1 ll)f8 12 .i.f4 a6 13
..

ing: t 2 ... .i.g6 13 li)g3 d51? (White has a small 'flb3?! (this allows cnunterplay; 13 d5!? li)g6
advantage after 13...l:le8 14 c3 .c7 15 li)5 14 .i.h2 cS 15 'li'd2, with a slight edge, is
i.f8 16 l:lxe8 l:lxc8 17 .i.f4 cxd4 18 li)3xd4 better, while also promising is 13 b41? �6
a6 19 .a4 �5 20 .i.g3 li)b6 21 .c2, Ste­ 14 .i.h2 a5 15 b5) l 3H,�6 14 .i.h2 .i.f8
f.'tnsson-Kholmov, Pardubice 2001) 14 dxc5 (Fischer assessed 14... b5!? 15 a4 bxc4 16
.i.xc5 15 .i.e3 .i.xe3 16 l:lxe3 d4 17 J:lc1 .i..xc4 d5 17 ..i.ft c5 as unclear) 1 5 :C2 b5 16
� with complicated play. 1ic2 ..i.b7 17 J:lae1 g6 with a complex posi­
b) 12 li)g3 is better: 1 2....i.xf3 (or tion, Fischer-Petrosian, 5th matchgame,
12....i.g6 13 .i.xg6 hxh>6 1 4 dS li)b4 15 c4 Buenos Aires 1971.
and White has a pleasant space advantage) 13 1 0 l:iJc3 e6 1 1 b4 l:iJbd7 1 2 .i.e3 t'iJf8 1 3
11fxf3 li)xd4 14 11fxb7 �8 15 ll)f5 li)x5 16 e4
.i.x£5 .i.fB 17 l:lxe8 'ifxe8 18 .i.d2 J!ll 1 9
.i.d3 d5 20 c4 and White has a strong initia­
tive, Ljubojevic-Smyslov, London 1984.
7 0-0 8 0-0 lieS
...

Black has alternatives but no path to


l.'tJUaJity:
a) 8...c5 9 llk3 li)c6 10 J:le1 a6 (to.. lbh4
.

doesn't solve Black's problems: White has an


edge following 11 .i.f4 d5 12 dxcS Lc5 13
.i.eS) 1 1 d5 li)a7 (after 11...li)b4 Karpov
suggt.-sts 12 .i.e4!? as a way to keep the initia­
tive) 12 a4 .i.d7 13 aS l:lc8 14 .i.f1 h6 15
.i.f4 .i.f8 16 J:lxe8 •xeS 17 .i.h2 •ds 1 8
li)d2 'flc7 19 �c4 li)xc4 20 li)xe4 and 1 3...85
White was a bit better in Karpov-Smyslov, White's initiative continues after 13 ...d5 14
Moscow 1972. cS li)e6 15 'ilc2 ..i.f8 16 bS.
b) 8...llk6 9 c3 J:le8 10 J:le I .i.d7 11 14 b6 t'iJg6 1 5 lle1 t'iJh5 16 •d2 .i.d7 1 7
lLibd2 .i.ffi 12 llk4 d5 (exchanging with lZab1 ecs 1 8 .i.f1
1 2...li)xe4 still leaves White with an edge 18 dS!? cS 19 �4 with promising play is
:1fter 13 .i.xc4 h6 14 .c2 'flf6 1 5 .i.e3) 13 also interesting.
ltig3 .i.d6 14 l:lxe8+ 'iVxeS 1 5 'ifc2 h6 16 1 8....i.f5 1 9 l:lb3
.i.d2 .f8 17 ltif5 .i.xf5 18 ..i.xf5 l:le8 19 1 9 g4? wouW be a naive blunder:
i.d3 li)dg 20 c4 and White had an initiative 19 ...i.xg41 20 hxg4 1ixg4+ and Black has a
.

in Bnmstein-SmysJov, Leningrad 1971. clear plus,


9 c4 c6 1 9 h6 20 �2 t'iJf6 21 .-r,2 d5 22 bxc6
••.

The alternative is 9..,li)bd7 10 li)c3 and: bxc6 23 cxd5 .i.b4

77
The Petroff Defence

lf23...cxd5 then 24 l:c1 ..d8 25 �b5 and This is a good move but White could also
White keeps an edge. cash in immediately with 38 I.Oa6 I.Oe4 39
24 llc1 cxd5 25 tDb6 WbB+ 26 Wg1 I.Oxb8 'l'xb8 40 i.d3.
lle7?! 38 . . .'iFd6 39 e5?1
This aUows White's advantage to grow. This allows Black to complicate. 39 I.Ob7?1
Black can limit the damage with 26 ..ltc8 27 J:lxb7 40 .:Xb7 I.Oxc4 with unclear play is also
lDc1 .:Xct 28 1Wxct 'lrd8 29 l2X!3, when unconvincing, but the simple solution is to
White is only A bit better. prepare e4-e5 by supporting the knight with
27 tDe5! i.d7 39 'l'c3 - White wins easily after 39 ...i.h2+
The tactical justification is seen after 40 �ht 1Wf4 41 e5.
27...I.Oxe5 28 dxe5 '1Vxe5 29 i.d4 'lre4 30 39....be5 40dxe5 'iFxc5+ 41 Wh1?!
�d3 Wf4 3 1 i.xfS 'lrxfS 32 .J:f3 •e6 33 Now White's advantage disappears en­
�xf6 gxf6 34 llg3+ �£8 35 1Wc2, when tirely. 41 Wh2 I.Oe4 42 l:c7 'iraS 43 i.b5
Black's weakened kingside causes terrible maintains a slight edge.
trouble. 41 ltla4 42 Wh2 1le6 43 1lc7 'iFa5
...

28 tDd3 ..txb5 29 axb5 ..td& 30 �5! 43..ltc6 44 .:Xc6 ...xc6 45 �d3 1Wg6 is
also good enough for equality.
44 ..i.d3

After 30 b6 a4 31 llbS a3 32 •a2 l:b7 33


l:cbt White's advantage, if any, is' very small
indeed. 44 . . .llg6
30...1rb6 31 li:la4 'iFdB 32 b6 llb8 33 The b-pawn is poisoned due to a back
�5 Ilea 34 11a1 tDh4 35 llxa5 tDt5 36 rank trick: If 44...:Cxb6? then 45 l:xb6 l:xb6
lla7 46 'l'f2! wins, while after 44.. ltbxb6? 45
36 �2!? I.Oe4 37 I.Oxe4 l:xe4 38 �c3 1i'f2! White has a clear advantage.
with a clear advantage is a good alternative. 45 i.xe4 dxe4 46 •d4 e3 47 llxe3
36...tDxe3 37 fxe3 i.g3 38 a4 'iFxb6 48 'iFxb6 llbxb6 %-%

78
3 li:lxe5: De via tions from the Main L ine

Summary
Firsdy, the passive continuations covered in this chapter (5 ... .i.e7 and a quick ...l0d6 after
5... d5) are not of great interest. The key position of the chapter is the one arising after S...dS 6
.Ll3 .i.e7 7 0-0 .i.fS 8 %let lesc6 9 c4. In response to 9...0-0, White successfuUy develops his
knight to c3 either immediately or after 10 cxdS ..xdS. In the case of 9...lesb4 White maintains
110 opening initiative by retreating with 10 .i.fl, but 1 0 cxdS also appears to be a good response

- the tactical blow 10 ..lC!xf21? is not a refutation. Probably Black should transpose to the line
.

6.. .i.e7 7 0-0 �6 8 c4 l0b4 9 cxdS lC!xd3 10 ..xd3 'tixdS 1 1 :C t -*..fS (see Chapter 2) by
.

means of to lC!xd3 1 1 'tixd3 'il'xd5.


...

1 e4 e5 2 li:lf3 li:lf6 3 lbxe5 d6 4 �f3 .!Llxe4 5 d4 (D) d5


S....i.e7 Gm11e J2
-

6 .i.d3 (D) i.e7


6 .i.g4 Game J 1
... -

7 0-0 .i.f5 8 .:e1 lOc6 (D) 9 c4 Gmne 30


-

9 lesbd2 Ga111t 29
-

5 d4 6 .i.d3 B. . . lilc6

79
CHAPTER FIVE I
3 ltJxe5: Fourth and Fifth
Move Alternatives

1 e4 e5 2 .!Llf3 �f6 3 �xe5 d6 4 �f3 idea. In the event of 5 ... �5 White retreats
.!Llxe4 his bishop to e2 and makes up for lost time
After 1 e4 eS 2 �0 �f6 3 �xeS d6 4 by attacking the knight with d2-d4. After
�0 �xe4 White sometimes declines to play S. . dS
. White continues the attack on the t..-4-­
5 d4 in favour of the other continuations, knight with 6 1i'e2, while if S...�f6 White
and these fifth move alternatives are the sub­ clears the way for the d-pawn by means of
ject of the fll'St part of this chapter. c2-c3 and .ld3-c2.
5 �c3 (Games 33-34) was played as fur 5 d3 (Game 39) is the most modest of
back as thc 1 9th century, while later on White's 5th move options. White offers
Nimzowitsch focused on it. Black the opportunity to transpose to the
5 1i'c2 (Games 35-36) was introduced by a Exchange Variation of the French Defence
1 3-ycar old Paul Morphy against LOwenthal after S.)t�f6 6 d4 dS where, as it is known,
(Nt:w Orleans, 1 850), and the line found an extra tempo in this symmetrical line
worthy adherents in the form of Emanuel promises White just a microscopic advan­
L'lsker and Jose Raul Capablanca. Owing to tage. However,. Black can equalise after
the success of such celebrities (sec, for ex­ 6....le7 foUowed up by ...c7-c5.
ample, Game 35) 5 'i6'e2 was considered to Moving onto 4th move alternatives, be­
be nearly a refutation of the Petroff Defence, sides 4 lDO White has two interesting op­
but now we know sevcral ways for Black to tions.
equalise. The fmt major game to witness 4 �c4
Kaufmann, a chess player from Vienna, was l hPaulscn-Schallopp, Frankfurt 1 887.
suggested 5 c4 (Game 37) as a way to pre­ The idea to transfer the knight to c3 is not
vent Black from supporting the e4-knight particularly impressive. In particular, Black's
with ... d6-d5. In principle it's still possible to attack on f2 after 4...lbe4 5 d4 dS 6 lDe3 1i'f6
play S ... dS, but an attack on Black's central- (Game 40) deserves serious consideration for
ist.-d forces with 6 �c3 gives White the better Black.
chances. Instead Black should continue his The Cochranc Gambit 4 lbxf7!? (Games
development with either s ...�c6 or s ...i.e7
. 41 -42), originated by John Cochrane in the
to avoid any problems. 1 840s, stands in rotal contrast to the other
5 .i.d3 (Game 38) is a comparatively new lines in the Petroff Defence. Staunton also

80
3 li::J x e5: Fo urth and Fifth Mo ve A lterna tives

analys1..-d this knigh1-for-two-pawns sacrifice 5 ...d5?!


in his Chess Praxi.r (1 860). Curiously enough, 11us is not a particularly good pawn sacri­
both Cochrane and the famous maestro con­ fice, but the alternatives covered here are also
centrated on 4...�f7 5 .i.c4+?1, which is not great. The best move, 5...l!lxc3, is con­
poor in view of 5...d5!. It nlight be appropri­ sidered in the next game.
ate to rename the Cochrane Gambit to the a) 5...i.f5?? 6 lllxe4?? .ixe4 7 d3 i.g6 8
Bronstein Gambit (he was the one who i.g5 i.e? 9 i.xc7 Wxe7+ 10 i.e2 lllc6 1 1
bn:athed nc.:w life into the sacrifice by indicat­ 0-0 0 was a comedy of error.; that led to
ing the possibility of 5 d4!) or the Vitolinsh C(jUal chances in Miles-Christianscn, San
Gambit (the Latvian 1 M played a number of Prancisco 1987. After the infinitely stronger 6
brilliant attacks with 4 lllxf7). After the piece We21 Black immediately resigned in Zapata­
sacrifice Black's king gets sruck in the centre Anand, Bid 1987.
and White's initiative can be lasting and dan­ b) 5...lllf6 doesn't promise full equality: 6
gerous. d4 i.e7 7 i.d3 0-0 and now:
Finally, Game 43 srudies a couple of rare b1) 8 ltle2 with a further branch:
3rd move options for Black. Strictly speak­ b1 1) 8 ...1le8 9 0-0 lllc6 10 c3 i.f8 1 1 llJg3
ing, 3...l!lxe4 was the move srudied at tl1e d5 12 Wc2 h6 13 lllh4 �h8 1 4 b4 l!lg8 1 5
very beginning of the Petroff Defence the­ l!lhf5 lllce7 16 l!le3 l!lf6 17 a4 l!leg8 1 8 b5
ory, Danliano considering it as far back as lllg4 19 lllxg4 (1 9 c4 aUows 19 ...:xe3! 20
1 51 2! After 4 We2 'ife7 Black gives up his fxc3 'irh4 21 h3 "irxg3 22 hxg4 i.d6 23 i.a3
knight but immediately wins it back due to -.nz+ 24 �f2 "irg3+ with a draw, Ulibin­
the pin on the e-flle, and he is left a pawn Akopian, Thilisi 1 989) 1 9...i.xg4 20 l!lf5 and
down with the hope of active play for his White has a bit of pressure.
pieces. Black's only other option is 3...'ife7, b12) 8...c5 9 h3 lllc6 10 c3 l:te8 I 1 ()..() b6
but hereWhite has more than one way to 12 'Wc2 i.b7 13 dxc5 bxc5 (probably
obtain a clear plus. 1 3...ltle5!? is stronger: 14 lllxe5 dxe5 1 5 :d1
,...---
.- ----------.. 'flc7 1 6 cxb6 'lt'c6 17 £3 Wxb6+ 1 8 �h1
Game JJ :adS gives Black play for the pawn, while 1 5
Alekhine-A.Rabinovich cxb6 .ic5 16 �h2 'lt'xb6 also seems to give
Moscow 1918 BL1ck decent counterplay) 14 lllg51 g6
(1 4...h6? lead to a fabulous end in the follow­
1 e4 e5 2 lL!f3 li::Jf& 3 truces d& 4 li::Jf3 ing game: 15 i.h7+ Wf8 16 lllxf7! �f7 17
li::Jxe4 5 li::Jc3 'flg6+ �f8 18 i.xh6! gxh6 19 lll f4 '1t'd7 20
..xh6+ Wf7 21 ...g6+ �f8 22 l!lh5 :ebB 23
:act lllxh7 24 •xh7 l!le5 25 f4 i.£6 26
Wh6+ i.g7 27 fxe5+ �g8 28 lll f6+ 1-0
L.Dominguez-Ro.Perez, Cuba 2003) 15 'lt'b3
llle5 16 lllxf7! ..i.d5 17 l!lxd8 ..i.xbJ 18 axb3
lllxd3 1 9 l!lc6 ..i.d8 20 lll6f4 lllxct 21 lllxct
g5 22 l!lfd3 and White retains some advan­
tage (L.Dominguez).
b2) 8 h3 is also good, fur example 8...l!lc6
9 a] lieS 10 0-0 h6 1 1 llet ..i.f8 1 2 llxe8
1Wxe8 1 3 l!lb5 ...d8 14 c4 a6 1 5 lllc3 l!le7
(1 5 ...d5 16 c5 .i.e6 17 b4 also looks better for
White) 1 6 d5 i.f5 1 7 .te2 i.h7 18 b3 'ifd7

81
Th e Petroff De fen ce

1 9 .i.b2 and White's chances were preferable bring the queen home. 17 c3! is stronger.
in Acs-Haba, Bled 2002. 1 7...'ifxd2 1 8 'iih8+ �e7 1 9 1Wxg7+ �d8 20
6 ..e2 �e7 7 lillle4 dxe4 8 •xe4 0-0 9 l:adt 1Wf4 21 g3 ..f8 22 �xe6+ ami White
�c4 wins (Aiekhine); or 1 7....i.e5 1 8 Wh5 �g8 1 9
Black obtains good play after 9 .i.d3 g6 1 0 Wf7+ �hs 20 1Wg6! �gS 2 1 d4 .i.f6 22
0-0 lbc:6 1 1 .tbS .tf5 1 2 •e2 :Cs 13 .txc6 'ifb7+ �f8 23 �xc6+ and White has a very
bxc6 1 4 d4 l:b8, when the bishops and the large plus.
development compensate for the pawn defi­ 1 7 ......f6 1 8 �5
cit. 1 8 'ifh8+?1 is weaker. after 18...�e7 1 9
9 ...�d6 10 0..() l:xe6+ Wd7 20 l:xf6 l:txh8 2 1 l:£7+ �7 22
White is aiming for rapid development. h3 g6 Black's chances in the endgame are
Alekhine gave the following line: 1 0 d4 l:e8 preferable.
1 1 liJcS .i.xeS 12 dxeS l'Dc6 1 3 .i.f4 �4 14 1 8 ...�8 1 9 lle3 �f4?
0-0-0 l:xeS 1S l:d8+! 'ifxd8 1 6 .i.xeS ..e7 17 Now I know we should not talk badly of
l:ct �xeS 1 8 "ffxeS ..xeS 1 9 llxeS �fB, the dead, and Rabinovich was truly one of
when the draw is nt:ar. the great fathers of Russian Chess, but here
1 0...lle8 1 1 •d3 1Llc6 1 2 b3! his defence was not particularly good. White
White needs to get his dark-S(]Uarcd is also very happy after 1 9 �4?! 20 llh3 g6
...

bishop into play. After something like 12 21 Wh7+ WfB 22 11i'd7 :e7 (22...�g8? 23
..c3 .i.g4 1 3 .i.dS .i.xf3 14 .i.xf3 �d41 l:h8+1 would be an unpleasant surprise) 23
Black would have good play for the pawn. �h7+ �g7 24 l'Dxf6 llxd7 25 �xd7 l:d8 26
1 2...•f6 13 �b211 %ld3 l'De2+ 27 �h1 l:lxd7 28 g3 (AJekhine).
However, Black can play 1 9 ...�eSI 20 'ifb7+
WfB 21 f4 �£7 22 'ifhs �g8. probably keep­
ing the position level.
20 Wh7+ Wf8 21 Wh8+ h7

The master of attack strikes again! Instead


of clinging onto his pawn, White launches a
strong counterattack.
1 3 ...•xb2 1 4 �5 .te6
Mter 14 ..g6 White has 1 S .i.x£7+ ..tg7 1 6
. 22 llxe6+1
.i.xe8 11'eS 1 7 ..c3 � 4 1 8 f4 and a very This is probably what Rabinovich missed.
nice position (Aickhine); 1 4...�8 1 S 1Wxh7+ 22......xe6
�fB 1 6 llaet .i.e6 17 c3 also looks very Or 22.. .'�d7 23 l:txf6 :xh8 24 llxf4 with
promising for White. a winning endgame.
1 5 �xe& fxe6 1 6 •xh7+ � 1 7 llae1 ?1 23 •xg7+ �d6 24 �xe6 llxe6 25 d4
This natural looking move allows Black to llee8 26 c4 ll8e7 27 •f8 lle4 28 •f51

82
3 liJxe5: Fourth snd Fifth Move A lterna tives

llxd4 in Nimzowitsch-Marshall, San Sebastian


28..lt7e6 29 1Wc5+ �d7 30 dS also wins 1 9 1 1 , when 1 8 'trxg4+ f5 19 'trc4 looks like a
for White. clear extra pawn to us) 16 'irxa6 bxa6 1 7
29 c5 mate ( 1 -01 :lxd 1 and White has fine compensation for
r-------. the exchange.
Game 34 b) 7 .i.e3 lbc6 8 'ird2 looks more sensible:
Khalifman-Atalik 8 ... 0-0 9 0-0-0 ltles 10 lbd4 cS (to ...a6 1 1 f4
Halkidiki 2002 lbg4 1 2 .i.d3 cS 1 3 �fS lle8 14 1lhe1 .i.£8
._______________.. 1 5 h3 ltlxe3 16 ltlxe3 gives a position where
1 e4 e5 2 ctlf3 l0f6 3 �xeS d6 4 ctlf3 the control over the light squares favours
ctlxe4 5 �c3 lLlxc3 White) 1 1 lLlbS 'tWaS (or l l ....i.e6!? 1 2 lLlxd6
The best move. 'irb6 13 b4 .i.xd6 1 4 'irxd6 1fxd6 1 5 llxd6
6 dxc3 cxb4 16 cxb4 .i.xa2 17 .i.f4 f6 when Black
Also possible is 6 bxc3, when the game has equalised, Nunn-Z'Iu.Polgar, Brussels
could continue 6 ... .i.e7 7 d4 0-0 8 ..id3 .i.g4 1 985) 1 2 a3 a6 1 3 ltlxd6 lld8 1 4 ltlxc8!
9 0-0 lbc:l7 to lZ.b1 �b6 1 1 c4 .i.f6 (or llaxc8 (but not 14 ...:Xd2? 15 ltlxe7+ �f8 1 6
1 1 ...c5 1 2 h3 .i.hS 1 3 dxcS dxcS 14 a4 1Wc7 llxd2 �xe7 17 lldS ltld7 1 8 .i.e2! when
with an unclear position) 1 2 .i.e3 llb8 13 c3 Black has too many problems: 1 8...b6 19 .i.f3
'ii'd7 1 4 .l:lbS a6 1 5 llaS dS 16 cxdS lLlxdS 1 7 ltlf8 - 19 ...1la7 20 lle1 �fB 21 .i.f4 - 20
.i.d2 c6 1 8 h3 .i.e6 and Black is okay, Up­ lieS+ 1-0 Van Der Wiei-Van Der Sterren,
ton-Dutreeuw, Batumi 1999. Wijk aan Zee 1 984) 1 5 'ire2 bS 1 6 llxd8+
6 -*.e7
.•• l:lxd8 1 7 .i.d2 ltlg6 1 8 �b1 b4 1 9 cxb4 cxb4
20 axb4 .i.xb4 21 .i.ct .i.e7 and Black has
fine com[>t:nsation for the pawn, Volokitin­
Kozakov, Lvov 2001 .

1 -*.f4
White has tried two other means of devel­
opment
a) 7 .i.d3 � 8 .i.e3 .i.g4 9 ..ie4 doesn't 7 . . .li1d7
I<K)k particularly appealing, even though this 7. ..li1c6 also looks sensible, for example 8
worked well in the following example: 'ifd2 .i.g4 9 .i.e2 'ird7 1 0 h3 .i.fS 1 1 0-0-0
9 ...'ird7 t O 'ird2 0-0-0 1 t 0-0-0 llhe8 1 2 0-0-0, avoiding the danger of opposite side
lLld4 d S (1 2. ...Ldt ? 1 3 ltlxc6 bxc6? 1 4 'ird3 casding. Now after 1 2 .l:lhe 1 lZ.he8 1 3 g4 .i.g6

and White wins) 13 lbxc6 Wxc6 1 4 .i.xdS 14 .i.bS a6 I S ..ta4 bS 16 .i.b3 ltlaS t 7 c4
'iVa6 1 5 'ii'd3 .i.xd1 (this is necessary; �xc4 1 8 ..txc4 bxc4 19 WdS Wbs 20 WaS+
I S...'iraS 16 f3 c6 1 7 'irc4 .l:lxdS was played 'ifb8 21 'ifxa6+ Wb7 22 Wxb7+ �b7 Black

83
The Petroff Defence

has equal chances in the ending. In this line 1 9....txg6? is bad on account of 20 lDeS
17...lDxb3+ is weaker: 1 8 axb3 1i'c6 19 lDd4 .tf5 21 .tbS :t£8 22 lDd7, when White has a
1i'b7 20 cxbS axbS 21 1i'a5 with an attack for clear advantage. Now Black loses after
White, Boricsev-Raetsky, Fribourg 2000. 22...lle8 23 ..xd5, so he is forced to play
8 •d2 22 ... d4, which is hardly ideal.
20 ll:le5 .tf& 21 ll:lg4

8 �5
•••

Black needs to get in ....tg4 to have a 21 ...d4?1


chance of equalising. After 8...0-0 9 0-0-0 A natural push, but it's tactical suspect.
lDcs 1 0 h3 lieS 1 1 g4 White develops some Black can improve with 21 .....a5!?, when
pressure: 1 t ...lDe4 12 ..e1 .i.f6 1 3 .i.e3 c6 after 22 � 1 the push is stronger: 22...d4 23
14 .i.d3 'it'aS 15 .lxe4! llxe4 16 .i.b61 .i.g5+ .i.c4 bS (the difference; 23...dxe3? 24 lDxf6+
(Black also suffers after 16 ...1i'a4 17 lld41 gxf6 25 ..xe3 still docs not work, though) 24
..xd4 1 8 .lxd4 llxet+ 19 :txe1 .i.e6 20 .ixe6+ llxe6 25 cxd4 'ifxd2 26 :txd2 cxd4
.lxf6 brxf6 21 lDd4) 1 7 lDxg5 .xgS+ 18 .i.e3 27 .txd4 :dB 28 lDxf6+ gxf6 with a likely
'if'g6 19 ..d2 dS 20 f3 :CS 21 h4 gave Black draw.
serious problems in the form of an advanc­ 22 .tc4
ing attack in Tseshkovsky-I.Zaitsev, USSR 22 lDxf6+ is also good for White:
1975. 22...1i'xf6 23 cxd4 cxd4 (23...lted8? 24 .i.c4
9 0-0-0 .tg4 10 .te2 0-0 1 1 h3 .th5 bS 25 .tgS! and White wins) 24 .txd4 lDxd4
1 1 ....i.e6 12 lDd4 ..d7 13 b3 looks better 25 ..xd4 .xd4 26 :txd4 llad8 27 llg..H
for White. :txd4 28 llxd4 .tf5 with a slight pull in the
1 2 g4 .tg6 1 3 h4 J:leB endgame (Atalik).
13 ....i.e4!? 14 hS lDe6 1 5 .ie3 c5, immedi­ 22 ...dxc37
ately establishing counterplay in the centre, A very risky move, even if there were no
also looks good. forced win. 22...dxe3? 23 1i'xe3 .tgs 24 f4
14 h5 .te4 1 5 l:hg1 ll:le& 1 6 .te3 c5 1 7 would also not work for Black, but after
g5 d5 18 g&! 22 ... b5 23 .i.xc6+ :Xe6 24 cxd4 •ds 25
White has no reason to wait. lDxf6+ brxf6 26 1i'a5 his position is still de­
1 8...hxg6 fendable.
Forced. After 1 R if6 19 gxh7+ �xh7 20
•• 23 1te21 .b&
li)gS+ lDxgS 21 .lxgS d4 22 f3 .tf5 23 .i.d3 Or 23 ...cxb2+ 24 Wb1 1i'c7 (24...1i'c7 25
White has a raging attack. .i.xcS! wins) 25 lDxf6+ gxf6 26 1Wg4 'ireS 27
1 9 hxg& fxg& :tdS! and White wins (Atalik).

84
3 liJxe5: Fourth and Fifth Mo ve Alterna tives

24 lt!xf6+ gxf6 25 bxc3 •c7 26 'ifg4! gave White the chance to irritate Black in
11Ve5 Hodgson-Barua, l .ondon 1 986.
7 .i.g5

27 lld5!1 1bc3
White wins in all lines: 27 ... i.f5 28 'ifxf5 7 �e6
..

or 27....ixd5 28 1i"xg6+ Wffi 29 i.xc5+! For some reason 131ack insists on keeping
when Black will get mated. the queens on, which is not necessary be­
28 'ifxe4 1fa1 + 29 �2 •xg1 30 lld7l cause the endgame promises Black guod
1 -0 chances. 7...1i"xe2+ is studied in the next
r----- game, while 7.../i)bd7 8 /i)c3 1rxe2+ (8...h6?!
Gatne 35 9 .ic3 /i)b6 1 0 0.0-0 .id7 1 1 1i"d2 U-0-0 1 2
Em.Lasker-Marshall llc1 is a tad uncomfortable for Black) 9
St Petersbutg 1912 .i.xe2 transposes to 7...9xe2+ 8 .ixe2 /i)bd7
.________________. 9 /i)c3 (see the notes to the next game).
1 e4 e5 2 lt!f3 etlf6 3 lt!xe5 d6 4 lt!f3 8 ltlc3 ltlbd7
lbxe4 5 'ife2 8 .../i)c6 is answered by the natural 9 /i)e4
This line is usually played with the hope of 0-0-0 10 /i)xf6 gxf6 1 I .ie3 d5 1 2 d4 .ibr4 13
a draw. However, many White players have 0-0-0 when White's position is preferable.
found that achieving the draw from an even 8. .. h6 looks dubious after 9 .ixf6 9xf6 1 0
ending is not always so easy. d4 i.e7 1 1 'Wb5+ /i)d7 1 2 .i.d3 (1 2 'iVxb7
5 .. -•e7 6 d3 0-0 1 3 'ifxc7 i!Db6 appears very risky for
The score from this position in our data­ White) 1 2...g5 1 3 h3 (or 13 ll:le4 'ti"g7 14 h3
base is actually 51°/co for White and 49% for with a slight edge - Keres) 13 ...0-0 1 4 'iVxb7
Black, which suggests complete equality. llab8 1 5 1re4 'iVg7 16 b3 (we think White is
However, it also shows that people play on better here) 1 6 ...cS? (1 6.../i)cS 17 9c3 .if6
from here... 1 8 0-0 llfe8 1 9 1rd2 g4 - Tanasch - with a
6 ...etlf6 messy position is necessary) 17 0-0 cxd4 18
The most common move, but not the /i)JS i.d8 19 .ic4 .!DeS 20 ..xd4 ..xd4 21
1 1nly one. For example, 6...lL!c5 7 lL!c3 i.g4 8 /i)xd4 LdS 22 .ixdS .if6 23 J:ladt and
.i.e3 (8 ll:lds Lf3!? 9 'ifxe7+ .ixc7 10 White enjoyed a large plus in Capablanca­
ltlxc7+ Wd7 1 1 .!Dxa8 .ic6 is not clear, but Marshall, St Petersburg 1 914.
should be fine for Black) 8...c6 9 h3 .i.h5 lO 9 0-0-0
hr4 .ig6 1 1 .tg2 lL!bd7 12 lL!d4 ltle6 13 f4 9 d4 leads to equality after 9...h6 1 0 i.h4
li:lxd4 1 4 .i.xd4 f5 1 S 0-0-0 'ii'xe2 16 ltlxe2 g5 1 l i.g3 �IS 12 /i)xdS .ixdS 13 0-0-0

85
The Petroff Defence

'ifxe2 14 .i.xe2 .i.g7. More interesting is 9 sors.20...�c4 is the only move, although after
�bS!? .!llb6 1 0 .Lf6 gxf6 1 1 g3 dS 1 2 0-0-0 21 l:tb3! 'ifgS+ 22 �b 1 (also strong is 22 f4!?
.i.h6+ 13 �b 1 with an unbalanced position, gxf3+ 2..� �b1 �2+ 24 'lt>at �xb3+ 25 cxb3
something that no one would have thought 'ifxg3 26 .!l:\x£3 .id6 27 llct .i.f4 28 �eSI
possible just a few moves ago. .ixeS 29 Wa7+! �cS 30 dxeS �d7 31 �xc7
9 . . .h6 10 .th4 g5 with a clear plus for White) 22...�2+ 23
1 0... 0-0-0 seems to be more accurate. Af­ �a1 �xb3+ 24 cxb3 .i.d6 25 Wa7+ �cS 26
ter 1 1 d4 gS 1 2 .ig3 �b6 1 3 1i'b5 a6 1 4 'it'aS �xd6+ l:txd6 (26...cxd6 27 f4! brxf3 28 �xf3
i.g7 the position is unclear. 1fe3 29 l:te1 and Black can no longer prevent
1 1 .tg3 .!Dh5 the deadly l:tcl+) 27 'ifa8+ �d7 28 'ifxh8
This looks like a loss of time, but Black White has every chance of winning even if
does not have it easy. 1 l ..ig7 12 �4 0-0 Black can put up some resistance.
1 3 h4 g4 1 4 �£5 ..d8 1 5 �xg7 �xg7 1 6 hS 21 c;h,1 .td6
is not a serious alternative. Black also loses after 21...'ife7 22 l:lb3
1 2 d4 �xg3 13 hxg3 g4?! l:d6 23 a4 :c6 24 aS, when White's attack is
This move deprives the bishop of the £5- conclusive.
square, which proves to be a problem in the 22 1lb3
later tactics. 1 3... 0-0-0 1 4 We3!? with a slight Or 22 l:tc3!? .!llc4 23 Wa7+ �cS 24
edge for White was better. �xd6+ cxd6 25 �a1 ! .:r.de8 26 llb3 and
1 4 �h4 d5 1 5 1rb5! 0-0-0 Black is busted.
1 5...Wb4? loses to the pretty 1 6 �xdS!. 22. . .1lhe8 23 a4! .tf5
1 6 'fra5
White should avoid 1 6 .!llxdS? .ixdS 17
..xdS 'it'gS+ 1 8 ..xgS hxgS, winning a piece.
16 ••• a6 17 .txa6! bxa6
Or 17...Wb4 18 Wxb4 .ixb4 1 9 .i.d3
.ixc3 20 bxc3 when White is simply just a
pawn up (Kasparov).
1 8 •••6+ Wb8 1 9 �b5 o!Db6 20 lld3

24 flJ.7
Or 24 aS .i.xc2+ 25 �xc2 l:le2+ 26 q;,b1
and it is aU over.
24....td7 25 a5 1i'd2 26 axb6 lle1 +
27 �a2 c6 28 �b5 cxb5 29 1ta7+ 1 -0

GaHJe 36
Aronian-Akopian
20 1i'g5+
••.
Ohrid 2001
After this we cannot find a defence for
Black. The last chance for was given by Kas­ 1 e4 e5 2 �f3 �f6 3 lbxe5 d6 4 lilf3
parov in his series of books MY GTMI Prr:deces- lbxe4 5 1i'e2 1i'e7 6 d3 lilf6 7 .tg5

86
3 liJxe5: Fourth and Fifth Mo ve A lterna tives

'lrxe2+ tt:'lxe4 1 3 dxe4 .i.xgS+ 14 tlligS We7 1 S f4


This is definitely the solid choice. �e6 1 6 lt)tt3 lbcs 17 ll'lf2 (1 7 f5 J:lbB 1 B
8 .txe2 J:ld4 b6 1 9 :Cd1 J:ld8 20 c4 aS also looks
leveQ 17 ...J:ldB 1 8 .i.f3 f6 19 h4 .i.e6 20 g4 aS
21 a3 h6 22 hS J:ld7 23 lt)h1 J:leB 24 lt)g3
�dB 1/z-'/z Spassky-Hort, Reykjavik 1 977.
b) 10 0-0 lt)a6 1 1 J:lfe1 lbc7 12 d4 dS 13
.i.d3 �6 14 .i.c3 0-0 1 5 J:lad1 .i.d6 t 6 .!i)eS
.!i)e8!? (White is better after the risky 1 6...cS
t 7 .!i)bs .i.bB 1 8 c3 a6 19 lt)a3 - Yusupov)
1 7 tt:'le2 f6 1 8 �f3 tt:'!Bc7 1 9 b3 .i.d7 with fuU
equality, Spassky-Yusupov, Toluka 1 982

8. . ..te7
8...lbbd7 is equaUy good after 9 lbc3 h6
and now:
a) to .i.h4 gS 1 1 .i.g3 lbhS 12 lbd4 (12
�iS �d8 13 d4 lbb6 is totaUy level too)
1 2...lbxg3 1 .3 hxg3 �f6 1 4 0-0-0 .i.d7 15
.l:ldel 0-0-0 16 lbd1 cS 17 lbf3 .i.g7 1 8 �e3
.i.c6 with equal chances, ApiceUa-Nikcevic,
France 2000.
b) to .i.d2 br6 (1 0...lbb6 1 1 lObS!? i!LlbdS - 10 ..th4
1 1 ...�d8!? - 12 c4 .i.d7 13 a4 c6 14 i!Llbd4 10 .i.f4 does not look very dangerous ei­
lbc7 1 S 0-0 .i.e7 1 6 b4 1ead to a white advan­ ther: what exactly is the bishop meant to be
tnge in Spassky-Yusupov, Moscow 19B1) 1 1 doing here? Anyway, one game continued
0-0-0 (1 1 lObs �dB 1 2 c4 .i.g7 1 3 0-0 :CB to...lbc6 1 1 0-0-0 .i.e6 1 2 d4 .!i)b4 1 3 a3
14 J:lfcl a6 I S lbc3 lbcS 16 b4 .!Lle6 17 d4 aS .!i)bds 14 tt:'lxdS .!i)xdS 1 s .i.d2 0-0-0 1 6
18 bS gS is another example of Black obtain­ J:lhe1 J:lhe8 1 7 h3 .!i)b6 1 8 b3 .i.dS 1 9 i:e3
ing equal play) 1 1 ....i.g7 12 d4 .!Llb6 1 3 h3 0-0 .i.e4 20 c4 li)d7 21 g3 cS 22 b4 .i.£6 23 lbg1
1 4 .i.d3 .i.e6 tS J:lhcl J:lfcB 1 6 b3 a6 1 7 .!Lle4 cxb4 24 axb4 .!i)b6 25 �b2 .i.fS with
lt)xc4 1 B .i.xe4 .i.dS 1 9 .i.xdS lt)xdS 20 c4 chances for both sides, McShane­
lt)f6 with complete equality, Westerinen­ Mamedyarov, Lausanne 2004.
Ractsky, Hafnarfjonlur 1 999. 1 0...liJbd7
9 liJc3 10 ...lbc6 is also fine, for example 1 1 0-0-0
Or 9 c4 h6 10 .i.f4 lbc6 1 1 lbc3 .i.fS 1 2 .i.d7 12 h.1 0-0-0 1 3 d4 J:lde8 14 .i.c4 J:lhfB
0-0-0 0-0-0 1 3 J:lhc1 gS 14 .i.c3 �g4 1 S lbdS l S .i.g3 .i.d8 16 dS .!Lle7 17 .i.h2 a6 1B a3
�xe3 16 lt)xe7+ �xe7 1 7 fxe3 J:lde8 with a .!i)g6 19 J:lhc1 J:lxe1 20 J:lxcl J:lc8 21 J:lxe8
level position, Spassky-Karpov, Hamburg .!i)xe8 22 Wd2 lt)h4 23 lllih4 1/z-1/z Shott­
1982. Anand, Wijk aan Zee 2000.
9 ...h6 1 1 liJd4 lOb& 1 2 a4 a6 13 a5 liJbd5 1 4
Another option is 9...c6 and now: o!Oxd5 liJxd5 1 5 .txe7 liJxe7 1 6 .tf3 c5
a) 1 0 0-0-0 lba6 1 1 J:lhc1 lbc7 1 2 llk4 1 7 .!082 liJc6 1 8 .txc&+ bxc6

87
The Petroff Defen c e

1 3 .i.d3 .i.c6 14 'W'c2 g6 1 5 a4 is one exam­


ple) 7 'lfc2!? ltlxc3 8 dxc3 dxc4 9 lbd4 .i.dS
1 0 .i.e3 lbc6 1 1 0-11-0 and White has good
compensation for the pawn in the shape of
rapid development.
However, s.. lbc6 seL1DS to be a strong al­
.

ternative. Now we have:

Black has equalised.


1 9 0-0-0 Ag4 20 f3 .i.e& 21 d4 cxd4
21 ...c4 22 ltlf4 �d7 23 dS cxdS 24 ltlxdS
would give White a real reason to play for a
win, even though the disturbance of the bal­
ance is very minor indeed
22 llxd4 �7 23 llhd1 llhd8 24 l£!f4
llab8 25 b4 .llb5 26 tnd3 lldb8 27 c4 a) 6 lbc3 will most likely transpose to the
llxb417 28 .!i)xb4 llxb4 29 h4 main game (see the note to Black's 7th
29 llxd6 l:txc4+ 30 �b1 l:tb4+ also leads move).
to a draw. b) 6 d4 dS 7 &3 .i.b4 8 1ib3?1 (8 .i.d2
29...g5 30 hxg5 hxg5 31 llxd6 llxc4+ 0-0 with equality is preferable) 8...'1fe7! (an
32 Wb2 l:tc5 33 l:t6d3 .llb 5+ 34 ¢'c3 improvement over 8....i.e6 9 cS!? ltlxcS 10
llxa5 35 Wd4 �6 36 �3+ We5 37 dxcS d4 1 1 .i.c4 dxc3 12 0-0 .i.xc4 1 3 1i'xc4
llh1 c5 38 llh5 WI& 39 l:th&+ Wg7 40 cxb2 14 i.xb2 0-0 1 5 a3 i.aS 1 6 .1Ld1 11t'e7
llh1 c4 41 l:tc3 l:ta2 42 llhc1 llxg2 43 1 7 lidS l:tad8 1 8 .J:tgs g6 1 9 h4 when White
l:ta1 g4 44 fxg4 llg3+ 45 Wd2 l:txg4 46 had good compensation for the pawn in
llxa6 WI& 47 lla5 llg6 %-% Velicka-Raetsky, CappeUe Ia Grande 1 996) 9
r----- .i.e3 lbas 10 11t'c2 dxc4 1 1 i.e2 i.fS and
Game 37 Black has emerged from the opening with an
Areshchenko-Mista advantage.
Cappelle Ia Grande 2003 c) 6 i.e2 i.e7 7 0-0 0-0 8 d4 i.f6 9 dS
'--------------• ltle7 10 ltla3 l:le8 1 1 ltlc2 with a further
1 e4 e5 2 .!i)f3 .!i)f& 3 .!i)xe5 d6 4 lllf3 branch:
lDxe4 5 c4 .i.e7 ct) 1 t ...h6 1 2 l:le l a5 13 l:lb1 i.f5 1 4 i.c3
Black has a couple of serious alternatives .i.h7 (an improvement over 14 ...1Wd7 1 5
to this very natural move. ltlfd4 i.h7 1 6 .i.g4 'lidS 1 7 11t'e2 c6 1 8 dxc6
S ...dS seems a bit risk')':. 6 lbc3 .i.e6 bxc6 1 9 o lbcs 20 J:lbdl 'il'b6 21 b3, which
(6 ...ltlf6?! leads to an French Defence, Ex­ is slighrJy preferable for White, Kholmov­
change Variation with the loss of a tempo! Raetsky, Vorone7.h 1 988) 1 5 ltlfd4 ltlfS 1 6
White should be able to organise a real ad­ lbxfS .i.xfS 1 7 i.d3 1Wd7 and Black i s equal.
vantage here: 7 d4 i.e7 8 cxd5 ltlxd5 9 i.c4 c2) Also strong is 1 t ...ltlf5 1 2 ltb1 h6 13
ltlxc3 1 0 bxc3 0-0 1 1 0-0 lbd7 1 2 :et ltlb6 lle1 i.d7 1 4 i.d3 ltlcs 1 5 ..tfl (1 5 :XeH+

88
3 lDxe5: Fo urth Bnd Fifth Mo ve A lternB tives

'1Vxe8 t6 lbe3 lbxd3 17 '1Vxd3 with an equal 7 0-0


.•.

position is better) 1 S...J:lxe1 1 6 'irxe1 •e8 17 We consider ...lDd7-cS to be the most reli­
'irdt ?l ..c4 18 b3 aS 19 .i.b2 .i.xb2 20 llxb2 able way to develop, but there arc other op­
l:le8 21 llket and here, with Black slightly tions: 7 ...llk6 8 .td3 lbeS 9 lbxcS dxeS 1 0
better, the players agreed a draw, Brodsky­ 'ifc2 .i.gS 1 1 0-0 .te6 12 J:let .i.xct 1 3
Ractsky, Cappclle Ia Grande 2000. l:axct •gs 14 lle3 0-0-0 1 5 J:lce1 f6 1 6 b4
6 lDc3 lld7 and Black has solved his problems.
6 d4 is also interesting, for example: Maroczy-Marshall, San Sebastian 1 91 1 .
a) 6...0-0 7 .i.d3 lbgS!? leads to wild play 8 -*.d3 l0d7 9 1ic2 g6
after 8 lbxg5 hgS 9 1ths l:e8+ t o �d1 9...h6 seems to be less harmonious. After
Let 1 1 'irxh7+ �f8 1 2 1th8+ �e7 13 10 .i.e3 .i.f6 11 0-0-0 (the unambitious 1 1
1!rh4+ �f8 1 4 1Wxd8 J:lxd8 l S Wxc1 dS and 0-0 J:le8 1 2 l:ad 1 lbeS 13 ltlxe5 .ixeS 14
we end up in an endgame where Black has llfe 1 9h4 1S g3 'iVhs gives equal chances)
enough counterplay for the pawn, Slobodjan­ 1 t ...lbc:S t 2 lbxeS .i.xeS 13 f4 .i.f6 1 4 Wf2
Van Der Sterren, Bundesliga 1 999. lle8 1 S h3 .i.d7 16 :he1 a6 1 7 J:le2 b5 1 8 c5
b) 6...dS 7 cxdS [1 llk3 lbxc3 8 bxc3 0-0 9 White was a bit better in Velimirovic­
cxdS 1WxdS transposes to 7 cxdS) 7...1Wxd5 8 Motylcv, Novi Sad 2000.
llk3 lbxc3 9 bxc3 0-0 10 .i.d3 cS 1 1 0-0 10 -*.e3
cxd4 1 2 lle1 .tf6 1 3 J:lb1 lbc6 1 4 .i.e4 ..xa2 1 0 .i.h6 J:le8 1 1 0-0 looks less aggressive:
1 5 9d3 h6 1 6 cxd4 .tg41? (or 1 6...J:ld8 17 1 1 ...lbcs 1 2 l:fc1 .tg4 1 3 lDd4 .tf8 14 .i.e3
.i.a3 .i.e6 18 llb2 ..c4 with an unclear posi­ (14 hf8 :xet+ 1 S l:xe1 'irxf8 1 6 lbbS
tion) 1 7 .i.a3 iLxO 1 8 gxf3 tLlxd4 1 9 .txf8 'ird8 is simply levd) 14 ...tLlxd3 1 5 'irxd3
l:lxf8 20 llxb7 1faS 21 1ffl .i.eS 22 IIebl f5 .tg7 16 h3 .i.d7 11 9d2 'irh4 1 8 .tgs 9hs
23 .c4+ �h7 24 .i.dS 'ifd2 and Black had 19 lbe2 h6 20 .te3 .ic6 21 lbf4 1ff5 22
good play for the exchange in Gajewsky­ .td4 a6 23 .ixg7 Wxg7 24 ltlds .ixdS 2S
Mista, Poland 2003. cxdS and a draw was agreed in G.Kuzmin­
6 lDxc3
.•• Munl7.ia, Cappelle Ia Grande 2003.
6 ...lbg5 is worse: 7 .i.e2 0-0 8 0-0 lbc6 9 10 ...l0c5 1 1 0-0-0
d4 J:le8 to lbds lbxf3+ 1 1 .i.xf3 .i.f6 12 1 1 .i.xcS dxcS 12 0-0-0 is interesting: fol­
.i.e3 .t£5 1 3 b4 a6 1 4 a4 h6 1 S .d2 and lowing 12...i.d6 13 h4 .tg4 1 4 bS!? .ixhS 15
White was better in Nisipeanu-Motylev, Bu­ l:xhS!? gxhS 16 .i.xh7+ �g7 the position is
charest 2001. highly unclear.
7 dxc3 1 1 ...lDxd3+ 1 2 •xd3 -*.f5 1 3 •d2 c5
1 3....i.e6!? 14 b3 aS with counterplay is
also Logical.
14 -*.g6 -*.xg5 1 5 lDxg5 1fa5 1 6 b3 b51?
Energetically played. White has an edge af­
ter l6 ...'1fa3+ t 7 1ib2 .xb2+ 18 �b2 llae8
19 l:d2.
1 7 cxb5 c4
The most precise. After l7 ...a6 1 8 b6
llab8 19 Wb2 :txb6 20 1Wf4 White enjoys a
slight plus because Black has no c..-asy way to
weaken White's Icing's position.
1 8 �2 lZab8 1 9 b4
19 1Wxd6?! is very risky. After cxb3 20

89
The Petro ff Defence

axb3 WxbS 21 b4 aS Black's attack has every 26 Wxd4 bxa2+ 27 �xa2 'lla6+ is equal.
chance of being successful.
1 9 .Axb5
.•
Game 38
1 9 ...'i'xb5 is another attractive option. Naiditsch-Timman
Following 20 WdS 1rxd5 (20...Wa4!?) 21 Dortmund 2002
.ILdS l:lfe8 22 lld2 the endgame is probably
equal, though not necessarily drawn. 1 e4 e5 2 �f3 llJf& 3 ofu:e5 d6 4 �f3
�xe4 5 .i.d3

20 �f3
After 20 ..xd6 ..a4 21 c;tlat 9a3 22 l:lc1 This somewhat anti-positional move has
l:lb6 23 ..e7 Da6 24 ..e2 �d3 25 'i'b2 ..a4 found some popularity over the last few
Black has good play for the pawn. years. However, it shouldn't seriously
20 .:fbs 21 �4 Lb4+ ! 22 cxb4
•.. threaten Black.
Lb4+ 23 ltlb3 5 ltlf6
...

Black has also tried:


a) S...l&s 6 �e2 and now:
at) 6 .. ie7 7 0-0 0-0 8 d4 �e4 (8 ...ll:k6
seems inferior: 9 l:let l:le8 10 .i.d3 �d7 1 1
h3 �f6 1 2 c4 gives White a slight edge, and
now Black should play 12 ... d5 instead of
1 2... c5?! 1 3 dS �fB 1 4 llk..'"-\ lllg6 1 5 ..c2
�d7 1 6 �d2 a6 17 a4 after which White is
very happy, Safin-S.Ernst, Dicrcn 2002) 9 c4
c6 10 'irc2 dS 1 1 .id3 �c6 with a complex
struggle ahead.
a2) 6... d5 7 d4 �e4 8 IJ-0 ..ltd6 9 c4 c6 1 0
�3 (1 0 l:le1 0-0 1 1 .td3 transposes to 5 d4
The only move. If 23 ci>a 1 ? then dS 6 .id3 .id6 7 0-0 0-0 8 c4 c6 9 ltet)
23...l:lb I +I wins. 1 0...0-0 11 W'b3 �xc3 1 2 bxc3 dxc4 1 3
23 '1rb51?
••. .ixc4 h6 14 l:let �7 1 5 ..c2 �b6 1 6 .id3
Black can keep the game alive with cS was level in Constantini-Raetsky, Bid
23 ..cxb3 24 a3 1i'e5+ 25 1i'c3, when it's still
. 2001 .
very complicated. b) S d S 6 'i'e2 'lle7 7 0-0 with another
.•.

24 'lrxd6 cxb3 25 .:d4 Ld4 %-% branch:

90
3 liJxe5: Fourth a nd Fifth Mo ve A lterna tives

12 i.e3 l:le8 1 3 lL!bd2 i.fB 14 li)e4 lL!d5 1 5


dxc5 li)xe3 1 6 lixe3 dxc5 1 7 ..e2 Wc7 1 8
llc1 l:lad8 with complete equality, Moro­
zevich-Topalov, Dorttnund 2001 . 14 .i.b3
ll\d5 1 5 .i.xd5 .i.xd5 1 6 c4 .i.xf3 1 7 lLlxf3
1i'f6 provides more imbalance, but it's not
clear who if anyone gains from this.
9 �c2 liJe5

b1) 7...ll\d6 8 1i'd1 1? (8 llc1 'ifxe2 9


llxe2+ .i.e7 10 lbc3 .i.e6 1 1 tl'kl4 Wd7 1 2
lLlxe6 fxe6 13 lL!b5 lL!xb5 1 4 .i.xb5+ c6 1 5
.i.d3 .i.f6 1 6 c3 c 5 lead to a drawish ending
in Lcko-Kramnik, linares 1 999) 8...'ifd8 (this
seems best; 8 ...g6 9 l:le1 .i.e6 10 .i.fl .i.g7 1 1
d4 0-0 1 2 i.f4 lbc6 1 3 c3 'ifd7 1 4 lDbd2
�-,rave White a slight plus in Morozevich­
Shirov, Astana 2001) 9 lDcl c6 10 llct+ .i.e7 White's pos1bon looks preferable after
I I 'l'e2 .i.g4 (1 1 ...lLld7 1 2 b41? lL!f8 13 b5 9 ... d5 1 0 d4 lL!e4 1 1 l:let f5 1 2 .i.f4.
llk6, as suggested by Romero, looks a bit 10 d4 lLlxf3+ 1 1 'tlfxf3 c6 1 2 �f4
better for White) 1 2 b3 tlkl7 1 3 i.b2 lbc5 1 4 White gained nothing after 1 2 l:let .i.e6
lL!a4 i.xf3 1 5 'l'xO lDxd3 16 'l'xd3 Yz-Yz 1 3 llkl2 1fd7 14 li)f1 l:lae8 1 5 lDg3 .i.d8 1 6
Morozevich-Kramnik, Wijk aan Zce 2000. i.g5 lL!dS 1 7 Wd3 f5 1 8 i.d2 g6 1 9 .i.h6
Black has no problems at all in the final posi­ l:lfl 20 11ff3, Lal.ic-Nikcevic, Paris 2000. In
rion. fact, the players agreed a draw in this posi­
b2) 7 ...lbc5 is less reliable: 8 llc1 i.e6 9 tion.
i.bS+ c6 1 0 d4 cxb5 1 1 dxc5 lbc6 1 2 �86
•..

(1 1 ...'1'xc5?! 1 2 lLlg5 'l'c4 1 3 lDxe6 'l'xe2 1 4 Here we recommend that Black should
llxe2 fxe6 1 5 lixe6+ � fl 16 l:le5 is awful for deviate with 1 2 ...1fb6!? 1 3 b3 d5 1 4 lLld2
Black) 1 2 .i.e3 a6 1 3 a4 b4 14 c3 a5 1 5 cxb4 i.e6 leading to a position with equal chances.
axb4 1 6 li)bd2 and White enjoys a plus, 1 3 lLld2 'tlfd7 14 J:lfe1 J:lfe8 1 5 �h2
Morozevich-Shirov, Sarajevo 2000. We suggest 1 5 l:le2, for example 1 5.. ..i.f5
6 h3 �e7 7 0-0 0-0 8 c3 liJc6 1 6 i.xf5 Wxf5 17 l:lae1 .tf8 1 8 lDe4 lL!xe4
Alternatively: 1 9 l:lxe4 and White's control of the e-tile
a) 8 ... d5 9 .ic2 lieS to d4 .id6 1 1 i.g5 annoys Black; or 1 5...c5 1 6 dxc5 dxc5 1 7
lLlbd7 1 2 li)1x12 h6 (12... lLlfB t 3 1lc1 l:lxel+ l:laet and White has a slight plus.
1 4 1fxel h6 1 5 .i.h4 lDg6 1 6 .i.g3 i.xg3 17 1 5 g6 1 6 c4 d5 1 7 c5 liJh5
•.•

fxg3 lL!fB 1 8 g4 wa s better for White in Shi­ 1 7 ...lL!e4 leads to trouble after 18 li)xe4
ruv-Anand, Leon 2001 - Black's pieces can­ dxe4 19 :Xe4 .i.d5 20 i.b3!, when the attack
not really get into the game) 1 3 .ih4 li)b6 1 4 against fl leads to a position with an extra
a4 a S and Black has nothing to fear. pawn for White.
b) 8...c5 9 i.c2 lbc6 to d4 i.e6 1 1 l:le1 h6 1 8 'irb3 f6 1 9 liJf1 �f8 20 liJe3 �h6

91
The Petroff Defence

himself. The main line runs 26 l:lxe8+ llxe8


27 '1Vxe8+ Wg7 28 ..e7+ Wh6 29 .i.e4 dxe4
30 '1Vxe4 1Wh4+ 31 Wgt ll:lh3+!, winning.
25. .:ed8 26 .td1 'irf5
.

Also possible is 26 •xd4 27 Lh5 hrxh5


...

28 g3 h4 and Black should be winning.


27 .txh5 gxh5 28 'ire6+ 'irxe6 29 l:lxe6
� 30 :as .tt4?1
A much bcner choice is 30...l:ldb8 31 b3
lt4, when Black has every chance to win.
31 g3 .th6 32 b4 .lld7 33 b5 .llb8 34 a4
.tg7 35 .llb 1 f5 36 .llb4 f4 37 a5 fxg3 38
fxg3
Alternatively 20.. f5 21 ll:lft JLg7 22 'ifc3
. 38 b6 axb6 39 axb6 :C7! 40 c6 lle1+ 41
.i.h6 23 f3 iLf7 with even chances. 'iPg2 gxf2 42 'iPxf2 llbe8 43 :a?+ Wg6 44
21 .!Dg4?! llxg7+ Wxg7 45 b7 llcl 46 bS'IV llxb8 47
lbis spells trouble. White should play 21 llxb8 llxc6 48 llds llf6+ 49 Wg3 l:lf5 gives
.td6 iLf4 22 hf4 ll:lxf4 23 ll:lft b6, when Black winning chances. In this line 43 lla2
nothing is decided yet. .i.. f6 is probably bad for White, although
21 .txg4
••• these complications are really difficult to
21 ....i.g5 is weaker, as after 22 '1Vf3 f5 23 aSSCS.'i.
ll:leS '1Vg7 24 llc2 lle7 25 llael White's po.'ii­ 38 •• -Ac7 39 l:ld6
tion looks a good deal preferable. 39 b6 also leads to a draw after 39 ... axb6
22 hxg4 'irxg4 40 axb6 1lc6 41 l:la7+ Wg6 42 l:td7 .i.xd4 43
22...ll:lf4 only leads to equality after 23 l:lxd4 llxc5.
'1Vf3 ll:le6 24 .c3 ll:lf4. 39 .txd4 40 .llxd4 .llxb5 41 ll4xd5 % -
••.

23 1fxb7 .tf4! 24 'irxc6 %


24 g3 would be strongly met by 24...ll:lxg3!
25 fxg3 .i.xg3 26 .i.xg3 '1Vxg3+ 27 Wft Wf4+ Game 39
with an advantage for Black. Dolmatov-Raetsky
24....txh2+ Podo/sk Voronezh 1992
1 e4 e5 2 .!Df3 .!Df6 3 .!Dxe5 d6 4 .!Df3
.!Dxe4 5 d3
lbis can lead to symmetrical play and is
rather harmless.
5 ....!Df6 6 d4 .te7
Or:
a) 6...d5 transposes to the Exchange
French, which is outside the scope of this
book
b) 6...g6 7 .i..d.3 1!i'e7+ (the most precise;
7 .. ..i.g7 8 0-0 0-0 9 h3 ll:lc6 to a3 ll:le7 1 1 c4
.i.f5 12 ll:lc3 :Cs 1 3 llct .i.xd3 14 'ifxd3
25 �1 '1Vd7 15 i.g5 gave White a small plus in Ni-
After 25 �xh2 ll:lf4! White cannot defend kolenko-Donchcnko, Moscow 2000) 8 '1Ve2

92
3 ltlxe5: Fourth and Fifth Mo ve Alternatives

1i'xc2+ 9 Wxe2 i.g7 10 ltk3 ltlc6 1 1 ltlb5 i.e2 f5 1 6 f4 ltld7 1 7 i.d3 ltlc7 1 8 1i'f3 i.f6
Wd8 and the endgame should of course end 19 ltld1 b5 20 c3 g6 21 i.d2 ltlb6 22 lbe3
in a draw. 'ii'd7 both sides have chances, Shtyrenkov­
7 ..i.d3 ..i.g4 Raersky, Voronezh 1989. lnstead of 1 2 ltk3,
7 ...0-0 allows 8 h3 when White may be 1 2 g4 i.g6 1 3 i.xg6 hxg6 14 ltlc3 ltlh7!? 1 5
able to gain a slight pull: 8...c51? 9 0-0 ltk6 1 0 a4 ltlg5 is interesting - maybe White i s a bit
%let h6 t t ltlc3 i.c6 1 2 i.b5 lbd5 t3 ltlxd5 better here
i.xd5 1 4 dxc5 (this is better than 1 4 c3 a6 1 5 1 1 ...dxc5 1 2 g4 ..i.g6 13 ..ixg6 hxg6 1 4
i.xc6 i.xc6 1 6 dxc5 dxc5 17 1i'xd8 i.xd8 1 8 •xd8 ..ixd8 1 5 ..if4 ..ia5!
i.c3, Kveinys-Schandorff, Germany 2001 ; Accurate play. After 15 ...ltlbd7? 1 6 gS
here Black can play 1 R...i.xf3 1 9 gxf3 b6 ltlh5 17 i.d6 i.aS 18 l:le7! White has gained
with a slightly more comfortable position, a serious advantage.
even though it is probably not enough for an 1 6 c3
advantage) 14...dxc5 1 5 i.f4 i.f6 1 6 ltlc5 Or 16 lld1 ltla6 1 7 ltla3 i.e? 1 8 i.xc7
and as so often in the Petroff the position is ltlxc7 1 9 lld2 l:lfd8 20 :lad1 llxd2 21 l:lxd2
more or less level but Black is a bit passive. (Panchenko-Tolstikh, Cheliabinsk 1993).
8 0-0 Now Black is no worse after 21 .-ltle6 22
8 ltlbd2 isn't intimidating: 8...0-0 9 ltln lObS b6.
cS!? 1 0 ltle3 i.h5 1 1 0-0 �c6 12 ltlf5 l:le8 1 3 1 6...Ad8
Jxc5 (1 3 c3 cxd4 1 4 cxd4 i. ffi 1 5 i.g5 h6 1 6
i.c3 ltld5 17 ltlg3 i.g4 1 8 h3 i.e6 equalised
fully in Strzelecki-Raetsky, Katowice 1 990)
1 3...dxc5 14 l:le1 i.ffi 1 5 l:lxe8 1i'xe8 1 6 i.g5
ttlc5 and Black has enough countcrplay.

1 6...ltlbd7?! is imprecise. After 1 7 ltla3


ltlds 1 8 i.d6 l:lfe8 1 9 ltk4 i.b6 20 l:lad 1 the
problems arc mounting up for Black.
1 7 lllbd2 lllc6 1 8 ltle4
White also has nothing after 1 8 lbc4 lLlds
8 ...0-0 9 Ae1 1 9 ltlxaS (1 9 i.d6 i.xc3! 20 bxc3 b5 21
9 i.f4 should be met by 9...ltld5 (9...ltlc6 i.xc5 bxc4 and Black is no worse) 1 9...ltlxa5
I l l c3 l:le8 1 1 ltlbd2 and White is a bit better 20 i.g3 f6 21 l:lad 1 �f7.
- Keres) 1 0 i.d2 i.f6 1 1 h3 i.hs 1 2 c3 c6 1 8...lbxe4 1 9 Axe4 AdS
wilh level chances. Black has neutralised whatever pull White
9...c517 1 0 h3 ..i.h5 1 1 dxc5 ever had.
This is not particularly ambitious. After 1 1 20 Aae1 f& 21 h4 Aad8 22 �g2 �7
d5 ltlbd7 (li ...ltlxdS?? loses to 1 2 g4) 12 23 A1e2 A8d7 24 .tea .tb& 25 a3 Ad3
ltlc3 tOes 1 3 i.e2 i.xf3 1 4 i.xf3 ltle81? 1 5 26 ltld2 %-%

93
Th e Petroff Defence

r------ bxc3 l:r.e5 with equal chances, Khainallin-


Game 40 Bezgodov, Cheliabinsk 2004. In rhis line
Smagin-Makarychev 1 1 .. ic5 is not so good after 1 2 .J.g5 .J.g4 13
Moscow 1987 'ifd2 (1 3 .i.x£6?! .i.xf2+! 1 4 �fl .i.xd1 -
14...1i'xf6 1 5 'ifxg4 .i.xel+ 16 Wxe1 1ib6
1 e4 e5 2 li:)f3 ll:)f& 3 li:)xe5 d& 4 li:)c4 with unclear play is also possible - 1 5 .J.xd8
Fortunately for our understanding of the .J.xe1 16 �e1 .J.xc2 17 .i.xc2 :Ud8 with
game, the eccentric 4 tZki3?! has not lead an endgame that is hard to assess) 1 3... .i.d6
great practical results. 4 ...c!Dxe4 5 'ife2 1i'e7 6 and Black is only a bit worse, but still, why
b3 c!Dc6 7 .tb2 .i.f5 s lik3 0-0-0 9 o-o-o ds choose to be so?
10 f3 c!Dc5 11 'ifxe7 .i.xe7 12 c!Dxc5 .i.xc5 1 3 5 d5 6 li:)e3 'Itt&!
•..

d4 .i.d6 14 Wb 1 c!Db4 1 5 .i.c3 a6 gave Black


a healthier position in Yandemirov-Raetsky,
Voronezh 1988.
4 li:)xe4 5 d4
•••

Or:
a) 5 'ife2 'ife7 6 c!De3 c!Df6 7 b3 c!Dc6 8
.i.b2 .J.e6 9 g3 0-0-0 10 .i.g2 d5 1 1 d4 (1 1
0-0 h5 12 h4 'ifd7 and Black has good coun­
terplay on the kingside) t t ...1Wd7 1 2 tZkl2 g6
1 3 h3 .i.h6 14 0-0-0 l:r.he8 with even chances,
Murey-Harikrishna, Pardubice 2002.
b) 5 c!Dc3 shouldn't lead to an advantage:
bl) 5 ...c!Dxc3 6 bxc3 (6 dxc3 is not very
dangerous: 6 ....J.e7 7 .J.f4 0-0 8 1i'd2 lDc6 9 lbis move seems to be the strongest -
0-0-0 .i.e6 1 0 .i.d3 .i.f6 and Black can't be Black develops quickly and annoys White in
worse) 6...g6 (also fme is 6...lDd7 7 d4 d5 8 the process. Alternatively:
c!De3 �£6 9 .i.d3 .i.d6 1 0 0-0 0-0 1 1 1Vf3 a) After 6... c5?! White gains a plus with 7
Aes 1 2 l:r.b1 c6 1 3 c4 dxc4 14 .J.xc4 1Vc7 1 5 .i.bS+ llk6 8 0-0 a6 (8....te7 was met by 9
h3 b5 1 6 .i.d3 .i.e6 1 7 c4 bxc4 1 8 .i.xc4 c4! dxc4 10 d5 lDJ6 1 1 .i.a4 bS 12 dxc6 bxa4
llad8 and Black has equalised, Gallagher­ 13 llM15 with some advantage in Jansa­
Delchev, Batumi 1999) 7 d4 .i.g7 8 .id3 0-0 Miralles, Paris 1 989) 9 .i.xc6+ bxc6 10 c4!
9 0-0 tZkl7 10 f4 �b6 1 1 �b6 axb6 1 2 f5 .i.e6 1 1 cxd5 cxd5 12 lDc3 lDf6 (1 2...lDxc3
lla51 and the inclusion of the queenside rook 13 bxc3 .i.d6 1 4 .ta3 leads to a rigid struc­
into the struggle guarantees Black decent ture with better chances for White) 13 f4
chances in this complicated position. 1 1 cxd4 1 4 'ifxd4 l:r.c8 1 5 �a4 .i.e7 1 6 b3 0-0
c!De3?! i'> weaker after l l ...c!Da41 1 2 .i.d2 c5! 17 .i.b2 'ifa5 18 l:r.ad1, Sanakocv-Raetsky,
(12...lt)b2?! 1 3 9£3 �d3 14 cxd3 f5 and Voronezh 1 987.
White had a slight edge in Ovetchkin­ b) 6...c6 7 tZk12 .id6 should be playable
Motylev, Russia 2004) 13 dxc5 lDxc3 14 'iff3 too, but Black was unlucky in the following
dxc5, when we rhink Black has the better game: 8 .i.d3 f5 9 0-0 0-0 1 0 f4 (10 c4 f4 1 1
chances. lDc2 lt)xd2 1 2 .txd2 n 1 3 g3 dxc4 14
b2) Even 5...lt)f6 is possible: 6 d4 .J.e7 7 .i.xc4+ Wh8 is complete!)' unclear, but
.i.d3 c!Dc6 8 d5 lDe5 9 lDxe5 dxe5 10 0-0 0-0 probably more unpleasant for White than for
1 1 llet lt)d7 12 .t£S lDb6 1 3 .i.xc8 llxc8 1 4 Black) 10....i.e6 1 1 g3 c5 12 c4!? lDxd2 (a
llxe5 .i.f6 1 5 lle3 l:r.e8 1 6 l:r.d3 .i.xc3 17 reasonable alternative is 1 2...cxd4 13 lDxd5

94
3 li)xe5: Fourth and Fifth Move A lternatives

llX5 14 �f3 lilid3 1 5 ..xd3 �c6 with a ll\d6 10 .ia4 ll\d7 Black's chances are no
complex position offering level chances) 1 3 worse.
.ixd2 dxc4 14 .ixc4 .ixc4 1 5 �xc4 .i.e7?1 9 l0xd5 •dB 1 0 Jle1 1rxd5 1 1 &3 •dB
(1 5...cxd4 was necessary, although 16 1Wf3 1 2 lhe4+
:t7 17 1i'd5 would have been a little uncom­ 1 2 .ig5 is inferior. White's main idea is to
fortable) 16 1i'b3 1i'd5 (maybe this is also not reach a murky position after 1 2...1ixg5 13
the best, but Black completely missed �xe4 1Vg6 14 �f6+ �d8 1 5 :C8+ �c7 16
White's next move) 17 �b6!! ..xb3 18 axb3 ll\dS+ �d7 17 'lre2 �6 18 :C1 (here Black
with a clear edge, Lukjanenko-Raetsky, Vo­ is mated after 1 8...b6?? 19 lld8+!). Instead
ronezh 2000. Black can play 12 ... .ie7 1 3 .ixe7 �xe7 14
7 .ibS+I? .D.xe4+ WfB when, according to Makarychev
A very ambitious, but not nece.'ISarily fool· and Smagin, the compensation is insufficient
ish. piece sacrifice. 7 ..e2 is less adventurous; and Black has slighdy better chances.
following 7....i:e6 8 c3 llX6 9 �d2 0-0-0 1 0 1 2....ie7 1 3 11re2
g3 (10 �xd5 .i.xd5 1 1 �xc4 ..g6 t 2 lDg3 13 .tg5 llX6 14 •e2 f6 1 5 d5 �5
.id6 gives Black excellent compensation for doesn't seem to cause Black the same kind of
his pawn) 1 0.. 1Wg6 (an improvement over
. problems.
10 ... �xd2 1 1 .ixd2 h5 1 2 .i.g2 1Wg6 1 3 0-0-0 1 3 ...�c6
�7 14 .i.f3 with a slight edge for White,
Smagin-Makarychev, Moscow 1 990) 1 1 .i.g2
f5 12 �f3 .id6 13 0-0 1Wf7 Black has good
counterchances.
7 c6 8 0-0!?
•••

An interesting alternative is 13....if51? 14


'it'xbS+ 'lrd7 1 5 l:te2 1ixb5 16 �xb5 �d7 17
�7!? Wxc7 18 llxe7+ �6 when the end­
game is very difficult to assess. However, it's
hard to believe that White should be worse
8 . ..cxb5 here.
Accepting the sacrifice is not the only pos­ 1 4 .ig5 .i.f5
sibility: Black is also all right after 14....ie6, for
a) 8 ... .id6 9 �xd51? 1Wg6 leads to a very example 15 .i.xe7 �xe 7 16 d5 (otherwise the
complex position, though 1 0 f3 �5 1 1 �f4 initiative disappears) 1 6...lilid5 1 7 lldt 0-0
.ixf4 1 2 .i.xf4 1i'xb5 1 3 ..ell? (13 fxe4 18 lilid5 Ld5 1 9 :C5! .i.c41 20 :Xd8 .i.xe2
1i'xb2 1 4 ..d2 1Wxat 1 5 c4 0-0 16 �3 21 l:txa8 l:txa8 22 l:txe2 with a drawn posi­
1i'xft+ 17 Wxft is not that clear) 1 3...0-0 1 4 tion. On the other hand, 14 ... b4 1 5 Jl..xe7
fxe4 1Vxb2 1 5 eno seems to favour White. �xe7 1 6 1Wb5+ Jl..d7 17 1Wxb4 promises
b) 8....ie6 is perfecdy possible. After 9 f3 White plenty of compensation.

95
The Petroff D efen ce

15 .i.xe7 lnxe7 16 .:.e5 0-0 a) 5 .i.c4+?! was the original idea, but it is
1 6...�£8 1 7 'ifxb5 'ifd7 18 ltac:1 would considered more or less refuted these days:
ensure White an initiative. S ... dSI (this is stronger than 5...�e6 6 .i.xe6+
1 7 llxe7 % -% �e6 7 0-0 Wd7 8 d4 WeB 9 c4 /l}bd7 1 0
/l}c3 - White has compensation here as
Black has no easy way to coordinate his
forces - 10...Wb8 1 1 f4 a6 12 a4 Wa7 1 3 b4
lieS 1 4 .i.eJ with a continued attack, Coch­
rane-Mohishundcr, Calcutta 1 848) 6 �b3 (6
exd5 .i.d6 7 0-0 l:lfB 8 d4 �g8 and Black is
close to winning - Bisguier) 6...�e6 7 e5
/l}e4 8 d4 c5 9 'iVf3+ WeB and Black has a
large advantage, as stated in JrhadJ�IIIng as
far back in 1 8611
b) 5 �3 is more inventive: 5...c51?
(5...1We8 6 d4 transposes to 5 d4 '1Ve8 6 /l}c3)
6 .i.c4+ .i.e6 7 .i.xe6+ Wxe6 8 d4 Wfl 9
After 1 7...b4 1 8 /l}d1 ..xd4 1 9 c3 bxc3 20 dxcS �6 10 'iVe2 1Wd7 1 1 .i.c3 dxcS 1 2 f4
/l}xc3 .i.g4! (Makarychev/Smagin) the posi­ lle8 1 3 e5 lDg4 (1 3.../l}d4 14 1Wc4+ 'ifdi 1 5
tion is level. 'iVxe6+ llxe6 1 6 0-0-0 is unclear) 1 4 lldt
r-------. 'irES 1 5 0-0 h5 1 6 .i.ct /l}d4 17 'ifc4+ Wg6
Game 4 1 and the advantage was probably with Black
Vitolinsh-Anikaev in Topalov-Kramnik, unares 1 999.
Riga 1982 5 . . .g6
The most important alternative, 5...c5, can
1 e4 e5 2 lLlf3 lnf& 3 lnxe5 d& 4 lL!xf7 be seen in the next game. Other tries include:
The Cochrane Gambit is probably not en- a) 5.../l}xe4? is bad because of 6 'ifh5+ g6
tirely correct, but it is also not that easy to 7 'IVdS+ Wg7 8 ..xe4 �6 9 dS .i.fS 10 'ifa4
refute. Vitolinsh was a great expert on this /l}eS 1 1 .i.e2 .i.e7 1 2 0-0 when White's posi-
� � � ���
4 �xf7
.•. b) 5 ... �e7 6 /l}c3 l:te8 is a very natural
way to develop, although after 7 .i.c4+ .i.e6
8 .i.xe6+ Wxe6 9 f4 Wfl 10 e5 dxe5 1 1 fxe5
White has a genuine attack. Now Black
should play 1 t .../l}c6!, for example 12 0-0 (12
exf6?! .i.b4+ 13 Wf2 ..xf6+ does not work,
while the slow 1 2 �c3?! �g8 1 3 .i.f2 /l}d5
14 /l}e4 .i.h4 15 0-0 �xf2+ 1 6 llxf2 .J:lfB
gave Black a clear edge for Black in Petrik­
Kujovic, Slovakia 2000) 1 2...'ifxd4+ 1 3 '1Vxd4
/l}xd4 1 4 exf6 .i.xf6 1 5 �g5 /l}xc2 1 6 l:lad 1
�3 17 .i.xe3 l:lxe3 1 8 lld7+ lle7 1 9 ltxe7+
Wxe7 20 /l}dS+ Wfl 21 /l}xc7 ltd8 and
White has some practical problems holding
5 d4 this endgame.
Or. c) 5...'ife8, taking prophylactic measures

96
3 IC.xe5: Fourth and Fifth Move A lterna tives

aWtinst .i.c4+, is a very recent idea: 6 liX:3 (6 dxeS 1i'xe5 1 1 Jle1 'IreS 1 2 .i.e3 1Wb4 1 3 a3
.i.d3 is punished by 6... c5! 7 dxcS dSI 8 liX3 '1Vg4 14 0) 10 h3 ltlh6 1 1 exd6 �g8 (White
dxt.-4 9 ltlxe4 .i.f5 10 0 .ixcS 1 1 1i'e2 .ixe4 wins after 1 1 ...cxd6? t 2 llet '1Vc6 13 .ie4) 12
1 2 .i.xe4 1i'a4, when Black is clt.-arly better) dxc7 (1 2 ltld5!? cxd6 1 3 ltlc7 1lc6 1 4 ltlxa8
6 ... d5 (6 ... c5 also looks fine) 7 eS .i.b4 b6 1 5 ltlxb6 axb6 1 6 c3 with a slight edge is
also pos!>ible) 12...ltlc6 1 3 d5 ltle5 (Vitolinsh­
Domuls, Riga 1983). Now after 1 4 .ixh6!?
.i.xh6 1 5 d6 Wf7 1 6 .i.e4 we're leaning to­
wards giving White a clear edge. This of
course is not all due to 7 .i.d3, but it does
confirm that it's a viable option for White.
7 . . .i.e& 8 d5 i.cB
Here we assess 8 ... .td71? 9 0-0 bS 1 0 .i.d3
.i.g7 as a bit better for Black.
9 0-0 i.g7 1 0 Ae1 llfB

and now:
c 1) 8 .i.e2 ltle4 9 0-0 .i.xc3 1 0 bxc3 ltlxc3
(J O_.'tlc6 1 1 c4!? also gives White play) 1 1
.ihS+ g6 1 2 'IVO+ �g8 1 3 1i'xc3 gxhS 14
:1b1 and the position was a complete mess in
11crnandez Romero-Andres, Seville 1999.
c2) 8 11'0 �g8 9 .id3 ltle4!? 1 0 0-0 (this
is stronger than 1 0 .ixe4 dxe4 1 1 'lfxe4
'iVc6! with an advantage for Black - Re­
inderrnan) 1 0....i.xc3 1 1 bxc3 ltlc6 12 .ixe4
dxc4 1 3 1i'xe4 1i'g6 with a complex position, 1 1 e5
Ucrgcz-Grimbcrg, Clichy 1 998. 1 1 f4 regains the piece but then the attack
6 lbc3 11t'e8 dissipates after 1 l ...�g8 1 2 eS dxeS 1 3 fxeS
Alternatively: ltlg4 14 d6+ �h8 1 5 e6 ltles 16 d7 (1 6 e7
a) 6 ...'j;g7 7 f4 We8 8 cS looks very dan­ ltlxc4 17 cxfB1i'+ 1i'xf8 18 dxc7 ltlc6 is very
�crous for Black, for example 8...dxe5 9 fxeS unclear play but we do not think Black
.li.b4 10 .i.c4!? .i.xc3+ 1 1 bxc3 with the idea should fear this) 1 6 ...ltlbxd7 1 7 exd7 Wxd7
l l ....i.g4? 12 0-011 .i.xd1 13 exf6+ �fB 1 4 1 8 1i'xd7 .i.xd7 with a drawish endgame,
.li.h6 mate! Popov-Grodzensky, correspondence 1 983.
b) 6....i.g7 7 .i.c4+ .i.e6 8 .i.xell+ �e6 9 1 1 ...dxe5 1 2 d6+ i.e& 1 3 llxe5 .bc4
f4 cllfl 1 0 e5 lieS 1 1 0-0 was played in Vito­ 1 4 1lxe8 llxeB 1 5 dxc7 lila6
linsh-Anikaev, Frunze 1 979. Now after After 15 ...ltlc6 16 .i.gS llac8 17 1i'O
l l ...ltlfd7 12 f5 dxeS 13 fxg6+ �g8 1 4 Wo llxc7 18 W£4 li:)eS 19 li:)e4 llc6 20 l:ld1
White would have a strong attack. White manages to generate threats.
7 .i.c4 + 1 6 .*.f4 llecB?I
7 .td3!? is another possibility, for example 1\ sad decision - Anikaev was probably
7 ... .i.g7 8 0-0 llfB 9 eS ltlg4 (the pawn­ afraid of ghosts such as ltlb5-d6. It would
w-:tbbing 9 ... dxe5? leads to a disaster after 10 have been better to leave a rook on e8, e.g.

97
Th e Petroff Defence

l6...:ac8 17 1Wd4 bS 1 8 1Wxa7 I?Jxc7 19 29 lDf5 lDef6 30 lDe7 + �8 31 lDxcS


i.xc7 ltt:7 20 .VaS l:lcxc7 with an unclear :XeS 32 •d6+ �g7 33 .id4 .J:le8 34
t:ndgame. �1 g5 35 .ba7 �6 36 11fc7 ..te& 37
1 7 b3 11fxb7 .i.f5 38 .i.e3 g3 39 ¢>g1
I 7 1Vd4! is simple and strong. AftL-r After 39 'W'b6 gxf2 40 .i.xf2 .i.d3+ 41
17 ... bS? 1 B b3 �15 1 9 .i.eS Black's position �gl l:lc2 Black has exccllem attacking
coll.:Jpst:s. chances.
1 7 .i.e& 1 8 clflb5 �
••• 39 t;},g4 40a4 lDhf6 41 ••a o-1
•••

1 8...1?Jd5 is a weak move: 19 I?Jd6+ �g8


20 lbxc8 J:lxc8 21 .i.e5! LeS 22 ..e1 and Game 42
White has a huge advantage. Novozhilov-Raetsky
Comspondence 1983
1 e4 e5 2 �f3 lDf6 3 lDxe5 d6 4 lDxf7
¢>xf7 5 d4 c5
Time has proven this to be the most test­
ing response to the Cochrane Gambit.
6 dxc5
This is the most commonly played move,
but 6 .i.c4+ is worth considering: 6 ...d5 7
exd5 .i.d6 8 0-0 lieS 9 lL!c3 .i.g4 (maybe
9 ... cxd4 10 1txd4 i.eS 1 1 1Wd3 Wg8, when
we prefer Black - we think!) 1 0 f3 .i.f5 1 1
dxc5 .i.xcS+ 1 2 �h 1 .i.d6 I 3 lL!b5 'ili'g8 1 4
19 11ff3?! �5 a6 1 5 lbd4 .i.d7 1 6 llk6 .i.xe6 1 7 dxe6
After the stronger 19 llc1 �g8 20 1Wf3 'ifc7 with a typically messy position, Stcll­
lbaxc7 21 lbxc7 lbxc7 22 c4 Black's uncoor­ wagt:n-I.Smimov, Heraklio 2002.
dinated forces promise White some kind of
edge.
1 9 ....i.xa1 20 .i.e5+ �8 21 .ixa1
�xc7 22 �
Vitolinsh could have offered better resis­
tance with 22 1Wc31? lbxb5 23 1Vh8+ �f7 24
1Wxh7+ �ffi 25 1Vxg6. Ht:re Kaemer and
Petkevich give the line 25....i.f7 26 1Vh6+
�e7 27 1Wg5+ with a draw, but after the
stronger 25 ...lbbc7 26 .i.b2 lidS we still be­
lieve Black can play for a win.
22 .i.d5 23 11fg3 t;},g7 24 c4 il.n 25 h4
.•.

ltX:eS 26 •e5 h5
Also strong is 26...b6!? 27 h5 :c5 28 1Ve4 6 11fa5+
.••

.1Lc8 29 h6 lDf6 30 1Vf3 lbges with clear After 6...c!L!c6 7 i.c4+ we have two possi­
advantage for Black. bilities:
27 g4!? hxg4 28 h5 lDxh5 a) 7...d5 8 .i.xdS+ i.d) (not the only op-
28...gxh5?? 29 I?Jf5 would revt:rsc the tion; 8...1?Jxd5!? 9 cxd5 lL!b4 looks very
trend! strong for Black) 9 i.xt:6+ ci>xc6 10 'ife2

98
3 l0xe5: Fourth a nd Fifth Move A lterna tives

1Wa5+ 1 1 ltlc3 .i.xcS 12 0-0 (12 1rc4+ 9ile7 was close to losing in Short-Shirov, Dubai
U 0-0 llad8 is unclear) 1 2...�f7 1 3 11'c4+ (rapid) 2002.
b2) 9 ... d5 is also interesting:
b21) 10 e5?! � 1 1 1i'g4+ �f7 12 'l'f5+
�e8 and we do not think there is enough
compensation, for example 13 �c3 llXI4 1 4
Wh3 ..d7 1 5 e6 'iltxe6 1 6 ..d3 �xc3 1 7
bxc3 ltle2+ 1 8 �h 1 �ct 1 9 l:axct � f7 20
llcct 'irc6 with a clear edge for Black, Vi[O­
linsh-Dautov, Minsk 1 988.
b22) 10 ltlc3!? d4 1 1 ltla4 (1 1 1re2?1 is
awarded a '!' by Forintos and Haag, but after
1 t ...dxc3! 1 2 Wc4+ �e7 n es 1Wd4! 1 4 cxf6+
*xf6 White has no compensation) 1 1 ...�f7
with chances for both sides.
and now:
at) 1 3...�g6?1 t4 ltld5 .i.d6 1 5 .i.f4 llad8
1 6 llad1 11'c5 1 7 11'b3 �xe4? (a mistake,
although Black was already in ttnuble) 1 8
.tcJ Was 1 9 •xb7 :Cs 20 b41? (20 1i'd7!
nlso win.<�) 20...�xb4 21 �f4+ .i.xf4 22
11fxc4+ 1Wf5 23 9xb4 and White was win­
ning. Vitolinsh-Raetsky, Naberczhnye Chelny
1 9R8.
a2) 1 3 ..'�f8 14 ltlclS ltleS 1 5 Wb3 ltlxc4
.

I 6 9xb7 lieS and nothing has been decided.


b) 7 ....i.e6 8 .i.xe6+ �xe6 9 0-0 with:

7 �c3 1txc5 8 .i.e3 ••5


8...•c7!? 9 ltlb5 WaS+ 1 0 c3 �6 is a
possible recommendation- White seems to
have difficulties generating an initiative.
9 .i.c4+ .i.e6 1 0 .i.xe6+ Wxe6

b1) 9...�f7 1 0 We2 1re8 1 1 llc1 dS


( I t ...Wt!6!?) 12 eS �c4 1 3 �c3 ltlxc3 1 4
hxc3?! (1 4 ..f3+ �g8 1 5 WxcJ Wc6 is
str onger but Black still looks pretty solid
,

here) 1 4.....e6 1 5 :b1 :bs 16 .i.c3 .i.e? 17


f4 :hfB 1 8 .i.d4 �g8 1 9 :n gS and White

99
The Petroff Defence

11 0-0 with his queen. The stronger option is 14


It 1i'e2 has also been played: t t....i.e7 1 2 g5!? and now:
g4 :CB (this and Black's next move are not a) 14...lbxe4 1 5 9g4+ Wf7 (15...�e71? 1 6
srricdy necessary) 13 0-0 l:lxc31? 14 bxc3 oltX6 llae1 i s also unclear) 16 'iff5+ WeB 17lDxc4
with a very unclear position, Zelinsky­ dxe4 1 8 1txe4+ .i.e7 19 f5 and the position
Volchok, correspondence 1 988. continues to be very difficult to as.c;css.
1 1 ....!tlc6 1 2 f4 AdS b) 1 4...dxe4 is not so good: 15 'ife2lDd5
1 2...J..e7 1 3 11t'e2 ..a6 t4 lbb5 rlif1 is also 16 f5+ �f7 (16...We71 may well be an im­
pretty messy. If push came to shove, we provement) 1 7 'ifhS+ Wg8 1 8 f6 gxf6 19 llf5
would probably choose to be Black. (Bielak-Borys, Poland 1 992). Now after
19...lbce7 (what else?) 20 gxf6 lbxf5 21 f7+
Wg? 22 •g5+ c;P,c.n 23 'il'xf5+ �e8 24 'ife5+
Wd7 25 lbxd5 White has serious attacking
chances.
14 ...¢17 1 5 g5 J.c5 1 6 gxf6 .i.xe3+ 1 7
�1 d4
Stronger than 1 7... dxe4 1 8 fxg7 Wxg7 19
Wg4+ wm, when Black's king is extremely
insecure.
18 fxg7 llhgB!
18 ...Wxg7?1 is inferior. After 1 9 f6+ Wf7
20 llf5 1i'xf5 21 exf5 dxc3 22 ..g4 �xf6 23
bxc3 Black's superiority is not as clear as it
1 3 g471 should be.
'This is too risky and is met by a strong
counter in the centre. 1 3 a3 Wf7 1 4 b4 Wc7
1 5 lDdS gives White the initiative, even
though the position is probably still okay for
Black. However, perhaps 13...'ifh5!? gives
Black the better chances.
1 3...d5

1 9 f6
'This leads nowhere. White can struggle o�
a while with 1 9 lbds llxd5 20 exdS ..xdS+;
21 'iff3 Wxf3+ 22 llxf3 llxg7, or t 91ih �
�g7 20 f6+ �h8. Although Black is dearl�
better, the g:1mc is not totally decided. j
19...dxc3 20 1i'e2 1i'c5 I
14 f5+?1 Also strong is 20...cxb2 21 llae1 'ifc5
:;j
This eliminates White's chances to invade Wxd ..xeJ 23 llxe3 lld61 24 eS lle6 25j

100
3 lllx e5: Fourth and Fifth Mo ve Alternatives

tlb3 b6, when Black is a knight up for noth­ 'ii'd2 'ifb6 181i'h6 1-0 'Iimofuev-Phoobalan,
ing. Goa 2002.
21 l:tf5 cxb2 22 l:taf1 � 23 1ih5+ 4 1We2 1We7 5 1fxe4 d6 6 d4 dxe5 7 dxe5
'tte6 24 1ih3 Wi'xe4+ 25 l:t5f3+ Wd6 26 7 1i'xeS?! is a weaker choice: 7.....xe5+ 8
*g2 �d4 27 l:txe3 1Wxe3 28 1Wxb7 l:tbS dxe5 .i.fS 9 c3 (91£lc31? .txc2 10 lDdS �d7
29 f7 1Wf3+ 0-1 I 1 .i.f4 l£lc6 12 .tbS l:ld8 13 llct is still
White resigned because of 30l:lf3 btW' 31 probably better for White) 9...1£Jd7 10 f4 (1 0
�1l:lbt32Wg2l:lg7. .tf4 0-0-0 11 1£Jd2 I£Jxe5 12 �xeS l:le8 and
Black is at least equal - Yusupov) t0...0-0-0
Game 43 1 1 .i.e3 f61? t2 I£Jd2 (12 .i.xa 7 fxe5 13 fxe5
Kholmov-Belousov I£Jxe5 14 .i.d4 :C8 15 �d2 g6 gives Black
Gorky 1974 serious threats) 12... fxe5 13 fxe5 I£Jxe5 14
0-0-0 I£Jct3+ 15 .i.xd3 :Xd3 and Black's posi­
1 e4 e5 2 �3 �6 3 lllxe5 lll xe4?1 tion is preferable, Kos-Potapov, Ceske Bude­
jovice 19tJS.
7 ...lllc6

This gambit/beginner's move is simply


bad.
3...1Vc7?! is also poor: 4 d4 (a decent alter­ 8 J.b&
native is 4 l£lf3!? I£Jxe4 5 �e2 dS 6 0-0 �e6 We think this is a good decision. Alterna­
1 .l:let •f6 8 d3 I£Jcs 9 I£Jgs �d6 to .i.o tively:
when White has a clear advanta�} 4...d6 a) 8 .tf4 g5!? 9 .tbS (9 .i.g3?? fS! 1 o 'lre2
(after 4 ...1£lxe4 5 �3 d5 6 0-0 ..f6 7 �xe4 f4 and Black winsQ 9....i.d7 10 .i.d2 .tg7 1 1
dxe4 8 1£Jc3 .i.fS 9 l:le1 White ends up a I£Jc3 0-0-0 12 0-0-0 .i.xeS and Black has
pawn ahead) 5 1£Jf3 'irxe4+ (S...I£Jxe4 6 �e2 certainly survived the opening.
i.g4 7 0-0 dS 8lte1 1£Jc6 9 c3 0-0-0 tO I£Jg5 b) 8 i£1dl? is interesting: 8...'ii'xe5 9
i.xe2 t 1 l:lxe2 also gives Black a serious 1Wxe5+ I£Jxc5 10 .i.f4 (10 ll'lbs .tb4+ 1 1 c3
headache) 6 .i.e2 �e7 (or 6...�f5 7 c4 1i'c2 8 .taS 12 .i.f4 f6 and BL1ck holds his own)
1i'xc2 .i.xc2 9 i£1a3 �e4 tO i£1bS I£Ja6 11 lO....i.d6 11 i.g3 .i.d7 12 IJ-0-0 0-0-0 13
i.gS with an advantage for White - all ll'le4 .i.c6 t 4 1£Jxd6+ cxd6 15 f3 llhe8 1 6
Black's pieces are passive) 7 0-0 0-0 8 c4l:le8 l:ld41? rl;c7 17 a4 and White had the advan­
9 l£lc3 W'fS to .i.d3 'ti'hs t I h3 c6 12 l:let fabFC in the endgame, Vasiukov-Chckhov,
lba6 13 .tgS (White has emerged from the Moscow l97S.
opening with a decent plus) 13...1£Jb4 14 8 J.d7 9 �c3 0-0-0
...

.ixf6 gxf6 1S ll'le4 .i.e6 16 lbg3 WaS? 17 After 9...1i'b4?! 10 .i.c4 �£laS It .i.d3

101
The Petroff Defence

Whi et is a pawn up for nothin g. es sary . Afte r 1 3 Wf4 <l.>b 8 1 4 lta dt White has
a b ig advanta ge. StiU , it is very diffic ult to
defe nd again st suc h a sac rifice . Th is, alo ng
wi th the ae sthe tic be auty of the idea,e nc our­
age d K holm ov take the r isk.
1 3...�7 14 1ta4 l0c6 15 •as+ �7
1 6 :ad 1 + We& 17 1txb7 1te6 1 8 Ade1
.te7
1 8..1lb
. 8 lose s to 19 .xc 7 llc 8 20 .g3
.g6 21 i.c S+ .i.e? 22 11 a3 1lg5 23 lle 3
f oU owe dby lif e 1.
1 9 .tc5 Ab8
19...1i d7? is me t wi thbye el gant 20 .xc 6!
11V xc 6 211l xe 7+ �f8 221l xc 7+, winnin g.
10 0-0 20 1ta6 ..d5?1
A sim pler way to play is to .i.f4 gS 1 1 Th is os
l esby force. A fter 20.. .Wf6 21 11V a4
i. g3 ( 1 1 e 6 fxe 6 1 2 i.e 3 i. g7 is unc el ar) �f8 221lxe 7 �xe 7 23 :C 1 White hase very
1 1 ...i. g7 1 20-00 - h5 1 3 h4! with ac el ar ad­ c hance of winnin g,b ut the game is stiU on.
vantage. 21 1ta3 1tg5 22 .:e3 Wf8 23 .txe7+
10.. .lt)xe5 rtJxe7 24 f4 1tf6 25 lZ.fe1 :ea
10...a6 1 1 i.c 4 lilie S 1 2 /lhls 11Ve 6 1 3
.i.gS 6 f 1 4 .i. f4 give s White good a ta t ckin g
c hance s, while 1 1 i. xc 6i. xc 6 12 Wf5+ '1Vd7
1 3 Wf4 should be be tter for White too.
11 .te3
1 1 .i.xd7+ �xd7 12 '1Va4 a6 13 �S We S
14 WaS also pro vide s White wi th an edge.
1 1 ....txb5
Or 1 1 ...a6 12 /lhls ( 12 .i.xa6 .i.c 6 13
Wf5+ .i.d7 14 .e4 i.c 6 with a re pe titi on­
Kh olm ov) 12...We 6 1 3 �f4 'lieS 14 .i.xd7+
�xd7 1S W£5 and White retain s the i nitia­
ti ve .
1 2lilllb5 &6?1 26 :Xe71
Missing ac hance . A re f r 12...a6 13 �4 g6 F orcin g a winn inge ndgame .
1 41l ef 1 i. g7 1 Sila d1 llhe8 B lac k would be 26 ....xe7 27 :Xe7 Axe7 28 g4 f5 29
aU ri ght. •c5 g6 30 gxf5 gxf5 31 1txf5+ Af7 32
1 3lillla7+ !7 1tc5+ �8 33 Wg2 h5 34 �3 Ah6 35
Th is et m ptin g sacrifice is not str ci tly nee- 1tg5+ Wh7 36 a4 1-0

102
3 l0xe 5 : Fourth and Fifth Mo ve Alterna tives

Summary
S ll:lc3 has been underestimated by theory, and Black should act carefully here. In Game 33
Alckhinc played fantastically to defeat Rabinovich, but the pawn sacrifice with 5...d5 is dubi­
ous. 5...ll:lf6 is quite passive, but Black appears to be on the verge of equalLo;ing here. The more
popular and sharper S...ll:lxc3 6 dxc3 �c7 normally results in opposite side castling and a fight
with mutual chances.
The easiest way for Black to cqualL<ie against 5 ...e2 is with S.....e7 6 d3ll:lf6 7 �g5 ..xc2+
8 J..xe2 ll:lbd7 and 9...h6, preparing the simple ...g7-g6 or the extended fianchetto (...g7-g5).
The continuation 5 c4 is more challenging than 5 d3, but Black's play in Game 37 demon­
Mrates that it's also fairly harmless. The modern 5 J..d3 is pretty safe for Black. However, the
pc 1sitions are quite untypical and have not been studied very much.
The sacrifice 3ll:lxc5 d6 4ll:lxf7 deserves consideration because it has been shown dmt it is
not easy for Black despite the extra piece. 5 d4 ..e8, in conjunction with either ... g7-g6 or ...d6-
d5, seems to be promising for Black, while 5...c5 and in particular 6 dxc5 ltk6 7 J.. c4 ..tc6 8
J..xe6+ Wxe6 has not lost its favourable reputation. Of course, the presence of the black king
in the centre puts higher demands on Black but, objectively speaking, Black's chances are pref­
erable in the Cochrane Gambit.
After 3...ll:lxe4 4 '1Ve2 '1Ve7 5 ..xc4 d6 6 d4 dxcS 7 dxe5llk6 White has better prospects.
lie has a pleasant choice between 8llk3 and 8 ..ib5.

1 e4 e5 2 liJf3 ttlf6 3 ttlxe5 d6


3 lDxc4- Game4J
...

4 ttlt3
4llk4 - Game 40
4ll:lxf7 Wxf7 5 d4 (D)
5... c5- Gatne42; 5... g6- Game41
6 ttlxe4 5 tDc3 (D)
.••

5 d3- Game J9; 5 .i.d3- Game 38; 5 c4- Game 37


5 ..e2 'fle7 6 d3ll:lf6 1 J..g5 (D)
7 ... i.e6- Game 35; 7.. ..xe2+- Gm11e36
.

5 ...ttlxc3 - Game34
5...d5- Ga111e 33

5 d4 5 ttlc3 7 �g 5

103
CHAPTER SIX I
3 d4: The Main Line

1 e4 e5 2 �f3 �f& 3 d4 lL!xe4 4 ..td3 d5 d'-"Vclops by means of 6 'ire2 'We7 7 .i.f4


5 �xe5 tLl d7 foUowed by 0-0 and llet.' The Black Army
In this chapter we begin our study of 3 d4, leader could hardly expect that many years
the Stcinir.t system. 1 e4 eS 2 .!bf3 .!bf6 3 d4 later his idea would rise like a phoenix from
.!bxe4 4 ..ld3 dS 5 .!bxeS is one of the key its ashes.
lines in the Petroff Defence. Having placed Up to the 1980s, S...lbd7 had been a dead
his knight on eS, White obtains a symmetrical branch on the spreading tree of the Petroff
position with an extrn tempo. In this chapter Defence. However, the year 1980 proved to
we look at Black's main choice, S...lbd7 be a crucial moment in the hi10tory of the
The authors have special feelings towards variation. In the Tilburg super-GM tourna­
the line with 5...lbd7. In 1990 in Voronezh a ment, Larsen sacrificed a pawn for counter­
small book (40 pages) was published in Rus­ play and the bishop pair against Karpov, tl1e
sian under the title A Fashiolltlble Vllliafioll i11 world champion at the time, with 6 'ifc2
the Petroff Difelice. It was our first serious joint .!bxeS! 7 .i.xe4 dxe4 8 1Wxe4 .i.e6 9 'irxeS
work... 'ifd7 (following the game Unzicker-Rogoff
In his first correspondence theme tour­ from Amsterdam earlier in the year). The
nament devoted to the Petroff Defence Danish GM could hardly have hopt:d for
(1982-1983), Alexander Raetsky faced the better publicity than he received after this
foUowing problem: how to initiate a compli­ impressive win. Players started to employ
cated fight after 3 d4 .!bxe4 4 ..td3 dS 5 s...lbd7 at major tournaments and now iliis
.!bxeS. At that time the main continuations is Black's main weapon against tile Steinitz
were S.....td6 and S.....te7. However, in both system - 3 d4. At present almost all the chess
lines Black only dreamed of 'beating off elite, players such as Kramnik, Anand,
White's attack and attaining a draw. But Morozevich, lvanchuk, Ponomariov and
Alexander had a burning desire for some­ Kasimdzhanov, arc advocates of 5...lbd7.
thing 'sharp' and 'interesting'. The main continuation is 6 .!bxd7 .i.xd7 7
In search of the active weapon let's refer 00- (Games 44-55). Black has a wide choice
to Bilguer. The omniscient f-ltmdbnch informs between the modest 7.....te7 (Game 55), tl1e
us: 'after s. ..lbd7, which was played in the queen lunbrcs 7...11f'f6 (Game 54) or 7...-.,4
Barry-Showalter match (1896), White "-asily (Games 49-53), and the fa10hionable 7....i.d6

104
3 d4: The Main L ine

(l1ames 44-48). (Game 58) is one of the numerous discover­


At the end of the 20th century, the sharp­ ies by Igor 7...aitsev, one of the most creative
rst line 7...9114 8 c4 0-0-0 9 c5 g5!? (Games grandmasters in tlte world. This novelty led
49-52) had been hotly debated for almost to a quick and amusing draw by perpetual
two decades - amazing for such a quiet check in Zaitscv-Karpov, Leningrad 1966
upcning! Even the representatives of the after 6...1i'e7 7 li)xh8 �3 8 ..td2 li)xd1 9
chess bt:au monde like Kasparov, Anand, :Cl li)xf2 10 .i.xh7 li)e4 11 J:lxe4 dxe4 12
lvanchuk, Judit Polgar participated in these .i.g6 ..td8 13 li)f7. Six years later, Gur­
interesting discussions. Despite the statistics gcnidze routed Bellin by employing 7 1i'e21?
lx:ing in White's favour, we believe that there ..txm 8 'iVhs �f6 9 o-o 'fkf7 10 'ikh4 gs 1 1
nrc possibilities to improve Black's play in all .ixgS! li)xg5 12 f4 with a mating atmck. But
the main lines (10 .ie3, 1 0 f3, 10 li)c3). this is too good to be true, and Game 58
At the lx.-ginning of the 21st century demonstrates Black's superior options.
7...i.d6!? is in the limelight. (It is surprising
how the theory is developing: in the above­ Game44
mentioned book of 1990 we gave only one Svidler-Ponomariov
reference out of 78 to the 'third-rate' move Moscon' 2001
7....i.d6) After the pretty typical 8 c4 c6 9
cxd5 cxd5 White is at the crossroads: he can 1 e4 e5 2 ll:lf3 ll:lf& 3 d4 o!Oxe4 4 �d3 d5
play either 10 �3 li)xc3 11 bxc3 0-0 12 5 li:lxe5 ll:ld7 6 ltlxd7
'i!Vhs (Games 44-45) or the immediate 10 Sixth move alternatives for White will be
1i'h5 (Games 46-47). In both cases Black has considered in Games 56-58.
reasonable play. 6 -*.xd7 7 0-0 �d&
...

Nowadays Black is not afraid of the tacti­


cal move 6 1Ve2 (Game 56), intending to use
the pin along the e-ftle; in fact, Black even
looks forward to it. Extensive practice has
proven tltat the pawn sacrifice after
6...li)xe5!? 7 .i.xe4 dxe4 8 1i'xe4 .i.c6 9 1i'xc5
1i'd7 is correct; easy development, two bish­
ops in the open, and awJ..:ward position of the
white queen create enough compensation for
Black.
The practical continuation 6 li)c3 (Game
57) opposes the knight on e4. But once all
ti1ur knights have been exchanged (6...li)xc3
7 bxc3 li)xe5 8 dxe5) White is unlikely to This is the solid alternative to 7 ...9114,
have an advantage, due to his broken pawn which is discussed in Games 50-53. T .ess
structure. On the other hand, grabbing mate­ common options for Black are studied in
rial with 6..ib4 7 0-0 li)xc3 8 bxc3 .i.xc3 9 Games 54-55.
llb1 provides White with a dangerous initia­ 8 c4
tive. Also in Game 57 is the unimpressive 6 Bli)c3 features in Game: 48.
0-U, which was often played in the earliest s c&
...

days of 5...�7. Black has easy play after 8...9114 transposes to 7 ...'ifh4 8 c4 .i.d6-
6...li)xe5 7 dxeS �5. see the notes to Game 49, while tL0-0 9
Finally, the unexpected sacrifice 6 li)xf7!? cxdS %le8 10 li)c3 9114 11 g3 li)xc3 12 bxc3

105
The Petroff Defence

11'h3 13l%b1 favours White - Black has no 14 i.xc4+�hS 15 i.gs •es t61Wh41ooks a
easy way to get his pawn back. bit better for White) 14 l%bl!? 1Wxd4 15
9 cxd5 llxb7 dxc4 16 .i.xc4+ 9xc4. Here the play­
Another option is 9 ltX3 and nuw: ers agreed a draw in Velicka-Rabiega, Berlin,
a) 9...0-0 10 'iihs lC!f6 11 -..,4 dxc4 12 1999. After 17l%xd7 1t'c6 White is forced to
.i.xc4 :Cs 13 .i.gS h6 14 .i..xf6 1Wxf6 t 5 play the adventurous 1S llxg7+1 Wxg7 19
1Wxf6 gxf6 is ftnc for Black. After 16 llfel i.b2+ i.eS (19...WgS 20 1Wg5+ Wfl 21 1Wf6+
.i.e6 17 lbc4 ll..c7 HI .L.c6 fxe6 19 lbcs WeB 22 We6+ W<.l8 23l%dt looks very dan­
.ll.xc5 20 dxc5 l%adR 21l%ad1 Wfl the play­ gerous for Black) 20 i.xe5+ WgS 21 l%<11
ers agreed a draw in Sveshnikov­ when he has substantial compensation for
Mikhalchishin, Kuibyshev 1986. the exchange.
b) 9...lbxc3 10 bxc3 dxc4?! (Black could 1 3.. :ec7 1 4 i.d2 :aea
still transpose to the main game with 10 ...0-0)
11 l%et+ ll..c7 12 .i.:u .ie6 13 .L.e7 1Vxe7
14 .ixc4 0-0 15 .lieS 1Wa3 16 .i.xe6 fxt.-6 17
11Ve1 and White had some pressure, Shott­
Hubner, Wijk aan Zce 1986.
9. .cxd5 10�3
.

1 5 -*.c2
Maybe White has a brighter future after 15
g3, for instance:
a) lS .. .:Xet+ 16 :Xet WaS 17 1r f3 'it'xa2
1Sll..f4 i.xf4 19l%e7l%f7 20l%xf7 Wal+ 21
Wg2 Wxfl 22 1Vxf4 when White has consid­
The main alternative is 10 Whs - sec erable activity for his pawn, and Black needs
Games 46-47. to defend precisely to survive.
1 O . .l2�xc3 1 1 bxc3 0-0 1 2 Wh5 f5
. b) 15. g6 16 Wf3 l%e4!? (this compelling
..

Ths is the most solid move. The pawn idea crops up more than once) 17 :Cct bS
sacrifice with 12..g6 is considered in the next 18 a4 a6 19 .ih6 :Cs 20 .i.xe4 dxe4
game. (20... fxe4?! 21 1Wf6 .i.fB 22 .i.xfB .:.XfB 23
1 3lZe1 1Wxa6 bxa4 24 c4 and there is no compensa­
Alternatively: tion for the exchange) 21 'lfe3 1Vc4 22 axbS
a) 13 11V f3 Wh8! is fine for Black, for ex­ ll..xbS 23 i.f4 .ifB 24 i.eS and it looks like
ample 14 ll..f4 ..c7 15 ll..xd6 1Wxd6 16 i.c2 White has a slight edge, Ponomariov­
:CeS 17 :ret l%ac8 18 h4 g6 19 h5 Wg7 20 Moro?.evich, Moscow (rapid) 2002.
g3 ..c6 21l%xeS :XeS 22 Wg2 1Wd6 23l%h1 1 5 ... g6 1 6 1rf3lZe4! 1 7 g3
i.bS 24 i.b3 l%e4 with equal chances, 17 ll..xe4? fxe4 1S 1We2 .i.xh2+ 19 �h1
Galkin-Macicja, Istanbul 2003. l%5 (Golubev) gives Black a strong attack.
b) More interesting is 13 c41Wf6 (13...dxc4 1 7 ...-*.b5 1 8 i.f4

106
3 d4: The Main Line

1 H .i.xe4 fxe4 19 'iVg4 .i.d7 offers Black 36 l:l.4b3 'iVxu4 37 Jle2 WgS is probably
compensation, while 18 .i.h6 l:lf7 1 9 l:lnb I better, though Black has all the chances.
'iVc6 is equally unclear. 36 ...Wg5 37 lle2 Wf4 38 llb31Vxd4+ 39
18 ....txf4 19 gxf41Vd6 Wg2 1l'c4 40 llf2+
19 ...'itxf4 20 .i.xe4 fxe4 21 'itxf4 l:lxf4 is Or 40 llbe3 'iVct 4 1 Wf2 1Wht and Black
also fine for Black, but probably just a dmw. wins.
20 .txe4 40 �g5 41 llf7 d4 42 h4+ Wxh4 43
•••

White no longer has a choice. Even llxh7+ Wxg4 441lg3+ Wf5 0-1
though objectively the position is probably
equal, White's game is harder to play. Game 45
20...fxe4 21 1Vg3 llxf4 22 :Sb1 .td7 23 Rublevsky-Vallejo
f3 Ohrid 2001
23 h3 b6 24 c4 'iff6 25 cxd5 l:lf5 26 Wb8+
c.tr>g7 27 llb3 'tWxd4 28 llbe3 (Baklan) is an 1 e4 e5 2 ll'lf3 ll'!f& 3 d4 ll'lxe4 4 .td3 d5
interesting possibility for White, who has 5 ll'lxe5 ll'ld7 6 ll'lxd7 .i.xd7 7 0-0 .td& 8
some countcrplay of his own. c4 c& 9 cxd5 cxd5 10 ll'lc3 lL!xc3 1 1
23 ... b6 bxc3 0-0 1 21Vh5 g6!7 13 Wxd51Vc7
A weaker option is 23...exf3?1 24 'it>f2 b6
25 lZ.e3 .i.f5 26 l:te5 Wa3 27 'iVxf4 'ilfxa2+,
when Black has not completely e'lualised.
24 lle31Vf6 25 Af17
25 l:lb2 'irfB 26 llf2 1Wh6 is equal accord­
ing to Golubev.
25 . . ..tg4!
This 'sacrifice' destroys White's position.
261Vxg4
26 h3 .i.xf3 27 �h2 llf5 28 a4 llgs 29
1i'h8+ Wg7 30 l:tf2 Wh6 is also discouraging
for White.
26...1lxg4+ 27 fxg4 1Vg5 28 llg3 b5 29
Af2 Wg7 30 Wg2 a5 31 llb2 b4 32 cxb4 141Vf3
axb4 33 h3 1Vc1 34 1lgb3 �h& 35 1lxb4 It seems that the text move is the most
1Vd1 36 Wf2?1 dangerous for Black, although it's not the

107
The Petroff Defence

only attractive option. In fact, White ha.., tried strongest is 16....i.eS! 17 f4 (17 dxc5? %lxc5
a few other moves here: 18 .i.f4l:l.xd5 19 .i.xc7 ll.xd3 favours Black
a) 14 h3 does not appear to be dangerous slighdy) 17....i.c6 18 ..c4 b5 19 Wb3 ..d71?
for Black: 14..ie6 1S Wf3 ...xc3 16 ll.bt 20 51? .i.g3 21 Wg1 with complex play,
(after 16 .i.h6 Black plays 16..llfd8 17 .i.gs LepeUetier-Marciano, Auxcrre 1996.
ll.e8 18 .i.f6 .i.fS 19 .i.xfS ..xf3 20 gxf3 gxfS c2) 15 h4 .i.e6 16 •o •xc3 17 •f6 .i.f8
with equal chances) 16.....xd4 (16....i.xa2 17 18 .i.xfB llxfB with a fmal split:
llxb7 ..xd4 18 .i.b2 ..f4 19 •xf4 .i.xf4 20 c21) 19 .i.e4 19....1:lad8 20 %lfd1 ..td5 21
ll.a1 .l:lfd8 21 .i.ft .i.e6 22 ll.axa7 lL'lc8 is ll.act 1fa3 22 hS ..d6 23 ..xd6 %lxd6 24
level according to Belikov) 17 .i.e4 We5 18 .i.xd5 llxd5 25 ll.c7 ll.xh5 26 d5 ILls 27
g3 .i.xh3 19 .i.b2 'fle7 20 ll.fe11lae8 21 .i.c2 ll.xb7 llhxd5 with a draw in Sutovsky­
(or 21 bb7 1ixet+ 221lxe1l:l.xet+ 23 ..th2 foridman, Medellin 1996.
.i.fS 24 •c3 :Cs 25 f4 ll.e2+ 26 .i.g2 .i.eS c22) 19 llfdt WaS (19..llad8!? should be
27 fxeS .i.e4 28 e6 ll.xg2+ 29 Wh3 llxb2 30 fine; maybe White continues with 20 .i.e4, as
exf7+ ll.xf7 31 ..xb2 with a draw- Belikov) 20 .i.xg6!? hxg6 21 hS leads to a draw after
21.....gS 22 J..b3 (G.Guseinov-Belikov, 21...gxh5 22 'lfg5+ �h7 23 ..xh5+ Wg7 24
Alushta 2001). Now Bclikov gives the 1t'g5+ Wh7 25 ll.act �) 20 .i.e4 Wds 21
following line as the most natural way to end ..xd8 ll.axd8 22 d5 (White should have a
the game: 22...• 51 23 1t'c3 .i.cS 24 :XeS puU here) 22....i.f5 23 .i.xfS gxfS 24 d6 ll.d7
..xbt+ 2s Wh2 W'fl 26 Lf7+1 Wxf7 27 zs Ads ll.rds 26 :tad1 Wg7 21 ll1d3 f6 2s
:fS+! �e611 28 ll.e5+ Wf71 29 .l:lfS+! with %lg3+ 1/z-1/z liviakov-Yusupov, Groningcn
perpetual check - a very attractive variation. 1994.
b) 14 h4 .i.e6 1S 9f3 ..xc3 16 .i.h6 14 . .ixh2+
..

ll.fd8!? 17 .i.gS (17 9f6?! .i.fB 18 .i.xfB 14.....xc3 15 .i.h6 ll.fe8 16 .f6 J..fB 17
..xd4! and Black is better) 17...:d7 18 .l:lfd1 .i.xf8 %lxf8 18 .i.e4 is preferable for White as
.i.f8 (18.....xd4 19 .i.e4 •es 20 g3 ll.c8 also Black cannor free himself easily. Black has
seems line) 19 ll.act •as 20 .i.f6 a6 21 .i.e4 compensation for the exchange after
.i.d5 22 "lfg4 .i.e6 23 •f4 J..d5 24 ll.c5 l5...•xd41? 16 J..xfB .i.c6 17 'ffh3 .l:lxf8 but
Wxa2 25 .i.eS with compensation for the he still has much to prove.
pawn, Nedev-Urban, Elista 1998. 1 5 Wh1 .*. d& 16 c4
c) 14 .i.h6 ll.fe8

16 .i.h6 Wd8!? 17 g3l:l.e8 18 %lfe1 :c8 is


with a further branch: less dangerous for Black.
c1) 15 llact .i.xh2+ 16 Wh1 and now 1 6 .Afe8
••

108
3 d4: The Main L ine

Maybe 16.....d81? 17 g3 .i.c6 18 d5 .i.d7 stiU give real chances of winning.


19 .i.b2 .l:lc8 is better. Most players would 59 ... llh5 60 d7 .i.xd7 61 llxd7 h2 62
prefer White, but is there reaDy an advantage? lldB+ ¢>b7 63 a6+ ci>xa6 64 llaB+ ¢>b7
17 c5 .i.fB 1 8 .i.e4 .i.e& 19 d5 .*.b5 20 65 lla1 h 1 'ir 66llxh1 llxh1 %-%
d6! 'irxc5 21 .*.d5 .*.g7 22 .i.a3 1rc3
After 22.....d4 23 .i.xf7+ Wh8 24 .i.xe8 GanJe46
.l:lxeB 25 .l:lad1 �4+ 26 Wgl .i.xfl 27 l:lxft Svidler-Yusupov
White's strong passed pawn promises him Bad Hombutg1998
w-cat prospects.
23 .i.xf7 + �hB 24 'irxc3 .i.xc3 2& .i.xeB 1 e4 e5 2 tl:lf3 lbf6 3 d4 lbxe4 4 .*.d3 d5
.i.xf1 26 llc1 llxeB 27 llxc3 5 lbxe5 lbd7 6 lbxd7 .bd7 7 0-0 .i.d6 8
c4 c6 9 cxd5 cxd5 10 'irh5
This alternative to 10 li:)c3 is perhaps
more dangerous for Black.
1 0...0-0
Black has alternatives, but they aU appear
to give White the edge.
a) 10...lbf6 11 l:lci+ �f8 (11....ie6? 12
l:lxe6+ Wf8 13 l:lxf6 and White wins) 12
�4 gives White a slight pull. Now a fantasy
line runs 12. � 13 lbc3 lbg4 14 .i.f4!?
..

.i.xf4 15 "ire?+ Wg8 16 ..xd7 .i.xh2+ 17


�h1 lDf6 18 lbxd5!lbxd7 19 lbxb6 lbxb6
20 Wxh2 and the endgame is wonderful for
This endgame is deeply unpleasant for White.
Black. b) 10.....f6 was suggt:sted by Igor Zaitscv.
27 ...lldB 28 llc7 �gB 29 llxb7 .*.c4 30
lte7 a6
30....i.xa2? 31 d7 .i.f7 32 .i.b4 wins in an
instant.
31 .i.b4 .i.b5 32 �2 lld7 33 lle5 �7
34 �3 .i.c4 35 a3 llb7 36 �4 .i.e& 37
J:la5 ll.b6 38 ¢>e4 llc6 39 .i.c5 �eB 40
'it'e5 ¢>d7 41 �6 .i.f5 42 f3 h5 43 .i.b4
J:lc2 44 llxa6 llxg2 45 lla7+ ¢>c8 46
J:lc7+ ¢>b8 47 .*.c5 llg3 48 f4 llf3 49
�e5 lld3 50 a4 h4 51 aS h3 52 llh7
.i.cB 53 .i.e7 +
53 l:lh8 h2 54 a6 is a simple win.
53 ...�a8 54 .i.d4 lla3 55 .i.b6 lld3 56 Now White has the following options:
�c5? b1) 11 1i'xd5 .i.c6 12 .ibS 0-0-0 13 �3
56 a6, intending 56 ...ixa6 57 d7, would
. (13 ..c4? .ixh2+! 14 Wxh2llkt6 gives Black
have won. the advantage) t3 .Lb5 14 'li'xb5 'li'xd4
..

56 ...llc3 57 .i.b4 llc4 58 .i.d2 llc5+ 59 and it is White who has to prove C(JUality.
ot>f&?l b2) 11 .i.e3 ..g6 12 'li'xd5 .ic6 looks like
59 We4 .l:lhS 60 l:lxh5 gxh5 61 Wf3 would enough compensation for the pawn to us.

109
The Petroff Defence

After 13 11t'a5 b6 14 'lra6 Black is of course ..i.c4 �6 17llxt2ibxc4 18 �xc4 •c8 with
fine, but now the simplest solution is some compensation) 16 ..f3 ..i.c6 17 .i.c4
14 ...llk51 15 .ixg6 hxg6 16 .if4!? (otherwise (here White should consider 17 i.xe4!? ll'te4
it's perpetual check) 16...t'tlxa6 17 i.xd6 18 ..d1 .i.d6 19 lbc3 Wh4 20 h3 ltxe3 21
0-0-0 18 .tg3 ltxd4. fxc3 .i.xg21 22 ..g4 1Wxg4 23 hxg4 .t.xft 24
c) 11 ltlc3 (the most challenging) Jlxfl with an edge - Y.Gonza.lez) 17...�6
t t ...11t'xd4 12 1Wxd5 ..xdS 13 lll xdS ltlcs 18 .tdS .txdS 19 'ifxdS � 20 �2 'irxb2
(after 13... 5 14 .lf4 .lxf4 15 t'tlxf4 0-0-0 16 21 llab1 11t'c2 22 llfct 1Wd3 23 ll!f3 ltac8
ltfe1 .lc6 17 lladl g6 18 f3 1l!f6 Black was and a draw was agreed in De Ia Paz-Y. Gon­
close to being level in Konguvel-Barua, San­ zalez, Santa Clara 2004.
gli 2000) 14 :C1+ llle6 15 .tgS 0-0 (Black is
not much worse after 15 ...h6 16 i.e3 .i.c6 17
lllf4 We7) 16 i.e7 .i.xe7 171llxe7+ Wh8 18
llad1 lL.dS 19 .le4 with a plus for White,
Svcshnikov-1. Zaitsev, Moscow 1991.

1 3...J.b4
Apart from 13...lbg5 (see the next game)
Black has a couple of enticing alternatives:
a) 13...ltc8?! is punished by 14 .txe41 Qess
strong is 14 ..i.M :Cs 15 li:kl2 .td7 16 1i'f3
1 1-.xdS 1i'h4 17 g3 ibxd2 18 .i.xd2 'lrxd4 19 i.c3
11 W k-ads to a draw after 11...g6 12 .g4 20 .xg4?1 120 1Wf6 ltxc3 21 .xc3 .i.c6
'irxdS .i.c6 13 �3 �xc3 14 bxc3 Lh2+ 15 22 f31 could give White a slight L-<ige; it is not
Wxh2 1Wh4+ 16 Wg1 .i.xg2 (lvanchuk). easy to tcU, but not too relevant either]
11 -*.c& 1 2 �5 g6
••• 20....i.xg4 with a draw, lvanchuk-Yusupov,
12.-1if6?1 13 llk3 ..xd4 14 .i.xe4 .t.xe4 Linares 1993) 14....i.xe4 15 llk3 lte8 (or
15 Ad1 .td3 16 'tth3 :t'd8 17 ltxd3 1Wxd3 t5.....i.f5 16 •o Wh4 17 g3 •xd4 18 Ad1
18 'lfxd3 .i.xh2+ 19 Wxh2 llxd3 20 i.e3 •c5 19 .i.e3 1Wc6 20 �5 with some advan­
gives White good winning chances, while tage) 16 .i.e3 (also strong is 16 .i.g51? ltxc3
13...g6 141th6 ltfe8 I S dS! (Yusupov) is also 17 Wh4! llh3 18 .i.xd8 .t.xh2+ 19 Wh1
good for White. However, 12. ..i.b4 is play­ l:lxh4 20 .i.xh4 .td6 21 .i.g5 with good win­
able, 13 .te3 :CS 14 a3 g6 t 51th3 transpos­ ning chances) 16....tf5 17 1Wf3 .i.b8 18 ltfe1
ing to the main game. bS 19 g4 (Iimman-Yusupov, Unarcs 1993)
1 3 trh3 and now following 19...b4 20 gxf5 bxc3 21
This move reaches the most critical posi­ bxc3 ltxc3 22 ltact llxcl 23 ltxc1 White
tion in the 7 ....td6 1ine. simply has an extra pawn.
t 31th6 Des 14 .i.e3 .tfB is an improve­ b) 13...1Wh6?! is probably not particularly
ment for Black: 15 1th3 .i.d7 (1 S...h5!? 16 good: 14W 1Wxd4 15 .i.xe4 ..i.xe4 16 11h4

1 10
3 d4: The Main Line

f5 17 Jlh6 llfe8?! (17...ltt7, with a slightly 17�


worse position, is a better option) 18 %lad1 17 .i.xe4 i.xc4 18 tbc3 .i.xc3 19 llxc3
(IK tbb5!? might offt:t a slight edge) 18...'.c5 :Xc3 20 bxc3 WeB 21 'irxc8 :XeS 22 i.d2
IIJ tbxe4 llxe4 20 1t'f6 .i.e5. Now the game i.d5 was nor particularly testing in Sad­
Jl;llac-Marciano, Cannes 1998 continued 21 vakasov-Kasimdzhanov, Lausanne 1999.
'1Vc6+? �h8 22 'lrd7 and here Black is fine 1 7 ._.i.b6 18 tOOl lDxf21

after 22..Wc6 23 b31le8. However, 21 lld8+ This seems to be the best option. Yusu­
l:lxJK 22 'irxd8+ Wt7 23 'IVJ7+ We7 24 pov docs not believe in Black's position after
'IVliS+ ..e6 25 'ifxb7+ We7 26 ..d5+ 'ire6 27 18...i.xd4?! 19 i.xe4 i.xe4 20 lldl i.xe3 21
'IVnH would have given Palac a laq,re advan­ 'irxe3 Wb6 22 'irxb6 axb6 23 lld4, and we
I;�Jt:. follow our bruru all the way here.
1 9 .bf2 ..txd4 20 lld1
The strongest move. Aftt:t 20 et)d 1??
:Ct+ 21 i.ft .i.b5 22 tbcJ .i.xf2+ 23 �xf2
1f'f6+1 Black has a decisive attack, while 20
Whl?! ..i.xf2+ 21 �f2 ..d4+ 22 Wft .i.d7
23 11'f31le3 is also unappealing.
20 .i.xf2+ 21 Wxf2 1tb6+ 22 Wf1
.•.

14 ..te3 .lleB 1 5 a3 .tas


15...Wa5?? is punished by 16 axb4 11'xa1
17 b5 .i.d5 18 tbc3, when Black must resign.
16 .llc 1 .Ilea
16...h5?! is weaker: 17 tbc3 ..i.d7 18 11'0
i.g4 19 'irf4 .i.xc3 (Svidler-Anand, Dort­
mund 1998) and here White can play 20 bxc3
g5 21 'irxe41lxe4 22 i.xe4 f5 23 i.xb7 llb8 22...Wxb2
241lab 1 f4 25 i.d2 with an edge (Anand). 22 .1le61? is another attractive possibility:
.

231rh4 l:lf6+ 24 'irxf6 (24 �e1 is met by the


convincing 24.....ta4! when White is in trou­
ble, e.g. 25lbxa4? 'irg1+ 26 �d2 'irxg2+ 27
i.e2 lld6+ with mate to follow) 24....i.xg2+
25 �xg2 'iVxf6 26 .i.e4 with a murky posi­
tion.
23 11rd2
Possible is 23 lbe2 llcd81? (23 ... .i.xg2+ 24
Wxg2 ltxe2+ 25 �h1 lie7 26 i.c4 favours
White) 24 %lab1 'iVe5 25 Wh3 'fff6+ 26 �g1
'1Ve5 when it's unclear whether Black is
worse here - for one thing ...llxd3 is a threat.
23._.11rxc3 %-%

111
The Petroff Defence

�xc2 �xe4 19 .i.h6 lle8 20 .*.b5 llt:S 2'1


Game 47 .*.d3 .i.c6 22llae1 llxe1 23llxe1 'ilff6 with
Macieja-Rozentalis equal chances, Blehm-Ribshrein, Pattas 1999.
Reno 1999 b3) 15 �h6

1 e4 e5 2 l0f3 �f6 3 d4 l0xe4 4 �d3 d5


5 �xe5 �7 6 �xd7 �xd7 7 0-0 .*.d& 8
c4 c6 9 cxd5 cxd5 10 11h5 0-0 1 1 •xd5
�c6 1 2 1ih5 g6 1 3 11h3 �51

with a further branch:


b31) 15....i.f4!? is an interesting option,
Black's point is that after 16 .i.xf8 'ilfxd4 17
1ih3lbg5 (17...11'xb2 18 �a3llxf8 19 .i.c4
l'l)gs 20 1Vg4 WeS is unclear) 18 Wh4 WxdJ
Time has shown that this is Black's best 19 11'xf4 .i.xg2 20 Wxg2 Wh3+ 21 �gt
path, after which he seems to be able to �f3+ 22 ...xf3 1fxf3 White has some prob·
count on equality. lems. Instead Holzke-Schandorff, Bundesliga
14 .ixg5 2001 continued 16 .txf4 1i'xd4 17 c&:3
Alternatively: 'lf'xf4 18 11'xf4 �xf4 1/a-1/a.
a) 14 'li'h6 �e7 15 .i.e3ltk6 16ltlc3 �f6 b32) 15...lle8 16 �3 with a final split
17 llad1 �xd4 18 �xd4 �xd4 19 l&4 b321) 16...�d4 17 ltad1 .i.eS 18 f4 f5 19
�xe4 20 .i.xe4 � 21 Dd2llad8 22llfd1 ..,3 (19 hf5?! Wb6 20 .i.xg6 �f5+ 21lltz
ltd7 is equal, Mikhalchi.<�hin-Pavasovic, Celje �xh6 22 'IVgS .tf61 23 '1Wxh6 hxg6 24 1Vxg6+
2003. More critical is 18 .i.e4!? �xe3 19 .tg7 and White has nowhere near enough for
11'xe3 'Wb6 20 .i.xc6 bxc6; this could be a the piece - Van dcr Sterren) 19....i.f6 20 .tg5
tiny advantage for White. �xg5 21 .tc4+ Wh8 22 fxgS Wb6 23 �ht
b) 14 1tg4 is White's main alternative. Af. 'illxb2 24llxd4 lle3! zs ll)dt 11'xd4 26 �e3
ter 14...� there are three options: lies 27 1ih6! 11'xc3 28 llxf5! with an equal
b1) 15 .i.e3 h5 16 'li'h3 .i.d7 17 •e .i.c6 position, Nijboer-Van der Srem:n, Rotterdam
(Black has equalised but now White goes 1999. The main point is 28 ..1fc1 + 29 lln.
.

crazy) 18 'lf'd1? 1fh4 19 h3? (although 19 g3 '1Vxc4 30 JlfB+ llxf8 31 'ilfxf8+ ..g8 32
1th3 20 f3 h4 is great for Black anyway) 1Wf6+ ifg7 33 1l'd8+ with a draw.
t9. ..lng5 20 d5 �h3+! 21 gxh3 •xh3 22 b322) Black can also try 16....i.f4 17 .Lf4
lte1 .i.h2+ 2.1 Wh1 .i.xdS+ 24 f3 .i.f4+ 25 11'xd4 18 .i.e4 f5 (or 18...�xf4 19 'ilfxf4 .i.xe4
Wg1 �xf3 0-1 Lautier-Gelfand, fo'JDE 20llfel .i.b1! 21 'IVct �f5 22 11'd1 1Vxd1 23
World Championship, Las Vegas 1999. llexd1 LiB 24 f3 �g7 1/z-112 Bologan·
b2) 15 �3 h5 16 'ii'h3 �xd4 looks fine Onischuk, Biel 1999) 19 'iVd1 (19 .txf5 leads
for Black, for example 17 �4 �2+ 18 to a draw after 19...�xf4 20 llad1 WeS 21 g3

1 12
3 d4: The Main Lin e

··� 22 Wl'xf5 gxf5 23 g:xf4 Wf7- Kaspamv)


UI,�·JCd I 20 llfxd1 fxe4 21 .i.e3 Wf7 22 b4
hi\ 11 n4 and White has a bit of pressure, Kas­
f'II'UV·Pikcr, Wijk aan Zee 1999.
lll., ••xg& 1 5lbc3 Aae8

This sacrifice leads t o a draw. Or a t least it


should...
21 �xh2 AM+ 22 �g1 -.6 23 g3
23 f3 .e3+ 24 llf2 �6 25 �fl .c3 26
*gt is quite a cute draw.
Ulnck has a reasonable alternative in 23 Ah1 + 24 �g2 Ah2+ 25 Wf31?
•••

I � lfcK, after which 16 lad1 is probably


... White wants to win, but maybe this is not
hu�1. J(, tiS!? has been played a few times, for in the cards. Safer is 25 �gl llht + with a
uurnplc 16 ....i.xd5 17 f4 •ds (17...•f6!? 18 draw by perpetual check.
f'tKIIS Wl'M+ 19 �h1 .xdS 20 lad1 .xa2 25...J:r.e8!
�1 (51? is very unclear- Van der Wiel) 18 Cutting off the king.
f)KtiS .i.cS+ 19 'iPht 'ilxdS 20 llact (Shirov- 26 ird7?
1\l•nmnik, 4th matchgame, Cazorla 1998) and After this mistake Black wins with a sensa­
lll'rc Black equalises after 20...Ile3 21 i.c4 tional attack. White should play 26 .1e41 f5
lcl4. The point is that following 22 .i.xf7+?! 27 We21 W'h5+ 28 �d2 fxe4 29 �ct 1i'f51 30
*117 23 ll.f3?! Wxf7 24 llxe3 9xe3 25 g4! (30 d6? llc8+ 31 �b l e3+ 32 �a1 e2 is
1Kh7+�f6 Black should win. winning for Black) 30.....f4� 31 Wbl llb8
11 lad1 Ae7 1 7 d5 -*.d7 18 � Axe4 32 'tie7 e31 with a very complicated position
1 I •xd7 AdS (Macieja/Rozentalis).
Another plan is 19..1Wc7
. 20 11xe7 llxe7 26.....,5+ 27 1Wg4 Axf2+ !
21 g3 lidS with an ending that should be
,Jmwn, l.uther-Delchev, Nova Gorica 2000.
ao •xb7
Black is fine after 20 11xd8+ 11xd8 21
1xc4 Wh4! 22 f4 .i.xf4 23 llxf4 ..xf4 24 d6
tfw11 25 .i.f3 1Wd7 26 �ht �fB 27 Lb7
�ticK the king acts as a blockader of the
-

l'"���ed d-pawn. Instead of 21...1i'h4, the game


I ull tcr-Aiterman, Recklinghausen 1998 con­
.

llnucJ 2t...1Wf6 22 g3 .xb2 23 llb'l •d4 24


An h6 25 llbct i.cS when we prefer White,
rvcn though lhe players agreed a draw.
IO .i.xh2+
...

1 13
The Petroff Defence

Macicja must have either missed or mis­ b1 ) 9 11'h4 .i.e6 (or 9 ...ll'lc4 10 'iVxd8+
evaluatl-d this sacrifice. llxd8 1 1 n- 1 1 llcl 0-0 1 2 .ixe4 dxe4 1 3
28 l:txf2 Wxd5+ 29 Wf4 1td4+ 30 ¢>95 llxe4 llfe8 1 4 llxc8+ llxe8 1 5 .i.e3 .if4 is
30 �11 :e3+ is winning for Black. equal- 1 t ...ll'lf6 12 .i.g5 c6 1 3ll'lc3 �f8 1 4
30 1txf21
••• llae1 h6 1 5 .i.xf6 gxf6 16ltld1 fj;}g7 and it is
No perpetual- Black is winning! not clear whether White has any advantage at
31 1tf4 all, 1iviakov-P.Nielsen, Bergen 2000) 10 ltlc3
White cannot escape, for example 31 �h4 .i.c7 1 1 lle1 (1 1 'irg3 0-0 1 2 ll'lb5 lC!cs 1 3
:e5 32 WeB+ �g7 33 'lfc3 f6 34 'lfc7+ �h6 .i.f4 c6 1 4ll'lc7lllxc7 1 5 .i.xc7 Wd7 is level
and there is no defence; or 31 �h6 'iVf6 32 - 1iviakov) 1 t.. 'it'd7
. 12 .i.ft a6 13 n ll'le4
1i'f4 :e5l!, winning instantly. 1 4 11h5 yfJ 1 5 'it'h6 .iffi 16 11'h4 .i.e7 17
31 1tc5+ 32 .tf5 1te7+ 33 �h6
•.• 11'h6 .if8 1/z-1/z Tiviakov-Lcko, Wijk aan Zee
Or 33 �g4 1i'e2+ and Black wins. 2001 .
33...f6 34 ..i.xg6 1tg7+ 0-1 b2) 91le1 + is a worthy alternative.

Game 48
Shirov-Kramnik
Be lgrade 1997

1 e4 e5 2 .!Df3 .!Df6 3 d4 .!Dxe4 4 ..i.d3 d5


5 .!Dxe5 .!Dd7 6 .!Dxd7 .txd7 7 0-0 ..i.d6 8
�c317

9 ...�f8 10 'iVe2 (10 11'h4 is harmless due


to 1 o ..ll'lg4! 1 1 'ifxd8+ llxd8 with equality:
.

1 2 h3lllf6 nll'ld2 g6 1 4ltln �g7 1 5ltle5


.i.c8 16 .igS h6 1 7 .i.f4 g5 18 .i.h21lhe8 1 9
lte2 lllgB 20 .Jt.et f 6 2 1 ll'ln Yz-Yz Movse­
sian-Akopian, FIDE World Championship,
Las Vegas 1 999) 10 ... c6 lt ll'ld2 tlc7 1 2ll'lf1
.i.g4 13 flc3 .ixn 14 'iVxn .ixh2+ 15 �ft
This move has been popularised by Shi­ .id6 16 .i.g5 '1t'd8 1 7 lte2 (an improvement
rov. 811'h5 has also been tested: over 1 7 :C3?! i!be41 1 8 .i.xc4 'irxg5 1 9 .i.d3
a) 8...c6 9 :e 1 0-0 1 0 .i.xe4 dxe4 1 1 D.xe4 g6 20 llael !J;g7 21 'it'h3 llad8 22 c3 h5 and
:C8 121lxe8+ (not 1 2 tlkl2?! 'W'f6 13 'iVe2 c5 White didn't have quite enough for his pawn,
14 c3 cxd4 15 cxd4 :ac8 when Black's de­ Sutovsky-David, Istanbul 2003) 17 ... h6 1 8
velopment ensures him of an initiative) .ih4 g5 1 9 .ig3 with chances for both sides.
12.. .'iVxe813 .i.e3'it'e4 1 4'it'd1 .i.g4 1 5 Wet However, we cannot sec how Black can ever
.i.f5 16 c4 1Wh4 1 7 g3 11'h5 and Dlack has be worse here.
good compensation for the pawn, :>Ge Jun­ 8 .!Dxc3
•••

Karpov, Buenos Aires 2001 . 8. ..'it'h4 has also been tried, e.g. 9 g3 i!bxc3
b) 8...lnf6 and now: to bxc3 tlg4 (10 ...11'h3 1 1 llc 1+ �fB 12 'iVn

1 14
3 d4: The Main Lin e

l:6 13 .i.f4 Jl.e7 14 J:lab1 b6 15 c4 favours 1Wb3 J:lad8 1 6 .i.d2 'if g6 with compensation
White) 1 t J:lcl+ *d8 (11...*£8 12 .i.e2 9£5 - Mikhalchishin) 14....J:lae8 15 J:lxe8 .i.xe8 16
1.� J:lb1 J:lb8 14 c4 dxc4 15 .i.xc4 h5 16 h4 is lle1 .i.h5 17 9g.l (17 'ifxh5 .i.xf4 18 9xf5
also a bit better for White) 12 .i.e2 95 13 9xf5 19 .i.x£5 .id2 is drawing) 17..Jle8 18
Ab1 b6 14 c4 dxc4 15 .i.xc4 J:le8 16 .i.e3 J:.xe8+ .i.xc8 19 Jl.xd6 cxd6 20 11'0 g6 21
.i.c6 17 d5 .i.d7 18 .i.ft h6 t 9 c4 with a 9xd5 9e7 and Black has play for the pawn.
sli�ht plus for White, Kamsky-Karpov, Pavlovic-Mikhalchishin., l..enk 1999.
FIDE World Championship (Game 6), Elista 11 .•• b6
1996.
9 bxc3 0-0
A recent game in this line continued
IJ ...Jl.c6 10 llb1 :bs 11 9h5 c6 (1t...g6 12
1irh6 *d7 13 .i.gS 9£8 14 1Vh4 h6 15 .i.f6
.i.c7 16 llfe1 looks slightly better for White)
12 .i.g5 .i.c7 13 .i.xe7 9xc7 14 f4 'iff6 15
l:.bc1 *d8 16 g4 g6 17 9h6

1 2 'irf3
Again White has other options:
a) 12 c4 dxc4 13 Jl.xc4+ *h8 14 ..igS
9e8 15 'iVh4 1i'g6 16 .i.d3 J:lae8 17 c4 h6 18
.i.e3 9f7 was level in Palac-Fridman, Pula
1997.
b) 12 :C1 c6 (this is even better than
12...1i'f6 13 •o c6 14 Jl.f4 b5 15 11fg3 .i.xf4
and now: 16 1fxf4 J:lf7 17 *rt l:ta£8 18 J:lc2 ..d8 19
a) 17...*c7?! 18 f5 ..i.d7 19 fxg6 'ifd6 20 J:lbet � when Black is close to equality,
l:txf7 hxg6 21 'iff4 'if'xf4 22 l:txf4 and White Shirov-Anand, Groningen 1997) 13 .i.gS
cvcntunUy converted his pawn in Shirov­ ..c7 14 c4 .i.e8 15 9h3 dxc4 16 .i.xc4+ (16
Bologan, Sarajevo 2004. .i.x£5 .i.xh2+ 17 *h1 J:lx£5 18 9xf5 .i.d6
b) Shirov described the position after 17 gives Black sufficient compensation - Salov)
'irh6 as the most fascinating he had ever 16....i.f7 17 .i.e6 Le6 18 l:txe6 J:lae8 with
analysed. The foUowing variation is a sum­ level prospecrs, Shirov-Kramnik, 2nd match­
mery of several page of his analysh.-: game, Cazorla 1998.
17 .. .i.xg4 18 51 *d7 19 h3 .i.h5 20 fxg6
. 1 2...c6 13 ..tf4 •c7 14 ..txd6 'irxd6 1 5
.if3! 21 J:le5! hxg6 22 J:.£5! 9xf5 23 .i.x£5+ Afe1 l:lae8
�-,rxfS 24 1i'f6 J:.bg8+ 25 *f2 .i.e4 26 1fxf7+ Black has completely equalised. In fact, it
�c8 27 9e6+ *c7 28 9e7+ �c8 29 h4 and is White who should be careful. For example,
White is a bit better. 16 c4 J:.xel+ 17 :Xe1 9b4 and White is
10 Wh5 f5 1 1 lZb1 alrt."lldy starting to encounter problems.
11 J:lc1 is interesting: 11...1i'f6 12 •o 1 6 AxeS :XeS 1 7 c4 •a31 18 h4?1
WhK 13 :b1 b6 14 ..if4 (or 14 'ifxdS .i.c6 15 After this move Black is probably slightly

1 15
The Petroff Defence

better. Possibly stronger is 18 1ff4 1fxa2 19 a) . 8 �2 lllxd2 9 1Wxd2 .i.d6 (not


1fc1 1Wa5 20 c3 bS 21 lla1 flc7 22 cxdS 9...1Wxd4?! 10 l:le1+ .i.e7 11 WaSt? b6 12
cxdS 23 11'c2 g6 24 ..-.,3 .i.c6 25 1tb4 when 'Wa6 with compensation - Yusupov) 10
White has good compensation for the pawn. 11'e2+ (after 10 :C1+? Wf8! Black wins a
18 11'xa2 19 llf1 dxc4 20 -*.xfS -*.xfS
••• pawn, 11 f4? losing to I t....i.xf4Q 10....i.e6
21 Wxf5 c3 11 f4 0-0 12 g3 1Vf6 13 f5 .i.d7 14 .i.e3 l:lfe8
21...a..'i 22 1Vd7 llf8 23 :C1 gives White and Black is no worse.
enough compensation. One point is that b) Blllc3 is also not very strong: 8...lllxc3
after 23 ..flxc2 24 1We6+ Black must acqui­
. 9 bxd 0-0-0 10 'itfJ i&.e6 1 1 .i.f4 (what
esce to a draw as 24...�h8?? 25 "lrf71 :gB 26 else?) 11....i.g4 12 'ite3 .i.d6 13 .i.xd6 llxd6
:CB wins for White. 14 llaet .i.d7 15 'itfJ 'Wf6 16 'Wg3 (or 16
22 Wd3 WdS 23 •xc3 lle4 24 Wg3 1Wxf6 llxf6 1 7 c4 dxc4 18 .i.xc4 .i.fS 19 .i.b3
llxd4 25 Wb8+ Wd8 26 Wxa7 h& l:ld8 20 dS c6 with a likely draw) 1 6..Jlc6 17
26 .:Xh4 27 11t'b7 'Wd6 28 "IreS+ *f7 ztJ
.. :eS!? l:r.xc3 18 .J:hdS .i.e6 19 llg5 g6 20 dS
1ff5+ is also a draw, but Kramnik could have .i.d7 21 'ife3 lla3 with unck·ar play,
continued playing for a win with 26...:C4 27 Klovans-Raetsky, Apolda, 1994.
1Vb7 h5 28 lla1 �h7. Note, however, that c) 8 g3
27...:Xc2?? is out of the question due to 28
llat.
27 Wa2+ �8 28 g3 bS 29 We& 1t'd5
30 1t'xd5 llxdS 31 lla1 llc6 32 lla2 �7
33 �1 %-%

Game 49
Anand-lvanchuk
Linares 1993

1 e4 eS 2 .!Df3 .!Df& 3 d4 tt.'\xe4 4 ..i.d3 d5


5 lLixe5 lLid7 6 lLixd7 ..i.xd7 7 0-0 1i'h4

and now:
cl) 8...'ifh3? looks natural but White can
play 9 f3 llld6 (9...lllf6?! is even worse after
10 llet+ .i.e6 11 :CS! gS 12 .i.f5 g4 13 .tgs
.i.e7 14lllc3!, when Black is in trouble, Pav­
lovic-Konguvel, Bcnasque 1998) 10 :CI+
i&.c7 11 lleS .i.c6 12 g4 ..h4 13 i&.gS, win­
ning the queen fur insufficient compensa­
tion.
c2) R flf6! 9 .i.xe4 dxe4 10 lllc3 1fg6 11
..

lllxe4 0-0-0 12 :C:1 hS and the weakening of


the light squares guarantees Black compensa­
This is the 'active' main line. tion for the pawn.
8 c4 8 0-0-0
•••

This is the principal move. The alterna­ Other moves arc weaker: �

tives fail to impress: a) 8...i&.d6 9 g3 ..h3 10 lllc3 lllxc3 11

1 16
3 d4: The Main L in e

hxd dxc4 12 ltc1+ Wf8 13 i.xc4 i.g4 14 the attack i n Rublcvsky-Ngucn Anh Dung,
'it'll3 should favour White. Moscow 2001) 16 l:l£2 hS 17 l:lb2 h4 with
b) 8 ...lt�f6 can be met by 9 llkl21?, for reasonable countcrplay.
example 9.....xd4 10 .!DO 'irb6 11 cxdS il.e7 9...g5 10�3
(11....!Dxd5 12 o!Dgsl h6 13 .!Dxf7 Wxf7 14 Other options will be discussed in Games
'it'h5+ WgH 15 ..xd5+ i.e6 16 ..e4 is great 51-52.
for White) 12 l:lcl 'i'd6 (12....!Dxd5 13 .ic4 1 0....ig7
c6 14 .i.g5 gives White a very strong attack)
13 i.gS h6 14 i.h4 'ifxd5 15 l:lct i.c6 16
'it'c2 'ild8? (Z.Almasi-Cs.Horvath, Hungary
1997) and now 17 l:lxc6! bxc6 18 .i.xf6 gxf6
19 o!Dd4 is a nice win. 16.. .'1Ve6 is stronger,
hut Black's position is still unenviable.
9 c5
9 cxdS is less critical. Black continues
IJ...i.d6 10 g3 (10 h3? .i.xh3! is simply too
dnngerous!) 10... .!Dxg3! 11 fxg3 i.xg3 12 'iVc2
'it'xd4+ 13 Wg2 i.xh2 14 Wxh2 9h4+ 15
'wt>gl 'iVg3+ 16 'ilg2 Wxd3 with great coun­
tcrplay for the piece.
9 o!Dc3 is also not to be feared. The best reply. The alternatives have not
been successful:
a) 10....!Df6?1 11 g3!? 'iVh3 (1t ...'iVxd4?1 12
c61 is a well-known trick) 12 'ilf3 ltlg4 13
Wg2 'i'xg2+ 14 Wxg2 h6 15 f41 .i.g7 (or
15...l:e8 '1 6 fxg5 hxg5 17 h3 o!De3+ 18 .i.xe3
:Xe3 19 i.f5 with a clear edge) 16 h3 .!Df6
17 fxg5 hxgS 18 g4 and White has a cll:ar
advantage, Tseshkovsky-Gagloshvili, Kras­
nodar 1997.
b) 10...f5?! 11 .!Dxd5 and now:
b1) 11....!Dxf2l? 12 l:lx£2 'lfxd4 13 .i.xg5
(we think that White should play 13 c6!?
..xdS 14 cxd7+ l:lxd7 15 l:ld2 .i.b4 16 .i.c4
Black can pL1y the following: ..c5+ 17 Wh1 i.xd2 18 i.e6 with a clear
a) 9....!Df6 10 cS .i.e7 11 .ie3 .!Dg4 12 .if4 edge) 13 ...i.xc5 14 �3 (14 WOI? i.c6 15
looks good for White. i.xd8 l:lxd8 16 i.xfS+ Wb8 17 lltll 'it'x£2+
b) 9...i.d6 10 g3 .!Dxg3!? 11 fxg3 i.xg3 12 18 'ifx£2 i.x£2+ 19 Wx£2 i.xdS 20 .i.xh7
'irc.J2 'iVxd4+ 13 ..e3 i.xh2+ 14 Wxh2 'iVh4+ looks great for White) 14...f4 15 i.xf4 l:lhf8
15 c;t>g1 ..g4+ 16 W£2 d4 17 'lfg3 dxc3 18 16 Wft ..xe3 17 i.xc3 .i.xe3 18 ..e2 i.x£2+
bxc3 i.e6 is equal according to Yusupov and 19 �ht ltde8 20 ..c2 and White had decent
lliibner, but to f4 may be better for White. chances to win the endgame, Wetlberg­
c) 9....!Dxc3 (the easiest) to bxc3 dxc4 1 1 Schneider, Torshavn, 1987. Even so, White
�xc4 i.d6 12 f4 f 6 13 .i.dS c6 14 g3 Wg415 should punish Black further with deviations
.if3 ..g61? (but not 15...'iVe6 16 'iVa4 a6 17 on move 13 or 14.
J:tbt hS 18 .i.d2 1t'f5 19 l:lb2 and White had b2) tt...i.c6 also does not work after 12

117
The Petro ff Defence

ltk3l:lxd4 (or 12...i.g7 13 g31i'h6 14 olllxe4 unclear) 20 1ib6 9e7 21 -'.&6+ Wa8 22'
fxe4 15 .i.xe4 l:lxd4 16 1i'f.3 and Black is in 9xc6+ Wb8 23 'ifh6+ Wa8 and now taking;
trouble) 13 .i.c3l:ld8 14 'Vc2 olllxcS 15 Lf5 the draw by perpetual check with 24 1i'c6+
.i.d6 16 g3 1fc4 17 .i.xgS l:ldfB 18 .i.xe6+ might make sense as after 24 Lc4 24...-.,7
o!ilxe6 19 .i.c3 and Black is simply a pawn the game could go either way.
down, Glck-Raetsky, ZeD am Ziller 1993. b) 13... 15 14 ..tg2 with a further branch:
1 1 g3 b1) 14....i.c6 15 .i.xc6 (15 d5 .i.xd5 16
This is the most critical line for Black. ..txdS c6 17 'ifa4l:lxdS 18 11xa7 f4 is unclear
Other options arc featured in the next game. according to Vladimirov - White has prob- ·

terns developing) 15...1i'xc6 16 .i.xgS l:lxd4 •


17 •c2 l:ld5 18 .ie3 gave White some ad­
vantage in Macieja-Stefanova, Krynica 1998.
b2) 14...f4 15 d5 (after 15 gxf4 gxf4 16,
•o ..tc6 17 1fg4+ �b8 18 dS- 18 ..txf4?
l:lxd4! would be a nasty surprise- 18....i.xd5i
19 .i.xf4 9f6 Black has exceUent compensa­
tion for the pawn) IS ...llhfB 16 l:le1 �b8 17
d6 cxd6 (Arencibia-Vladimirov, Leon 1991)
and now after 18 11fxd6+ •xd6 19 cxd6 .i.d4
20 gxf4 gxf4 21 .lle4 l:lde8 Black can be·
happy with his counterplay.
1 2 �xe4
11 1rh3
.•. The older line is 12 olllxdS l:lhe8
Another possibility is 11...'ifh6!? 12 olllxe4 (12....i.g4? 13 .ixe41 .i.xd1 14 .i.f5+! and
(12 o!ilxd5?! .llhc8 13 o!ilc3 o!ilxf21 14 l:lxf2 White wins- 1-J.Olafsson) 13 Wf3 (13 .i.xe4?.
.i.xd4 with a clear edge for Black- Yusupov) l:lxc4 14 .i.xgS ..tc61 1 5 .ixd8 .i.xdS 16 f3
12...dxc4 13 .i.xe4 l:lxd4 17 'lfct l:ld3 gives Black a decisive
attack) 13.....tl5 14 �3 (14 o!ilc3?! l:lxd4 15
lle1 -or 15 .ixe4?Il:ldxc4 16 ..txgS ..txc3
17 bxc3 l:le2 and Black has aU the threats-
15....i.g4! 16 11fxt7 l:lffi 17 WxfB+ .i.xffi 18
..tn 11'hs 19 olllxe4 .i.l5 and Black's material,
gains give him the advantage) 14....i.g6

and here Black has two choices:


a) 13.....th3 14 'iff.3!? .i.xfl 15 ..txb7+
�b8 16 .i.xgS!? (16 Wxft -'.xd4 17 h4 'ife6
18 .i.xgS f6 19 .i.f4 is very messy) 16 ...Wxg5
17 Wb3 .i.c4 18 Wb4 aS 19 'ifxaS c6
(19...�b7 20 c6+ �b8 21 1lxg5 .i.xd4 also

1 18
3 d4: The Main L in e

and now: LcS i s equal - Howell) and here Black must


a) 1 5 c6 is risky: 1 5 .. .:xd4 1 6 .i.c2 hS 1 7 make a choice:
c:xh7+ 'itlb8 1 8 Act '1Vd7 1 9 lLlfl g4 20 1i'g2 b 1) The dubious 1 8 ...Ad7?1 is met by 1 9
ilkS 21 i.e3 .i.t.-4 22 f3 .i.xb7 and it was •a4:
Ulack who had the attack, Smagin­
H.Oiafsson, Sochi 1988.
b) 15 d5 lLld2 (also good is 15...lLlxc51? 16
.i.fS+ .i.xf5 17 lL!xf5 .i.f6, witl1 chances for
both sides) 16 .i.xd2 .i.xd3 17 '1Vxf7 .i.xb2
(17 ... i.xf1 ? 18 ..xg7 Axc3 19 :Xfl Ad3 20
.i.xg5 is better for White) 1 8 d6 1i'd7 1 9
1Wb3 'ilbs with unclear play, Joscliani­
l lowell, Spijknisse 1989.
12 ...dxe4 13 .ixe4 .ib5 14 .ig2!

bt l) 19 ...'itlb8 20 b4 W'd3 21 ..tf4+ WaS


22 .i.d6 (this looks great for White so Black's
reaction Lo; understandable) 22....ZZ.xd6 23 cxd6
1i'xd6 24 bS cxbS? (24... c5 25 ...c4 flc7 26
Act with a clear edge - Howell) 25 '8'xb5
llb8 26 ltb1 'irc7 27 a4 i.c3 28 ltct flc5 29
1ib3 .i.d4 30 .i.xb7+! 1 -0 Gcller-J.Howcll,
Reykjavik 1 990.
b12) 19 ... a6 20 .i.xc61 bxc6 2 1 1Wxc6+
wds (21 ...�8?! 22 Wb6+ l:tb7 23 1i'd6+
Anand's improvement over a previous �c8 24 1i'xa6 �b8 25 c6 and White wins;
).,'llme against lvanchuk, albeit with colours 21 ...ltc7?! 22 1i'xa6+ �b8 23 c6 ltc8 24 ltd1
n:versetl. With 1 4 .i.g2 White obvious does 1ic2 25 lld3 ltxc6 26 'iVa7+ �c8 27 1i'd7+
not worry about losing the exchange; the 1-0 Hracek-Haba, Czech Republic 1998) 22
attack is much more important, and Black lte 1 'irf6 23 'ira4 1i'xb2 24 c6 .i.c3 25 cxd7
hsls no way to force exchanges. .i.xe1 26 1Wc6 .i.xf2+ 27 .i.xf2 W'b1+ 28
The alternative is 1 4 .i.xg5 :Xd4 and Wg2 'Wb2 29 �f3 with a clear plus for White
now: - Hracek.
a) 1 5 'Wb3 Axc4 1 6 1i'xb5 h6 17 .i.e3 b2) 18 ...:hd81 1 9 h4 (or 19 'ira3 Ddt 20
l:h4! 18 Afd1 1txh2+ 19 'itlfl 'irh3+ 20 �el 1ixa7 ltxft+ 21 .i.xfl .i.c5 22 1i'a8+ �c7 23
l:e4 (20...l:th5 21 1Wa4 a6 22 c6 Wb8 is un­ 1Wa5+ with a draw - Yusupov) 1 9... .i.e5 20
clear) 21 c6 Axe3+ 22 fxe3 'irxg3+ 23 �e2 .i.xd4 .i.xd4 21 �h2 �c7 22 flc4 .i.xcS 23
'irg2+ 24 We I ...g3+ 1/z-'12 Dolmatov­ b4 .i.b6 24 b5 ltd4 25 Wb3 ltd3 26 'iVc4
Akopian 1988. lld4 and Black escaped with a draw, lvan­
b) 1 5 i.g2 'irf5 (or 15 ...'irc6 16 1Wf3 .i.c6 chuk-Anand, Roquebrune 1992.
1 7 1Wb3 .i.xg2 18 Wxg2 ...dS+ 1 9 1Wxd5 14 ...1rf5 1 5 .ie3
l:xd5 20 .i.d .i.xb2 21 Aab1 .i.g7 and Black 1 5 'irb3 c6 1 6 ltd l l:txd4 17 ltxd4 .i.xd4
i� only very slightly worse) 1 6 1i'b3 c6 1 7 18 .i.c3 is equal - Akopian.
i.c3 .i.xf1 1 8 llxfl (1 8 .i.xd4 .i.xd4 1 9 Lfl 1 5 ....ixf1 1 6 .ixf1 Zlhe8

1 19
The Petroff Defence

1 6...c6 1 7 Wa4 looks good for White: game via 22...1lxJ4! 23 .i.xd4 1Wxd4 24 Wb3
1 7... h5?! 1 8 1i'xa7 llxd4 19 llel l lld7 :C7 25 .i.fl 1i'xc5, with chances fur both
(19...1ldd8 20 .i.a6! and winsQ 20 b4 1Wg4 21 sides.
b5 and White has a very strong attack. 22 .hd4?
...

1 7 1ra4 ¢'b8 18 l:ld1 c6 More resilient is 22...1le7 23 llb4!? f5 24


1 8. .1lc6 19 d5!? llxd5 20 .i.h3! 1i'e4 21
. d5 f4 25 d6 lled7 26 .i.xa6, when Black is
:Xd5 1i'xd5 22 .i.xe6 1Wxe6 23 c6 also a..__ still just about breathing.
surcs White of an edge. 23 l:lxb7+1 �xb7 24 •xa6+ �8 25·
1 9 l:ld3 1re4 20 l:la3 a6 21 .i.d3 1tb6+ �8 26 •xc6+ �b8 27 'Wb6+
�a8 28 .i.b5 1 -0

Game 50
J.Howeii-Makarychev
Fnmze 1989

1 e4 e5 2 li)f3 ll:)f6 3 d4 .!Oxe4 4 .i.d3 d5


5 .!Oxe5 .!Od7 6 li)xd7 .txd7 7 0-0 1ih4 8
c4 0-0-0 9 c5 g5 10 lilc3 .i.g7 1 1 .!082

21 ...1rg4?1
Soon after this game Black's play was re­
paired to some extent with 21 ...1Wd5 22
.i.xa6 .i.xd4 23 .i.xb7 .i.xc5! 24 .i.xc6 .dl +
25 Wg2 Wxa4 26 llxa4 llxe3 27 fxe3 lld2+
28 �f3 llxb2 (Hemandez-J.HoweU, Matan·
zas 1 993). This position should lx: within the

'drawing range', although it will take Black


some time and technique.
22 l:lb3! 1 1 �xd5 .i.xd4 12 i.e3 .i.xc3 1 3 fxe3
docs not seem frightening for Black after
1 3 ..ie6 (but not 1 3...�xc5?! 1 4 .c2 .i.e6
1 5 �xc7 Wxc7 1 6 11Vxc5+ �b8 17 llac1 with
attacking chances for White) 1 4 �7+ �b8
1 5 �f5 .i.xf5 16 llxf5 llhc8 when Black has
good counterplay. 1 1 i..c3 allows the trick
1 1 ...�xc5!?, when 12 �xd5 �xd3 1 3 1i'xd3
Wb8 looks equal.
1 1 1lhe8
...

1 1 ... f5 is an attractive alternative. Follow­


ing 1 2 f3 Black has a choice:
a) 12...�f6 1 3 i.e3 f4 14 i.. f2 11Vh6 1 5
1rd2 l:lhe8 1 6 llacl �>8?! 17 b4 �7 1 8 b5
22 i..xa6?! aUows Black back into the (Black has clearly wasted his chance for an

120
3 d4; Th e Main L ine

1111�\ck) 18 ...�b8 19 ru ll!f5 20 c6 .i.c8 21 13 'ii'e t 'ii'xet 14 l:lxe1 h6 1 5 .i.d2 lllg8 is


cxh7 �xb7 22 .i.xf5 .i.xf5 23 lba4 with a equal. Now after 16 l:ladt ?l liJc7 17 .i.e3
clc:tr plus, Tiviakov-Raetsky, Makhachkala lbf5 18 ..tf2 �3 1 9 l:lct l:te7 20 b4 l%de8
IIJK7. Instead of 16 ...lbg8?!, Black needs to Black was already better in Vidarsson­
RCI quickly with 1 6 ...g41? 17 1rxf4 gxf3 1 8 Ractsky, l-lafnarfjonlur 19%.
gx £3 .i.h3 1 9 l:lfd1 lbh5 20 1rxh6 .i.xh6,
IIKsuring him of reasonable counterplay.
b) 1 2...l:lhf8

1 3 .llxe21
•.

An idea of Akopian's.
14 'it'xe2
and now it's White who has to make a de­ 14 .i.xe2 is strongly met by 14...�g41
ci:;ion:
b1) 1 3 fxe4?! (this reckless move hasn't
been tried) 13 ... fxe4 14 l:lxf8 (not 14 .i.c2?
J:lxfl+ 1 5 1rxf1 l:tf8 and the game is already
over) 1 4 ...l:txf8 1 5 .i.e3 exd3 16 1Vxd3 l:tf6
and Black has the advantage.
b2) After 1 3 'ii'e 1 Black should exchange
with 1 3 ...1Vxet. Instead 13 ...'ii'h5?1 proved
futile after 14 fxe4 dxe4 1 5 .i.c4 f4 1 6 d5
'it>b8 17 c6 .i.c8 1 8 cxb7 .i.g4 t 9 lbxf4!? gxf4
20 .i.xf4 and White was virtually winning,
Mannion-Kobese, Yerevan Olympiad 1996.
b3) 1 3 a4 l:tde8 with a final split:
b31) 14 g3 lllxg31? 1 5 hxg3 (15 lbxg3 and now:
..ixd4+ 16 �h 1 f4 1 7 lbe2 .i.h3 gives Black a a) 1 5 fxg4? ..i.xd4+ 1 6 �hl .i.e5 17 .i.f4
good attack) 1 5...'ii'h3 1 6 c6! (16 l:lf2?! .i.xd4! (even worse is 1 7 h3? 1lg3 18 �g1 'ifh2+ 1 9
17 lllxd4 1Vxg3+ 18 Wfl f4 is very bad for � f2 .i.g3+! 20 �e3 - 2 0 W f3 .i.h4! -
White) 16....i.xc6 17 l:tf2 .i.xd4 1 8 lbxd4 20...1le8+ 21 �d3 .i.bS+ and Black wins)
'irxg3+ 19 �fl 'ii'h3+ with a draw. 1 7 ....i.xf4 1 8 l:lxf4 gxf4 19 1rxd5 l:lcHI 20
b32) 1 4 .et 'ii'xet 1 5 Axel f41 16 fxe4 •d 1 .i.c6 21 1i'f1 :C3 22 :C1 1re7 23 .to
tlxc4 1 7 .i.c4 f3 18 .i.e3 fxe2 1 9 l:lxe2 c6 20 .i.xf3 24 gx£3 lle2 and Black had an edge in
dS cxd5 21 .i.xd5 h6 22 l:td2 l:td8 with an Sherzer-Halasz, Budapest 1990.
level ending, Sax-Salov, Brussels 1 988. b) 15 .i.f4 lbf2!? 16 .i.g3?! (16 J..xg5
1 2 f3 li)f6 1 3 .td2 lbh3+! 1 7 gxh3 1Vxg5+ 18 �h1 1Vf4 with

121
The Petroff Defence

unclear play is stronger) 16...llhd1 17 ..i.xh4 lle8. Although Black is better, the game is
..i.xd4+ 18 �ht lt)e3 (we prefer Black's posi­ still open.
tion after the stronger 18...lbxb2!? 1 9 ..i.xgS 21 .. Jlf8 22 ..txh7?1
l:le8) 19 ..i.xgS l:le8 20 ..i.xc:3 llxe3 21 l:lfe1 Now it is all over. 22 c6 bxc6 23 llee1
.i.xb2 22 ..i.fl .id4 with an unclear endgame, .ie6 was the last chance.
Movsesian-Raetsky, Pardubicc; 1 992. 22.....te6 23 Ae1 ..te3 24 g4 AhB 25
.i.f5 �dB 26 Le6 fxe& 27 ..txe6 Lh2
28 Ad1 Ah1 + 0-1
White resigned because of 29 �e2 l:lxd1
30 Wxd1 c6 31 gS ..i.xc5.

Game 51
Kasparov-lvanchuk
Debrecen 1992

1 e4 e5 2 lLlf3 lbf6 3 d4 lbxe4 4 ..td3 d5


5 lbxe5 lbd7 6 lbxd7 .i.xd7 7 0-0 Wh4 8
c4 0-0-0 9 c5 g5

1 4...c!bh51?
14...Wxd4+ 1 5 .te3 WeS 16 Wd2 is !>imply
great for White.
1 5 1if2 •xf2+ 1 6 �xf2
1 6 :Xt2 ltlf4 17 .i.xf4 gxf4 1 8 c6 .i.c6 1 9
cxb7+ Wb8 with unclear play (Yusupov)
is probably a better option.
1 6. . .1i�f4 1 7 ..txf4 gxf4 18 Afe1 .txd4+
1 9 �1 ..txb2
19 ... ..i.xc5 20 J:leS ..i.e6 21 J:lhS allows
counterplay.
20 Aab1 ..td4
1 0 ..te3
White has a couple of alternatives here:
a) 10 g3 Wh3 11 lbc3 t5!? 12 .ie2 l:Zg8 13
..i.f3 .ie8 14 a4 �6 15 lla3 1i'f6 16 b4 ..i.g7
17 lbe2 hS with chances to both sides,
Pilipovic-Nikcevic, Tivat 1995.
b) 10 lbd2 lbxd2 (also possible is
lO ... i.g7 11 ltlf3 �5 12 lbcs Wxd1 13
l:lxdt .i.e6 14 i.c2 f6 1S ltl.n t5 16 lbe5 f4
when Black is no worse) 11 i.xd2 llg8 12
.i.c3 (or 12 l:lct !? Wxd4 13 ..i.c3 1i'h4 14
..tf6 l:le8 15 f4 i.e7 16 c6!? with unclear
play, Pinkas-Kuczynski, Wroclaw 1987; one
21 Ae7?! continuation is 16...i.xc6 17 .it5+ Wb8 18
More resilient is 2 1 l:lect .ic3 22 l:lc2 llxc6! bxc6 19 1ib3+ �a8 20 i.d7 .icS+ 21

122
3 d4: The Main L in e

"'h I .i.b6 22 .i.xc6+ 'iPb8 23 .i.xe8 llxcB


wilh a messy position) 1 2...g4 1 3 b4 Wb8 1 4
b 5 1ff6 1 S .i.d2!? ..i f5 (1 S...1i'xd4 1 6 .i.c3
l'c5 17 1fa4 gives White good compensa­
tlon) 1 6 .i.xfS 11fxf5 17 a4 11fd3 1 8 .i.f4
I'Kd1 1 9 llfxd1 llg6 20 aS WeB with an
c;umplicated struggle, Komeev-Raetsky,
Cannes 1 994.

and now:
ct) n...Whs t4 lbbs Wb8 1 s 11'et !? g4 1 6
'ireS 'irxeS (Black should play 1 6....i.xbS 1 7
1i'xhS lbxhS 1 8 .i.xbS gx f3 1 9 gx f3 .i.e7 20
:ret ..if6 with a slighdy worse position) 1 7
dxeS lbhS 1 8 fxg4 .i.xg4 (Khait-Raetsky,
Upetsk 1 993) and now after 19 c61? bxc6 20
lbd4 Wb7 21 lbb3 Black's position looks
10 Ae8
... uncomfortable.
Black should probably look carefuUy at c2) 1 3...'1'h61? is stronger: 14 l:tet?! g4
rhc alternatives to this: (Black now has the initiative) 1 5 c6 .i.xc6 1 6
a) 10....i.g7 1 1 f3 lbf6 transposes to 10 f3, .i. f5+ Wb8 1 7 fxg4 .i.d6 1 8 •o 1i'g7 1 9
while I t ...lbxcS?! 1 2 dxcS .i.xb2 1 3 lbd2 l:te2?1 ( 19 .i.h4 is better, although Black's
.1Ka1 14 'lrxa1 is fantastic for White. position remains preferable after 1 9 .:d£B 20
..

b) 10...lbf6 is certainly a playable altcma­ gS lbd7) 1 9 ... hS 20 gxhS J:lxhS 21 .i.h3 l:tdh8
rive: 1 1 lbd2 J:lgs (the main point) 1 2 lbf3 22 Wf1 &4 and Black was close to winning
WhS 13 lDeS 11fxd1 1 4 :axd1 .i.e6 l S f4 in Werner-Raetsky, Cappelle Ia Grande 1 999.
gxf4 16 .i.xf4 (1 6 :xf4 .i.h6 1 7 :o .i.xe3+ Instead of 14 l:te1?1, White shouW play 14
i H :xe3 lbe4 1 9 :n with a slight edge,
, lbbS Wb8 1 S 11'd2 g4 16 11'aS .i.xbS 17
�hould be considered) 1 6 ...lDe4 17 c6 f6 1 8 'iVxbS (but not 1 7 .i.xbS?? gxf3 and Black
cxb7+ 'iPxb7 1 9 lbc4 f5 (l.Gurevich-Barua, has a winning attack) 17 ...c6 1 8 11'a5 .tg7
Hastings 1 993/94). Now White could with chances for both sides.
probably have been slighdy better after 20 1 1 .!i)d2 .i.g7 1 2 �ta -.s 1 3 lt)xg5
li)e3 .i.g7 21 .i.eS .i.xcS 22 dxeS .rigS 23 This leads to a slighdy dull position, but
J:lr4. one that is preferable for Black. 1 3 .i.xe4
c) lO .. f5 is also enticing - we see no fault
. :Xe4 (13 ...dxe4 1 4 lbxgS irg6 1 S dS h6 16
with this move. For example, 1 1 f3 lbf6 (not c6 is also reaUy messy) 1 4 lbxgS l:lg4 1 S f4
l l ...f4?! 12 fxe4 fxe3 1 3 g3 Wh3 1 4 cSI hS?I (IS lbm llxg2+1 16 Wxg2 .i.h3+ 1 7 Wh 1
1 5 c6! and Black was in deep trouble, .i.g4 and Black wins) 1 S...h6 1 6 lbf3 J:lg8 is
Emcste-Goldmane, Riga 1 989) 1 2 lDcJ (1 2 certainly very unclear.
c6 .i.xc6 13 ..ixf5+ i.d7 1 4 g3 'lfhs 1 S 13 •xd1
•••

i.xd7+ :xd7 1 6 lbc3 .i.g7 gives both sides 1 3....i.g4?! is strongly met by 1 4 f3 lbxgS
chances) 12 ... f4 1 3 .i.f2 1 S fxg4 1Wh6 1 6 .i.f5+ Wb8 1 7 .i.f4 and

123
The Petroff Defence

White is dominating the board. 27 ....i.e& 28 b4 d4 29 a4 a5 30 b5 .i.b3


1 4 llfxd1 lDxg5 1 5 .i.xg5 31 lla1 �c8 32 �2 h6
32 ...h51? is probably stronger: 33 lit>e2
Wd8 34 c;t>dz c;t>e7 35 .i.c2 .ixc2 36 c;t>xc2
llg6 37 c;t>d3 llxg3 38 lla2 c;t>d6 39 c;t>xd4
.J:lgS and Black should be okay.
33 We2 � 34 �2 �e7 35 .i.c2 .i.c4.
36 llh1 d371
In time trouble lvanchuk pushes his
pawn, but he only succeeds in weakening it .
36. .1le6 37 .i.d3 .i.b3 38 llh4 lld6 39 .i.c2
.

i.c4 40 lle4+ c;t>f8 41 g4 is slightly better for


White (Kasparov).
37 .i.d1 �8 38 llh4 .i.a2 39 lle4 llg6
40 g4 h5
1 5....i.g4 Or 40 ..:d6 41 llc3 .i.b I 42 c;t>c3 and
.

1 5 ....i.xd4 1 6 c6 .i.c6 1 7 cxb7+ Wxb7 1 8 White has a great advantage.


i.bS Aeg8 19 Axd4 AxgS 20 llct i s a bit 41 f41? hxg4 42 f5 lld6 43 llxg4 .i.d5
unpleasant for Black. 44 llh4 �g7 45 g3
1 6 lld2 .i.xd4 1 7 c6 .i.e5 18 .i.b5 b6 1 9
.i.h4 llhg8
The alternative 1 9 ...d4 20 Lg3 .i.g7 2 1
.i.a6+ WbB 22 a4 would be dangerous for
Black (Kasparov).
20 .i.a6+ � 21 .i.g3 .i.xg3 22 hxg3
lidS 23 lld4 lld6 24 llc1 .i.c8 25 .i.d371
25 .ixc8 is a better move: 25...'�xc8 26 b4
a6 27 a4 h5 28 bS and White has some pres-
sure.
25 ...llg41? 26 llxg4 .i.xg4 27 f3

45 ...�6?1
This loses and should be avoided, though
4S....ia2 46 llc4 Cit>f8 47 lle3 is also nice for
White.
46 llh8 'ite5 47 .Z:c8 �d4 48 llxc7
llh6 49 l:ld7 llh2+ 50 �1 1 -0

Game 52
Lastin-Najer
Elista 2000

27 .i.xh7 d4 28 .i.d3 .i.c6 29 b3 .i.dS 1 e4 e5 2 ltlf3 ltlf6 3 d4 ltlxe4 4 .i.d3 d5


gives Black sufficient countcrplay (Kas­ 5 ltlxe5 ltld7 6 ltlxd7 hd7 7 0-0 'ifh4 8
pamv). c4 0-0-0 9 c5 g5 1 0 f3 ltlf6 1 1 .i.e3

124
3 d4: The Main L in e

hxg3 1l'h3 17 l:td1 .i.g7 1 8 .i.ft 1-0,


Tomashevic-Kondali, correspondence 1 991)
13 .i.f2 'lfh6 14 'l'a5 Wb8 1 5 lLlcl l:te6 1 6 a4
(or 16 ltlb5 .i.xbS 1 7 1i'xb5 ll'lb5 18 g3 '..f6
19 Jlae1 c6 20 'ifa5 l:the8 and Black has de­
cent counterplay) 1 6...:C8 17 Jla3 g4 1 8 ll'lb5
l:ta6 19 Wb4 brxf3 20 ll'lxc7!? :Xc7 21 .i.xa6
fxg2 22 Wxg2 lllh5 with a real mess, Mor­
gado-Gottardi, correspondence 1 997.
1 2 l003
1 2 g3 is worth considering: 12...'1Fh5 1 3
ll'lc3 D.he8 14 l:te 1 g4 1 5 f4 ll'lc4!? 1 6 li)xe4
{16 1Vc2 ll'lxc5! 17 dxc5 d4 1 8 .i.f2 dxc.3 19
11 .tg7
.•. bxc3 .i.c6 gives Black a fine game) 1 6...dxe4
In this rabiya Black has many interesting 1 7 .i.e2 1Vg6 1 8 Wb3 (1 8 dS .i.c6 19 .i.xg4+
c 'PP<>rtunities: �b8 20 f5 1Wf6 21 d6 cxd6 22 cxd6 1Wxb2
:•) 1 1 ...g4 12 g3 Wlh5 13 f4 :Cs 14 Jlet isn't clear ac aU) 1 8....i.c6 1 9 .l\.c4 'if£5 20
.i.g7 1 5 �3 Jle7 1 6 Wd2 Jlhe8 17 b4 .i.f5 :Cd1 (20 .i.xfl l:te7 21 dS .i.xd5 22 .i.xdS
( 1 7...tDc4 1 8 .i.xe4 dxe4 19 b5 Jld8 20 Wb2 'l'xdS 23 'l'a4 a6 shouldn't be worse for
looks very dangerous for Black) 1 8 c6!? bxc6 Black) 20....i.d5 21 c6 bxc6 22 lLct llc6 2.3
I IJ ll'la4 .i.xd3 20 Wxd3 ltle4 21 Jlact with 1Wa3 .i.xc4 24 :Xc4 Wbs 25 l:tdct and White
compensation, M:usuura-Kapelari, Sao Paulo had fantastic play for the pawn in Tiviakov­
1 997. Rozentalis, Groningen 1997.
b) I t ...llg8 1 2 ll'lc3 g4 looks very att:rac­ 1 2 ...Dhe8 1 3 .i.f2 1ih6 14 irb3
livc:
b1) 13 Wet ?! is strongly met by t3 ...g31 1 4
hxg] l:txg3 1 5 lLlez (even worse is 15 Wd2?
.ixcS 16 dxc5 :dgS 17 l:tfd1 d4 1 8 c6 dxe.3
1 9 cxd7+ Wd8 0-1 Dolmatov-Makarychev,
Palma de MaUorca 1 989) 1 5....i.d61 1 6 'l'f2
( 1 6 cxd6? l:txg2+! and it is all over) 1 6 ....i.h3
1 7 ll'lxg.3 .i.xg3 18 'l'c2 l%g8 with a fantastic
nrtuck.
b2) 1 3 g.3 Wh3 14 f4 lllh5 1 5 Wet (weaker
is 1 5 .i.f2? ll'lx£41 16 gxf4 g3 1 7 .i.xg3 l:txg3+
1 8 hxg3 ..xg3+ 19 Wh1 'ifh4+ 20 Wg1 .i.g7
nnd Ulack wins, despite being a rook down)
I S .. .lle8 1 6 'l'f2 ltlf6 1 7 Jlae 1 hS with an Another try is 1 4 ll'lb5 Wb8 1 5 a4 ll'lh5 1 6
unclear game. Wd2 a6 1 7 l:tfcl!? li)f4 (not 17...axb5? 1 8
c) 1 t...l:te8 12 'lfd2 (12 .i.f2 1Wh6 1 3 ll'lc3 axbS b6 1 9 1Wb4 with a winning attack) 1 8
g4 1 4 f4 - 14 l:tct c6 1 5 b4 gxt1 16 Wxt1 .i.fl l:le6 1 9 'I'aS l:tc6 20 lL3 .i.c8 21 lle 1
lL'l,.,r4 also gives Black considerable counter­ ..tf6 22 1Vd2 'ifg7 with chances for both
play - 14 ...11'xf4 1 5 .i.h4 .xd4+ 16 Wh1 sides, Zulfurgarli-Bayramov, Baku 1 998.
.ig7 with unclear play - Makarychev) 14 ...g4 1 5 �b5 �8 1 6 1lfe1?1
12 ....i.g7 (12...:Xc3? is too optimistic: 13 White should have played 1 6 .l\.g3 lieS 17
Wxd lllhS 14 g3! ll'lxg3 1 5 'IreS l:tg8 16 .i.eS; after l 7...g.3!? 1 8 .l\.xg3 ll'lhS 1 9 .i.f2

125
The Petroff Defence

/l)f4 Black has brood play for his pawn. 20...�xa7


1 6 ...gxf3 1 7 ltld61? After 20 ..�a8? 21 1i'b6! Black cannot de­
.

White is trying to complicate matters; after fend himself.


1 7 gxO /l)h5 18 �hl 1i'd2 Black is in con­ 21 1ta3
trol.
1 7 . . .cxd6 18 J..a& b6?!

21 .. .i.b7
21 ...1la8? would be met by 22 1Wa51 .i.b7
This makes no sense - Black probably 23 llact!! l:lxct+ 24 ..i.fl+ 'it>b8 25 'ii'c7+
overlooked the note to White's 20th move. <tla7 26 llc3 and White wins.
After 18 ....i.c6 1 9 .i.xb7 .i.a4 (even 22 .txb7 + �b7 23 111b4 + � 24 'lra5 +
1 9....ixb7 20 c6 Axel+ 21 llxel l:ld7 22 �7 25 M»5 + �7 26 '1ra5 + %-%
cxd7 �xd7 23 1Wc3 a6 is comfortable for
Black) 20 1i'b4 (20 ..xa4? /l)g4 and Black Gan1e 53
wins) 20 ..a5 21 'irb6 li)d7 22 1Wxa5 llxel+
. Anand-Hubner
23 l:txe1 Wxb7 24 11fxd8 dxc5 Black has the Dortmund 1992
advantage.
1 9 cxb& i.e& 1 e4 e5 2 ltlf3 ltlf& 3 d4 ltlxe4 4 J..d3 d5
5 lZ':ixe5 lbd7 6 ltlxd7 J..xd7 7 0-0 1Wh4 8
c4 0-0-0 9 c5 g61?

20 bxa7+
Much stronger is 20 1Wc3t lbg4 21 h4 tbe5
22 dxe5 dxe5 23 llact and Black is under This move keeps the g5-square vacant as a
attack. possible retreat square for the knight.

126
3 d4: The Main Lin e

10 c!i.:lc3 .i.gS+ � f8 1 7 1Wc2 according to Rozcntalis)


I 0 f3?! ll)g5 1 1 J.e3 .*.g7 1 2 g3 'ilh5 gives 14 lL!c6+1 (the move Rozentalis overlooked?)
Black considCfllble counterplay. 1 4...Wc8 15 lL!xa7+ Wb8 1 6 lLlc6+ Wc8 1 7 f3
10 .tg7 , , g3
.•• Ld4 18 .i.c3 llxd3 19 ..xd3 lllxg3 20 .i.f41
Or: 1 -0 lvanchuk-Rozentalis, Debrecen 1 992.
a) 1 1 llk2 :hetl

1 2 i.e3
and now: White has alw tried 1 2 lL!xe4 dxe4 1 3
a l ) 1 2 a4 /.l)g5!? 1 3 :a3 h6! 1 4 J.c2 (1 4 .ixe4 i.h3 1 4 1Wb3 11fa6 (or 1 4...c6!? 1 5 d5
i.c3 �6 1 5 1i'd2 with unclear play is proba­ cxd5 16 J.xdS Ld5 17 "tlfxd5 :dB 18 "tlfb3
hly better) 14 ...llxe2! 1 5 1i'xe2 �6 1 6 1Wd1 9c6 19 f3 J.xfl 20 Wxfl :d 5 with unclear
li:)xd4 and Black had enonnous play for the play) 1 5 :et :xd4 16 .i.f4 J.e6 17 9f3 c6
exchange, I vanchuk-Rozentalis, Dcbrecen 1 8 .id6 l:.d2 19 Wf4 l:r.d4 and the position
1 992. was very messy, Sax-Skembris, Burgas 1992
n2) 1 2 .i.e3 J.h6 1 3 .i.xh6 9xh6 1 4 1Wct
and White might have a very slight edge.
b) 1 1 '*.e3 lLixc5 1 2 g3 1i'h3 13 .i.e2 (1 3
dxcS d4 14 lLld5 dxe3 1 5 fxe3 - 15 c6? exf2+
I (l llxf2 .i.d4 17 cxd7+ Ld7 is bad for
White - 1 5...-*.e5 1 6 Wc2 �b8 leads to an
unclear position) 13 ...h5 14 lL!xd5 J.a4! 1 5
b.l .D.xd5 1 6 bxa4 :es (but not l 6 ...:hd8?!
17 i.f3 llxd4 1 8 Wc21 with a clear edge for
White, J.Polgar-Skembris, Moscow 1 994) 17
'i'c2 lL!e4 18 :ret Wd7 and the position is
very complex.
1 1 ..f&
...

1 1 ...'ilh3?! was cleverly refuted by board 1 2 .tf5?


...

nne of the 2004 Olympiad Champions, 1 2...lllg5 is necessary. Now 1 3 lL!xd5?


Ukraine: 12 lL!xdS J.g4 1 3 lLle7+! (not 1 3 lllh3+ 14 Wg2 .i.c6 15 .g4+ �b8 1 6 Wxh3
'i'b3?! i.f3 1 4 lLif4 lL!g51 and the position is :xd5! gives Black a clear advantage for Black
less clear - Ivanchuk) 1 3 ...Wb8? (Black (Anand). Instead White plays 13 f4 when
should settle for a bad position after Black has the foUowing tries:
LL�d7 14 1Wa4+ �c7 1 5 .i.xe4 :xd4 1 6 a) 1 3..."ti'e6?1 1 4 l:r.c1 lLle4 1 5 J.xe4 clxe4

127
The Petroff Defence

16 dS 1i'a6 17 .i.c.I4 with a considerable ad­ 19 a4 g5 20 a5


vantage (Yusupov). 20 f3? would aUow 20...c!Llxg3 21 .i.xg5
b) 1 3 ...1i'e7?! 14 lle1 �6 1 5 .i.ft ! (a nice c!Llxft 22 .i.xd8 .llxd8 23 Wxft 1i'a6 when:
refutation of Alexander's idea; I S c!LlxdS 1We8 BL'lck suddenly has counterplay.
gives Black enough counterplay, Sasha's 20 c& 21 c!ilc3 ltlxc3 22 bxc3 h& 23 a&
•.•

preparation running 16 f5 gxf5 1 7 .i.xf5 .i.a4


18 1i'g4 llxdS 1 9 .i.xe6+ fxe6 20 1i'xg7 l:g8)
1S...c6 16 'i'a4 Wb8 17 f5! (another strong
move from the strongest player of our little
region) 17 ...gxf5 1 8 .i.f4+ WaS 19 .lle3! c!LlxcS
(19 ...'i'f6 would fail to 20 'i'xa7+!! �xa7 21
c!LlbS+! cxb5 22 ltt3 mate) 20 .llxe7 c!Llxa4 21
c!Llxa4 .i.xd4+ 22 .i.e3 .i.xe3+ 23 :Xe3 and
White went on to win with his extra piece,
Delikov-Raetsky, Voronezh 2004.
c) 1 3...c!Llh3+! (the best move) 1 4 �g2
.llhe8 1 5 9d2 (Leko-S.Farago, Dudapest
1 993). Now after 1 S ....i.g4 1 6 b4 hS Black
has counterplay, but it's difficult to teU Now White is winning.
whether it is enough for equality. 23 . . .f5 24 J.h3 l:lfS 25 a7+ �cS 26
1 3 lilb51 J.h3 '8b1 g4 27 J.f1 �7 28 ..i.d3 h& 29
Black is in a bad way. After 1 3... a6 14 J.f4 l:tf7 30 itc2 ..i.fS 31 'lle2+ 1 -0
c!Llxc7! �xc7 (14...9c6 1 5 c!LlaB! is horrible
too) 1 S .i.f4+ �d7 1 6 .i.e5 ...e6 17 .i.xg7 Game 54
.llhg8 18 .i.eS c!LlxcS 19 .i.xf5 ...xf5 20 g4 Elizarov-Raetsky
9e6 21 .lle 1 �4 22 f3 c!Lld6 23 1ib3 White Belorechensk 1989
won in Har Zvi-Lev, Tel Aviv 199S.
14 ltlxa7 + �8 1 5 ltlb5 J.xf1 1 6 J.xf1 1 e4 e5 2 lilf3 lilt& 3 d4 c!l:lxe4 4 J.d3 d5
5 ltlxe5 ltld7 6 ltlxd7 J.xd7 7 0-0 itf617
This is probably a little bit better for
White, but it's stiU playable.
8 J.xe4
Probably the critical move, but White has­
some alternatives:
a) 8 .i.e3!? is also a decent option: 8 ... .i.d6.
9 c4 c6 10 cxd5 (1 0 llk3 c!Llxc3 l 1 bxc3 dxc4
1 2 Lc4 0-0 13 9hs .llfe8 is fine for Black)
l O...cxdS 1 1 Wb3 .i.c6 12 .i.xc4 dxc4 1 3
llk3 0-0 14 dS .i.d7 1 S c!Llxe4 (IS 11'xb7?1
1We5 1 6 g3 .i.h3 with an attack on the light
S<JUares) 1S....i.xh2+ 16 Wxh2 1i'h4+ 1 7 �gl
White has a clear edge despite the missing 1i'xe4 1 8 1i'xb7 .i.e6 1 9 .llfd1 (Macieja­
exchange - Black's king is simply too weak. Rowson, Duisburg 1 992). Now after
1 &. . .:heS 1 7 ita4 ita& 18 itb4 Was 19 . ...11fb8 20 9c6 ll.d8 21 lld4 .i.xdS 22
Or l B ...1i'c6 19 .i.f4 :C7 20 WaS and .llxe4 .i.xc6 23 .lle7 White might have
White wins. slightly better chances in tllis endgame.

128
3 d4: The Main Lin e

b) 8 c4 is very aggressive, but not clear at 'ife3 lld6 19 .xa7 l:txc6 20 .aS+ �c7 21
nil. Black plays 8...1Wxd4 llab 1 .C.b6 22 :Xb6 �b6 23 .C.b1+ �c5 24
1i'a3+ �d4 25 .e3+ �e5 26 i.xg6+ 1-0
Kotronias-Atalik, Pucarevo 1 987.
8 dxe4 9�3 0-0-0 10 lllxe4 1tg6
...

and now:
b1) 9 ltk3 �c5 1 0 Xlel+ i.e6?1 (1 0...�d8
iN necessary; after 1 1 i.e2 1Wxd1 1 2 l:txdt
dxc4 13 i.xc4 White has compensation, but 1 1 f3
nothing more) 1 1 .C.xe6+! �xe6 12 cxd5 i.c5 The alternative 1 1 /i)g5 is probably best
1 3 i.e3 •e5 1 4 dxe6 i.xe3 1 5 i.bS+ �f8 1 6 met by 1 1...£6, for example 1 2 �f3 i.h3!?
.f3 .xe6 17 fx.e3 c 6 1 8 .L4 �18 1 9 i.b3 (1 2... h5 13 c4 h4 14 l&1 h3 15 g3 i.g4 16 f3
'fkc7 20 .C.ft f6 21 llk2 and Wh_ite had a clear i.cS 1 7 i.e3 llhe8 1 8 i.f2 i.f5 1 9 d5 i.d6
nlgc in the game Tseshkovsky-Bareev, Kiev, 20 1i'd2 gives White a slight edge, S7jebcrt­
1 1)R6. Raetsky, CappeUe Ia Grande 1 999) 1 3 �e1
b2) 9 cxd5 is probably less dangerous: i.c5 14 d l:the8 15 i.e3 i.d6 with compen­
11...0-0-0 1 0 .c2?1 (10 i.xe4!? .xe4 1 1 �c3 sation for the pawn.
'fkh4 1 2 i.e3 i.d6 13 g3 with unclear play
was necessary) 10 ... �5 1 1 i.c4 i.d6 1 2

Ac3 'ireS 1 3 g3 i. f5 1 4 'ird2 1We4 1 5 �a3


.C.hd� 1 6 llfd1 i.h3 17 i.ft i.xft 1 8 :Xfl
Q)d3 and Black dominated events in Chudi­
novskikh-Raetsky, Briansk 1995.
c) 8 � is pretty tame:
cl) 8 ....xd4 9 �xe4 dxe4 1 0 i.xe4 1Wxd1
I I llxd 1 0-0-0 12 i.g5 f6 13 i.f4 i.cS 1 4
l:c.l5 i. b6 1 5 llad1 g6 16 a3 c6 1 7 .C.Sd3 i.g4
I K :Xd8+ llxd8 1 9 .C.xd8+ �xd8 1/:�-1/z Ken­
�is- Rozentalis, Vilnius 1 984.
c2) 8.. �d 9 bxc3 0-0-0 is possible, but
.

White seems to be slightly better: 10 'irh5 11 h5


•••

'fkc6 1 1 i.d2 g6 1 2 'iff3 f6 13 l:tfel 'lff7 1 4 Black ha.c; no path to absolute equality, for
c4 dxc4 1 5 i.e4 c6 1 6 d5 i.e8?1 (16...i. f5 1 7 example:
i.x£5+ gx f5 1 8 .xf5+ 'ird7 19 1i'xd7+ �xd7 a) I 1 ...f5 1 2 �f2 i.bS 13 .C.e1 .i.d6 14
211 .C.ab1 is better for White, but the game �h3 .C.dc8 1 5 i.f4 :Xel+ 1 6 1i'xe1 Xle8 17
WI 1uld still be undecided) 17 dxc6! l:txd2 1 8 'ifg3 1i'xg3 1 8 hxg3 is a bit better for White.

129
The Petroff Defence

b) 11 ...-*.5 12 c3 h5 (or 12...i.xe4 1 3 fxe4 lion for dte pawn) t 8 .!Dc3 'ifb4 19 l:la4 'ii'e7
...xe4 1 4 :Xf7 i.d6 1 5 .*.g5 lldf8 16 llxf8+ 20 d5 g5 21 .ig3 f5 22 gxf5 .tf7 23 ltc4 with
ltx£8 1 7 .d2 and we prefer White due to the a dose-to-winning advantage for White,
extra pawn) 13 'ifa4 'ii1b8 14 .*.f4 h4 1 5 llae1 A.Ivanov-Kochicv, Kosttoma 1 985.
h3 16 g4 i.d7 (Klovans-Ro?.entalis, USSR b3) 16 ...a5!? seems to be necessary, al­
1 985). Now after 1 7 'ifc4 llc8 18 �5 i.d6 though after 1 7 'lrb5!? llxd4 1 8 1ixb6
19 .*.xd6 ..xg5 20 �e7 i.b5 21 .*.xg5 .*.xc4 l'lxdl+ 19 l:lxd1 cxb6 20 .i.e5 dte endgame is
22 llf2 .*.xa2 Black is worse but he has &JOOd more pleasant for White.
drawing chances in the endgame. 12 f6 1 3 ..th4 �
•••

1 2 ..tg5
12 .*.f4 is also strong after 1 2..h4 13 'ifd3

Or 1 3...•£7 14 1t'd2 g5 1 5 .tf2 h4 16 d5


h3 1 7 g3 and White is better.
and now: 14 c4 g5 1 5 ..tf2 f5 16 .!i)c3 h4 1 7 d5 g4
a) 1 3 ... 'ifb6 14 a4 .i.e6 1 S llfd t h3 1 6 aS 1 8 ..td4 .i.d6
..c6 1 7 a6 b6 1 8 c4 f5 19 l'Df2 (dtis is White cannot be allowed to play 1 9 f4, for
stronger than 19 dS?l fxe4 20 ..xe4, Raetsky­ example 1 8...:g8 19 f4 .td6 20 ...d2 c5 21
Kvt:inys, correspondence 1 987) 19 ...hxg2 20 dxc6 .ixc6 22 t'Dd5 and White has all dte
dS ..cS 21 Wxg2 and White is clearly better. chances.
b) 1 3...h3 14 g4 'fib6 15 a4 i.e6 1 6 llfd1 1 9 fxg4
and here Black must choose between dte 1 9 .ixh8? ...e3+ 20 l'lf2 g3 21 .i.d4 gx£2+
following options: 22 �ft 1t'f4 is bad for White.
b1) 16 ... 5 17 aS 'ifc6 (or 17 ... fxe4 18 axb6 1 9 :hg81
••.

exd3 1 9 bxa7 �d7 20 a8'ii' lLa8 21 llxa8 Not 19 ...fxg4 20 .txh8 (20 l'lf6!? 1Whs 21
dxc2 22 llct i.b3 23 .*.eS c5 24 lL.3 c4 25 c5 .ieS 22 c61 is equally strong) 20...'ifxh8 21
l:la8 and the advantage is definitely widt llk4 .ixh2+ 22 �xh2 'ifeS+ 23 �g1 Wxe4
White) 1 8 gxf5 i.xf5 19 a6 b6 20 .i.g3 ..g6 24 Wd2 and White the advantage, and a sub­
21 'ii'c4 lld7 22 d5 Le4 23 'ii'xc4 l'lh6 24 stantial one at dtat.
lla4 .i.cS+ 25 'ii1f1 and White was clearly 20 gxf5?
better in Raetsky-Sivets, correspondence. In For some reason White decides to assist
fact, following 2S...:C7? 26 d6! it was already Black's attack with his next two moves. Here
rime to resign. he should play 20 c51 .i.cS 21 .i.xe5 11fc3+ 22
b2) 16 ...a6?! 17 aS 'ii'xb2?! (17 ...'ii'c6 is bet­ l'lf2 'irxe5 23 'iffl fxg4 24 c6 bxc6 25 ...a6+
ter, even dtough after 1 8 ...c3 'ii'xc3 1 9 �b8 26 dxc6 .i.c8 and Black is only a bit
.!Dxc3 .i.b4 20 ltk4 Black has no compen.<�a- worse in this messy position.

130
3 d4: The Main L ine

23 g3 is met strongly by 23....txg3! 24


hxg3 :xg3+ 25 �h2 :gZ+ 26 �h 1 :Xd5! 27
cxd5 ..d6 with mate to come.
23 . . .11rh4 24 Wh1 llxg4

20. . . c5!
131ack grabs his chance instantly. l11e key
Idea is 21 .i.f2 .f41 22 .i.g3 ..g5 23 lbe4
Wg3+ and the queen's dance wins a piece.
21l... .i.xh2+1? 21 Wxh2 .d6+ 22 Wh1 ll.xg2! 25 ..i.f6
h:;tds to a draw as 23 .i.g1 ?? Wg3! 24 ll.f3 After 25 .i.f2 ..g5 26 11t'f3 :lxc4 White is
l:.xgt+ 25 ..xgt 1i'xf3+ 26 ..g2 ..xf5 is finished.
clc;trly better for BL1ck. 25 ...1lg1 + 1 26 Wxg1
21 dxc67 Or 26 :xgt ..c4+.
A terrible mistake, and suddenly Black has 26 ..tc5+ 27 �1 •xc4 0-1
.••

ufantastic attack. After 21 lbe4! cxd4 22


Wxd4 :dtll the game would have remained Game 55
unclear. Tiviakov-Miles
21 ...i.xc6 22 lilll6 l.inam 1998

1 e4 e5 2 ltlf3 ltlf6 3 cl4 ltlxe4 4 ..i.d3 d5


5 ltlxe5 ttJd7 6 ltlxd7 ..i.xd7 7 0-0 ..i.e7
11tis is a bit passive, and so is 7. ..lt'lf6, for
example:
a) 8 .i.g5 i.e? 9 c3 c6 1 0 lbd2 0-0 11 ..c2
h6 12 .th4 li)h5 13 .i.xe7 Wxe7 14 ll.fet
..d6 15 li)f3 li)f4 16 .th7+ �h8 17 .tf5
.i.xf5 1 8 Wxf5 was very slightly better for
White in GcUer-Smyslov, Moscow 1 981,
although here the players agreed a draw.
b) S ltct+ .te7 9 .tf4 (or 9 Wc2 .t� 10
f4 g6 1 1 li)c3 'ird7 12 f5!? gxf5 - t2 ... Lf5?
22 . . .h31 13 .txf5 gxf5 14 11t'e5 would of course be
The right way to attack. After bad - 13 .i.f4 lbc4 14lbb5 .i.d8 with un­
.txh2+?1 23 �xh2 ll.xd5 24 cxd5 ..d6+
22... clear play) 9....i.g4 (9...0-0? is met strongly by
2S Wh 1 l:.xg2! 26 .i.c5! Wxe5 27 Wxg2 10 .i.xc7 Wxc7 1 1 :xe7 '1Vb6 12 li)c3 ..xb2
Wg3+ all Black has is a draw. 1 3 lbxd5 li)xd5 14 l:lxd7 li)c3 15 ...ct
23 g4 ..xct+ 16 llxctlbxa2 17 l:la1 li)b4 18 .i.e4

131
The Pe tro ff Defence

and Black is struggling, possibly in vain) t 0 hxgS. Now 19 d5!? should en.or;ure some ad­
'il'd2 0-0 t t li:lc3 c6 12 o!ik2 ...d7 13 lbg3 vantage. Black has to be careful, for example
li:lh5 1 4 .!fuh5 hh5 1 5 .i.f5 ...d8 (not 1 9..Jife8 20 lZxe8+ lZxe8 21 dxc6 .ixc6 22
1 5......xf5?! 16 :Xc7 lZfc8 17 lZact lZxc7 1 8 h41 1i:lb5 23 li:lxb5 .i.xb5 24 hxgS with a clear
:Xe7 with a clear edge, a s t 8...b6? t 9 llc5 edge for White.
...g6 20 llg5 is winning) 1 6 :C3 i.g6 17 10 c5 0-0 1 1 .i.f4 c& 1 2 b4
Lg6 fxg6 1 8 1lac1 lZf7 1 9 g3 with a slight
edge for White, Palac-Zaja, Pula 2000.
8 c4
8 lZe1 doesn't really work: 8...0-0 (8...1i:lf6
transposes to 7...lbf6 above) 9 .i.xe4 dxe4 10
D.x'--4 .tf6 1 1 :C t i.c6 1 2 .i.c3 ...d5 1 3 0
l:tfe8 1 4 �3 (14 c3?! i.M! 1 5 .i.f2 .i.xf2+
1 6 Wxf2 :Xet 17 ...xe1 :Cs 18 ...d1 .tb5
with a strong initiative) 14......d6 1 5 Wh1
ltad8 1 6 Wd2 i.xd4 1 7 .i.xd4 1i'xd4 1 8
:XeS+ .ixe8 19 ...xd4 lZxd4 and Black is
ever-so-slightly better (Skatchkov).

It's advisable for White to get going here,


Svidler-Yusupov, Ka7.an., 1997 continued
more slowly with 1 2 h3 lZe8 13 :Ct o!L'ld7 1 4
•d2 (or 1 4 b4 .i.g5 1 5 .i.g3 •f6 and Black
is okay) 14 ....if6 1 5 .i.c2 lt'lf8 16 D.e2 b6 1 7
Le 1 bxc5 1 8 dxcS .i.d7 1 9 lZxe8 .*.xeS 20
b4 li:le6 21 .ld6 g6 22 f4. Here Svidler gives
the line 22... a51 23 f5 b>Xf5 24 .txf5 .ig7 25
.ixe6 fxe6 26 llxe6 'ii'h4 27 o!be2 and as­
sesses the position as unclear.
1 2 1Wd7 1 3 1Wc2 g6
••.

8 lLif6
•.•

Or 8 ...c6 9 cxd5 cxd5 1 0 o!L'lc3 lt'lf6


(IO...Ii:lxc3 1 1 bxc3 Wc7 12 :Ct is good for
white; 12 ......xc3 1 3 .td2 ...xd3? 1 4 lZxe7+! is
even winning!) 1 1 .i.g5 .i.c6 12 .i.c2 (1 2 lZet
0-0 1 3 •n looks even stronger) 12 .. 0-0 1 3
.

n :es 1 4 •d3 g6 1 5 :ret •d6 t 6 •d2


li:lh5 17 .i.xc7 lZxe7 1 8 lZxe7 ...xe7 1 9 lZc1
'iff6 20 lZc5 and White had an edge in Alek­
seev-Skatchkov, St Petersburg 2002.
9 ll:lc3 .i.e&
Zapata-Perdomo, Colombia 1 998 contin­
ued 9 ...dxc4 10 .i.xc4 0-0 1 1 :Ct c6 1 2 .*.g5 14 J:lfe1
h6 1 3 .th4 lDd5 1 4 .txe7 li:lxe7 1 5 'ilrh5 14 b51? :res (14...o!L'lh5 1 5 .tM {;jjg7 1 6
li:lf5 t 6 l:tad I �6 t 7 .i.b3 1i'g5 18 Wxg5 a4 .if6 17 �2 also l1x>ks a touch better for

132
3 d4: The Main L ine

WhiLe) 1 5 a4 li)hS 1 6 �e3 gives White a Or 34 :Xc6 i.xd4 3S :Xa6 l:lxa6 36


Hlight plus. Now in J.Polgar-Van der Sterrcn, .i.xa6 .i.e8 37 li)b6 1WeS and the endgame is
Wijk aan Zce 1 998 Black went astray with bad for Black.
I Cl ...i.d8?! 1 7 aS a6 1 8 bxa6 bxa6 19 li)a4!
. 34...1te7 35 Lc6 .i.xd4 36 lbb6 .i.c3 37
i.xa5 20 li)b6 .i.xb6 21 cxb6 i.f5 22 i.xf5 1i'g5
.xfS 23 l:lfcl Vxc2 24 :Xc2 aS 2S l:lxc6,
�iving White a clear plus.
14 . . .lbh5 1 5 .i.e5
l S .i.h6 l:lfe8 16 bS .i.f6 17 bxc6 bxc6 1 8
.:�4 li)g7 with the idea o f ...ll)f5 gives Black
�ufficiem countcrplay.
1 5...f6 1 6 .i.g3 lbxg3 1 7 hxg3 i.f7 1 8
b5 1tfe8
18... f5 19 l:labt .i.f6 20 li)e2 l:.fe8, with
chances for both sides, is probably a better
c 11>tjon.
19 a4

37 'fre57
•••

37...l:le1 38 li)xdS .i.xa5 39 �3 is obvi­


ously nice for White, but at least it's not over!
Now, though, White strikes a winning blow.
38 lbxd51 1tb7
Or 38...'.xd5 39 Vxe7.
39 lbxc3 11'xc3 40 11'f6 11'xa5 41 ltc8
.i.g8 42 i.c4 1 -0

Game 56
Sorokin-Raetsky
Kra.modar 1984
1 9...i.d8?!
19 ... f5 is stiU correct. 1 982, Vorone1.h State University, De­
20 ltxe8+ 'lfxe8 21 aS a6 22 bxa6 bxa6 partment of Mathematics, 23 years before the
23 'lfd2 f5 24 ll:la4 .i.f6 25 lte1 '1Vd8 26 publication of J>ehvjf Defmce by Everyman
ll'lb6 Chess. Wasting no time during a physics
Or 26 ..tn ..tg7 27 l:tbl Vf6 28 l:tb4, lecture, the future author makes an impormm
when White is more comformble. discovery on his pocket chess computer (a
26 ...1ta7 27 ..if1 �g7 28 ltb1 h5 29 version of the computer software Fritz 0.03):
11rc3 'lfha 30 ltb4 �7 the Petroff Defence is stiU a]jvc!
Black should probably play 30... h41? 31 He has just found a beautiful combina­
1,rxh4 Vxh4 32 g3 Vhs, although White does tional refutation of ECO'r critical assessment.
keep an edge. The bible of that time insists on '14 �12
31 'lfd2 'lfe8 32 lba4 '1Vd8 33 ltb6 'lfh8?! with a slight edge' as in the game Mortensen­
33...Vc8 is a better option, although after Borik played at the t 980 Chess Olympiad,
14 1ib4 White has a strong pressure. but his intuition whispers 'it's not that sim­
34 La6 ple' and 'Eureka!'...

133
The Petroff Defence

We had to wait for two years until my uni­


versity analysis was included into all books
on the Petroff Defence. My 'co-author' was
the future GM Sorokin.
1 e4 e5 2 o!i:lf3 o!i:lf6 3 d4 �xe4 4 -*.d3 d5
5 o!Lixe5 o!Lid7 6 ••2

and now:
a) 9 WI? 1fd4 1 0 'irbS+ c6 1 1 ..xb7
ltc8 1 2 .ie3 11'xe5 t 3 .xa7 .ib4 1 4 'iVd4 ·

and White was a bit better in Suetin-Radulov,


Athens 1 984.
b) 9 11rxe4 is less clear: 9._.id5 tO 'irg4 hS
This is not really considered dangerous 1 1 1Wh3 'ire7 12 f4 ..e6 t3 ..xe6+ fxe6 1 4
now. � f2 (1 4 0-0 .icS+ 1 5 �h1 0-0-0 1 6 lbd2
6 .1tlxe5
•. :thffi gives Black excellent play) 1 4...g5!? 1 5
Another option is 6 ...1Ve7 7 .ixe4 (I fxg5 ( 1 5 �c3 .icS+ 1 6 .ie3 .ixe3+ 17 �xe3
lDxf7!? ITansposes to 6 �xf7 'iVe7 7 1t'e2) .*.xg2 18 lthgt gxf4+ 19 �f4 llg8 looks
7...dxe4 8 .if4 lbxe5 9 .ixe5 and now: k.-vel) t 5.. ic5+ 1 6 �g3 (1 6 .ie3 0-0+ 17
a) 9 .. f6 is weak: 10 .ig3 f5 1 1 �3 c6 1 2
. �c2 .ixe3 18 �xe3 ltfS 1 9 lbc3 :XeS+ 20
0-0-0 g6 (even worse is 1 2...1Vg5+?! 1 3 �b1 �f4 ltfS+ 21 �e3 ltxg5 and Black is no
.ib4? 14 �bS f4?1 - t 4_.cxb5 1 5 1t'xb5+ worse) 16 ... h4+ 1 7 �h3 .ie4 and Black has
�f7 1 6 1Vxb4 f4 1 7 h4 is more resilient al­ good counrerplay.
though White keeps an advantage - 1 5 �7+
�ffi 1 6 1t'xe4 and Black resigned in Dcly­
Malich, Pees 1 964) 1 3 .ie5 .ih6+ t 4 �b 1
0-0 1 5 h4 and White clearly has all the fun.
b) 9....if5 10 �3 0-0-0 1 t 0-0-0 'iVe6 12
'iVc3 h5 1 3 h3 f6 1 4 .ih2 g6 1 5 �b1 .ih6 1 6
1Vg3 lth7 17 lthe1 'irb6 18 ltlxe4 :Xd4 1 9
ltxd4 1Vxd4 with level chances, Karpov­
Hort, Amsterdam 1 980. More energetic is t 4
.ixc7!? �xc7 15 d5 11rc5 16 1Vxa7 .id6 1 7
ltlbS+ �d7 (17...�c8 18 ltd4 is dangerous)
1 8 1Vxb7+ �e8 19 exg7 lth7 20 ltlxd6+
'iVxd6 21 Wg3 1Vxg3 22 fxg3 with a very
diffiCult endgame for both players. s •xe4 �e6
7 �xe4 dxe4 8..id6 9 dxe5 'fle7 to 0-0 0-0 1 1 .if4
A weaker option is the move 7....ie6?1 8 ltc8 12 :kt f6 13 �2 fxe5 14 .ig3 (or 14
dxe5 dxe4 .ixc5 .ixe5 1 5 'iVdS+ �h8 1 6 f4 .ie6 17

134
3 d4: The Msin Lin e

..xeS 1tb4 with a slight edge for White) sides, Krakops-Raetsky, Apolda 1994.
1 4... .i.d7 1 S lbf3 .i.c6 16 1i'c4+ Wt7 17 1 0 ....tb4+
..xt7+ �xt7 1 8 lbxe5+ .i.xeS 19 .i.xeS and 10...0-0-0 1 1 ..a5
White was a pawn up in Suetin-Bcx, Biel
1 99S.
9 1rxe5 ..d7

gives Black a wide, but unsatisfactory,


choice:
a) t t ....i.d5 1 2 lbc3 .i.xg2 1 3 llg1 b6 1 4
10 .te3 1i'xa7 .i.b7 1 S 0-0-0 .i.d6 1 6 h4 f6 17 d5
White has two main alternatives here: l:the8 18 lld3 h6 19 .i.xb6! cxb6 20 lba4
a) to lbc3 0-0-0 1 1 .i.e3 .i.b4 12 0-0 f6 t 3 .i.c7 21 d6! .xd6 22 llxd6 llxd6 23 b3 and
1Vg3?! (1 3 1fe4 AfS 1 4 Wf3 .i.xc2 1 5 :act Black was much worse, Udalov-Raetsky,
11rf5, with equal chances, is better) t 3.. ..i.xc3 correspondence 1982.
14 bxc3 h5 1 5 h4 g5 1 6 f3 (1 6 hxg5 h4 1 7 b) 1 t ...Wc6 1 2 lbc3 'lt'xg2 Qnstead of
1Ihz h3 1 8 g3 fxgS 1 9 .i.xg5 lldg8 20 f4 12 ...b6 13 1fa6+ �b8 14 lbb5 .i.c4 1 5 a4
1fc6 also gives Black a strong attack) .i.b4+ 1 6 c3 .i.d6 17 'lt'xa7+ �c8 18 0-0-0
1 6....1ldg8 17 llt2 Wc6 18 .i.d2 g4 19 f4 .i.c4 1i'xg2 19 d51 .i.xdS 20 llhgt 1i'xh2 21 aS
and Black has an attack. The game Karpov­ with a raging attack, Hort-Short, Bundesliga
l .arsen, Ttlburg 1 980 continued 20 dS .i.xdS 1 986) 1 3 0-0-0 a6 1 4 dS .i.h31? 1 S .i.f4 lld7
21 f5 :C:8 22 a3 lle4 2.1 %let l:the8 24 llxe4 1 6 llhet .i.c7 1 7 .i.g3 and White's position
Axc4 25 �h2 1i'c5 and Black was clearly looks preferable.
better. c) 1 1 ...�b81? is probably best: 1 2 lbc3 b6
b) 1 0 0-0 0-0-0 1 1 .i.e3 .i.b4 1 2 c3 f6 13 1 3 'lfa6 .i.b4 (13...Wc6 1 4 lbb5 1l'b7 1 5
1fg3 with a further split: 1Wxb7+ �xb7 16 0-0-0 is better for White) 1 4
bl) 1 3....i.d6 14 .tf4 .i.f8 1 S .d3 gS 16 0-0 .i.xc3 15 bxc3 Wc6 and Black has some
.i.c3 h5 17 lbd2 (we prefer 1 7 c41? .i.fS 18 compensation for the pawn.
ti'd2 h4 19 dS g4 20 lbc3 h3 21 g3 and 1 1 c3 i.d6 1 2 1ta5
White is slighdy better) 1 7 ... h4 1 8 f3 .J:lg8 19 White has other options, but none that is
c4 f5 20 f4 gxf4 21 llxf4 Wg7 with an un­ comforting:
clear position, Magcm Badals-Macieja, Ba­ a) 1 2 � 0-0 1 3 0-0 .i.g4 14 Wh4 llfe8
lumi 1 999. 1 5 tnd2 f5 16 h.l .i.e2 17 life 1 'Wb5 1 8 c4?!
b2) 1 3 ....i.e7 t4 lbd2 hS 1 5 f3 h4 1 6 ..t2 .i.xc4 1 9 b3 .i.dS 20 .i.f4 llxet+ 21 llxe1
hJ ( 1 6...g5 1 7 c4 is better for White - Yusu­ 1i'd3 22 .i.xd6 Wxd2 23 lieS was agreed
pov) 1 7 g.l 'Wb5 1 8 b3 llhe8 1 9 c4 •aS 20 drawn, Movsesian-Haba, Pardubice 1998,
n3 'li'c3 21 dS .i.fS with chances for both although of course Black is to be preferred

135
The Petroff Defence

here. Instead White should play 18 b3 .i.d3 Other moves played here include:
1 9 c4 ..d7, although Black retains excellent a) 16 :ret b6 1 7 •a6 .i.xh2+ 1 8 Wxh2
compensation. :xe3 1 9 Wgt 1le6 20 "lffl :ea 21 :xe3
b) 1 2 ..xg7 0-0-0 1 3 �2 was played in ..xc3+ 22 Wf2 Wd3 and Black was slightly
00-Khalifman, Sochi, 1 984. Now after better, Barcenilla-Ye Rongguang, Beijing
1 3 .. .'Vc6!? 1 4 f3 :hg8 1 5 1fxh7 :Xg2 Black 1 992.
has excellent play for the pawns. b) 1 6 .i.f2 b6 1 7 1fa6 Lh2+ 1 8 Wxh2
1 2...•c6 13 f3 1fh6+ 1 9 Wgl 1fxd2 looks good for Black,
bast..-d on 20 c4?! .i.xf3! when White is in
trouble.
18 Jie2 1 7 c4
.•

1 7 :ret is met by 1 7 ..b6 t R 1fa6 :Xd2 1 9


.

.i.xd2 .i.c4 20 d 5 .i.xh2+ 2 1 Wh1 ..xdS and


Black will win on the kingside.
1 7 ....i.xh2+

1 3 0-0?! �d5 gives Black a clear edge.


One game continued 14 f3? b6 1 5 1fa6 .i.c4
1 6 d5 1fxd5 (16....i.xh2+ 17 Wf2 .i.g3+ 1 8
Wgl 1fxd5 19 .a4+ b S 20 .c2 0-0 also
wins for Black) 1 7 ..a4+ bS 1 8 ..d1 •es 1 9
W f2 'irxh2 20 f4 .i.xfl 0- 1 Klinger-Wolff,
Baguio 1 987.
1 3 .i.d5 14 llXI2 0-0 1 5 0-0
..• 1 8 Wh 1 ?
1 5 Wf2 b6 1 6 .a6 f5 is also unappcrising The last chance was with 1 8 Wxh2 1t"d6+
for White. 19 Wht "llg3 20 D.gt WxgS 21 11fxc7 (not 21
1 5 ...:fe8 �fl Wh4+ 22 �2 .i.xf3! 23 1fxc7 .i.e4 and
Black should win) 21 ...lLd2 22 cxdS :xd4
2..'\ :ge 1 :d2 24 1fh2 ...xdS, although the
endgame a pawn down leaves few drawing
chances.
1 8 ...•g6 1 9 trxd5 .i.f4 20 g4 .i.xg5 0-1

Game 57
Dolmatov-Mamedyarov
Moscow 2002

1 e4 e5 2 ll:}f3 ll:}f6 3 d4 ll:}xe4 4 .i.d3 d5


5 ll:}xe5 lDd7 6 ll:}c3
White also has a tame option in 6 0-0
1 6 .i.g5? �xeS 7 dxeS �cS 8 �c3 (8 .i.e2 .i.e7 9 .i.c3

136
3 d4: The Main L in e

0-0 10 f4 f6!? 1 1 cxf6 l:xf6 12 c4 .le6 1 3 White hasn't quite got the most from his
CKdS 'ii'xdS 1 4 lbc3 1i'xd1 1 5 l:axdl c6 1 6 position.
:02 aS 1 7 b3 l:ff8 1 8 l: fd l l:ae8 lead to a2) 9 llb1 .lxd4?! (9. ..c!Dxe5, transposing
even chances in Wedberg-Rozenailis, Vasby to 6...�xe5, is better) 1 0 �xd7 J.xd7! 1 1
2000) :Xb7 0-0 ( 1 1 ....lc6? loses to 1 2 .la31 .lxb7
13 llel+ �d7 1 4 .lf5+ �c6 1 5 Wxd4 and
mate is imminent) 1 2 .lxh7+ �xh7 1 3 Wxd4
and White has a dear edge.
b) 6...�xe5 7 dxeS .lb4 (7...�4?! 8
.lxe4 dxe4 9 �5 'ild8 10 .lgS 'ilc.l7 1 1 e6
fxe6 12 �5+ g6 1 3 lbf6+ �f7 14 �4 gives
Black serious problems, while 7...lbc5?1 R
�xd5! �xd3+ 9 1i'xd3 J.e6 10 �f4 obvi­
ously favours White) 8 0-0

and now:
a) IL.c6 (the solid choice) 9 f4 f51? 10 llk2
( 1 0 exf6?1 is weaker: 1 0...1Wxf6 1 1 f5 - 11
llc l+ .lc7 1 2 ..c2 �xd3 1 3 cxd3 �fl and
us White has no �5+ coming, Black is bet­
ter - 1 1 ...�xd3 1 2 Wxd3 .le7 13 �2 0-0 1 4
ltlg3 'ilf7 1 5 .le3 b6 and Black was slightly
hettcr, Rozentalis-Turov, Montr'->al 2001)
I O .le7 11 .le3 0-0 with chances for both
•••

sides. and now:


b) 8 ...�xd3 9 .xd3 c6 10 �2 f!fJ 1 1 Wg3 b1) 8 ... .lxc3 9 bxc3 .lc6 with a funher
.i.g7 1 2 �f4 0-0 (12....lxe5?! looks risky: 13 branch:
J%c 1 f6 1 4 lDd3 0-0 1 5 �xeS fxe5 1 6 l:xe5 bt l) 1 0 ...el!? �c5 (1 0...f5 1 1 exf6 ...xf6
"nd Black is weak on the dark squan.-s) 1 3 1 2 .lxe4 dxe4 1 3 •xe4 0-0 14 .ta3 l:fe8 1 5
lbh5!? gxh5 1 4 .lh6 .lg4 1 5 .lxg7 �xg7 1 6 Wxb7 1Wxc3 1 6 .lb2 1t'c4 and White is only
h3 •gs 1 7 f3 f5 18 exf6+ l:xf6 1 9 fxg4 l:affi slightly better) 1 1 .lbS+ c6 1 2 .la3 �d7 1 3
20 :Xf6 l:xf6 21 :Ct hxg4 22 hxg4 :n with J.d3 c5 14 f4 g6 15 c4 1Wc7 16 cxd5 J.xd5
level prospects, Tiviakov-Van Wely, Leeu­ 17 c4 R.e6 1 8 l:b1 with the initiative for
wardcn 2002. White, Reefat-1-lossain, Dhaka 2003.
6 lllxc3
... bt2) to f4 f5 1 1 exf6 Wxf6 1 2 f5!? (12
Black has a couple of valid alternatives ttl R.xe4 dxe4 1 3 �5+ 9f7 14 'iWbS+ c6 1 5
1his natural move: Wb4 J.dS 16 a4 b6 was equal in Palac­
a) 6 ... .tb4 wins a pawn, but is quite dan­ Arkhipov, Belgrade 1 988) 1 2 ... .tf7 (not
gerous: 7 0-0 �xc3 8 bxd .lxc3 and now: 12 ...lxf5?1 1 3 J.bS+ c6 1 4 1t'xd5 l:ffi 1 5
at) 9 .la3?! is too optimistic: 9...�xc5 .la3 lld8 16 9c4 and Black is under serious
�Jut not 9 ... .lxa1?1 10 �c6!? bxc6 1 l Wc2+ attack) 1 3 J.xe4 dxe4 14 J.e3 0-0 and Black
lbc::S 1 2 .xeS+ .lc6 1 3 l:xa1 with an attack) is probably not worse in this complex posi­
I U dxcS .lxa 1 1 1 •xa1 and we feel that tion.

137
The Petroff Defence

b2) 8...lt�xc3 9 bxc3 .i.xc3 10 llb1 0-0 Or 10 llb1 ..d7 1 1 .i.gS 0-0-0 12 0-0 h6
(1 0...1t'e7 1 1 llb3 .i.xe5? - Sax gives 1 3 Le7 11'xe7 1 4 11'e2 (14 l:lb3 cS 1 5 l:lbS
1 t ....i.b4 12 f4 with unclear play - 1 2 lle1 c4 16 .i.fS llhe8 with equal) 1 4...1ic5 1 5
leaves Black in trouble; Sax-Nunn, Brussels 1id2 d4 1 6 cxd4 llxd4 17 .e3 llhd8 and
1985 concluded 12 ..0-0? 13 1rh5 f5 14 .i.f4 Black was no worse in Kremcnietsky-Pripis,
t -O) I t .i.xh7+ Wxh7 12 .d3+ Wg8 1 3 Moscow 1 977.
'lfxc.l d4 1 4 'iWg3 ..d7!? 1 5 c3 d3 16 lld1 10 g6 1 1 •t3 f5 1 2 exf6
•..

11'g4 17 ..xg4 (17 llxd3 .xg3 18 llxg3 .i.fS 1 2 llbt 'ireS 1 3 .i.e3 0-0 14 0-0 cS also
19 lla 1 llfd8 gives Black enough play for the gives Black decent counterplay.
pawn) 1 7 ....i.xg4 1 8 f3 .i.c8 19 llb3 cS is 1 2 -*.xf& 1 3 0-0 0-0 1 4 J.a3
.•.

level, Roiz-D.Fridman, Pardubice 2002. 1 4 l:lb1 c5 1 5 llxb7 1ic8 16 llb1 .i.xc3


7 bxc3 c!bxe5 offers chances to both sides.
7 ... .i.d6 transposes to 5....i.d6 6 0-0 0-0 7 1 4 .i.xc3 1 5 JZad1 .i.d4+ 1 6 c;th1 c5 1 7
•.•

llk3 lbxc3 8 bxc3 lbd7. c3 ...a5


8 dxe5 J.e7

After 17 ....i.xc3 18 .i.xc5 l:lf6 19 .i.c4


Other moves: 1ia5 20 llxd5 .i.xd5 21 1ixd5+ 'it'h8 22 .i.d4
a) 8....i.e6 9 ltb1 1l'c8 10 .i.gS h6 1 1 .i.h4 ..xd5 23 .i.xf6+ .i.xf6 24 .i.xd5 the game
.i.cS 1 2 0-0 c6 13 Wht gS 14 .i.g3 11'd7 1 5 f4 would end in a draw (but not 18..1tf7 19
gxf4 1 6 .i.h4 .i.e7 1 7 :Xf4 0-0-0? (Movse­ .i.xg6 hxg6 20 11'xc3 when Black has vulner­
sian-Wc:.:glarz, litomysl 1995) and now 18 able dark squares).
.i.a61 would have won after 1 8...bxa6 19 11'fl 1 8 cxd4 •••3 1 9 .i.f5! ••6 20 .i.xe&+
..c7 20 1Wxa6+ 'it'd? 21 llb7. Instead Black ••e6 21 dxc5 JZad8 22 l:lfe1 •c6 23
should play 17 ..ixh4 1 8 1lxh4 11'e7 with an l:le7 l:ld7
unclear position. 23 ..1lde8 24 l:lxc8 llxe8 25 f5 'IVxcS 26
.

b) 8....i.c5 9 0-0 .i.e6 (9...1ih4?! loses a fxg6 hxg6 with unclear play was a more chal­
pawn to 10 .i.bS+ c6 11 1Wxd5) 10 D.bt .i.b6 lenging try.
1 1 Whs h6 12 Wh1 •e7 1 3 f4 J!fl 14 1t'e2 24 l:lxd7 •xd7 25 l:lxd5 •a4 26 f5
0-0-0 1 5 a4 •cs 16 llbS 11t'c6 (or 16. ..xd.. •••2 27 h3 l:lxf5 28 l:ld8+ Wg7 29
17 .i.d2 •c6 1 8 a5 .i.cS 19 l:lfb1 with an •xb7+ Wh& 30 •e4 ••1 +
attack) 1 7 aS!? a6 18 axb6 axbS 1 9 bxc7 with Or 30...1lxc5 31 •f4+ llg5 32 •f8+ 'it'hS
compensation for the exchange, Reefat­ 33 ..f3+ with perpetual check.
Vakhidov, Dhaka 2003. 31 Wh2 ••5 + 32 1rxe5 :XeS 33 l:lc8 a5
9 'Wh5 .i.e& 1 0 f4 34 Wg3 a4 35 Wt4 l:le1 36 :as l:lc1 %-%

138
3 d4: The Msin Line

15...'ifxh8 1 6 .i.xe6+ �d6 17 .i.xtiS (or 17


Game 58 h3 :CB 1 8 .i.f4 gS 19 .if5 gxf4 20 J:lxe8
Timofeev-Raetsky 'ifxd4 21 .txe4 fxg3 and Black wins)
Correspondence 1982 17...'ifxh2 1 8 .1xe4 'ifgt+ 1 9 We2 'lfxct
•------• when White's position is critical.
1 e4 e5 2 Q)f3 Q)f& 3 d4 Q)xe4 4 .i.d3 d5 b2) 1 2 .ig8! 'ffh4 (or 12...lDe4+!? 1 3 �d1
15 ttlxe5 Q)d7 6 it:lxf7 .ixg8 14 lDg6 'ifxeS 1 5 lDxe5 .id6 and
This sacrifice is an attempt to achieve a Black has some compensation for his pawn)
draw directly from the opening, but Black 1 3 .if7+1 �dB 14 :Xe6 'i!fgS+ 1 5 �c2 'lfxg2
has enough resources to play on. and the position is completely unclear.
e �xf7
... c) 7 'lfe2
Also to be considered is 6 ...'1fe7!?, which
is bit risky but playable. Now we have a
a
many of variations to look atl

and now:
ct) 7...�f7?! 8 'lfhS+ �f6 (after B �e6..•

9 .ltxc4 dxe4 1 0 dS+ �d6 1 1 .tf4+ lDe5 12


a) 7 lDcs is good for Black: 7 ...lDxe5 8 llk3 the attack also seems to be very danger­
dxe5 •xeS 9 'lfc2 .td6 I 0 llx12 llk5 1 1 ous) 9 0-0 'iff7 1 0 Wh4+ g5 1 1 .ltxg5+1
..xeS+ .i.xeS 1 2 lD£3 lDxd3+ 1 3 cxd3 .i.f6 lDxgS 1 2 f4 We6 (1 2...'1fe7 1 3 fxg5+ �g7 14
with a better endgame due to the two bish­ lDc3 also gives White a winning attack) 13
c ops ami the stronger pawn structure. fxg5 'lfg7 14 lDc3 lDb6 15 J:lf6+ 'iPcl7 16
b) 7 lDxh8 lDc3+ 8 �d2 lDxdl 9 J:le1 11af1 �e8 1 7 J:lf7 Wxf7 1 8 J:lxf7 �f7 19
liJxf2 10 .ltxh7 (but not to :Xe7+? .ixe7 1 1 g6+ �c6 20 'ifet+ 1 -0 Gurgenid7.e-Bcllin,
lLic3 lDf6 and Black wins) tO...lDc5!? Thilisi/Sukhumi 1977.
(IO...lDc4+ 1 1 J:lxe4 dxe4 12 .ig6+ �dK 13 c2) 7...'1fxf7 8 f3 lDl'lf6 9 lDd2! Qcss ener­
liJf7+ �e8 1 4 lDd6+ 'h-'12, I.Zailliev-Karpov, getic is 9 fxe4 .i.g4 1 0 'ife3 dxe4 1 1 .i.xe4
L eningrad 1966, is safer) 1 1 J:lxeS .ie6 and 0-0-0 12 0-0 11fcl7!? 13 .i.d3 - but not 13 c3?
now: :Cs 14 �12 lDxc4 1 5 lDxe4 .tf5 16 .l%f4 gS
b1) 12 .i.g6+?! �d7 1 3 .tf7 (1 3 .i.f5? and White is on the ropes - 1 3.. Wxd4 14
.

loses to 13...'1fg5+ 14 �e2 'ifxct 15 .ixe6+ 'ifxd4 J:lxd4 1 5 .i.e3 .i.c5 with level chances)
Wc6 16 .i.xcl5+ �b6 - Yusupov) 13 ...lDc4+ 9 ...'1fh5 10 fxe4 ..tg4 1 1 'ife3 dxe4 t 2 lDxe4
1 4 �e1 (14 �d3 is met by 1 4...llk5+1! 15 0-0-0 with a further split:
clxc5 .i.f5+! 16 :Xf5 'lfe4+ and 81ack wins) c21) 13 lDxf6 gxf6 1 4 0-0 .i.d6 1 5 g.� (or
1 4...'ifh4+ 1 5 g3 (Smerclon-Solomon, Gold 15 h3 llhg8!? 16 'lfh6 - but not 1 6 hxg4
< :oast 1999) and now Black can play Wh2+ 17 �f2 'lfh4+ 1 8 �e2 J:lde8 1 9 h4

139
The Petroff Defence

:lxg4 20 J.£5+ �d8 21 .Lg4 Wxg4+ 22 .13+ Wc6 could still transpose to the game,
�d3 :xe}t 23 J.xc3 f5 and White has a lot though it looks risky with the king on f6) 9
of defending to do - 1 6..ixh3! 17 Wxh5 Wh5+ with a repetition.
:Xg2+ 1 8 �h 1 l:th2+ with a draw) 15 ....ZZ.he8
(15...Ade8 16 '1Vh6 would benefit White).
After 1 5 ...:he8 Black has enough play, for
example 1 6 Wh6 J.c5! 17 dxc5 (17 'Wxh5
.i.xd4+ 1 8 �g2 hh5 is equaQ 17 ...Wxc5+
1 8 �h 1 •d5+ and Black delivers perpetual
check.
c22) 1 3 0-0 is stronger: 1 3...lnd5 (after
13...tbxe4?! 14 .i.xe4 .i.d6 1 5 h3 l:[hf8 16
J.d2 White is simply a pawn ahead) 14 9g5
li)b4 1 5 h3 (15 tb£2 tbxd3 1 6 tbxd3 D.xd4
W"dS played in Zhao Zhong Yuan-Solomon,

Gold Coast 1 999; following 1 7 1i'xh5 J.xh5


18 .i.e3 lle4 1 9 l:tact the position looks 8 1i'e2
even) 1 5.....te2 1 6 .i.xe2 .xe2 1 7 'Wf5+ �b8 8 .i.xe4?1 is less reliable: 8 ...dxe4 9 d5+
1 8 .i.f4 and we slightly prefer White. �e7 10 ..tgS+ tbf6 1 1 tbc3 .irs 1 2 0-0-0
�d7 1 3 Wh4 J.d6 and Black is a piece up.
A more serious option is 8 .g4+ �d6
(8...Wf7 draws) and now:
a) 9 .if4+? Wc6 lO .e6+ (on t o .i.xe4
luckily for Black he has 1 O li)f61 1 1 'il'g5 h6,
•.•

winning) 1 o....i.d6 1 1 .ixe4 tbf6 and White's


queen is trapped.
b) 9 ..txe4?l dxl.-4 10 1i'xe4 (10 .if4+ We7
1 1 1i'g3 .!£Jf6 and Black is close to winning)
10....!£Jf6 1 1 .i.f4+ Wd7 12 1i'e2 .i.b4+ 1 3 c3
lle8 1 4 ..te5 .i.d6 and Black is certainly better.
c) 9 'il'e2 11h4 10 g3 'il'e7 1 1 c4 when
White has compensation, but is it enough?
7 'tih5+ We61? 8 .i.d6 9 f3 1i'h4+
.••

The adventurous approach. Also possible


is 7 ...rile7 and now:
a) 8 1i'xd5?l is known to be bad after
8 ...tbdf6 9 1ib3 (9 'ilfc5+? �fl 1 0 .i.xe4
.i.b4+ 1 1 c3 J:lcs and Black wins) 9 ... .i.e6 1 0
Wxb7 �f7!? (Yusupov) 1 1 0-0 (11 .i.xe4
l:tb8 12 •c6 l:lb6 1 3 •as 1i'xd4 and Black
wins) 1 t ..id6l 1 2 f3 .ixh2+! 1 3 �h2
•d6+ 1 4 �g1 'ilfxd4+ 1 5 Wh2 •e5+ 1 6 �gl
tbg3 and Black has the advantage.
b) 8 .e2 �fl (8...C�d6? 9 .i.f4+ Wc6 10
.Lc4 dxe4 11 tbc3 a6 12 1i'c4+ �b6 13
.i.xc7+! and White wins - Olthof; 8...Wf6 9

140
3 d4: The Msin L ine

10 g3? Another option is 1 8....td61? 19 cxd5 l£if6.


This is already the losing mistake - White 1 9 llf1 i.f4 20 llxf2 llxe3+ 21 Wf1 g5
must move his king:
a) 1 0 �ft .i.xh2 1 1 c4 c6 12 ltkl2 .i.e5 1 3
:xh4 li)g3+ 1 4 �f2 l£ixe2 1 5 �xe2 .i.f6 1 6
:xh7 l:txh7 1 7 .i.xh7 .i.xd4 would probably
lead to a draw.
b) 10 �d1 �fl (or to...lC!c5 1 1 fxe4
lbxd3 12 exdS+ l12 cxd3 %lf8 is a mess]
1 2 ... �d7 1 3 1fxd3 1Wg4+ 14 'lrt1 'ifxd4+ 15
..td2 with chances for both sides) 1 1 fxe4
ll'lf6 (after 1 l.. .dxe4?! 12 l:tft+ l£if6 1 3 J..c4+
rize7 14 h3 White is better) 1 2 .!0d2 dxe4 13
lC!xe4 i..,_t4 14 lCigS+ �fB 15 l£lt1 1Wh5 and
Black has good play for the pawn.
10 ...i.xg3+ 1 1 �d1 The endgame is horrible for White and, in
After 1 1 �ft .i.f4 White is dojng very correspondence chess. almost impossible to
badly. hold.
1 1 ...lDdf6 1 2 fxe4 �7 1 3 •g2 i.g4+ 22 'Dc3 dxc4 23 'De2 lle4 24 llc1
1 4 i.e2 llhe8 25 'Dxf4 gxf4 26 llxc4 c6 27 J:lc5
14 �d2 .i.f4+ 1 5 �c3 .i.h3 1 6 '1Vf3 .i.xct llxd4 28 llf5+ �g6 29 ll5xf4 llxf4 30
1 7 l::txct dxc4 18 i.c4+ �fB 1 9 'lrg3 11rxg3+ llxf4 lle5 31 llg4+ Wf6 32 l:tg2 a5 33
20 hxg3 l:te8 also gives Black a material edge. b3 l:tc5 34 �1 llg5 35 J:lf2+ �6 36
14 ...i.xe2+ 1 5 �xe2 11rxe4+ �d2 J:lh5 37 �3 b5 38 J:ld2 c5 39 llg2
Or 1 5... lC!xc4 1 6 l:tfl+ �8 1 7 ltkl2 .xh2 J:lh3+ 40 �c2 �5 41 lla2 b4 42 �2
I R 1Vxh2 i.xh2 1 9 lClxe4 l:te8 and Black �c& 43 a3 a41 44 axb4 llxb3+ 45 Wa2
should win, though it will take some work. cxb4 46 J:le4 �c5 47 J:le5+ �4 48 lla5
1 6 11rxe4 �xe4 1 7 c4 llae8 1 8 i.e3 'Df2 J:lh3 49 llxa4 Wc3 0-1

141
The Petroff Defence

Summary
In the main line Black has very much switched to 7 ....i.d6 recendy. However, the annotated
games prove that 7 ..11rh4 remains interesting and gives Black excellent counterplay after 8 c4
.

0-0-0 9 c5 g6, and especially after 9...g5. Nevertheless, the truth is that 7 ...'1fh4 leads to ex­
tremely sharp positions that are not to everyone's taste.
As Game 54 shows, 6 ltlxd7 .i.xd7 7 0-0 ..f6 allows White to win a pawn after 8 .i.xe4
dxe4 9 llk3 1i'g6 10 ltlxe4; Black has some compen.ution, but no more than that. However, if
we insert the moves 7...'1fh4 8 g3 and now 8...'1rf61 9 .i.xe4 dxe4 1 0 ltlc3 'lrg6 t 1 lLlxe4 0-0-0,
Black has a full compensation due to the weakened light squares on the kingside.
The brave 6 ltlxf7 (Game 58) has not been played much rccendy. Black should accept the
'gift' with 6...'1Pxf7 and following 7 'lfhS+ he should move forward with 7 ...We6! - Black•s
chances arc preferable in the arising complications. On the other hand, White has a plus aftet
6.....e7 7 'lrc2! 'lrxf7 8 f3 ltlf6 9 lLld2 1i'h5 1 0 fxe4 .i.g4 1 1 ..e3. The assessment of the Zait­
sev-Karpov game has changed and become more precise. At present the best line is considered
to be 6 ...'1rc7 7 ltlxh8 lLlc3+ 8 Wd2 ltlxd1 9 l:te1 ltlxf2 10 .i.xh7 ltle5 1 1 1lxc5 J.c6 1 2 .i.gS
1i'h4 l3 .i.f7+ �d8 14 :xe6 'lrg5+ 1 5 �e2 'lrxg2, which is 'a hell of a mess'.

1 e4 e5 2 lilf3 lilf6 3 d4 lilxe4 4 .tell d5 5 lilxe5 lild7 (D) 6 lilxd7


6 'lre2 - Game56; 6 lLlc3 - Game 57; 6 lLlxf7 Game 58
-

6 . . . .i.xd7 7 0-0 .td6


7...'1rf6 - Game54; 7 .ie7 - Game55
.

7...1i'h4 8 c4 0-0-0 9 c5 (D)


9...g6 - Gt1111e5J
. 9. .g5
10 J.d - Ga1nt 51
10 f3 - Game52
1 0 lLlc3 .i.g7: 1 1 g3 - Gt111e1 49; 1 1 lLle2 - Game50
B c4
8 lLlc3 Game 48
-

B ..c6 9 cxd5 cxd5 (D) 10 lDc3


.

10 Wbs 0--0 1 1 1i'xd5 .i.c6 12 1ih5 g6 1 3 11'h3: 1 3....tb4 - Gt����t 46; 1 3. lbg5 - Game47
..

10 ...lilxc3 1 1 bxc3 0-0 1 2 1rh5 f5 - GanH 44


12 ...g6 - Gt111e1 45

5 .ltid7
.. 9 c5 9... cxd5

142
CHAPTER SEVEN I
3 d4: 5 .i.d6 and
. . .

Fifth Move Alternatives

1 e4 e5 2 �f3 tl)f6 3 d4 ltlxe4 4 i.d3 d5 the comer is the main question of this key
5 �xe5 position, one which is studied in Game 59.
In the previous chapter we looked at the 5 ...lbc6 attacks both White's centralised
main line with 1 e4 e5 2 lbf3 lbf6 3 d4 lbxc4 knight and the d4-pawn. l lowever, this is not
4 .id3 d5 5 lbxe5 lbd7. Now it is time to tJ1c best idea as White can simply swap tl1c
:;rudy other 5th move options for Black. knights to inflict doubled pawns on his op­
Until quite recently the symmetrical re­ ponent. Now 6 lbxc6 bxc6 7 0-0 �e7 trans­
sponse 5...�d6 (Games 59-62) was as popu­ poses to 5 ....i.e7 6 0-0 lbc6 7 lbxc6 bxc6 (sec
lar as 5 ..lLld7. After mutual castling White
. Game 63), while 6 li)xc6 bxc6 7 0-0 .i.d6
:macks the centre, but this can be done in transposes to 5...�d6 6 0-0 lbc6 7 lbxc6
more than one way. If White develops his bxc6 (sec Game 61). However, Bil!,ruer
lJUeen's knight for this purpose, he should demonstrated that White docs not have to
prefer 7 lbd2 (Game 61) to 7 lbc3 (Game castle; 7 'iWe2 creates problems for Black
62) because it avoids doubled pawns. along the e-file, and that's why 5 ..lbc6 is
.

However, the main way of attacking the almost out of usc.


centre is with 7 c4 (Games 59-60). Black has The continuation S...�c7 (Game 63) docs
numerous responses against this. Strengthen­ not challenge the c5-knight and is dictated by
ing the d5-pawn with 7 ...c6 is a bit passive, the wish to complete dL-vclopmcnt as soon as
while the counterattack on the d4-pawn with possible. A bit passive, 5 ... �e7 has never
7...l'lk6 (Game 60) isn't sufficient to equalise drawn much attention. However, even after
in view of 8 lbxc6 bxc6 9 c5 �e7 I 0 �12 or tl1e most natural 6 0-0 0-0 7 c4 c6 it is un­
1 0 lbc3 - the pawn doubling is unfavourable clear how White can 1-,rain a visible advantage.
ti1r Black. A sharp and critical position (de­ Probably White should consider Stcinitz's
spite a queen exchange) is reached after recommendation of7 ltcl !?.
7...�xc5 8 dxc5 lbc6 (!L�e6 docs not
ClJUalise) 9 cxd5 (the only try for an advan­ Gaf!Je 59
tage) 9 ...'iWxd5 10 'iWc2 lbb4 1 1 �xe4 lbxc2 Van Der Wiei-Mellado
1 2 �xd5, and this has been considered a Elgoibar 1998
tabiya of the whole 5...�d6 system for many
years. The fate of Black's knight driven into 1 e4 e5 2 tl)f3 tl)f6 3 d4 �e4 4 i.d3 d5

143
The Petroff Defence

5 .lt»te5 .i.d6 i.xd3 12 Wxd3 dxc4 1 3 Wxc4 ltla5 14 We4


...e7 and Black is okay (Kapengut).
c) 9 .i.f4 lllb4 to cxd5 (or 10 ll:la3 lllc5 1 1
i.bt dxc4 1 2 lllxc4 i.c6 1 3 b3 ..xd1 1 4
:xdt :adS with equality - Yusupov)
lO.. :tl'xdS t 1 .i.c2 .i.f5 t 2 llla3 l:tad8 1 3
.i.c4 1i'c6 1 4 -.,3 aS 1 5 Ldt 1Wg6 1 6 :xd8
ll.xdS 17 lld 1 ltf8 18 We3 h6 19 ll:lb5 ltlc2
20 We2 c6 21 lild6 lllc5 with an unclear
game, Raetsky-rilatov, cortespondcncc t 982.
9 11'xd5 1 0 .c2
.•.

Other options are considered in Game 63.


6 0-0 0-0
Sixth move alternatives for both sides are
studied in Game 61.
7 c4 .i.xe5 8 dxe5 lbc&
8... .i.e6 is less reliable. White can put
Black under immediate pressure with 9 cxdS
1i'xd5 1 0 1i'c2 f5 1 1 exf6 lihf6 1 2 liX.1 1i'e5
(1 2...9c6 1 3 b3 lilbd7 1 4 .i.a3 ltlc5 15 .i.bS
1IVb6 1 6 Let also looks promising for
White) 1 3 lllc4 ll:lg4 1 4 lllg3 hS (14 ...h6? 1 5 1 0 'lff3? is much weaker as Black has
.i.xh61 lilxh6 1 6 llae1 9f6 17 :xe61 would 1 0 ... .i.f5l 1 1 'lfxf5 1fxd3 12 lllc3 lllc5 1 3
t."'ld the show) 1 5 .LIZ .i.dS (1 5...h4? 1 6 9hs :res 1 4 f4 •d4+ 1 5 �ht :a<�s 16 •f3
:net and White takes on e6 again) t 6 .i.c3 1i'c4 when his position is preferable. To
1i'g5 1 7 .i.h7+ �h8 1 8 ...g6 1i'xg6 19 .i.xg6 make matters worse, White only needs to
with a clear plus for White, Liberzon- play l 1 :C 1 ?! for Black to grab the advantage
1-lcnnings, Debreccn 1 968. with 1 t ...LdS 1 2 i.xe4 .i.xe4 1 3 ...g3 .i.xb1
9 cxd5 14 :xb1 lllxeS 1 5 .i.e3 :res, when Black
White needs to try to refute Black's set-up should convert his pawn, Zelcic-Ascic, Rabac
altogether in order to play for an advantage. 2003.
Other moves are less critical: 1 0. . .tnb4
a) 9 f3 lllcS 1 0 cxdS lllxd3 1 t 1i'xd3 ltlb4 Absolutely the main line, but not the only
(1 1 ...lllxe5 1 2 Wd4 :Cs 1 3 lilc3 .i.f5 1 4 .i.e3 move in the position because Black can also
fie7 also looks okay for Black) 12 -.,3 play 1 o... .tf5. For example, 1 1 lllc3 lllxc3 1 2
lllxdS 1 3 lldt c6 1 4 lllc3 1l'b6+ 1 5 ..xb6 .i.xf5 llld4 13 .i.xh7+ �h8 14 'lfd3 li:ke2+
lllxb6 1 6 b3 Ji.e6 17 �f2 .rl.fd8 1 S :xd8+ 1 5 �h1 'ifxe5 16 9h3 lilf4 17 .i.xf4 (17
ltxdS is drawish, Raetsky-Kuznctsov, corre­ 'ifh4 g5 18 1i'h6 9g7 1 9 ..xg7+ �g7 20
spondence t 9S3, .te4 c6 is very close to equal, if not simply
b) 9 f4 .i.f5 (maybe an <.oven safer way to cquaQ 1 7...•xf4 1 8 Ldt 'irh6 (1 S...Ilfd8?1
equalise is 9...ltlb4 to cxdS WxdS 1 1 .i.xe4 1 9 .tc4+ 9h6 20 'ifc3 lllc6 21 :d5 and
1i'xe4 1 2 ltlc3 9g6 1 3 .i.e3 .i.g4 1 4 llldS!? Black was under pressure in Mi.Tseitlin­
lllxdS 1 5 ..xdS b6) 1 0 lllc3 ltlxc3 1 1 bxc3 Kondali, correspondence 1 990) t 9 1ixh6

144
3 d4: 5 . . �d6 and Fifth Mo ve Alterna tives
.

wch6 20 IZ.xd4 �h7 21 lld7 and White has a should accept an inferior position with
preferable rook endgame. However, whether 1 6...bxc4.
this is enough to win is hard to tell. 14 i.f4
1 1 �xe4 �xc2 1 2 i.xd5 i.f5 14 .i.e4 �xa1 1 5 �3 is the alternative,
t2...lt�xa1 gives White a slight advantage but probably less critical (1 5 .tf4 simply
after 1 3 i.e4 (13 c6 lbc2 1 4 exf7+ llxf7 1 5 transposes): 15 ... £5 16 exf6 .i.h3 17 llet
lldl .t f5 1 6 .i.xf7+ *xf7 17 ltk3 c6 1eads to llae8 18 .L12 (1 8 .i.e3 IZ.xe4 19 �xe4 �2
immediate equality) 1 3..l[c8 14 l&3 llxc5 20 llct �xe3 21 fxe3 c6 22 �gS .tf5 23 f7+
1 5 .i.d2 llxe4 1 6 �xe4 �c2 1 7 llct i.f5 1 8 *h8 with equality, Tal-1imman, Reykjavik
f3 li)d4 1 9 llxc7 .lxe4 20 fxe4 b6 21 *f2. 1 987) 18..l[xe4 1 9 �xc4 �c2 20 :C1 li)d4
The endgame is uncomfortable for Black, 21 llxc7 �f3+ (21 ...gxf6?! 22 .i.e3 �2+ 23
lhough not necessarily impossible to defend. Whl looks better; 21 ..1lf71? 22 llxt7 �f3+
Note that after 21 ...�6 White retains the 23 c;i;>ht �xf7 24 fxg7 *xg7 25 .i.e3 b6 26
pressure with 22 lle7! *£8 23 .i.b4, when �12 �xd2 27 .i.xd2 aS 28 f3 a4 29 ¢>g1
23 ...�5 loses to 24 llc7. .i.t.-6 30 a3 '12-1/z Tai-Karpov, Milan 1 975 -
13 g41 Black will place his king on f7 and White can
never make any progress) 22 �h1 �xd2 23
llxg7+ c;i;>h8 24 lbgs .i.f51 (the most precise;
24.. .:Xf6 25 IZ.xh7+ Wg8 26 1Z.xh3 IZ.xf2 still
gives White chances - Gipslis) 25 �f7+ llxf7
26 llxf7 .i.e4+ 27 f3 .lxf3+ 28 c;i;>gt Wg8 29
lld7 lDc4 30 Wf2 .i.c6 3t llg7+ wm 32
llxh7 li)d6 33 We3 �e8 and the endgame is
drawn, Raetsky-Bclomestnykh, correspon­
dence 1 982
14 ...�xa1 1 5 .i.e4

1 3 . i.xg4
..

Or 1 3....i.g6 14 f4 and now:


a) 1 4....td3? 1 5 .Ddt i.a6 16 .le4! (16 e6
�xa 1 17 exf7+ Wh8 1 8 .le3 :ads 19 lbc3
also looks very promising for White)
1 6...�xa1 1 7 .le3 .i.e2 1 8 :Ct .i.xg4 1 9
�3 and White has good winning chances.
b) 14 ... c6 1 5 .i.c4 bS with a further split:
b1) 1 6 .le21? hS 1 7 f5 .th7 1 8 gS �xal
19 c6 fxe6 20 b>6 .i.xg6 21 fxg6 llxfl+ 22
�fl probably leads to a draw, but this line 1 5 llct has also been tried a few times,
is by no means forced and Black needs to do but it is less dangerous for Black: 1 5 ...c6 1 6
defend accurately. .le4 (or 1 6 .i.g2 f6!? 1 7 exf6 IZ.xf6 1 8 .i.e3
b2) 1 6 f5 �xal?, as in Ginda-Witt, Galati llg6 19 *h 1 .tc<, 20 �d2 .lxa2 21 llxa 1
1 973, should have lost quickly to 17 ..i.d3 i.dS 22 .txdS+ cxdS and Black should not
l:lad8 1 8 .i.e4 lld4 19 �2 when Black will have any trouble at all) 1 6 .. £5 (or 1 6 ... f6 1 7
.

end up with material losses. Instead Black �3 fxcS 1 8 .lxeS llae8 1 9 f4 gS 20 llxa1

145
The Petroff Defence

gxf4 21 .i.d4 a6 with compk-tc equality) 17 Oddly enough the bishop is best placed
cxf6 :Xf6 1 8 .i.d (Sveshnikov-Tischbierck, here, where it seemingly has little influence.
Budapest 1 988). Now Black could have con­ The reason for this is found in the foUowing
tinued 1 8 .. .:C:8 1 9 lL!c3 lL!b3 (1 9...bS 20 .i.g2 line: 1 8 .i.e6 g5! 1 9 ..LgS l:lae8 20 .i.c4 (or
llg6 21 �h1 b4 22 lL!e4 .i.fS 23 l0g3 :xg3 20 .i.h6 :xc6 21 .i.xf8 .i.h3 22 :C3 %Zg6+ 23
24 hxg3 lLlc2 25 i.xc6 looks pnmusing for llg3 �g8 24 .i.e7 f4 25 ltxg6+ hxg6 26 lL!a3
White - two bishops and a weakened black f3 27 .i.c5 b6 28 .i.c.l4 .i.£5 29 b4 ru 30
queenside) 20 axb3 a6 21 i.d3 .i.fS and lL!xc2 .i.xc2 '/z-'/z OU-Khalifman, Vilnius
Black's position is at worst marginally infe­ 1988) 20...b5 21 .i.ft f4 (Rm:entalis­
rior; in fact we think it's equal. Ivanchuk, Minsk 1986). Mter 22 .i.f6+ llxf6
1 5...f5?1 23 exf6 .i.fS 24 f7 llf8 Black is no worse.
After this move we cannot find a route to 1 8 .. .llfd8 1 9 �d2
fuU equality. Probably Black should try This is the critical move. 1 9 f3 .i.hS 20
lS ...f61? 16 lL!c3 fx�:5, when his results have lLhl :U4 21 .i.e3 lZ.b4 22 lL!c4 (22 llxa1 f4
been encouraging: 23 .i.f2 llxb2 24 lL!c4 llc2 2S lL!d6 b6 is not
a) 1 7 .i.g3 :ad8 1 8 l:lxa1 .i.£3 1 9 .ixe5 clear) 22...l:la4 23 lL!a3 leads to a repetition,
J:td2 gives good compensation (Yusupov). Sax-Yusupov, Thessaloniki 1 984.
b) l7 .i.e3 .i.f3 1 8 l:lxa1 .i.x�:4 1 9 lL!xe4
b6 20 b4 aS 21 b5 :ad8 22 l:.ct ll£7 23 a4 h6
24 Wg2 J:td3 and Black had sufficient
counterplay in Kasparov-1imman, Paris
(rapid) 1 991 .

1 9 ...h6
1 9...llxd2?1 has an awful score. FoUowing
20 .i.xd2 :dB 21 .i.c3 :d i+ 22 llxd1 .i.xd l
White should play the foUowing plan: 23
.i.fl I (23 f4 lLlc2 24 Wf2 �PB 25 a4 aS 26
1 6 -td5+ .i.xa5 lL!d4 27 .i.ft .i.b3 1/z-1/.t Kasparov­
Much weaker is 16 .i.xb7?! l:lab8 1 7 .i.dS+ Anand, Unares 1991) 23 ...g6 24 .i.c4 �g7 25
�h8 1 8 f3 (18 lL!a3?1 l:lfd8 19 .i.c4 J:[d4 20 b4! (25 �:6+ Wf8 26 .i.f6 'iPe8 27 e7 lLlc2 is
.i.e3 lL!c21 21 .i.xd4 lL!xd4 and Black's more unclear - Yusupov) 25...W 26 .i.b3 when it
active pieces mean that he enjoys an advan­ looks impossible for Black to improve his
tage in the endgame) 1 8. ..th3 19 lld1 Lb2
. position. Black seemed to lose th�: foUowing
20 e6 l:.g2+ 21 Wh 1 l:.c8 22 .i.c4 llg6 23 endgame without any real chance: 26 ...�f8
.i.g3 f4 24 e7 l:ld6 25 llxd6 cxd6 26 .i.xf4 27 .i.d2 We7 28 f3 �e8 2<J �f2 b6 30 .i.f4
.i.d7, when the responsibility of holding this cS 31 bxcS bxcS 32 c6 c4 33 i..xc4 lL!a3 34
endgame lies entirely with White. .i.dS .i.a4 3S .i.e3 a6 36 .i.c5 lL!c2 37 i..c4
1 6...�8 1 7 llc1 c6 18 ..tg2 aS 38 �g3 lLle 1 39 Wf4 i..c6 40 .i.e2 h5 41

146
3 d4: 5 . . . �d6 and Fifth Move Alterna tives

h3 �dB 42 �gS .i.e8 43 c7+ �c7 44 f4 li)g2 27 h5 :.aea 28 .i.d3 �f3


45 .i.f1 �c6 46 .i.f2 �d6 47 .i.xg2 1-0 Mru­ Also possible is 28...b6 29 i.f2! (29
gala-Wisc, correspondence 1 999. .i.xf5?? l:txeS! wins for Black). Now White
20 h4 :.d3 21 .i.f1 ! has the edge after 29....i.c4 30 �c4 .i.dS 31
.i.xdS cxdS 32 l:tdt .
29 �2 .i.xh5 30 :.h1 .i.g4 3 1 .i.xa7
:.Xe57
This is simply a blunder. Black needs to
play 31 ...lla8 32 i.d4 l:ta4 33 �c3 :xf4+ 34
�g3 l:ta4 35 �c2 liaS 36 .i.b3 when after
36...llxe5 37 i.xeS White has some chances
to win the game, though nothing is clear.
32 fxe5 :.Xe5 33 lbh6+ ! �g8 34 .i.c4+
�8 35 :.h8+ r/;e7 36 lZ.b8 b5 37 :r.b7+
wdB
37...�f6 38 .i.ft l:td5 39 :C7 also wins for
White.
This was Timman's improvement on his 38 .i.b6+ �c8 39 :.c7+ Wb8 40 :.Xg7
play in an earlier matc�>ame with Yusupov: 1 -0
21 Axa1 gS! 22 hxgS hxgS 23 .i.xgS (23 .i.fl
hrxf4 24 .i.xt.l3 l%d8 25 c6 l:txd3 26 e7 .i.hS Game 60
27 li.)b3 b6 28 ltet .i.cB and Black is fine - Shirov-Vusupov
Yusupov) 23 .Jlg8 24 .i.f6+ Wh7 25 li.)ft f4
.
TerApe/ 1997
26 �h2 Jlg61 27 :C t l:th6+ 28 �gl l:tg6
with equality, Timman-Yusupov, 2nd match­ 1 e4 e5 2 �f3 c!Lif6 3 d4 o!bxe4 4 .i.d3 d5
game, Linares 1 992 5 �xe5 .i.d6 6 0-0 0-0 7 c4 �c6
21 ...:.d4 22 .i.e3 :.d5 23 :.Xa1 :.Xe5 24
.!Dc4 :.e6
In the stem game Black played 24...l:td5
25 .i.g2 l:tb5?1 26 l:tc1 l:td8 27 .i.xa7 l:td1 28
l:lxd 1 �xdl 29 �d4 f4 30 .i.c4 and White
enjoyed a clear superiority, Timman­
Yusupov, 6th matchgame, Linares 1992.
L ater Black improved with 2S..ltdd8 26 l:te1
'it>h7 27 �h2 �hS 28 .i.cS b6 29 .i.e? l:td4
30 lbcl6 l:tc:12 31 �xc6 l:tb8 32 Wg3 f4+ 33
Wxf4 llxf2+ 34 �g3 ltxb2 with n:asonable
drawing chances even though it is stiU a
tough defence, Bucher-Girtz, Biel 1998.
25 f41 .i.f3 This particular variation, like so many in
25 ...:C4 26 .i.d3 l:tac8 looks tticky, but af­ the Petroff, is a speciality of Yusupov's.
lCr 27 �f2! l:t4e6 28 �5 White has the ad­ 7...c6 is also possible, after which White
vantage. has a wide range of choices:
26 c!Lie5 .i.d5 a) 8 1i'c2 1Wh4 9 li.)f3 1i'hs 1 0 cxdS cxdS
26...�h7 27 Wf2 .i.hS 28 l:tc1 also fa­ 1 1 ..ixe4 dxc4 1 2 'ifxc4 li.)c6 and Black has
vours White. compensation (Yusupov).

147
The Petroff Defence

b) 8 cxdS cxdS 9 'irc2 .IL:8 10 f3 (10 R.xe4 R.d7 13 .i.d2 lr.b8 14 l:tab1 .i.£6 15 �h1 h5
dxc4 1 1 .xe4 .f6 1 2 .i.£4 �c6 looks fine t6 �2 h4 17 h3 'l'c8 18 J..a6 'Wd8 t9 .i.d3
for Black) to ...�f6 1 1 .i.gS h6 1 2 .i.h4 �a6 ..c8 20 .i.a6 the game finished with a draw
1 3 a3 -*.xeS 14 dxeS lr.xeS t S R.f2 and White in Kaspamv-Yusupov, Horgen 1995. 1 8
has compensation for his pawn, although it is �f41? � 19 lr.fet is interesting, but we
difficult to see how Black should ever end up believe that Black is okay after the foUowing
being worse. line: 19...-*.gS!? 20 .i.a6 ..dB 21 �xe6 lr.xe6
c) 8 llXJI seems to give White the edge af­ 22 :Xe6 R.xe6 23 .i.xgS ..xgS 24 ..xc6
ter 8 ...�xc3 9 bxc3 and now: .i.xh3! 25 R.n ! .i.xg2+! 26 .i.xg2 ..5 27
c1) 9 ... R.e6 10 £4 .i.xeS 1 1 fxeS dxc4 1 2 l:lg1 h3 - the game will end in a draw.
-*.xh7+!? Wxh7 1 3 1Wh5+ �g8 t4 R.gs •as b) 10 l0d2!? looks stronger: 1 0.. ..i.f6 1 1
1 5 .J:r.£3 �7 1 6 lr.g3 looks very dangerous �xc4 dxe4 1 2 .i.xe4 .i.a6 (t2....i.xd4 1 3
for Black. .d3 �h8 14 .i.g51? .i.xf2+ 1 5 lr.xf2 .xgS
c2) 9...lbd7 1 0 £4!? (tO �xd7 R.xd7 trans­ 16 Lh7 looks slightly better for White) 1 3
poses to S...llkl7 6 �xd7 R.xd7 7 0-0 .i.d6 8 :C 1 .i.xd4 1 4 1i'c2 lr.eR 1 5 .i.e3 .i.xb2 1 6
c4 c6 9 W �xc3 10 bxc3 0-0) 10 ...�f6 1 1 .i.xh7+ �h8 (Sanche-.r.-Morgado, correspon­
cS J..e7 t 2 5 J..d7 1 3 g4 gave White an at­ dence 1 978) and now after 1 7 •xb2 �xh7
tack in Korneev-Y.Hernandez, Mondariz 1 8 l:ad1 .e7 19 1i'c3 we think White has
1 997. some pressure.
c3) 9....i.xe5 10 dxeS dxc4 l t R.xc4 fle7
(1 t ...'Wxd1 1 2 lr.xd1 J.. 5 1 3 .L3 lr.eB 1 4 f4
is clearly better for White; it gets even better
after 14 ...�7? 1 5 e6! when White was win­
ning in Maroczy-Marshall, Paris 1900) 1 2 a4
lr.d8 1 3 1Wh5 l:te8?! (13...g6 was sadly neces­
sary although after 14 J..gS gxhS 1 5 .lxe7
:CB 1 6 J..d6 Black is in a bad state) 14 J..a3
'l'd7 1 5 l:adt '1'5 16 W'h4 .xeS 17 f4 'Wf6
(17 ...1i'e3+ 1 8 �h1 .i.e6 1 9 .i.d3 h6 20 aS!
and Black's queen is trapped) 18 .x£6 gxf6
t9 .J:r.£3 J..e6 20 lr.g3+ �h8 21 J..e7! hS 22
J..x f6+ t -0 Chigorin-Lebedev, Moscow 1900.
S lOxc& 1 0. . .f5
This is the right path to an opening advan­ Black seems to be unable to equalise here,
tage. Mter 8 cxdS �xd4 9 .i.xc4 (9 �c4 and it's not due to a lack of trying:
1Wh4 1 0 �xd6 t'Oxd6 t 1 �c3 .i.5 and Black a) 10....i.5 1 1 f3 �xc3 1 2 bxc3 .i.xd3 13
has equalised - Euwe) 9 ....ixe5 1 0 �3 .IL:8 'l'xd3 aS (13....i.g5?! 14 ..a6! .i.xct 1 5
1 1 lr.e1 eo5 12 •n g6 1 3 -*.£4 ..i.x£4 1 4 Let .d7 16 lr.fel lr.fe8 17 lr.xc8+ l:r.xe8 18
'ffx f4 �6 1 5 :C3 J..d7 1 6 lr.act 5 1 7 .LI3 h3 h5 19 .xa7 and Black did not have
..f6 t8 a4 lr.xe3 1 9 lr.xe3 lr.e8 Black is very enough for the pawn Shirov-Hiibner, Frnnk·
close to equality, Zelcic-Pavasovic, Nova furt [rapid[ 1996) 1 4 l:tb1 lr.e8 1 5 .i.d2 and
Gorica t 997. White is a bit better.
s. . .bxc& 9 c5 j.e7 1 0 &3 b) to....i.f6 1 1 .c2 (1 1 �xe41? dxe4 1 2
This seems to be the soundest way of .i.xe4 transposes to 10 �d2 .i. f6 l 1 �xe4
playing for White, but it is not the only way: dxc4 1 2 .i.xe4) t l ....i.xd4 1 2 �c4 dxe4 13
a) After tO f3 �gS t t �c3 :Cs t 2 •a4 R.xe4 W'h4 (1 3...R.a6?! 14 .J:r.dt with a plus;

148
3 d4: 5 . �d6 and Fifth Mo ve A lternatives
. .

14....i.xf2+? 15 Wxf2 'i'h4+ 16 g31 - 16


�g) ?l llae8! is Black's point - 16. ..xh2+ 17
..

.tg2 and White is winning - Kasparov) 14 g3


..f6 1 5 .i.e3! .ixe3 (1 5 ....i.xb2 1 6 llab1 .i.eS
1 7 .i.xh7+ Wh8 18 .i.e4 should also favour
White) 16 fxe3 1i'h6 1 7 llf4 g5 1 M llf2 .ih3
1 9 .if5 and in this position White is some­
what better, lvanchuk-Yusupov, Hoq,�n
I IJ95.
1 1 f3 tDg5 1 2 1Wa4 �d7 13 .i.f4

1 7 o!Llxf4 1 8 g3 o!Llh3+ 19 �g2 1Wh6 20


..•

�d5 f4?
Black should play 20 .:ae8 21 �7+ �h8
..

22 ..i.xf5! llxe7 23 .i.xd7! (but not 23 llxc7?!


9d2+ 24 Wxh3 llxf5 25 llxf5 .ixf5+ 26 g4
Lg4+ 27 Wg3 .ih51 28 lle5 ..d3+ and
Black attains a draw) 23. .ltxft 24 Wxft llf7+
.

25 llfS llx5+ 26 .ixf5 and White is only


slightly better (Shirov).
21 �e7+ �h8 22 ltlf5!
Or 1 3 llk2 .if6 1 4 ti)f4 'lfe7 1 5 .i.d2 g6 It is important that it's the knight that
1 6 lLe1 ..g7 17 .i.c3 hS 18 ..i.a6 llab8 1 9 goes to 5: after 22 .i.f5? fxg3 23 hxg3 g6 24
ti)cB h4 20 ti)e5 with a slight advantagt:, .ixd7 ..d2+ 25 �xh3 l:lxft White is missing
Timman-Yusupov, Bth matchgume, Linares his bishop in defence.
1 992. 22...1i'g5
13 �f6 14 llae1 lDe6
••. 22. .ixf5 23 llxf5 1lae8 24 ..d t l wins for
..

14 ..ll:lf7 is punished by 1 5 ti)e2 g6 1 6


. White according to Shirov - the knight is
'liaS, when Black cannot defend himself simply ttapped.
properly. 23 o!Llh4 1Wh6 24 �f51
1 5 �e5 .be5 1 6 :Xes -...4
16. . ti)f4 1 7 .ib1 tl)g6 1 8 lleet Ae8 was
.

an alternative, although White's position still


looks more promising.
17 f41?
Shirov, the great calculator of complicated
lines, naturaUy takes the chance to enter this
wildly complex position. He suggests that
after 17 .i.xf5 llxf5! (17 ...ti)xc5? 18 ..i.xh7+
Wxh7 1 9 dxc5 is good for White) 1 8 llxf5
lt:lxcS t 9 dxcS ..xa4 20 �a4 .ixf5 White
has a slight edge.
After 17 f4 Black has little choice but to
take the challenge. 24. . .fxg3

149
The Petroff Defence

The last chance was probably 24....i.xf5 25 0-0-0 cS with a very complex position in
lbxf5 wrg6 26 'itc2 'iig4 27 'ii'd1, when which White might have an edge) 9 wrxd3 c6
White retains a large advantage but Black can 10 0-0 .i.g4 11 lbg5 (the less obvious 11
struggle still. .i.d2 llXI7 12lbd4 � 13 h3 .i.e6 14lbxe6
25 hxg3 g6 26 i.xd7 1rd2+ 27 �h3 fxe6 is not easy to evaluate) 11...h6 12 'itd4
l:lxf1 28 i.xc6 l:laf8 29 i.g2 .i.e6 13lLlxe6 fxe6 14 c4 1fe7 15 .i.d2 c5 16
Also winning is 29 .lie7 l:te1 30 Axe 1 Wg4 d4 and Black has no obvious worries in
'itxe1 31 'itb3 Wg7 32 g4, but the game con­ this unclear position, Raetsky-Y.Aleksandrov,
tinuation looks safer despite the complex correspondence 1983.
lines. d) 6 c4
29.. Jl1 f2
Black also loses after 29...l:t1 fl 30 l:le8 gS
31 .i.d5 llf6 32l:.xf8+ llx£8 33lbf3, but not
32 lbm g4+ 33 Wxg4 l:lg6+ 34 Wh3 :h6+
35 lbh4 l:lxh4+! 36 gxh4 'itd3+ whc::n Black
escapes with a draw.
30 1rd7 g5 31 l:le7 1rc2 32 ..i..e4 l:lh2+
33 �Lh4+
Black's attack also runs out of steam after
33...'fle2+ 34lbf3 34...h5+ 35 Wxg5 We3+ 36
Wg6 ltg8+ 37 ll.g7 1fxe4+ 38 Wh6 'ife3+ 39
lDgs, when despite his extra exchange Black
is 'out of bullets'.
34 gxh4 'ife2+ 35 �g5 l:lg8+ 36 This allows Black to act quickly with
�6 1-0 6 ....i.xe5 7 dxeS lbc6 8 0-0 lbcS 9 cxdS
r------. 'itxdS 10 .i.bS 'ifxdl (weaker is 10....i.e6 11
Game6 1 li'xdS .i.xdS 12 lbc3 0-0-0 13 .i.e3 lbe4 14
Timoscenko-Yusupov lbxcl5l:.xd5 15 .i.xc6 bxc6, when White had
F'mn� 1979 an endgame edge in Lindoerfcr-Meijers,
Schwabisch Gmund 1998) 11 llxd1 .i.d7 12
1 e4 e5 2 ll:!f3 lDf6 3 d4 lbxe4 4 i.d3 d5 lbc3 lbc6 13 f4lbcd4 14 .i.a4 0-0-0 and here
5 llJxe5 i.d6 6 0-0 we can see no argument supporting a white
White has a extensive list of alternatives: edge.
a) 6 lbc3 lbxc3 7 bxc3 0-0 8 0-0 ll:X17 6 ...0-0
transposes to Game 62. Or:
b) 6 'irf3 0-0 7 0-0 cS!? 8 .i.xe4 dxe4 9 a) 6 ....i.xe5 7 dxcS lbc5 is a slightly dubi-
Wxe4 'fle7 10 lle1 lle8 11 lbd2 cxd4 12 ous favourite of the solid GM from Uthua-
lbet1 (12lbec4i.b4 13 'iixe71lxe7 14 :xe7 nia, F�uardas Rozentalis. Now 8 .i.e2! is
.i.xe7 1Slbc4lbc6 would give Black a pref- probably the right path for White here, for
crable endgame) 12 ...lbc6 13 1fxe7 llxe7 14 example 8.. 0-0 9 b31? lbc6 (9...c6 10 .i.a3
.

lbe4 .i.c7 15i.d2 .i.b>416lbc5 .i.d6 17lbe4 'iie7 11 f4 aS 12 lbc3 lblxl7 13 wrd4 gives
.i.c7 with a draw, Raetsky-Mironov, corre- White a powerful initiative) 10 L3 b6 11 f4
spondencc 1985. lte8 t2lbc3 d4 13i.f3 .i.b7 14 .i.xcS bxcS
c) 6lbd2 .i.xeS 7 dxe5lbcs 8lbf3lbxd3+ 15 lba4 wrc7 16 We2 ll:XI8 17 'ifb5 and
(or 8....i.g4!? 9 h3 lbxd3+ 10 Wxd3 .i.xf3 11 White has strong prt:ssure, C.l-lanscn-
'irxf3 0-0 12 'irg3 Wh8 13 i.gS 'ii'd7 14 Ro:r.entalis, Malmo 1997.

150
3 d4: 5 .. . �d6 and Fifth Move Alternatives

b) 6 ...Qk6 7 lfu.c6 bxc6 8 c4 Wlt4?! equality but Black needs to prove it! In Slo­
(fL0-0, transposing to 6...0-0, is preferable) 9 bodjan-Forintos, Germany 1992, Black failed
g.� 1i'h3 (9 ...'ii'f6 is refutoo by to i..xe4 dxe4 to do so and foUowing 12...ltld7 13ltla3 i..£6
l tltlc3 i..f5 12 f31 ..g6 13 fxe4 i..h3 14 e51 14llk2ltle5 15ltlxd4ltlxf3+ 16ltlxf3 b6 17
i..xft 15 exd6 i..h3 16 'ii'e2+ 'l'e6 17 'ii'xe6+ i..g5 i..b7 18 i..xf6 .llx£6 19 :C5 White held
i..xe6 18 dxc7 i..xc4 19 i..f4 with a clear a clear advantage.
advantage for White - analysed by the very c) Lltld7 8 Le4 (8 ltlxd7 leads to an
reliable duo Yusupov and Dvoretsky) to c5 immediate draw after 8...i..xh2+ 9 Wxh2
i..g4 11 i..e2 i..xe2 12 1i'xe2 i..e7 13 f3ltlf6 1i'h4+ 10 �g1 1i'xf2+) 8...dxe4 9 ltlc4 ltlb6
14 i..f4 and Black has achieved little witlt his 10 ltlxd6 1i'xd6 11 llxe4 i..f5 12 l:leS Wg6
kingside actions - White has a slight advan­ and Black has good counterplay for the
laJ.,>e. pawn.
c) 6...c51? allows White to gain an edge 7 �xeS 8 dxe5 �c5 9 l0b3
••.

with 7 i..bS+ltld7 8 dxc5 i..xeS 9 9xd5 0-0! 9 'ii'h5 does not look dangerous here,
(9...1i'e7?1 10 c6! would be an unpleasant 9...ltlxd3 10 cxd3 c5 being the must natural
surprise) to ..xe4 'ii'c7 11 i..d3 g6 12 'ii'h4 reaction. Now 1 1 b4?! looks unjustified:
ltlxc5 13 i..e 2 i..f5 when Black has some 11...cxb4 12 a3 bxa3 13 i..xa3 l:le8 14 i..d6
compensation for the pawn, but hardly ltlc6 15 l:la4 'ii'd7 16 l:lf4 1Ve6 17 d4 'ii'g6 18
enough. White could even consider 14 Qk3!? 1i'f3 i..c6 with a clear edge for Black, J. Pol­
here. gar-Kamsky, Groningen 1993. After the
7 lDd2 more sensible 11 ltlb3 'ii'c7 12 i..f 4 i..e6
7 ltc 1 is a respectable alternative, after Black shouldn't be worse.
which we have the foUowing possibilities: 9 l0xd3 1 0 '1Fxd3 tOe&
•.•

a) 7...i..xe5 8 dxe5 Qk6 9 i..£4 f5 to f3 1 1 �f4


ltlc5 11 i..b5 ltle6 with unclear play (Yusu­ 11 f4 should probably be met by 11...f6 1 2
JX>v). ex£6 1Vxf6!? (12.. .l:lxf6 13 f51? ltle5 14 'ii'h3
b) 7 ... c5 8 c4!? cxd4 9 cxd5 i..xe5 (9 ltlc5
.
.. .ffi 15 g4ltlc4 16ltld4 is very complex and
10 i..c4 :C8 11 i..f4 i..f5 12 llXl2 looks difficult to evaluate) 13 .xdS+ i..c6 14 'ii'c5
hrood for White, as after 12 ... £6 13 ltlc6! llad8 15 i..d21? 1i'xb2 16 i..c3 •xc2 17
Ulack is under attack) 10 i..xe4 f5 11 i..O i..xg7! 1i'xc5+ 18ltlxc5 l:lfe8 19 i..c3 (Ljubo­
1i'd6 12 h3 giving a position balanced on the jevic-Razuvaev, Amsterdam 1975), and now
edge between a slight white plus and �-quality. 19...i..c4! 20 ltlxb7 i..xft 21 ltlxd8 i..xg2 22
In other words, something along the lines of �xg2 .:.Xd8 leads to an endgame that Black

757
The Petroff Defence

will have to defend, but rtln defcm.l with ac­ 23 :ata 24 .i..e3 1tg6 25 �2 lL!f5 26
...

curate play. i.f47


11 1id7 1 2l:tad1
••.

12 .i.g3 leads to a hannless position after


12...b6 13 :ret �b4 14 'ifd2 c5 15 �4
'it"g4 16 c3 cxd4 17 cxb4 d3 18 'it"xd3 Wxb4
19 a3 1t"c4 20 ..d2 (1/2-1/2 Griinfeld­
Yusupov, Amsterdam 1982).
1 2...1fg4 1 3 1i'e3 liJe7 14 h3
14 'ifc5 �fS 15 g."\ c6 docs not cause any
real problems for Black.
14 1tg6 1 5 c3
..•

15 1t"c5 invites Black to play a pleasing


piece sacrifice with 15....ixh3!? 16 .ig3 �f5!
17 gxh3 �xg3 18 fxg3 'ii'xg3+ 19 'it'h1
1t"xh3+ 20 Wgl 'it"g3+ 21 �h1 l:lae8!? (of Tougher resistance was possible with 26
course a draw was available) 22 :d3 Wh4+ l:lft, though after 26 ....ia4 27 llc I Wc6 28
23 �g2 :e6, when we fear for the safety of �2 .ibS 29 c4 .ia6 Black is very close to a
the white king. winning position.
15 b6
... 1 6 e6?! 26 lL!h4 0-1
...

Game62
Anand-Mishra
India 1988
1 e4 e5 2 lLJf3 lL!f6 3 d4 �xe4 4 i.d3 d5
5 lL!xe5 i.d6 6 0-0 0-0 7 lL!c3 lL!xc3
The colourful 7...f5!? 8 f3 .ixeS 9 dxcS
�xc3 10 bxc3 11c7 1 I .D.ct .ic6 12 a4 cS
lead to unclear play in Geller-Yusupov, Vil­
nius 1980.
8 bxc3 &iJJJ7

I\ poor strategic mistake. Black will enjoy


the open f-filc and develop very l]uickly. 16
�d4 cS 17 �f3 .i.b7 18 �h4 We6 would
give Black plenty of counterplay, but this was
stiU the best option.
1 6 c5 1 7 exf7 + l:txf7 1 8 l:tfe1 i.d7 1 9
•••

1fg3 1fc6 20 i.g5 lLJf5 2 1 1ff3 h6 22


i.c1 lLle7 23 1fe271
Here the queen is not very weU placed. 23
Wg3 is stronger; foUowing 23 ...1%af8 White
.

has 24 .if4 �f5 25 'ilfh2 gS 26 .ie5 with


some counterplay, even though Black is bet­
ter after 26...'ilfa4. ll..c5!? has also been suggested: 9 ..h5 g6

152
3 d4: 5 ... �d6 and Fifth Move Alternatives

10 'lt'h6llk61? 11 .i.gS .i.e? (1 l...f6 12 .i.xg6 ised, A. Sokolov-Schandorff, Bundesliga


- 12 o!ilig6? llt71 - 12.....c7 13 .i.xf6 llxf6 2001.
14 .ixh7+ ..xh7 15 ..xf6 might favour b) 11 .tfl :C8 t21Whs 1i'd7 13 J.e3 1i'c6
White slightly) 12 .i.xe7 ..xe7 l 3 llae1 1ff6 14 llab1 a6 1S llb4 b6 (1S...�e4 16 c4 .i.e6
14 �xc6 bxc6 1S dxcS ..xc3 16 f4 ..xc5+ 17 cxdS .i.xdS 18 c4 .i.e6 19 .i.d3 would
17 �h1 f5 with unclear play (Yusupov). give White real threats) 16 llh4 h6?! (Chibur­
However both 14 f4!? and even 14 .i.bS!? danidze-Schussler, Haninge 1988), and now
look like reasonable improvements for White had the chance to play 17 c4! dxc4 18
White. .i.xc4 .i.e6 19 .i.xh6! .i.xc4 20 .i.e3 with a
91le1 close-to-winning position - 20... £6 21 cxf6
Or: .xf6 22 1lxc4, if forced, simply leaves White
a) 9 li:)xd7 .ixd7 transposes to 5...�d7 6 a piece up. Instead of 16...h6, Black should
�d7 J.xd7 7 0-0 J.d6 8 etk3 �xc3 9 bxc3 play 16...11t'g6 17 1i'xg6 hxg6 when maybe
IJ-0. White has a small plus.
b) 9 f4 cS 10 c4!? cxd4 11 �xd7 (1 1 1 0 dxe5 .ic5 1 1 1ih6 g6 1 2 1ih6 lieS
cxdS?I �f6 12 .tc4 .i.cS 13 1rf.3?!- 13lbd3 1 3 �g5 .if8 14 ....4 1rd7
'ikc7 only leaves Black slightly better - 14 ... .te7 1S .tf6 eSt? 16 c4 d4 does not
13....tf5 14 lle1 WaS! 1S .i.b2 d3+ 16 �h1 look clear at all. All these positions are really
d2 and White is in deep trouble, Kupreichik­ difficult to evaluate, as time after time every­
Dvoretsky, Leningrad 1974) 1 L..i.xd7 12 thing hangs on one move.
cxdS Was 13 1Whs f5 14 •a .i.bS and Black 1 5 .lab1 1lb8 1 6 c4 d4 1 7 f4
should not be worse.
9 . .li�xe5
9....txeS I 0 dxeS llks is a respectable al­
ternative,

1 7 •c6?
...

First of at� it is not obvious what the


queen is doing here. Secondly, Black should
not give White a free hand to continue his
after which White can play: attack on the kingsidc. After 17...'iVg4 18
a) 11 llbl �xd3 12 cxd3 b6 looks very •f2 cS Black would at worst only be slightly
sc 11id for Black. I r is unlikely that White can worse.
slrueeze an advantage out of the position, for 1 8 �f6 �g7?
example 13 :b4 f5 14 cxf6 (14 d4 .i.e6 15 Black should not voluntarily part with his
.lla3 f4 16 llb2 llf7 with chances for both only defensive piece. After 18...h6 19 1Vg3
sides) 14.....xf6 1S l:lf4 .i.fS 16 .i.d2 Wd6 17 Wh7 20 :n .tfS 21 .i.xfS gxfS 22 'llh3 •d7
1i'f3 .id7 18 1Whs c6 and Black had equal- 23 llf3 things look bk"llk for Black, but this is

1 53
The Petroff Defence

still favourable to the game. 6 lbd2 might be a good way to tight for
1 9 f5 11t'c5 an advant:tf.,re. Now Black has two main ways
Or 19....txf6 20 exf6 .i.d7 21 l:tftl and togo:
the double threat of fxg6 followed by .i.xg6 a) 6...lbd6 7 'iff3 (l 111h5 g6 8 .i.xg6?1
combined with 'ti'h6 ends aU speculation does not work on account of 8 ... fxg6 9lbxg6·
about the result. .i.g4! to 'ifxg4 l:tg8 II 'ilh5 Axg6 1 2 ltl£3
20 -*.xg7 'it>xg7 21 •f6+ �g8 22 e6 1 -0 lbd7, when White doesn't have enough for
the piece) 7 ...c6 8 ltlftl? (Kapcngut) 8... 0-0 9
Game63 lbg3 lbd7 10 ltlxd7 'ifxd7 1 1 0-0 1tg4 1 2
Aagaard-Legky 'ifxbr4 .i.xg4 1 3 .i.f4 and White might have a
Budapest 1996 very tiny edge.
b) 6... ltlxd2 7 .i.xd2 ltlc6 (1...0-0 8 'ifh5
1 e4 e5 2 �f3 �f6 3 d4 �e4 4 .i.d3 d5 g6 9 1i'h6 ltlc6 1 0 lfuc6 bxc6 1 1 0-0-0 l:te8
5 �xeS -*.e7 1 2l%dc1 :bs 13 :CS looks better for White)
8 lbxc6 bxc6 9 0-0 0-0 10 'iVhs g6 1 1 'ti'h6
llb8 1 2 b3 .i.f6 1 3 c3:c8 t 4:fe1 :xe l+ 1 5
:xe1 .i.e6 1 6 ..f 4 .i.e7 1 7 'ii'gJ .i.d6 1 8 .i.f4
.i.xf4 19 1txf4 'ifd6 20 1if6 and White had a
bit of pressure in Hort-Spassky, Reykjavik
1 977.
6 ...0-0
The saf est choice. Others include:
a) 6...lLld7 7 .i.f4 (l c4 is less dangerous:
7...lbxc5 8 dxcS c6 9 cxd5 1txd5 10 Wn
.i.fS! 1 1 ltlc3 t£lxc3 12 'ifxf5 br6 1 3 'iVh3 :d8
with unclear play- Yusupov; 7lbxd7 .i.xd7
ttansposes to s. .ltld7 6 t£lxd7 .i.xd7 7 0-0
.

This is a bit passive and Black can easily .i.e7) 7 ...t£lxe5 8 .i.xeS 0-0 9 c4 c6 10 1tc2
end up in a slightly worse position. and White has an edge according to Euwc.
Another sideline that is no longer popular b) 6...t£!c6 also doesn't fully equalise: 7
is 5 ...ltlc6 6 ltlxc6 bxc6 7 'ife2!? (l 0-0 .i.d6 ltlxc6 bxc6 8 c4 0-0 9 t£lc3 ltlxc3 10 bxc3
ttansposes to 5....i.d6 6 0-0 ltlc6 7 ltlxc6 dxc4 1 1 .i.xc4
bxc6, and 7 0-0 JJ..e7 transposes to 5....i.e7 6
0-0 ltlc6 7 lbxc6 bxc6) 7...'ife7 8 0-0 g6
(8...lLld6 may be safer though after 9 :c 1
1txe2 1 0 ltxe2+ .i.e6 1 1 ltld2 �d7 1 2 ltlf3
f6 1 3 c3 .ifS 14 ltlc1 g5 1 5 .i.xfS+ lDxfS 16
lLld3 White retains a very slight edge) 9 .i.xe4
..xc4 1 0 1td2!? (this looks more logical than
lO 1txt:4+ dxc4 1 1 :e1 f5 1 2 f3 .i.g7 1 3 c3
0-0 1 4 .i.f4 c51? 1 5 dxc5 :b8, which gave
Black considerable compensation in PiUs­
bury-Schlechter, Munich 1900) IO....tc7 1 1
:e1 1tf5 12 'iff4 and White has St.-vera1
threats.
6 0-0 and now:

154
3 d4: 5 . . . .J.d6 and Fifth Move Alternatives

bt) 11....LI6 12 1fd3 lZ.b8 13 h3 gives b) 8...c6 9 lZ.e1 ll'lbd7 10 .i.g5 dxc4 11
White a s(jght edge. Holzke-Yusupov, .i.xc4 ll'lb6 (11...ll'ld5?! is punished by 12
Hundesliga 2000 continued 13...c5?! 14 dxc5 .ixd5 .ixg5 13 .ixt7+1 :Xt7 14ll'lxt7 �xt7
.lxc5, an(l here White can play 15 .ixt7+! 15 1Vh5+ *fB 16 li'xh7 ll'lf6 17 'lt'h8+ll'lg8
<bxt7 16 'tl'c4+ .i.e6 17 'tl'xc5 'ird6 18 'ifxd6 18 d5 .id7 19 llad1 with a strong attack,
cxd6 19 .ia3 with good chances of convert­ Yurtaev-D.Frolov, Tomsk 1998) 12 .ib3
ing the extra pawn into a full point. ll'lbd5 13 li'f3 .ic6 14 llad1 WaS and Black
b2) 1t ....tf5 12 .if4 .td6 13 .tg3 .JZ.e8 14 is only slighdy worse (Yusupov).
'ira4 .ie4 15 lLe1 .ixg3 (15 ... h5?! is reck­ 8&3
less: 16 f3 .i.d5 17 llxe8+ Wxe8 18 lle1 1t'd8 8 cxdS cxd5 9 .i.xc4 dxe4 10ll'lc3 .ifS 11
19 .i.xd5 cxd5 20 1Vc6 .i.xg3 21 hxg3 lZ.b8 l:le1 ll'ld7 12 ll'lxd7 1Vxd7 13 ll'lxe4 lZ.ad8
22 :C5 with a clear plus for White, Ivan­ gives Black excellent compensation for the
chuk-Yusupov, Novgorod 1995) 16 fxg3 pawn, a draw being the likely result.
i.d5 17 .i.d3 'irg5 18 'ifc2 h6 19 Wf2 and 8 1fc2 ll'lf6 9 c5 is less forcing and there­
here we believe there is no argument about fore also more ambitious: 9 ...ll'lbd7 toll'lc3!?
White's edge. ll'lxe5 11 dxe5ll'ld7 12 .ixh7+ *h8 13 .if5
7 c4 ll'lxe5 14 lle1 .i.f6 (14...f6 is wt:aker; after 15
Maybe the best idea for White is 7 llel!?, .i.e3 .i.xfS 16 1fxf5 d4 17 :Cd1 i.xc5 18
as first suggested by Steinitz. After 7 ... f6 8 lL\.14 ..e7 19 1Vh3+ *g8 20 ll'lxc5 dxe3 21
ll'ID f5 9 c4 .i.c6 10 cxd5 .ixd5 11 ll'lc3 1fxe3 White enjoys slight pressure) 15 .if4
ll'lxc3 12 bxc3 ll'lc6 13 ll'lc5 ll'lxe5 14 llxe5 lZ.eS 16 lZ.e3 .i.xfS 17 1fxf5 lL\g6 18 llh3+
White was a bit better in Ed.Lasker-Kupchik, *g8 19 1Vh5ll'lh4 20 .i.g3 1fe7 with a com­
New York, 1915. plex batde, Mi.Tseidin-Karasev, Leningrad
7 ...c6 1970.
B �c3 9 bxc3 dxc4
...

Also possible is 9...ll'ld7 10 cxd5 (10 f4


dxc4 1 I .i.xc4 transposes to 9...dxc4)
to...ll'lxe5!? 11 dxe5 ..xd5 12 1fe2 .tc6 13
l:d1 1Wa5 with decent counterplay.
10 .i.xc4 liJ47

The passive 7...ll'lf6 does not give real


chances for equality after 8ll'lc3 and now:
a) 8 ...ll'lc6 9 cxd5!? ll'lb4 10 .ic4 ll'lbxd5
It Wb3 c6 (11....ie6?! is a silly invitation: 12
WVxb7 .JZ.b8 13 9xa7 .JZ.as 14 �7 .JZ.b8 15
1fc6 ll'lb4 16 1ff3 J.xc4 17 ll'lxc4 1fxd4 18
ll'lc3 ..e5 19 ll'lfS and White is much better, 1 1 ltlxf71?
G.Guscinov-Ekdyshman, St Petersburg A surprising sacrifice. 11 f4 is generally
2000) 12 lZ.e1 .i.d6 13 .i.g5 widu plus. recommended: 12.)t)f6 12 l:lb1 (12 1fb3

155
The Petroff Defence

llkls 1 3 f5 f6 1 4 .!i)g4 bS 1S .i.e2 aS with 21 lld4 � 22 .txe7 1lxe7


unclear play is not something Black should Black is a bit better here.
avoid) 12...1Vc7 13 9b3 .i.d6 14 .i.a3 .i.xa3 23 h3 lld7 24 llxd7 + l0xd7 26 lld1 �5
lS irxa3 .i.fS 16 libel .i.e4 17 lL\g4 .!i)xg4 26 ••3 ltld7?1
18 llxe4 l:lae8 19 llxe8llxe8 20 irxa7 'ii'e7 This inaccuracy allows White's queen to
with chances for both sides, Popiel-Von penetrate Black's queensidc and create suffi­
Gottschall, Hanover 1902. cient countctplay to draw. Though White
1 1 .. Jtxf7 1 2 .txf7+ might have been able to save the draw any­
Also possible is 12 '1Vb3!? 'ii'ffi 13 l:let way, it was definitely worth torturing him
.i.d6 14 .i.g5ltJb6 lS .i.e7 .i.xe7 16 .i.xf7+ with 26...ltJg6 27llct 'ii'cS with a slight, but
'i'xf7 17 llxe7 1i'xb3 18 axb3 �ffi 19 lieS enduring, edge.
with a double-edged endgame. 27 1i'f4+ lL!f6 28 -.& a5 29 lld6 •xc4
1 2 �xf7 1 3 Wb3+ �8 14 lle1 b6!
.•. 30 llxb6 •xa2 31 ec7+ .td7

The only move, but good enough to f�ght White also achieves a draw after 31.../lkl7
on equal terms. 32lla6 'i'bt+ 33 �h2 'i'e1 34 'ii'f4+ ltJ£6 3S
1 6 c4 lla7+.
lb.is might be too rash a decision. After 32 llxf6+ �f6 33 •xd7 ft1 + 34 �2
lS .i.a3!? .i.xa3 16 'ii'xa3+ cS 17 c4 the posi­ � 35 .d8+ � 36 .g8+ �6 37"
tion is less clear, though it is difficult to be­ •d8+ �c6 38 11'c8+ �5 39 •d7 + Wc4
lieve that Black should be worse. 40 •xg7 1lrd6+ 41 g3 a4 42 •xh7 a3:
1 5 ...lL!f6 1 6 .tb2 .i.f5 1 7 d5 ..d7 1 8 43 •c2+ Wb4 44 �1 + �a5 45 ft3·
Jlad1 Ilea 1 9 dxc6 •xc6 20 .ta3 .te6 �%-%

156
3 d4: 5 ... Jl.d6 and Fifth Move Alternatives

Summary
After 1 e4 eS 2 �£3 �f6 3 d4 �xe4 4 .i.d3 dS 5 �xeS, the move S...i0c6 is unattractive in
vit:w of the strategic problems Black faces after the doubling of the pawns. In our view, an
interest in S....i.e7 will be maintained; this continuation is not ambitious and docs not have any
positional flaws.
After S....i.d6 interest in the forced line 6 0-0 0-0 7 c4 .i.xeS 8 dxeS �c6 9 cxdS 'ifxdS 10
'tl'c2 �b4 t 1 .i.xe4 �xc2 12 .i.xdS will die away as it is thoroughly analysed and docs not
appear to be more promising for White in terms of gaining an advantage than, for instance, 7
ltlc.l2!?.

1 e4 e5 2 �f3 �f6 3 d4 �xe4 4 il.d3 d5 5 �e5 (D) il.d6


S....i.e7- Gan1e 6J
6 0-0 0-0 (D) 7 c4
7 �2- Gmne 62
7 �3 Gt1111e 61
-

7 Jl.xe5
...

7...�c6 Game 60
-

8 dxe5 (D) - Game 59

5�e5 6 ... 0-0 8dxe5

157
CHAPTER EIGHT I
3 d4: Fourth Move
Alternatives

1 e4 e5 2 �f3 lt!f6 3 d4 lt!xe4 It is more na tu ral to arrack the centralised


Af te r the basic opening moves 1 e4 eS 2 knight with 5lL!bd2 (Game 66), which is one
lLif3 lLif6 3 d4 lL!xe4 the position arising of the most fashionable variations in the
after 4 .i.d3 dS 5lL!xeS is considered to be a Petroff Defence these days. Black can swap
tabiya. This brief chap ter cove rs deviations on d2, place his knight to c5, improve i ts
by White and Black on move four. position by means of ... f7-f5 or ... .i.fS, o r
Afte r 4 .i.d3 i t first appears tha t Black ca rry on with the development via 5...lL!c6 o r
must defend his knight wi th 4...d5. Howeve r, S... .i.e7 (p ro tec ting the yueen i n case o f an
in 1993 g randmaster Jacob Mu rey demon­ exchange on e4). Game 66 shows tha t all(!)
strated a pa radoxical al ternative in 4...lL!c6!? these con tinuations are yui te reasonable.
(Game 64). White can cap tu re the knigh t but
is unable to retain i t (5 .i.xe4 d5 and ... e5-e4, GaPJe64
or 5 d5 lL!c5 6 dxc6 e4). Since 1993 Morey's Tiviakov-Forintos
idea has been employed on many occasions Porto San Gimgio 1994
by m<xlem h'Tllllll masters.
4 dxe5 is often played instead of 4 .i.d3. 1 e4 e5 2 lDf3 ll)f6 3 d4 lDxe4 4 ..id3
Curiously 4 ... .i.c5, trying to sci7.c the initiative lt!c6!?
by a ttacking f2, is unpopular. Prom our pe r­
spective, Jines from opening manuals aimed a t
persuading readers of Whi te's advantage are
no t particularly convincing (see Game 65).
Tiu: continuation 4...d6 transposes to the
Philido r Defence ( 1 e4 c5 2 lLif3 d6 3 d4
lLlxe4 4 Jxe5lL!e4) and is not discussed he re.
In the main line wi th 4... d5 the move 5 .i.d3
(Game 65) is identical to 4 .i.d3 d5 5 dxeS,
which is pret ty harmless regardless of Black's
response (S...lL!c5, S... .i.e7 o r S...lL!c6). The
whi te pawn on e5 is almost no inconvenience
and often becomes an object of a ttack.

158
3 d4: Fourth Move Alternatives

This move was a shocking novelty when it t2.!Dc3 c6 13 :C1 ..c7 14 g3lDe6 15 1Wd3
first appeared in 1993. Nowadays, of course., i..e5 16 .i.e3 d5 (Sadvakasov-Koneru, Jodh­
ir has lost its surprise value. pur 2003) and now 17 .i.d2 l:lab8 18 l:labt
6 ..txe4 .i.f6, with a level position, is a possible con-
White has a large number of options: tinuation.
a) 5 .!Dxe5 .!Dxe5 (5....!Dc5 6 .i.c4 .!Dxe5 7 5 ...d5
dxe5 d6 8 1i'f3 1Wd7 9 .i.e3 c6 10 Lc5 dxc5
II 0-0 favours White, whilst 5...d5 transposes
In 4 .i.d3 d5 5 .!Dxe5 .!Dc6) 6 .i.xe4 d5! 7
dxe5 dxe4 8 11'xd8+ �cl8 9 .!Dc3 .i.b4 10
.ig5+ �e8 11 0-0-0 .i.g4 12 l:d4 .i.xc3 13
hxc3 h6 14 llxe4 .i.h3! (a typical Shirov
move, here played against him) 15 gxh3 hxg5
16 l:tg1 llxh3 17 llxg5 �fB 18 .J:leg4 :C8 19
llxg7ltxe5 20 llg8+ �c7 with a likely draw,
Shirov-Timman, Wijk aan Zee 1998.
b) 5 dxe5 and now:
b1) 5...d5 6 exd61Dxd6 7 0-0 .i.e7 8.!Dc3
Jtg4 9 .!Dd5 .!Dd4 10 i..e2 (10 :C1 i..x£3 11
hrxf3lbc6 12 f4 f!f.l13 f5.!Dxf5 14 i..xf5 gxf5 6 ..txh7
15 'ti'£3 leads to a wildly complex position 6 .!Dxe5 dxe4 7 .!Dxc6 bxc6 8 0-0 .i.d6 9
with chances for both sides; note 15....!Dd4?? :Ct 0-0 10.!Dd2 (10 l:lxe4 i..f5 11 :C1 1i'h4
16.!Df6+ �fB 17 i..h6 mate!) to...lbc6 11 h3 12 g3 'ti'h3 looks dangerous) 10...f5 11 .!Dc4
.i.h5 12 :C1 0-0 13 .!Dxc7+ 11'xc7 14 .!Dg5 .i.di 12 b3 Wh413.!Dxd6 cxd6 14 1fd2 gives
..i..xe2 15 1i'xe2 Wf6 16.!Dxe6 .J:lfe8 17 Wg4 a dynamicaUy balanced position, Hracek­
Axc6 18 l:lxe6 fxe6 19 c3 l:lf8 with level Barua, Moscow 1994.
chances, Yakovich-Makarychev, Elista 1995. 6 i..g5 leaves Black with a wide range of
b2) Also fine is s....!Dc5 6 0-0 .!Dxd3 7 choices:
'iixd3 d6 8 ..i..f4 dxe5 9 '1Vxd8+ 1Dxd8 10 a) 6...f6 7 .!Dxe5 dxe4 8 11Vh5+ (8 .!Dxc6?!
i.xe5 lbc6 11 :C1 i..d7 12 i..xc7 .J:lc8 13 bxc6 9 .i.e3 .id6 10 c4 0-0 11 c5 .i.e7 12
..i..e5 llxc2 14 lDbd2 i..c6 15.!Db3 .i.x£3 (an .!Dc3 f5 13 Wb3+ �h8 14 g3 f41? 15 gxf4- or
improvement over 15...i..b4?! 16 .!Dfd4 .J:lc4 15 .ixf4 ..xd4 16 .i.e3 We5 with attacking
17 :Ccl l:lxct+ 18 .J:lxcl �d7 19 a3 i..e7 20 chances according to Frolyanov - 15..ig4
.!Dxc6 bxc6 21 l:ldt+ �c8 22lDa5 c5 23lDc6 16 llg1 .i.f3 and Black had an attack in Yur­
with a clear edge for White in Bezgodov­ taev-Frolyanov, Moscow 2003) 8...b>6 9.!Dxg6
Cs.Horvath, Ljubljana 1995) 16 gxf3 i..b4 17 hxf!f.l10 1fxg6+ �d7 11 i..xf6 (11 1Vf5+ �e8
l:tc4 i..c5 18.!Dxc5 l:lxc5 with an <.>tJual posi­ with a draw is of course possible) 11..lth6 12
tion. Wxh6 i..xh6 13 i..xd8 �xd8 14.!Da3 and we
c) 5 d5 was meant to give White an edge evaluate this endgame as slighdy favourable
once upon a time, but we cannot see how: for White.
5....!Dc5 6 dxc6 e4 7 cxb7 (7 .i.c4 ex£3 8 b) 6...1Vd6 7 dxe5 Wh4+ 8lDc3 dxc4 9 a3
cxd7+ i..xd7 9 Wxf3 1Ve7+ 10 .i.e3 .ic6 11 WaS?! 10 .!Dd4 .!Dxe5 11 0-0 .i.d7 12 .!Dxe4
i.d5 .i.xd5 12 WxdS 1i'e6 13 1i'xe6+ .!Dxe6 li)g6? (Black was in a bad way, but resistance
14 .!Dd2 0-0-0 15 0-0-0 was drawn in V. Ko­ was still possible) 13 :Ct .i.e7 14 b41i'b6 15
rnv-Frolyanov, Russia 2003) 7..ixb7 8 .i.e2 ..txe71Dxe7 t6.!Dc5 l:ld8 17 We2 1-0 Palac­
cxf3 9 .i.x£3 i..x£3 10 1fxf3 .i.d6 11 0-0 0-0 Kos. Fcldbach 1997. Instead of 9...'ifas.

159
The Petroff Defence

.
Black can try 9.. 1Wxb2!? 10 ll:kls .tcs 11 llc1 (or 9 dxc5? 'iVh4 10 h3 .txh3! and Black:
llb1 1Wxa3 12lbxc7+ �fB 13lbxa8 exf3 14 wins) 9...lllf3+ 10 gxf3 .te6 1 I 'iVd3 llh3
'iVxf3 'ttxf3 15 gxf3 lbxe5 with a very un­ with a very strong attack for Black.
clear position. 7 ...i.g4 8 i.f4 11d7
c) 6...1Wd7 (the normal move) 7 .td3 e4 8 8... g51? is also enticing: 9 .tg3 f5 10 exf6
0-0 f6 9lle1 1Wxf6 11 lbc3 .i.xf3 12 'iVxf3 .xf3 13 gxf3
.i.b4 14 0-0 .i.xc3 15 bxc3 0-0-0 with an
equal ending; or 9 1t'd3 .i.xf3 to gxf3 (Aiek•
seev-Bezgodov, Hoogcvecn 2002) and aftCJ:
10..llh6 11 .tg3 .i.g7 12 f4 gxf4 13 .i.xf4
llc6 Black should have sufficient counter­
play.
9 �bd2 11f5 1 0 i.g3 0..()-0 1 1 0-0 i.c5
1 2 a3
12 llct lZ\ct4 13 lllxd4 .i.xd4 14 lZIO
i.xb2 15 llb1 .i.c3 161lc3 d4 17lld3 11'g6
18 h3 lldh8 gave Black a strong attack in'
Svcshnikov-Pavasovic, Nova Gorica 1996.
1 2...i.b6
with another branch: Or 12...�51? 13 c3 (13 b4? does not
.
ct) 9. .fxg5 to c4 .tb4 11 lbc3 (11 cxd5 work on account of 13...i.d4 , with a clear
.txel 12 'lrxe1 lbe7 13 .txe4 g4 14lbc5 advantage for Black) l3...d4 14 1Wa4 d3 with
'iVd6 15 lba3 a6 with an unclear game is substantial counterplay.
given by Murey) 11...0-0 12 cxd5 lbxd4 13 1 3 b4 �5 14 l:.e1 �d4 1 5 a4 a6 1 8
lbxd4 exd3 14 llX.o6 llt7 15 1fxd3 favours l:.a3 �f51
White.
.
c2) 9. ..te7 10 .tf4 cxd3 11 'ilfxd3 0-0 12
lbc3 .tb4 13 llc2 (Tunman-Hiibner, France
1993) and after 13...1Wt7 14lbb5 .taS the
chances are level (Hubner).
6 ...Axh7

1 7 �f1
17 .tf4 would be met by 17...g51 18 .i.xgS
J:lgs with a strong attack (Forintos).
1 7 ...trucg3 1 8 �xg3 i.xf3 19 •xf3,
'W'xh2+ 20 �1 g6 21 l:.d3 I
Or 21 We2 1Wh4 22 llh1 11'c4+ 23 '1Vd3
7 dxe5 llxh1 24 1Wxc4 dxc4 25 lbxh1 llc8 and th ,
7lbxc5?! is weak: 7...'ttc7 8 0-0 lllxc5 9 endgame favours Black.
J
1 60
3 d4: Fourth Move Alternatives

21 ...1rh41 22 llxd5 irc4+ 23 lld3 llxd3 ..t.gS d6 10 1i'e3 1i'd7 11 �c3 h6 12 ..t.h4 gS
24 •xd3 1rxb4 25 e671 13 �S �ffi and Black is under attack, but
'Ibis accelerates an uncomfortable posi­ nothing is clear) 7 'iVdSlbxh1 8 .i.hS fle7 9
tion into more trouble. After 2S l:te4 1WcS 26 ..t.gS .i.f2+ 10 We2 'l'e6 11 �c3 h6 12
..Wd2 :J,8 White is still only 'somewhat' ..t.d8!? lbc6 13 1Wxe6 dxe6 14 .i.xc7 with a
worse. position almost impossible to understand
25...1rf4 26 l:le2? without dedicating days or weeks of your life
to it!
b) 5 1i'dS ..t.xf2+ (S...�xf2? 6 WxcS �xht
7 .i.gS f6 8 cxf6 hrxf6 9 ..t.h4 is probably los­
ing for Black)

The last chance was 26 1i'f3 'l'x£3 Zl gxf3


l:th2 28 � fxe6, though Black has every
chance of winning the endgame.
26 ...1rxg31 27 lle3 lZh1 + 28 We2
..g4+ 0-1 and now:
bl) 6 �d11? f5 7 .i.c4 .z:tf8 (after 7...'fle7 8
Game65 llk3 c6 9 1i'd3 ..t.cs to :n bS 11 .i.b3 �f2+
Nevednichy-Collas 12 .z:txf2 .i.xf2 13 'ii'xfS White has a strong
Montpellier 2003 initiative for the exchange) 8 lblxl2 c6 9
�xe41? (after 9 'ii'd3 dS 10 cxd6 'ii'xcl6 11
1 e4 e5 2 li:lf3 /t)f6 3 d4 li:lxe4 4 dxe5 �xe4 'iVxd3+ 12 ..t.xd3 fxe4 13 Le4 ..t.fS
An unusual move. One further option is 4 14 .i.xfS llxfS the position has rapidly turned
�xeS, when 4 ...d6 S �£3 transposes to 3 drawish) 9. .cxdS 10 �d6+ �e7 11 ..t.gS+
.

�xeS d6 4 �f3lbxe4 S d4 and 4...dS S ..t.d3 .z:tf6 (1t...'.tc6?! 12 ..t.xd8 l1xd8 13 �gS+
1ransposes back to 4 ..t.d3 dS S �xeS. �xeS 14 �dn+ Wd4 1S �xd8 �xc4 16 JZ.ft
4 . d5
. . and White has the advantage) 12 .i.xdS �c6
4...d6 transposes to the Philidor Defence 13 exf6+ gxf6 14 �xfS+ �e8 15 �d6+ �e7
(I c4 eS 2 �f3 d6 3 d4 �f6 4 dxeSlbxe4), 16 .i.f4 with very unclear play. Of course
which is outside the scope of this book. We White could take a draw by perperual with 16
will say, however, that White is generally �fS+.
1hought to have a slight advantage. b2) 6 <.te2 f5 7 �3 (Black would get
4.....t.cS!? leads to very sharp play: away too easily after 7 exf6 �xf6 8 WeS+
a) S .i.c4 �f2 (S....i.xf2+!? 6 �e2 ..e7 7 �f8 9 .i.gS ..t.b6 10 �3 �6 It 1i'f4 h6 12
WJ3 f5 8 �3 c6 9 ..t.c3 .i.xe3 10 1i'xe3 with .i.xf6 'iVxf6 13 'ilrxf6+ gxf6, after which
unclear play is also possible) 6 ..t.xn+ �f8 White cannot fully justify the loss of a pawn)
(or 6...�n 7 ..dS+ �e8 8 'ifxcS �xh 1 9 7...c6 (7...�xc3+ 8 bxc3 .i.h4 9 �xh4 'iVxh4

161
The Petroff Defence

10 g3 'ile7 11 �f2/tk6 12 .lc4 'ilc6 13 'ilcS 6 0-0 .i.g4


with unclear play might be an improvement) 6....i.e7 is also safe: 7 l0c3 .i.f5 8 'iletl?
81Wd3 0-0 9 lh'te4 fxe4 10 Wxe4 dS 11 exd6 l0b4 9l0d4 (9l0xe4 dxe4 10 .i.xe4l0xc2 11 j
.l:te8 12 ...xeS+ ..xe8+ 13 Wxf2 -'.g4 .i.xc2 .i.xc2 is simply equal, but 9 e6!?l0xd3
(13....i.e61? looks safer; after 14 .i.d3 the to cxd3l0xc3 I 1 exf7+ Wxf7 12l0e5+ �g8
position remains very unclear) 14 -'.c4+ �h8 13 ..xc3 .i.f6 might give White slight pres­
IS .J:r.e1 WfB 16 :C7l0d7 17 -'.d2 and White sure) 9....i.g6 to f4 cS (or IO...l0xc31? 11 fS
had good attacking chances in Lozcnko­ l0cxa2l? 12 fxg6 fxg6 13 l0e6 Wd7 14
Titlianov, Sverdlovsk 1974. lOxg7+ Wd8 15 .i.h6 with a very complex
5 .i.d3 position) 11 f5 l0xd3 12 cxd3 cxd4 13 fxg6
Sl0bd2 is considered in the next game. hxg6 14l0xe4 dxe4 15 1Wxe41Wb6 16 b3 0-0
5 exd6 makes no sense: 5 ... -'.xd6 6 .i.e2 17 .i.b2 l:lad8 with level chances, Neved­
0-0 7l0bd2 -'.f5 8 0-0l0c6 9l0xe4 -'.xe4 10 nichy-Piket, World Championship, New
lOgs .lg6 11 .id3 •f6 12 .i.xg6 1fxg6 13 c3 Delhi 2002.
ltad8 14 1Wb3 b6 and Black was a bit better, 7 1Llc3
Karpov-Noakh, Leningrad 1966. 7 l0bd2 l0xd2 8 .i.xd2 l0d4 9 .i.e2
5.-.lDc:S l0xe2+ 10 'ilxe2 'ild7 11 h3 .i.xf3 12 ..xf3
.i.. c5 13 c4 0-0-0 14 .l:tadt •e6 15 .i.g5 ltd7
16 .llxd5 .l:txdS 17 cxd5 was agreed drawn,
Romanishin-Makarychev, Frunze 1985.
7 �c3 8 bxc3 .i.c5
•.•

Or 8....i.e7 9 .i.£4 0-0 10 h3 .i.hS 11 llbl


'ilc8 12 g4 .i.g6 13 .i.xg6 (Kumosov­
Skatchkov, St Petersburg 2001) and now best
is 13...hxg6 14 1Wxd5 1We6 15 .l:tfd1 .l:tad8 16
1Wb3 l0a5 17 1Wa4 b6 with an unclear posi­
tion. Even so, the text move seems more
prudent.
9 .i.f4 0-0 1 0 h3 .i.h5

This is the most active move, but Black


has reasonable alternatives:
a) s...lOc..c; 6 0-0 l0xd3 7 ..xd3 .i.e? 8
l0c3 c6 9 l0d4 0-0 to f4 f6 1 1 .i.d2 l0a6 12
a3 -'.cS 13 .i.e3 (13 Wh1 .i.xd4 14 •xd4
.i.f5 15 .l:tactl0c7 16l0e2 fxeS 17 fxe5l0e6
18 'ilb4 .d7 gave Black good play in Schif­
fers-Mason, Breslau 1889) 13...'ile7 14 b4
.i.b6 15 exf61bf6 is unclear.
b) 5... .i.e7 6 0-0 0-0 7 h3 f61? 8 c4 c6 9
'iVc2lOa6 to .i..xe4 (10 cxd5 cxdS 11 -'.xe4
dxe4 12 IZ.d1 'ilc7 13 ...xe4 .i.e6 is messy)
10...dxe4 11 'iVxe4 fxeS t2l0c3 ...c7 13l:<.tt 1 1 g471
.i.f5 14 11'xe5 Wxe5 15l0xe5 .i.cS and Black This i.<> very risky. A more modest ap­
has sufficient counterplay for his pawn, proach with 11 ltb l .i.b6 12 llbS �h8
Drazic-Lanzani, Milan 2003. (12..l0c7? 13 .i.xh7+! would be bad news

162
3 d4: Fourth Move Alternatives

Indeed for Black) 13 g4 .i.g6 14 .i.xg6 fxg6 A valid alternative is 18...lla41? 19 .i.b5
15 llxd5 'fle7 was called for, reaching a posi­ lla3 20 'irxd5 :Xc3, when it is probably
tion with chances for both sides. more pleasant to be Black.
1 1 .....tg6 1 2 .IZ.e1 ..tb6 1 3 ..tg3 19 �xf2?
An understandable mistake as White is
scared of allowing his king to go too far into
the open. However, this is the worst of the
two evils and Black now has a very strong
attack.
The alternative is 19 Wxf4 ...ffi+ 20 .te;
'fld6+ and now:
a) 21 lle5lilie5 22 .L.f2 �xg4+ 23 Wxg4
h5+ 24 Wh3 .i.x£5+ 25 Wg2 .ie4 offers
Black compensation, but it is very likely that
White wiU survive the attack.
b) 21 WgS! suggests to Black that he
should settle for a draw, as after 2L..'flc7+ 22
1 3 ...f5 Wf4 .i.xh4?! (22.. .1t'd6+ is perpetual check)
13...d4 is weaker: 14 cxd4lilid4 15 �xd4 23 �xh4 1fxh4 24 llht 1ff6 25 1fxd5 l:l.d8
Wxd4 16 .i.xg6 hxg6 17 'flxd4 .i.xd4 18 26 1t'c5 it is not so obvious that Black has
%lad1 .i.b6 and White has many ways to se­ sufficient compensation for the exchange.
cure the advantage in the endgame. The White king looks strange, but Black's
14 �2 ¢'h8 1 5 .i.h4 •ea 1 6 e6 txg4 1 7 piccc:s are not that active.
hxg4.1Z.f4 19 .. ...te4 20 ..tg5
Black can also play the safer 17 ... .taS! 18 20 .i.e2 1i'xe6 21 .i.g3 llf7, with ... l:laffi
11td2 d4 19 �xd4 (not 19 e7? llxf31 20 .ixg6 to follow, does not help White at aU.
hxg6 21 Wxf3 .i.xc3 22 'flf4 .ixat 23 llxa1 20 ....1Z.xg4
�xe7 24 9xc7 �6 and Black has the ad­ Or 20... llxf3+!? 21 Wxf3 .i.xf3 22 Wxf3
vantage) 19....ixd3 20 11Vxd3 �xd4 21 1Vxd4 d4 and Black has goo d chances. However,
1rc6+ 22 'fle4 .ixc3 �' 1rxc6 bxc6 when the game continuation makes more sense.
nlthough White will have enough compensa­ 21 llg1 •xe6 22 .IZ.xg4
lion for a draw, he will never be able to win.
18 �g3 ..txt2+ !?

Black's attack is also irresistible after 22


.i.xc4 dxe4 23 �h4 e3+ 24 We2 llf8.

163
The Petroff Defence

22...1rxg4 23 .i.e2 llf8 24 1Wg1 .i.xf3 0-1 should play 10...'1Ve71? 11 lbbS lbba6 12
..xdS lidS 13 'ffc 4 1Wxe5 with a messy posi­
Game66 tion) 11 lbb5lbe6 12 c4 a6 13 cxdS axbS 141
Smirin-Aiterman dxe6 ..xe6 15 .i.xbS+ c6 112-lfz Volokitin�
Haifa 1995 Mikhalchishin, Portomz 2001. In the final:
position 16 .i.e2 lZXI7 17 f4 l:lxa2 18 l:lxa.Z
1 e4 e5 2 li)f3 ll)f& 3 d4 li)xe4 4 dxe5 d5 11'xa2 19 0-0 appears to offer White some
5 li)bd2 chances.
c2) 6lbxe4 .i.xe4 7 .i.d3lbc6 8 0-0 .i.e1
9 l:le1 .i.xd3 10 1i'xd3 '1Vc.l7 11 .i.f4 0-0-0 12
a3 Wg4 13 .i.d2 f6 14 .i.c3 d4 15 ..tb4 l:lhe�
(t s....i.xb4?! 16 axb4 lbxb4 17 Wc4 lbc6 t!i
b4! with a clear plus for White, e.g.J
IS...c;ilbS?I 19 bS lbxeS 20 lbxeS fxeS 21
l:lxa7! 1-0 Glek-Mikhalchishin, Zurich 2001)�
16 .i.xe7 :Xe7 17 exf6 gxf6 18 l:lxe7 lbxe"f!
19 fle1 .!Og6 20 h3 'iWd7 21 l:le4 with a sligh�
edge for White (Giek).
d) s....!Ocs 6 .!Ob3

5 . .i.e7
..

Black has many alternatives to this modest


move:
a) S...ltk6 6 .i.bS .i.d7 7 1ie2 a6 8 .i.xc6
.i.xc6 9 tbd4 lbxd2 10 .i.xd2 .i.d7 11 0-0-0
11'h4?! (11...1le7 12 l:lhcl 0-0-0 13 f4 may be
a bit better for White, but the advantage is
not that obvious) 12 e61 11xd4 13 cxd7+
Wxd7 (Mainka-Schmidt, Gennany 1996) and
now after 14 .tgs 1i'a4 15 110 White would
have a very promising attack.
b) 5...15 6 exf6 lbxf6 (6...1i'xf6?! 7 lbxe4 and now:
dxe4 8 .tg5 1l'd6 9 1We2 would be bad for d1) 6....!0xb3 7 axb3 ..te7 8 .i..d3 .!Od7 9
Black) 7 c4!? d4 (7....i.d6!? 8 cxdS 0-0 9 .i.e2 c3 .!Ocs I 0 .i.c2 ..tg4 11 b4 lbe6 t 2 1i'd3
lbxdS looks equal) 8 .i.d3 .i.e7 9 0-0 0-0 10 '1Vd7 13 0-0 .i.hS (the weakening 13...�,>6?{
lte1 cS!? (an improvement over 10. lbc6?! 11
.. was played in Ivanchuk-Kasimd7.hanov
a3 aS 12lbgs .i.d6 13 tbdf3 h6 14 cS! hxgS Elisra 1998; now after 14 .!Od4 0-0 15 .i.. a4 c6
15 cxd6 11xd6 16 lbxgS with a clear edge in 16 h3 .!Oxd4 17 hxg4 .!Oe6 18 f4 .i..d8 19
Sveshnikov-Raetsky, Podolsk 1992) 11 b4!? Wh1 White would have enjoyed a clear edge)
cxb4 12 1i'c2lbc6 13 cS and White has com­ 14 lbd4 .i.g6 15 .!05 .i.d8! 16 :dt c6 17
pensation for the pawn. .!Od6+ ..xci6 18 exd6 .i..xd3 with equal
c) s....i.t5 might not lead to equality, for chances (Kasimdzhanov).
example: d2) 6....i..g4 is also sound: 7 h3 .i..hS S
cl) 6lbd4!? .i.g6 7 h4!? .i.cS 8lb2b3 hS 9 .i.e2 .!Oc6 9 .!Ofd4 .i..xc2 10 ..xc2 .!Oxb3 11
lbxcS lbxcS 10 .i.e3 1Wd7 (maybe Black .!Oxb3 ..d7 12 .i..f4 0-0-0 13 0-0-0 11'c6 14

1 64
3 d4: Fourth Move Alternatives

lld2 �c7 with level prospects, Kharlov­ and now:


Akopian, Moscow 1991. More complex is 8 a) 8.!tlg5 isn't dangerous: 8.. ..i.f5 (8 ....!tlc6
c4 .!tlxb3 9 'A'xb3 �xO 10 Wxb7 (10 gxO 9 .!tlxe4 ..if'S 1() .ld3 lDxeS 11 c!iXI6+ cxd6
ltlc6 1 I cxdS .!tlxcS with unclear play is pos­ 12 �x£5 0-0 13 0-0 �b6 14 .i.f4 �cS 15
sible as weU) 10....!tld7 with unclear play, for .llad1 J:lae8 16 d left White with some
example 11 gxO .!tlxeS 12 �e2 c6 13 cxdS chances in lordachescu-Chi Fengtong, Ulan
..icS 14 dxc6 0-0 and Black has good com­ Bator 2002) 9 .lc4 ..i.xg5 to .lxgS �7 11
pensation for the pawn. �f4 .!tlb6 12 .lb3 aS and we think Black wiD
c) s....!tlxd2 6 'A'xd2!? (6 �xd2 �e7 gives equalise.
n standanl position that has not been investi­ b) 8 .!tld4 R....id7 9 �f4 .!tlc6 10 0-0-0
).,'llted very much; a possible continuation is 7 gS!? (after 10...0-0 11 .lc4 .!tlxd4 12 .llxd4
.idJ .ig4 8 h3 .i.hS 9 0-0 0-0 to .llcl lbd7 .l£5 13 e6!? fxc6 14 lld7 White has obtained
11 c3 ltlcs 12 .i.c2 l&6 13 �f5 c6 14 g4 an unplt:asant pressure) 11 .lgJ hS 12 .lbS
.ig6 15 'A'c2 �gS 16 .llad1 �xd2 17 'A'xd2 h4 13 e6 fxe6 14 Lc6 (14 lDxc6?! hxg3 15
fle7 with unclear play, Kotsur-Zulfugarli, .!tlxd8 .llxd8 16 �xd7+ .llxd7 17 .ll xd7 �d7
lsL1nbul 2000) 6...�e7 (or 6...c5!? 7 �d3 18 fxg3 Wd6 and Black's activity gives him a
tlk6 8 cJ h6 9 0-0 �e6 10 :C1 .ic7 1 I h3 good endgame) 14...bxc6 15 .leS ltfB 16
1Wd7 with chances for both sides) 7 'A'f4 (1 .!tlb3 Itt'S (16...:Xf2 17 lDcS .lc8 18 .!tlxe4
.i.bS+ c6 8 .i.e2 0-0 9 0-0 f6 10 c4 fxeS 1 1 llxg2 19 D.hg1 l:lxg1 20 llxg1 looks very
cxdS cxdS 12ltlxe5 .!tlc6 leads to equality: 13 strong for White) 17 .ld4 .le7 18 g4 llf7 19
.!tlxc6 bxc6 14 b3 cS 15 L3 .i.b 7 16 llact .i.c3 and White's position was preferable,
nc8 17 �g4 D.c7 18 D.cet d4 19 1t'd3 ..LI6 Zviagintsev-Stohl, Rethymnon 2003. f-low­
20 �c1 Wh8 21 f4 :C7 22 .llxc7 ..xe7 23 ever, instead of 12...h4, Black should proba­

..id2 �e4 1/2-'12 Svidler-Kramnik, Dos Her­ bly play 12....!tlxd41? 13 .ixd7+ �d7 14
manas 1999) 7...0-0 8 'A'g3 �f5 (8 ...f6?! 9 ltxd4+ We6 15 h3 (15 h4!?) 15 ...h4 16 .lh2
..th6 :n 1o 0-0-0 ltlc6 11 �c4 �e6 12 c3!? 17 fxe3 .i.e? with compensation for the
.i.xdSI �xdS 13 e6! �xc6 14 .llxd8+ :Xd8 pawn through the safe, yet active, king.
15 .i.f4 gives White a clear edge) 9 c3 lieS 10 6 c6
...

..ie2 .!tlc6 11 0-0 .d7 and Black seems to be


no \vorse.
6..i.b5+
More critical is 6 .!tlxe4 dxe4 7 'A'xd8+
.i.xd8

Also possible, although appearing less


logical, is 6...�d7 7 Wc2 c6 8 �d� ltlcs 9
.!tlb3.!tlxd3+ 10 cxd3 cS with unclear play.
7 i.d3&5
7....!tlxd2 8 .lxd2 cS 9 c3.!tlc6 10 0-0 0-0,

165
The Petroff Defence

with level chances, is simpler. 20 lle1 e5


8 �e2 .i.g4 9 llXI4 .i.xe2 1 0 •xe2 tLlbd7 20...l:ld3? would lose to 21 cS+! WxcS 22
1 1 0-0 tLle6 .i.e3+ Wd6 23 .i.cS+! WxcS 24 llacl + with
11 ...0·0 12 f4 lbe6 13 ib2b3 l0xd4 14 mate to follow- a very nice line.
l0xd4l0c5 is 'calmer', although the position 21 b3 lld3?1
remains terribly double-edged. 21...l:lf6! is best, although this can be
1 2 �xe& fxe& 1 3 •g4 tLlxe5 strongly met by 22 f4!? (22 .i.a3+ We6 23
Black must take accept the challenge. ..xc6+ Wf5 24 'ird7+ We6 25 g4+ Wxg4 26
13...Wf7 gives White the initiative after 14 1i'xe6+ :Xe6 27 cxdS .i.xa3 28 dxe6 Wf5
l0f3 1i'c7 15 .i.gS l:lhe8 16 l:lfe1 - Black's looks fmc for Black) 22.. .llxf41 (22. e4 is
. .

king is not really safe. weaker due to 23 .i.a3+ We6 24 f5+! l:lxf5 25
14 Wxg7 tng& 1 5 tLlf3 Wd7 1i'xc6+ Wfl 26 l:lfl !l0f4 27 l:lxf4! l:lxf4 28
After 15...1i'd6 16 lOgs .i.xgS 17 .i.xgS 1i'c7 and Black is lost) 23 cxdS 1i'c8 24 �
..d7 18 ..c3 White is slighdy better accord­ .i.d8 25 .i.a3+ WxdS with unclear play. Our
ing to Greenfeld. computer prefers Black, but that assessment
16 c4 can change in an instant!
22 .i.a3+

1 6. .••g8?1
Black should probably play 16...Wf8 17 22 .•• c5?
11fc3 llgs 18 :C1 .tb4 19 �3 dxc4 20 Black is better off playing 22. .We6. After
.

:dt+ Wc7 21 Wc2 .i.cS with a very unclear 23 cxdS+ l:lxdS 24 Wxc6+ .i.d6 25 l:lad1
position (but not 19....i.xe1? 20 ..xb7+ Wd6 l0e7 26 l:lxeS+ WxeS 27 .i.xd6+ Wf5 Black
21 cS+ WxcS 22 .i.e3+ Wd6 2.1 .i.c5+1 and seems to survive miraculously. However, the
mate is imminent). stronger 23 .i.xe7l0xe7 24lLe5+ WxeS 25
1 7.d4 llf81? 1i'xe7+ Wd4 26 l:le1 gives White a very dan­
17....i.d6 18 cxdS exdS 19 :C1 looks to gerous attack.
give White a slight edge. 23 .i.xc5+ ! wxc5 24 llac1 dxc4
1 8 Wxa7 .:Xf3 1 9 1i'xb7+ Wd& 24... Wd6 25 cS+ We6 26 11fa6+ would pick
After 19 . We8? 20 ..c8+ .i.d8 21 .xc6+
. . up the rook.
Wfl 22 cxdS White's position is crushing. 25 .:Xc4+ 1 -0

1 66
3 d4: Fourth Mo ve A lternatives

Summary
I�xpcrience has proven that in the case of 4 dxe5 d5 5 ltlbc12 (and 5 i.d3) Black maintains
e<.Juilibrium in various ways. However, like in some other systems of the Petroff Defence, it is
difficult for Black to move from a solid equalisation to seizing the initiative. That's why we
recommend trying the sharp 4....tc51? - we can't fmd any advantage for White after either 5
1Wd5 or 5 .tc4.
Murey's brilliant discovery (4...ltlc61?) looks artificial and poor at ftrst sight. In fact, it is un­
clear how White can gain <.-ven a slight advantage. In any case, 5 i.xe4 d5 6 .txh7 (as in Game
64) cannot be recommended for White since it gives Black active counterplay.

1 e4 e5 2 �f3 �f& 3 d4 �xe4 4 dxe5 (D)


4 .td3 ltlc6 (D) - Gan1e 64
4...d5 (D) 5 .i.d3- Game 65
5ltlbd2 - Gan1e 66

4 dxe5 4... &6 4... d5

167
CHAPTER NINE I
3 d4: Black Plays 3 ... exd4

1 e4 e5 2 lLlf3 ll:Jf6 3 d4 exd4 numerous ideas from Morozevich, S ... f5!?, is


In this chapter we wiD discuss 3...cxd4, an enticing option (sec the notes to Game
Black's main alternative to 3...l£lxe4. foirsdy 67). Bacrot appears to have chosen a correct
though, let's briefly go over a couple of other reply, with his position being preferable after
tries for Black. The move 3...d6 simply trans­ 6 .i.c4 .i.cS 7 ..xeS!?lL\xcS 8 .i.gS.
poses to the Philidor Defence, which is out­ Let's get back to the main line S ... dS 6
side the scope of this book. The symmenical exd6 t£lxd6. The specific features of this
reaction 3...d5 drew players' attentions as far position are the two open files in the centre,
back as 1900 (Pillsbury-MarshaiQ, but became the queen on d4 that is vulnerable but puts
especially noticeable after the Stein-Bronstein pressure on the enemy camp, and the black
game (Ibilisi 1966). At the same time, this knight on d6 that has J.,rained stability but
game contributed to the decline of 3...d5. It prevents Black's development. To avoid con­
has been proven that White doesn't have to fusion with the transposition of moves, we
transpose to the other lines of the Petroff collected all the lines where White declines to
Defence with either 4 dxeS or 4 l£lxe5 l£lxe4 5 play the most natural move 7 l£lc3 in Game
.i.d3. Instead, the continuation 4 exdS! cxd4 5 69. In this case it is interesting to contrast the
.i.bS+ breaks the symmetry in White's favour typical fianchetto ....i.f8-g7 with the ma­
(see the notes to Game 67). noeuvre .i.e1-d2-d.
Moving onto 3 ...exd4, let's leave aside the Black has various ways to protect himself
Urusov Cr.unbit (4 .i.c4), since this is the after 7 l£lc3 l£lc6 8 ..f4, three of which arc

Bishop Opening (1 e4 eS 2 .i.c4 l£lf6 3 d4 studied in Game 67. The arrangement wid1
cxd4 4lL\f3), or the Two Knights Defence Of ...JJ..e7 and ....i.e6 is effectively met by the
4 ...l£lc6 is played). Instead the games in this queenside castling plan. The attack on the c2-
chapter arc devoted to the tabiya after 4 eS pawn by ....i.f5 is not popular in view of 9
llk4. .i.b5, but White doesn't seem to have a sub­
White's main choice is the obvious 5 stantial advanl:3b>c. The main part of Game
'i6'xd4 (Game 67-69). After S...dS 6 exd6 67 concentrates on the surprising 8. ..t£1f5!?­
t£lxd6 Black faces a dreary prospect of fight· the knight moves for the fourth time our of
ing for equality with only a slight chance of eight to prepare attacking the enemy queen
sei;dng the initiative. This is why one of the with ....i.d6. This idea was rriet.l for the ftrst

168
3 d4: Bisek Plsys 3 . . . exd4

time during the Kasparov-Karpov match c3 dxc3 9 1t'xd8+ �xd8 10 l0xc3 �e7 11 0-0
(New York, 1990). A fourth option, the very lld8 (Stein-Bronstein, Thilisi 1966) and here
fashionable fianchetto of dte black bishop Stein could have obtained a sb.eable advan­
with 8 . g6, is covered in Game 68. Here cas­
. . tage with 12ll)a4 .i.b6 13 b3 .i.c7 14 .i.e3.
ding long looks more risky for White than 4 e5
against 8....i.e7, but White still stands better. 4 .i.c4 transposes to the Bishop's Open­
The paradoxical S .i.b5 attracted our at­ ing (2 .i.c4ll)f6 3 d4 exd4 4ll)o).
tention (fal employed it a few times) and is 4 �
...

covered in the annotations to Game 70. The only sensible move. If 4...ll)d5?! then
Black can turn the game into the Berlin De­ 5 '1Vxd4 c6 6 .i.c4lt!b4 7 0-0 lt!xc2 8 .i.xf7+!
fence of the Ruy Lopez by means of 5 .lik6,.. �f7 9 ..c4+ �e8 10 .xc2 and White's
transposing to a line that is harmless for advantage is clear-cut.
Black (1 e4 e5 2 ll)f3 ll)c6 3 .i.b5 ll)f6 4 d4 5 1Vxd4
exd4 5 e5 ll)c4). A wormy alternative is 5 ...c6
6 1Wd4 'tWaS 7 c3 ll)xf2 (Keres) 8 0-0 ll)h3!
(F.uwe) - anomer in-between move that de­
stroys White's pawn chain.
The main part of Game 70 concentrat<..'S
on me interesting move 5 1t'e2. Steinitz dis­
cussed dtis move in his book Modmt Chess
lflslnlctor (1889). White attacks the centralised
knight from the e-ftle in order to prevent
Black from strengthening the knight with
either the d- or the f-pawn. However, after
s....i.b4+ 6 �dl !? (Steinitz) it is possible to
improve support the knight with 6...d5 7
exd6 f5, when Black has the initiative. That's 5 d5
•..

why White prefers the calmer 6 ll)bd2, while Morozevich has risked the remarkable
Black often retreats with 5 ...ll)c5 after 5 1t'e2. 5 ... f5!? here:
Black's play in the opening in Game 70 is still
considered to be a perfect model.

Game67
Cabrera-Collas
Malaga2003

1 e4 e5 2 �f3 l0f6 3 d4 exd4


3...d5?! is inferior:. 4 exdS! (4 dxcS ll)xe4
transposes to 3...ll)xe4 4 dxe5 dS) 4...exd4
(after 4...e4 5 ll)e5 ll)xd5 6 .i.c4 .i.e6 7 0-0
.i.d6 8 .i.xd5 .i.xd5 9ll)c3 c6 10 '1Ve2 White
has a clear advantage- Yusupov) 5 .i.bS+ c6
6 dxc6 bxc6 (6...1t'a5+ is critical but White is a) After 6 cxf6 ll)xf6 7 .i.g5 lt!c6 8 Wh4
much better after 7 ll)c3 bxc6 8 ll)xd4! cxb5 1t'e7+ 9 .i.c2 �+ 10 'irxb4 .i.xb4+ 11
1J 'IVO 1Wc7 10 ltldxb5!? 1Wc7+ 11 .i.e3 .i.b7 lt!bd2 0-0 it is doubtful that White has any
12 'irf4 - Znosko-Borovsky) 7 .i.e2 .i.c5 8 advantage.

1 69
Th e Pe troff De fen ce

b) 6 .i.c4 .i.cS 7 'ifxcSI? (the queen sacri­ �xc2 1 5 :tel .i.d3 1 6 tnds 'ird8 17 1i'd4
fice is, of course, only temporary; instead 7 �a6 (after 17....tg6 Sax gave 1 8 0-0 lDfS 19
1i'd5 .i.xt2+ 8 We2 1i'e7 9 lDbd2 lDf6 10 'ireS lDxe3 20 lDxe3 as clearly better for
1i'd3 dS 1 1 .i.bS+ c6 1 2 Wxf2 /0g4+ 13 We2 White) 18 lDxc7 l:tb8 (Sax-Yusupov, Rotter­
cxbS is very complicated) 7 lDxcS 8 .i.gS
..• dam t 988). Now White should play 1 9
1i'xgS 9 lDxgS lDc6 10 0-0 lDe6 1 1 ltlxe6 (1 1 lDxa6! WaS+ 20 ltlb4 ltlf5 (not 20...l:txb4? 21
ll)f3 b6 1 2 .ids .i.b7 13 c4 0-0-0 14 lDc3 h6 'ifxd6! lld4+ 22 b4, winning immediately) 21
1S l:tad1 gS gave Black good countetplay in l:tcSI lDxd4 22 LaS l:txb4 23 0-0 with good
Bacrot-Moro7..evich, Biel 2003) 1 l ...dxe6 12 winning chances.
f4 a6 1 3 lDd2 and Black's position is slightly b4) 9...9e7+
cramped.
6 exd6 lLlxd6 7 lL!c3 �6 8 11f4 lLlf5
Other than 8 ...g6 (sec Game 68), Black
has two inferior options:
a} 8 ....i.e7 9 .i.d3 (also good enough is 9
.i.c3 .i.e6 to 0-0-0 0-0 1 1 lDgS .i.xgS 1 2
1i'xgS 1i'e7 1 3 9xe7 tDxe7 1 4 .i.cS l:tfd8 1 5
�e2, when White's bishop pair will be use­
ful) 9 ...�e6 10 .i.e3 �f6 1 1 0-0-0!? .i.xc3 12
bxc3 11'£6 (after 1 2...�xa2 13 c4 'if£6 14
:bet White keeps the initiative) 1 3 ...xf6
gxf6 1 4 ltld4 �d4 1 5 .i.xd4 Wc7 16 l:thc1
b6 17 f4 c5 1 8 .tf2 f5 1 9 g41? and Black was
under severe pressure, Parkanyi-Krivolapov, with a further split:
Gyongyos 1 998. b41) 10 Wf1 .te4 1 1 .i.xc6+ .i.xc6 1 2
b) 8. .. �f5 9 .i.bS lDe s 'ire6 1 3 lDxc6 bxc6 1 4 'ir f3 'ifc4+ 1S
Wg1 .i.e7 1 6 b3 1i'a6 17 �b2 0-0 and White
has a slight edge. Instead Klovans-Harman,
correspondence 1 967 continued 1 2... 0-0-0?!
1 3 lDxc6 bxc6 14 'ifa4 lDbS 1 S 1i'a6+ Wb8
1 6 .tc3 9b4 17 'ifxc6 tnd4 1 8 11fa6 .i.cS 19
a3 'ifb7 20 1i'xb7+ Wxb7 21 l:tct �b6 22 g3
when White was simply a pawn up.
b42) The simple 10 .i.e3 is also promising:
l O...lDxbS 1 1 lDxbS 9b4+ 1 2 'iVxb4 �xb4+
1 3 c3 .td6 (not 1 3 .. i.aS?! 14 b4 �d3 1 S a4
a6 16 lDbd4! .tb6 1 7 lDxc6 bxc6 1 8 .txb6
cxb6 19 lDeS 0-0-0 20 lDxc6 l:the8+ 21 Wdl
lldS 22 tnd4 and White is a pawn up) 1 4
and now: lDxd6+ cxd6 1 S 0-0-0 .i.e6 1 6 :Xd6 .txa2 1 7
bl) lf 9 ...�xc2? then 1 0 lDcs wins in­ .tcS! 0-0 1 8 l:txc6 bxc6 19 .i.xf8 Wxf8 20
stantly. tnd2 .tdS 21 8 l:te8 22 Wc2 and White's
b2} 9 ...lDxbS? 10 lDxbS 1i'e7+ 1 1 �fl l better structure gave him the edge in Matu­
and c7 drops. lovic-Kholmov, Sochi 1 968.
b3) 9..i.e7 10 .i.xc6+ bxc6 1 1 llks 0-0 9 .i.d2
1 2 lDxc6 1i'e8 1 3 lDxe7+ 1i'xe7+ 14 .i.e3 White has a range of options, but Black

1 70
3 d4 : Black Pla ys 3 . . . exd4

should be fine in all cases: A bold and unclear <:JUeen sacrifice. l11e
a) 9 lDbS .i.b4+ 10 c3 .laS 1 1 1i'e4+ ..e7 simple 1 3 .i.bS leads to equality after
12 1i'xe7+ lDcxe7 1 3 .lf4 lDdS 1 4 .ics f6 1 S 1 3 ... lDxb5 1 4 lDxbS 0-0 1 5 llhe1 1i'f6 1 6
0-0-0 a6 1 6 llxd5 axbS 1 7 .lf4 c6 and the �3 .i.c6.
JX)Sition is level. 1 3....i.xa4 1 4 �xf5 1t'd7 1 5 �xg7 + �dB
b) 9 .i.c4 .id6 1 0 ..e4+ .le7 (1 0.....e7 is Instead 1 S ...We7?1 16 llhet Wf6?1 walks
what White wants: 1 1 .id2 ..xt.-4+ 12 �xe4 into trouble: 1 7 �hS+ �g6 1 8 g4 and Black
0-0 1 3 0-0-0 J.e6 1 4 .i.xe6 fxt.-6 1 5 l:hel will lose too much material in escaping from
l:r.ae8 1 6 �fgS eS 1 7 .lc3 a5 1 8 a4 b6 1 9 f3 the mating net.
h6 20 lDxd6 lDxd6 2t lt)c4 lDf5 22 b3 with a 1 6 .tg5+ �c8 1 7 ZZ.d41 ZZ.g8?1
pleasant advantage, Bc:rzinsh-Nciksans, Riga
2003) 1 1 .ld2 0-0 1 2 .idS .if6 13 0-0-0
lDcd4 1 4 g4 lle8 1 5 ..f4 lDxf3 1 6 'Wxf3 lDd4
17 1i'g2 c5 (the more passive 17...c6 leaves
White with the initiative after 1 8 .le4 .le6
1 9 gS .leS 20 .i.e3) 1 8 .i.e3 'Wb6 1 9 lDe4
i..e7 20 c3 .le6 21 .ixe6 lDxe6 was unclear
in Tiviakov-Ye Rongguang, Groningen 1 997.
c) 9 .ibS .i.d6 1 0 ..e4+ ..e7 1 1 .ig5 f6
12 .id2 .id7 1 3 0-0-0 1i'xe4 1 4 lDxe4 ole7
1 5 g4 a6?! (more accurate is 1 5...lDd6 16
lDxd6+ .lxd6 1 7 :del+ �fB 1 8 l:btg1 lle8
1 9 llxe8+ ¢>xe8 with equality - A7mai­
parashvili) 16 .lxc6! Qnstead 1 6 .i.c4 lDd6 17 Now White regains the sacrificed material
lDxd6+ J.xd6 1 8 l:lde 1+ was agreed drawn in with interest. Black should play 17 .....c6 and
Kasparov-Karpov, World Championship now:
!Game 1 01, New York 1 990) 1 6 ....i.xc6 1 7 a) Not 1 8 l:txa4? ..xg2 1 9 .ig4+ fSI and
llhe1 .lxc4 1 8 llxe4 lDd6 1 9 lle2 � f7 20 Black wins more material.
�4 lDc4 21 .if4 and White has an edge. b) Also weak is 18 �xa4?! 1i'xg2 1 9 .i.g4+
9....td6 1 0 1t'a4 •e7+ Wb8 (19... 5?! 20 lDxfS ..xht+ 21 Jldt
..xh2 22 lDxd6+ �b8 23 lDfl is unclear) 20
:Ct hS! 21 lDxh5 bSI 22 lDc3 llxh5! and
Black is much better.
c) Best is 1 8 :hd ll 1 8..ltg8 1 9 .i.g4+
Wb8 20 .i.h6 'Wxg2 with a massively compli­
cated position.
18 .tg4 ZZ.xg7 1 9 .txd7+ .txd7 20 .tf4
ZZ.xg2 21 �d6 cxd6 22 �d6 �7 23
ZZ.hd1
The game has simplified leaving White
with a useful extra pawn.
23 .tc6 24 ZZ.f6 ZZ.xh2
••.

Or 24...:m 2S h4 and White keeps his


10....i.d7 1 1 0-0-0 lDcd4 1 2 llet+ �£8 13 material advantage.
'ifc4 with equality is also acceptable. 25 lbf7+ �6 26 1i:ld5+ �.5
1 1 .te2 .td7 1 2 0-0·0 �cd4 1 3 1i:lxd4!? Now the king is in danger. Black had to

171
The Petroff Defenc e

take the knight. even though the rook ending a) Getting n."3dy for kingside castling with
is probably lost. 9 J..bS is also reasonable: 9 ...J..g7 10 0-0 0-0
27 a3 :h8 28 b4+ Wb5 29 :d4 1 1 J..xc6 bxc6 12 J..e3 l:.b8 1 3 l:lab1 aS 1 4 a."
(it's unclear after 1 4 J..cs llc8 1 5 llkl4 J..e S
16 'irf3 J..b7 17 .J:lfd1 •gS) 14...l:.e8 15 J..a7
l:.b7 1 6 J..cS ( 1 6 J..d4 gives Black the option
of 1 6...g5!? 17 .cl f6 HI h3 .i.f5 with a
sharp position) 1 6...J..f5 1 7 1i'a4 ll:k4 1 8
l:.lxlt ti'c8 19 J..d4 li)xc3 20 .i.xc3 J.. xc3 21
bxc3 l:.bS and White may have a tiny advan­
tage, Romanishin-Smyslov, Leningrad 1 977.
b) 9 .i.d2, like 9 J..e3, prepares castling
queenside, but this time preventing the pos­
sibility of doubled c-pawns, although of
course the bishop is slightly less active:
9...'..c7+ 10 J..e2 .i.e6 1 1 0-0-0 J..g7 1 2 h4
29 ••• 85 h6 1 3 l:.he1 0-0-0 14 .i.d3 ti'f6 1 5 1i'xf6
The king is also caught after 29...a6 30 c4+ .i.xf6 1 6 .i.xg6 .J:.dg8 (this is the right rook;
Wa4 31 �b2 .i.xd5 32 llxdS b6 33 .J:.ff5 b5 instead 1 6...l:.hg8 17 hS J..g4 18 J..h7 l:.h8 19
34 1lf6 liaS 35 cxb5 axb5 36 .J:.dd6. ..td3 J..xh5 20 .i.e2 J..g6 2·1 lbd5 ..tg7 22
30 a4+ Wxa4 31 �c3+ ¢'a3 32 b5 .i.c3 .i.xc3 23 li)xc3 l:.he8 24 li)h4 gave
:h1 + 33 ¢'d2 1 -0 White an edge in Geller-Smyslov, Moscow
r------. 1 991) 17 hS ll:k7 1 8 .i.f4 J.. xc3 19 bxc3
Game 68 li)xg6 20 hxg6 l:.xg6 21 li)h4 llhr4 22 g3 h5
Stefansson-Yusupov and the position is level.
Eupen 1994 9 ....i.g7 1 0 0-0-0
1 0 .i.d3 is a worthwhile alternative. For
1 e4 e5 2 �f3 �f6 3 d4 exd4 4 e5 � example, lO....i.e6 1 1 0-0-0 1i'f6 1 2 �
5 11xd4 d5 6 exd6 �xd6 7 .!003 .!006 8 ti'xf4 1 3 J..x f4 0-0-0 (Yusupov assessed
1rf4 g& 9 ..ie3 1 3....i.xc3 1 4 bxc3 0-0-0 1 5 li)xe6 fxe6 1 6
l:.het l:.hfB 1 7 .i.g3 .J:.de8 1 8 f3 as a bit bet­
ter for White) 1 4 l:.he1 .i.xc3 1 5 bxc3 .i.xa2
16 c4 (Black has an edge after the weaker 16
J..xd6 l:.xd6 1 7 c4 .J:.d4 1 8 lDxfl llg8 19 �5
lDxcS 20 .J:lxeS .J:.gu8) 16 ...h6 1 7 lbe4 lDxe4
(less accurate is 1 7 ...lba5 1 8lbxd6+ cxd6 19
cS! d5 20 .J:.e7 .J:.d7 21 llxd7 �xd7 22 J..bS+,
as in Raetsky-Rodionov, correspondence
1 982; now 22...�c8 23 lt:t .i.c4 24 l:.c8+
.l:lxe8 25 .i.xe8 would have been marginally
in White's fitvour) 1 8 .l:lxe4 l:r.he8 1 9 llde 1
:Xc4 20 llxe4 g5 21 .i.d2 ..ib3 22 h4 and
White has reasonable compensation for the
Preparing to castle long without worrying pawn.
about the prospect of doubled pawns. Alter- 1 0 . . ...ie6
natively: This is more flexible than 1 (J...0-0, which

1 72
3 d4: Bisek Pla ys 3 . . . exd4

perhaps commits the king a move too early: a winning attack for White.
1 1 h4 h6 12 .i.cS!? .i.e6 13 .i.bS a6 14 .i.xc6 28 lle2 •xf3 29 1i'xf3 llxf3 30 �2 c6?
bxc6 1 5 .id4! (the greedy 1 5 .ixd6 cxd6 1 6 Now White wins by force. 30...gxh5 31
l:lxd6 allows Black dangerous counterplay �13 l:.xd3 32 cxd3 �18 woukl have given
after 1 6 ...'iWb6 l7 lDd4 ltab8) 1 5...f6 1 6 .icS Black some drawing chances.
l%f7 1 7 llhel ..Ll7 1 8 �4 WeB 19 'ifg3 31 lbe4 lbb5 32 c4 lbc7 33 lld7 lba6 34
�b7 20 �a4 g5 21 'ifb3 �aS 22 'ifd3 and hxg6 l:lef8 35 g7 Jl8f7 36 lld8+ �xg7
White had a powerful bind in lvanchuk­ 37 llg2+ �h7 38 lbg5+ 1 -0
Akopian, Lucerne 1997.
1 1 lbg5 Game 69
Bonch-Osmolovsky - Baranov
Moscow 1954

1 e4 e5 2 lL'lf3 lbf6 3 d4 exd4 4 e5 lbe4


5 •xd4 d5 6 exd6 lt!xd6 7 j.d3
White's other alternatives to the usual 7
�c3 are:
a) 7 .if4 ltlc6 8 'ilfd2 'ii'e7+ 9 .ie2 tlJe4
1 0 'ilfe3 ltlb4 1 1 Wet .if5 (chasing material
with 1 t ...1Wc5 is risky: 12 0-0 �xc2 - or
1 2...'ilfxc2 1 3 'ii'c: t !? with an unclear position
- 13 tlJc3 tlJxc3 1 4 1Wxc2 ltldS 1 5 We4+
.ie6 1 6 ltac1 and White has a dangerous
A promising alternative is 1 1 .ibS!? .i.xc3 initiative) 1 2 0-0 0-0-0 with a double-edged
12 bxc3 0-0 13 �gS fle7 14 .icS when position.
White's activity compensates for his weak­ b) 7 .igS
ened strucrure.
11 0-0 1 2 h4 �e5 1 3 •f3 h5
•..

Grabbing a pawn with 1 3....ixc3 14 bxc3


.ixa2 leads to great danger after 1 5 hS tbcs
1 6 Wf4.
14 lbxe6 fxe6 1 5 �3 •f6 1 6 �g5
�f4+
Or 16 ...flxf2 1 7 .id3 .i.xc3 1 8 bxc3 tlJf5
19 .i.c4 and White's initiative continues.
1 7 �xf4 •xf4+ 18 �b1 Aae8
If 1 8...lClf5 then 19 f3 consolidates
White's edge.
1 9 �e2 lbe5
Taking the f-pawn is still risky: 19 ... Wxf2 and now:
20 .llhft 1fc5 21 llxf8+ �xf8 22 g4 with a bl) 7 ...ltlc6 8 Wc3+ (or 8 'Wc3 f6 9 .if4
powerful attack. 1fc7+ 10 .i.e2 .ie6 1 1 0-0 0-0-0 and Black
20 lZhe1 .J:lf5 21 f3 1ib4 22 a3 �6 23 has equalised) 8....i.e7 9 lDc3 lDf5 10 .ixe7
g4 l:lf4 24 gxh5 lbdc4 25 �xc4 lbxc4 26 lDcxe7 1 1 WeS 0-0 1 2 lld1 ltld6 1 3 .i.d3
b3 lbxa3+ 27 �c1 •f2 ltlg6 1 4 flg3 We7+ 1 5 'it>J2 ltlf5 1 6 .i.xf5
If 27...Wc6 then 28 'ifg2 llf7 29 /.De4 with .i.xf5 17 ..t>ct lladB 18 .!lkl4 .i.c8 19 llhe1

1 73
The Petroff Defence

� was level in Konstantinopolsky­ ..xg7 1i'd5 20 .i.xg6 and the king is in real
Smyslov, SvenUovsk 1 993. danger) 1 1 11xf6 .i.xf6 12 .i.c3 �xc3 1 3
b2) Black should seriously consider the di­ c!bxc3 and White may have a tiny advantage.
rect 7 ... f6. For example, 8 .i.f4 c!lk6 9 11t"d2 8 i..e3 .ltlf5?1
�fS 10 .i.e2 11t"e7 1 1 0-0 0-0-0 1 2 lte1 lfr4 1bis wastes too much time - instead
1 3 11t"ct gS 14 .i.d3 �h6!? 1 5 llk3 (the wild Black should develop with 8.. ..i.f5. For ex­
t S li)h4!? leads to unclear play after 1 5...gxh4 ample, 9 ill c!bc6 1 0 11'f4 .i.xd3 1 1 cxd3
16 �xh6 llhg8; note that White must avoid 'iWe6 12 0-0-0 .i.e7?1 (instead 12 ...1fg6 1 3
1 7 1i'f4? ltxg2+! 1 8 Wxg2 :gS+ 1 9 �1 lthe 1 0-0-0 is unclear) 1 3 d4 11'f5 14 d 5 c!bb8
/bg3+) 15 ...gxf4 16 .i.xc4 .i.xe4 17 l:lxc4 1 5 Wxf5 c!bxf5 16 .i.f4 c!ba6 1 7 g4! c!bh4 18
flg7 1 8 11'f1 .l:thg8 1 9 .1Le1 1i'g4 20 �h1 c!bxh4 Lh4 1 9 d6! 0-0-0 20 c!bbs c6 21 d7+!
�f8 was unclear in Stcinitz-PiUsbury, St Pe­ :Xd7 22 c!bd6+ l:lxd6 23 .i.xd6 and White
tersburg 1895. However, after 12 ..'iWf7!? 1 3 was much better in Spassky-Kholmov,
.i.d3 .i.g4 Black has a slight plus. Rostov on Don 1 960.
9 i..xf517 .i.xf5 1 0&3 Wb4
White's lead in development is significant
whichever way Black plays:
a) 10....i.xc2 1 1 :ct llk6 1 2 'ff£4 c!bb4 13
0-0 and the initiative continues. For example,
1 3...c!bd3?! 14 'ti'c4 c!bxct 1 5 Let and, with
c!bd5 also threatened, White is winning.
b) 1 0...c!bc6 1 1 .f4 1lt"b4 1 2 0-0-0 and c7
is again impossible to defend.
1 1 fke5+ i..e6 1 2 0-0-0 cltlc6 1 3 fkxc7
Ac8 1 4 fkf4 1ra5

7 fke7+
•••

Also interesting is 7 ...c!lk6 8 11'f4 f!!> 9 0-0


�7 and now:
a) 10 llc!t+ leads nowhere after 1 0....i.e6
1 1 c!bgS 0-0 1 2 c!bxe6 fxe6 1 3 .g3 (Hiibncr­
Scgal, Dresden 1969 continued 1 3 11t"g4?!
•f6 14 11Vg3?! c!bb4 1 5 ltfl c!bfS 16 11'xc7?
.1Lc8 1 7 1i'xb7 c!bxd3 1 8 cxd3 c!bg3! 1 9 hxg3
ltxcl! 20 ltxct 11'xf2+ 21 �h1 .xg3 22 1l'e7
1i'c3! and Black won) 1 3....i.d4 14 :C2 (not
1 4 ltfl? c!be4! 1 5 .i.xc4 .i.xf2+! and Black
wins immediately) 14 ...c!bf5 1 5 flh3 c!bcs and 14 ...•xf4 is met by 15 ..i.x£4 .i.b4 1 6 c!bc4
Black has excellent counterplay. and White is a pawn up.
b) This sacrificial to .i.d2!? is the only war 1 5 ..g5
to cause Black trouble: 10...11'f6 (White has Simple is best. Instead 1 5 lOgS .i.xa2 1 6
good comp<.>nsation if Black tries 1 0....i.xb2 c!bxh7?1 leads to trouble after 1 6. ..1Lh71 1 7
1 1 .i.c3 .i.xa 1 , for example 1 2 .i.xa I O-Il 1 3 We4+ .i.e6 1 8 Wxh7 1Wat+ 1 9 c!bbt c!bb4
.i. f6 'iWd7 14 1ih6 c!bes 1 5 .i.g7! c!bxg7 16 when Black's attack is very dangt:rous.
c!bgs :es 1 7 1Wxh7+ �f8 18 flhs+ �e7 19 1 5 11ra6 1 6 .l:he1 10b47
•••

1 74
3 d4: Black Pla ys 3 . . . exd4

Now Black is losing. 1 6...h6 1 7 1Ph5 i.e7 5 tDc5


.•.

1 8 i.gS! 0-0 1 9 i.xe7 lhxc7 20 lhd4 is un­ This simple move is more reliable th:m
pleasant but at least Bl.1ck can play on. 5...j.b4+
1 7 c!i)d4 Zlxc3

and now:
Allowing a beautiful finish. 17 ...f6 1 8 a) Steinitz suggested the strange 6 Wd1,
'irh5+ f!!l 19 11b5+ 11'xb5 20 �lxbS would for example 6...d5 7 exd6 f5 8 tbgs (after 8
have lost slowly and painfully. dxc7 11'xc7 9 lhxd4 lhc6 10 c3 lhxd4 1 1
1 8 1Vd8+1! cxd4 i.d7 Black has a tremendous initiative)
A mating combination d1at exploits the 8. . .0-0 9 lhxe4 fxe4 10 11'c4+ Wh8 1 1 dxc7
power of discovered checks. (Lipschutz-Showalter, USA 1 896 continued
1 8 Wxd8 1 9 lDxe6+ We7
..• 1 1 11'xb4 .i.g4+ 1 2 i.e2 i.xe2+ 1 3 'itxc2
Or 1 9 ...We8 20 lhxg7+! i.xg7 21 i.g5+ lhc6 14 ..el?! Wfxd6 1 5 �d1 L8 1 6 b3 c3
�fB 22 :ld8 mate. 1 7 .ta3 'lrf4 and Black had a wonderful posi-
20 �g5+ f6 21 lDd8+1 1 -0 tion) 1 1 ...'ifc7 1 2 cxb811' .l:txb8 and Black
.------., obviously has a powerful initiative - Porreca.
Game 70 b) 6 lhlx12 lhxd2 7 i.xd2 'il'e7 8 0-0-0
Tal-Kholmov lhc6 9 i.xb4 Wxb4 10 11'c4 b6 11 lhxd4
AlmaAla 1968 i.b7 1 2 i.bS 0-0-0 1 .3 a3 lhxd4 (or 1 3 .. ."ti'c5
1 4 lhxc6 i.xc6 1 5 i.xc6 dxc6 1 6 f4 and
1 e4 e5 2 tDf3 tDf& 3 d4 exd4 4 e5 tDe4 White has a defmite cdbrc) 1 4 axb4 lhb3+ 1 5
5 1i'e2 cxb3 i.xc4 1 6 .l:td4 i. £5 1 7 f4 and despite his
A rare alternative to 5 ..xd4. White has bi?.arrc quecnside structure White was a litde
another unusual try in 5 .tbS c6 (5...lhc6 better in Rodrique?. Andrcs-Ginzburg, San
transposes to the Ruy l .opez: 1 e4 c5 2 lht'3 Martin 1 995.
�c6 3 i.b5 lhf6 4 d4 ed4 5 e5 lhc4) 6 'ifxd4 6 c!i)xd4 lDc6 7 �e3
'ifa5+ Qess active is 6...lhc5 7 i.c4 lhe6 8 This is more testing than 7 lbxc6 dxc6 (the
'1Ve4 dS 9 exd6 i.xd6 1 0 0-0 when White has com.'Ct recapture; instead 7 ... bxc6 8 .i.c3 'irh4
a tiny edge) 7 c3 lhxf2!? 8 0-0 (White should 9 1i'c4 'lrxc4 1 0 i.xc4 lha4 1 1 i.b3 lhb6 12
consider 8 'ifxf2 1fxb5 9 a4 11'd5 1 0 i.e3 lhd2 aS 1.3 a3 a4 1 4 .ta2 i.n6 15 .txb6 cxb6
.te7 1 1 0-0 0-0 1 2 lt:la3 with good play - 1 6 llk4 left White a touch better in Spassky­
Yusupov) 8...lhh3+! 9 b'Xh3 'lrxb5 10 11'f4 Vistiniet7.ki, Tallinn 1 959) 8 lhc3 i.£5 9 i.e3
11'd3 t l lhd4 'ifg6+ 12 �hl d5 with a com­ h51? 1 0 f4 l&.:4 1 1 lhxe4 i.xe4 1 2 1if2 ..d5
plex posicion, Zapata-Castro, Colombia 1 999. and Black has good counterplay.

1 75
The Petroff Defence

7 �xd4
.•. Instead the slow 1 7...0-0 is marginaUy in
Safer than 7...l£lxeS?I 8 f4 � 9 l£JbS l£la6 White's favour after 18 .i.d3 .i.g4 1 9 .:dfl
to .i.d4+ 'ile7 1 1 f5 1i'xc2+ 12 .i.xc2 l£lh4 h4 20 l£ld2.
13 f6 g6 14 0-0 when, despite the exchange of 1 8 ..td3
queens, White has a dangerous initiative.
8 hd4 1ih4
A more active approach than 8...l£le6 9
.i.c3 dS 10 exd6 'Wxd6 1 1 l£ld2 .i.d? 1 2 l£lc4
'ife7 (or 12 ...Wcs 1 3 l£les with a promising
attack, e.g. t3..J:d8? 1 4 1i'f3 'We? 1 S 0-0-0 c6
1 6 .i.c4 hS 1 7 l:the1 l:th6 1 8 ..i.b4 was already
winning for White in Rac::tsky-Varlamov,
correspondence 1 983) 1 3 l£leS 0-0-0 1 4 W'e3
and White has a pleasant edge.
9 ..te3 itb4+ 1 0 c3

1 8 ...f6?!
Black should continue the qucensidc pawn
storm with 1 8...a61? 1 9 l:tdf1 c5 when he has
a strong attack.
1 9 exf6 ..txf6 20 ..tc5
20 l£ld41? .i.d7 21 f5 with the initiative is
also good.
20-....tf5 21 llge1 <i>t7 22 �2
And here the simple 22 l:lc3 aS 23 l:ldc1 is
promising.
22 . . �xc5
.

1 0...1re4 Black had to be very careful: 22 .. l[hc8?1


Not 1 0 ...l£ld3+? 1 1 1i'xd3 1Wxb2 1 2 WbSI allows a nasty trick after 23 l£lxc4 .i.xc4 (not
9xa1 1 3 .i.c4 L7 14 0-0 0-0 1 5 'ifb3 when 23 ...dxe4? 24 .i.xe4 ..Le4 25 Wb3+ .i-dS 26
the queen is trapped. l:txd5 l:txet+ 27 .lld t+ l:lc6 28 l:td7+ and
1 1 f4 d5 1 2 �d2 White wins) 24 l:lxe4 and White has an edge
Or 1 2 exd6 Ld6 13 l'Od2 We6 1 4 f5 1i'e7 because 24... dxe4? loses to 25 Wb3+ l:te6 26
1 5 f6 gxf6 with an unbalanced position. .i.xb5 .i.e5 27 l:tft !.
1 2...1Vg6 1 3 lDf3 c6 1 4 0-0-0 ..te7 1 5 23 ..txf5 1rxg2 24 b4
llg1 h5 1 6 1rf2 �e4 TI1e aggressive 24 l:lg1 allows Black to es­
Black can also consider 1 6 . b6, for exam­
. . cape with a safe edge after 24...'Wxh2 25
ple t 7 llXI4 � 1 8 1i'c2 0-0 19 .i.d3 cS 20 .i.g6+ We7 26 l:tdet+ �dB 27 11f5 Wc7 -
l£1£3 .i.f5 21 llXI2 l:tfd8 22 c4 d4 23 l£lxe4 the king is no longer in danger.
(not 23 .i.xe4?1 d31 24 1i'c3? .i.xe4 25 l£lxe4 24 ...lDb7 25 1i'd3 llae8 26 llg1 1te2
1i'xe4 26 lldel b5! when Black was much 27 ..tg6+ <i>t8 28 llge1 •xe1 29 llxe1
lx:tter in Ractsky-Matsukcvich, correspon­ llxe1 + 30 Wc2 �6 31 lDf3 lle7 32
dence 198S) 23... dxc3 24 l£lf6+ gxf6 25 .LfS � llh6 %-%
1i'h6 with a murJ..l' position. Tal gave the explanatory variation 33 l£lc6
1 7 1rc2 b5!? l:lg6 34 'ifg6 :CZ 35 Wd3 l:th2.

1 76
3 d4 : Bla ck Pla ys 3 . . . exd4

Summary
Statistically 3...l0xe4 has not scored overwhelmingly better than 3 ... cxd4. However, the �"Xam­
ples shown here after 3...cxd4 were mainly played more than a decade ago.
The continuations 4 e5 lbe4 5 .ib5 and 5 '1Ve2 do not worry Black any longer, but with 5
11xd4 the situation is more serious. It appears that after 5 ...d5 6 exd6 l0xd6 7 c!lk3 c!lk6 8 1i'f4
itlf5!? Black maintains equilibrium. However, with the symmetrical pawn structure and easier
development, White has many aggressive possibilities that Black needs to be ready for. Conse­
quently, the general interest in 3 ...cxd4 has dropped, but Morozevich's experiment with 5 ..f5 is
.

interesting.

1 e4 e5 2 l0f3 l0f6 3 d4 exd4 4 e5 l0e4 (D) 5 •xd4


5 1i'e2 - Ca1ne 70
5 ...d5 6 exd6 l0xd6 (D) 7 l0c3
7 .id3 Game 69
-

7 lOc& 8 ..f4 (D) l0f5


... - Ca/lle 67
8.. g6 Ca1ne 68
. -

4. . . l0e4 6. . . l0xd6

1 77
CHAPTER TEN I
Third Move Alternatives
For White

1 e4 e5 2 lbf3 l0f6 Knights Defence, but 3...lbxc4 seems to be


After 2. ..lt)f6 White has two equally im­ strongest move. A fter 4 lbc3 Black can re­
ponant continuations: 3 d4 and 3 lbxe5. turn to the Two Knights again with 4...lbc6
Theory pays much less attention to other (by the way, this variation is favourable for
white possibilities, but statistics prove that Black). However, it is even more promising
White often chooses 3 .t.c4, 3 d3 and espe­ to hold onto the extra pawn with 4 ...lbxc3 5
cially 3 lbc3. In this case the positions dxc3 f6. After this Black's pawn chain be­
reached are not always in the domain of the comes a formidable force. I t is especially
Petroff Defence, as other openings can also useful to activate the pawn chain with ...c7-c6
arise. and ...d7-d5, neutralising the bishop on c4.
After 3 lbc3 (Games 71 -72) Black is of White should aim to undermine the chain
course able to transpose into me Four tactically by taking advantagt: of the insecure
Knighrs (with 3...lbc6). However, Pillsbury position of Black's king (especially with the
employed the move 3....i.b4, Alekhine calling push f2-f4). Objectively speaking, Black's
this continuation the 'Ruy Lopez for Black'. chances are preferable in these sharp posi­
The extra tempo lbgt-f3 plays a significant tions. However, he must be very careful:
role here, but it has to be used properly. 4 there are numerous examples of Black facing
.i.c4 (Game 7Z) is best met by 4....i.xc3 5 mpid defeats in this line.
dxc3 d6 with comfortable development for After the modest 3 d3 (Game 74) we have
Black's pieces. More promising is 4 lbxe5 a position that also arises from the Alekhine
(Game 71), after which it is easy for Black to Defence (1 e4 lbf6 2 d3 e5 3 lbf3) - in effect
win his pawn back but it is more difficult to White is playing the Philidor Defence with an
achieve an equal position. After 4...0-0 5 extra tempo. In this manoeuvring battle
lt:k-13 .t.xc3 6 dxc3 lbxe4 7 .i.e2 the modest White is rarely able to gain an advantage.
7...d6 can put Black under long-term posi­ Black can maintain the symmetry with 3...d6,
tional pressure. Black should play 7...d5 and but 3...lbc6 followed by ...d7-d5 is more
... c7-c6 with 1,•ood prospects of equalising. promising. Black can develop the dark­
The continuation 3 .i.c4 (Game 73) offers squared bishop to c5, e7 or g7, although he
Black the interesting opportunity to play normally delays the choice until move six
3 ... lbf6 with the transposition into the Two (after 4 c3 d5 5 lbixl2 aS 6 .i.e2). ln brcneral

1 78
Third Mo ve A lterna tives for White

Black docs not have any difficulties and often and Black is fmc) 1 0...a6 1 1 .i.e2 b5 12 1if4
maintains a small space advantage. .i.b7 1 3 .i.f3 lLldf6 14 tlJcs lLlxcS I S .i.xcS
.i.xf3 16 9xf3 lL!e4 1 7 .i.e3 lle6 1 8 9f5
Game 71 lieS 1 9 .g4 'it'cS 20 .i.d4 'lfxg4 21 hxg4
Svidler-Akopian lle6 and a draw was agreed. [f anyone should
World Team Ch., Lucerne 1997 feel relieved in tlte final position, it is White.

1 e4 e5 2 ll:lf3 ll:lf& 3 ll:lc3


White is trying to escape to the l'our
Knights, but Black will not allow it!
3 ....tb4

7 d5
...

Very natural. Black has also played 7...l:.eS


8 0-0 d6 and now:
a) 9 llcl lL!c6 to .i.fl .i.f5 1 1 f3 lLlf6 1 2
..t.gs lL!c:s 13 'iVdz h6 t4 .i.h4 b..s 1 5 .tf2
4 ll:lxe5 lLldS?! (1 S ...lL!c4 16 .c l l:xc1 17 lLlxe1 lL!b6
The critical move. 4 .i.c4 is discussed in 1 8 a4 would have left White only slightly
the next b>ame, while in Ragozin-Kan, Lenin­ better) 16 lL!xeS dxeS 17 .l:lad 1 tlJf4 1 8 l%xc:SI
grad 1 936 White: played 4 �S lllids S exdS 'iVxd2 1 9 llxe8+ l%xe8 20 .llxd2 and Black
e4 6 tlJd4 0-0 7 c3 .i.cS 8 d3 .i.xd4 9 cxd4 didn't have any compensation for dte pawn
1Vc7 10 .i.e3. Now after 10...d6!? 1 1 .i.e2 deficit, Lau-Ractsky, Bad Ragaz, 1 994.
.i.f5 1 2 llct c6 Black should be okay. b) 9 lLlf4 tiJc6 (or 9 ...lL!d7 1 U c4 lL!ffi 1 1
4 0-0
... a4 aS 1 2 .i.e3 .i.d7 1 3 lL!ds .i.c6 1 4 l%e 1
Also possible is 4...'tle7 S �3 (S tlJg4!?) lLlg6 1 5 .i.ft h6 16 f3 lL!f6 11 Wd2 b6 1 s b3
S ... .i.xc3 6 dxc3 tlJxe4 7 .i.e2 dS 8 0-0 0-0 9 and White has an edge in SoUeveld-Pikct,
lLlf4 c6 (but not 9...lld8? 10 lLlxdS 'it'es 1 1 Bundcsliga 2002) 1U c4 ..tf5 1 1 lL!ds h6 12
c4 c6 1 2 .i.f4 1i'e6 1 3 .i.g4! f5 14 .i.xf5 and b3 lL!e7 1 3 .i.b2 lL!xd5 1 4 1i'xdS 'ifg5 lS
White wins - Alckhine) 1 0 c4 dxc4 1 1 .i.xc4 .i.d3 (1 S f4 1Vg6 16 'iVxb7 lL!cS 1 7 'iVf3
.i.f5 1 2 'it'e2 1le8 1 3 llet 1i'd7 14 .i.c3 and .i.xc2 leaves Black in a solid condition)
·

White was slightly berter in Alapin-Aiekhinc, 1 5...c6 1 6 Wd4 l%e6 17 h4 'ifg6 1 8 l:ae1
Carlsbad 191 1 . llaeS 1 9 lle3 and White is slightly better,
5 ll:ld3 .txc3 6 dxc3 ll:lxe4 7 .te2 Chudinovskikh-Raetsky, Orel 1 992.
This is the critical line (although there is a 8 0-0 c& 9 ll:lf4 Ae8
limit to how critical such a position can be). Black is also doing okay after 9 ... .i.f5, for
Instead Galdunts-Ractsky, Aachen 1994 con­ example l 0 c4 d4 1 1 .i.d3 lieS 1 2 f3 lL!d6 1 3
tinued 7 .i.c3 lieS 8 1Wf3 d6 9 0-0-0 lLld7 1 0 c 5 i.xd3 1 4 Wxd3 tiJbs 1 5 a4 lL!c7 1 6 c 3 (1 6
h3 (or 1 0 lLlf4 lL!df6 I t h3 .i.d7 1 2 .i.c4 .i.c6 b4 aS 17 bS cxbS 1 8 axbS lL!ct7 gives Black

179
The Petroff D e fence

l."XccUcnt countcrplay) 16... dxc3 17 'irxc3 a5 i.xg4 16 'irxg4 c5 17 i.h6 rfl 1 8 Wft may
18 �h5 f6 19 i.h6 liXI5 20 'irc4 g6 21 �3 be slightly better for White) 14 i.xe6 i.xe6
�7 22 J:tfet f5 23 .i.d2 :Xet+ 24 :Xel 15 �xe6 llxc6 16 Axe6 fxc6 17 i.d2 'irf6
..fB and Black has equalised, AJams­ the game is level, Adams-Hubner, Dortmund
Rozentalis, Copenhagen 1997. 1996.
1 1 . ..1tla6 12 c3
12 0 �c5 13 lle l li)xd3 14 cxd3 i.fS
might even preferable for Black.
1 2....if5!?

1 0 c4
to i.c3 is best met by 10...�16 (10...�d7
11 c4 dxc4 12 i.xc4 �5 13 'irxd8 :Xd8 1 4
.i.c2 i.fS 15 g4 i.d7 1 6 f3 �6 17 l:tad1
li:XIc4 18 i.c1 i.e8 19 b3 �b6 20 .lb2 gave An interesting pawn sacrifice. After 12..c5
White a bit of pressure in Biro-Chetvt:rik, 13 ..c2 �f6 14 i.d2 White has a pull.
Nagykanizsa 1 995) 1 1 i.d3 i.fS 12 'ifh5 g6 1 3 g4 .tg& 14 f3 �5 1 5 .txg& hxg& 1 6
13 'ifh6 .lxd3 14 cxd3 �fS 15 'iFh3 � 16 cxd4
:ae1 'irf6 17 .i.d2 �5 18 d4 llk4 19 Act Black can answer the alternative capture
Ae4 with level chances, Benjamin-Yusupov, 16 ..xJ4 with 16...'irc71?.
Munich, 1994. 1 6... b5
10 d4
.•• The alternative 16...1if6!? 17 d5 llad8 18
h4 �h7 19 g5 1ie5, with play for the pawn,
is also strong.
1 7 cxb5
17 ..d3 is powerfully met by 17...'1fd6 18
cxb5 �b4 19 1Wb3 'ifxd# 20 Wht cxb5 21
a3 �a6 22 1ixb5 llk5 when Black is very
active.
1 7...cxb5 1 8 d5
Or 18 h4 �6 19 �xe6 Axe6 20 i.g5
1id6 21 ..d2 llaeB 22 i.f4 1id5 with suffi­
cient play for the pawn.
1 8 llc8 1 9 d&
.•.

19 i.d2 :C4 20 �g2 �c7 gives Black


1 1 .id3 good play, while after 19 h4 �h7 20 g5 �b4
After 11 i.f3 li)g5 12 i.g4 �6 13 Ac t Black's active pieces also promise him a
�6 (13...li)xf4 14 llxe8+ 1i'xc8 1 5 i.xf4 pleasant game.

180
Third Mo ve A lterna tives for White

with 23...1lc21 24 i.e3 (24 .xbS lOx£3+1 25


J:lx£3 lldt+ and Black wins) 24 .. .lldd21 25
l:lf2 1ldt+ 26 Wg2 llxf2+ 27 .Lf2 llxa1 and
Black has all the chances.
24 llxf3 l:ld1 + 25 �g2 11h4. 26 .!i)e2
llc2 27 llf2 •xg4+ 28 /t)g3 l:lxf2+ 29
�xf2 •d4+ 30 �2 %-%

Game 72
Lev-Alterman
Ramal-Can 1992

1 e4 e5 2 /t)f3 ll)f6 3 lbc3 .ib4!? 4 .ic4


1 9 .Ac6 20 d7 lle7 21 •d5 llc5!
.•

This is stronger than 21 ......xd7?1 22 ...xg5


...d4+ 23 Wh1 lles 24 Wh4 g5 25 Wg3 gxf4
26 i.xf4, after which Black looks a bit worse.
22 ..d6?
22 'il'b3 llxd7 23 i.e3 lle5, with a com­
plex position, was the right path.
22 ...llxd7
Black misses a direct win with 22...1lc2! 23
1Wd3 (23 Wxa6 l0xf3+! 24 llxf3 l:let+ 25 J:lft
Wxd7! and Black wins; or 23 h4 l:lxd7 24
1We5 l0xf3+1 25 llx£3 llc7 26 lt.I3 Wc7 and
White )uses his queen) 23...l0b4 24 Wxb5
llxd71 25 i.c3 (25 Wxb4 lld l ! and White has 4 . 0-0
. .

no defence) 25...1ldd2!! Black has reasonable alternatives here:


a) 4..ixc3 5 dxc3 d6 6 ...c2 l0bd7 7 h3
lOcS 8 i.d3 0-0 9 0-0 lle8 10 c4 i.d7 1 t b4
l0c6 12 llet aS 13 bS c6 14 c3 1Vc7 with
e'Jual prospects, Gaponenko-R.1etsl-y, Kras­
nodar 1995.
b) 4...d6 5 0-0 i.g4 6 h3 i.hS 7 g4 i.g6 8
d3 c6 (safer is 8.,.i.xc3 9 bxc3 l0c6 10 a4 0-0
1 1 l:lbt l:lb8 when Black is close to equality)
9 l0c2 l0bd7 10 c3 .ia5 1 1 b4 i.c7 12 i.b3
lOfB 1 3 ltlg3 l0c6 14 g5 lOhS 1 5 i.xe6 fxe6
1 6 l0xe5 dxeS 1 7 lLlxhS 0-0 and Black had
good compensation in Chemyshov­
Yusupov, Ohrid 2001 .
(maybe it was this brilliant move the play­ 5 0-0
ct"S overlooked) 26 Wxb4 l0h3+ 27 l0xh3 5 tOds is rather unexciting: 5 ...t0xd5 6
J:lg2+ 28 Wh t llxh2+ 29 �gl llcg2 mare. .i.xdS c6 7 i.b3 dS 8 0-0 ..i.g4 (or 8...dxe4 9
23 •xa6 �xf3 + ? lOxeS tlX17 10 d4 lOxeS 1 1 dxeS Wxd1 1 2
Black could stiU have gained the advantage l:r.xdt .i.g4 13 l:ld4 llfd8 14 l:lxd8+ l:[xd8 15

181
The Petroff Defence

.i.e3 with level chances) 9 h3 .i.hS 10 d3 One line continues 7...c6 8 llc 1 dS 9 l:lxe4!
dxe4 11 dxe4 .!LXI7 12 We2 'flc7 with an dxc4 (9...dxe4? to J..xf7+ �h8 1 1 "ifhs J.. f5
equal position, Spiclm:mn-Marshall, Buda­ 1 2 lL\g6+! J..xg6 1 3 J..xg6 leads to mate) 10
pest 1 928. :04 1ie7 t 1 lDxc4 and White was a pawn up
in LGuliev-Smougalev, Moscow 199S.
7 i.g5
Practice has also seen 7 1ie2 lDbd7 8 .i.gS
h6 9 J..h4 lDcs to l:ad1 We7 I 1 lbd2 gS 12
J..g3 Wh8 13 h4 (dlis docs not seem logical)
1 3...J..d7 1 4 hxgS lucgS I S f3 Wg7 16 Wf2
llh8 1 7 l:l.h1 lDhS 1 8 �fl 1if6 and Black
had the initiative in Kofidis-Alterm:m, Ko­
motini 1992.
7 . . . h6 8 i.h4 i.g4
8. ..g5 c:m be met by 9 �xgS!? hxgS 10
.i.xgS with the idea of meeting 10 ... J..e6 with
1 1 .i.d3 �bd7 12 f4!, after which White has
5 ....i.xc3 a very strong attack - the absence of Black's
Black usually uses this chance to inflict dark-squared bishop is really felt.
double pawns on White, but he can also try 9 h3 .i.h5 1 0 'ird3 �7 1 1 b4 'ireS
s ... d6!?, for example 6 d3 .i.g4 7 h3 .i.e6 8
lbc!S .i.xdS 9 exdS h6 10 c3 .i.aS 1 1 d4 exd4
(our improvement over 1 1 ...e4 1 2 �h4 c6 1 3
dxc6 �xc6 1 4 � f5 d S 1 S .i.b3 .i.c7 1 6 f3
when we prefer White, Priehoda-Chetvcrik,
Martin 1 996) 12 �xd4 �bd7 1 3 �f5 �S
14 .i.b3 J..b6 with more or less level chances.
6 dxc3

Also interesting is 1 l ...gS 1 2 J..g3 �b6 1 3


.i.b3 .i.g6 1 4 :re 1 aS when Black has good
counterplay.
1 2 'ire3 i.g6 1 3 lL\d2 lbh5
Or 1 3...c6 1 4 l:tfc1 dS 1 S exdS �xdS
(tS...cxdS 16 .i.bS a6 1 7 .i.xf6 axbS 18 .i.xeS
�xeS 1 9 1WxeS 1ixcS 20 l:txeS would leave
Black with a tedious endgame in which he is
Black has nothing to fear after 6 bxc3 slightly worse) 16 .i.xdS cxdS 1 7 c4 d4 1 8
lDxe4 7 �xeS dS R .i.b3 aS 9 a4 l:leR 1 0 .!LXIJ 1W f3 and the position i s rather unclear.
lDc6. 14 a4 Wtl8 1 5 a5 f5 1 6 exf5 i.xf5 1 7 a6
6 ...d6 b6 1 8 g4 •us 1 9 i.d5 lbf4 20 .i.xa8
6...�xc4?! 7 �xeS L-. better for White. llxa8 21 Wh2 i.xc2

182
Third Mo ve A l ternatives for White

Black has escaped from the opening with a) S...d6? loses to 6 lbgS .i.e6 7 .i.xc6 fxc6
adequate play for the exchange. 8 11ff3 (Bilgucr).
22 c4 l:lf8 23 .ig3 h5 24 f3 Wh7 25 b) S ....i.e7 is met strongly by 6 lbxe5 0-0 7
.i.xf4 llxf4 26 �3 1ih6 27 Aae1 c5 28 •n i.. f6 8 tDg-4 i..e7 9 lbh6+l? gxh6 10
b5 ltlf8?1 .i.xh6 with a powerful attack.
28...lbf6!?, with good play, would have c) S ...c6!? leads to equality after 6 lbxeS dS
been better. 7 Wc2 i..e6 8 .i.d3 llkl7 9 f4 lbxeS 10 ..xeS
29 ltle4 .ixe4 30 fxe4 ltle6 31 gxh5 'ild6 1 1 1Wc2 .i.c7 1 2 0-0, as in San Claudio­
'lrxh5 32 Wg2 'lrf7 33 1i'd3 'lrg6+ 34 Bonari, Mislata 2001 .
'lrg3 1i'h6 35 llxf4 exf4 36 1i'g4 t£!d4 37
h4?!

6 0-0
6 lbh4 g6 7 f4 is probably less dangerous
37 �h2! would have kept an edge for (I 0-0 transposes to 6 0-0 g6 7 lbh4). Play
White. continues 7 ... c6! 8 f5 dS 9 fxg6 (9 .i.b3 �f7
37 . . .f3+? to c4 d4 is better for Black) 9 . ..dxc4 1 0 'ifhs
Black misses his chance. After 37...lbc2! Wd7 1 1 g7 .i.xg7 12 ..g4+ Wd6! (12...Wc7?!
the game would most likely have ended in a 1 3 11'xg7+ lbd7 1 4 .i.h6 'iVe8! I S 0-0-0 l:r.g8
draw. 1 6 ...xh7 %lh8 is 'only' equal) 1 3 'iVxg7 ...f8
38 Wh3 Wh7 39 llg1 f2 40 llf1 1i'e3+ 14 ...g3 (or 1 4 ..xfB+ :Xf8 1 S .i.e3 .i.c6 1 6
41 Wg2 l002 42 llxf2 ltle1 + 43 Wt1 0-0-0+ .i.dS and Black dominates th e board)
ltld3 44 llf7 1i'e1 + 45 Wg2 ...d2+ 46 14....i.e6 1 S .i.c3 lbd7 16 0-0-0+ Wc7 17
Wg1 •c1 + 47 Wh2 'lrb2+ 48 1tg2 llhfl llg8 1 8 •rz b6 t 9 lbf5 J:ld8 and Black
•e5+ 49 1tg3 1tb2+ 50 Wh3 Wg8 51 has a small plus, Tribushevsky-Raetsky, cor­
1tf3 1 -0 respondence 1982.
,.-------------.. 6 . . .l006
GaHJe 73 Or:
Morphy-Barnes a) 6...c6? is now strongly met by 7 lbxeSI
umdon 1858 d5 8 'ifhs+ We7 9 .i.d3 with a winning at-
_______________. tack.
1 e4 e5 2 ltlf3 ltlf6 3 .ic4!? b) 6...g6 7 lbh4 ..e7 8 �h I c6 9 f4 dS 1 0
A romantic gambit. .i.b3 e4 i s worth thinking about - we prefer
3 ...ltlxe4 4 t£Jc3 ltlxc3 5 dxc3 f61 BL1ck here.
This should lead to an advanta�;,>e for c) 6 ...'ile7 7 lle1 d6! (l...c6?! is met by 8
Black. Altc�tivcly: lbxcS!? fxcS 9 'fibS+ g6 10 .xeS dS I 1

183
Th e Petroff Defence

1txe7+ .i.xe7 12 .i.gS 0-0 13 il..xe7 dxc4 14 1 1 ... .i.e6 1 2 11h5+ g&
.i.xf8 �fB t 5 llad1 and the endbrame looks 12 ...i.f7 13 'li'g4 g6, with a slight edge,
good for White) 8 �4 c6 9 f4 g6 1 0 P.i dS was also poNSible.
1 1 .i.d3 'ilg7 also looks better for Black.
d) 6...d6 7 l0h4 g6 8 f4 'ile7 9 f5 'flg7
should be better for Black. Now 10 .i.e3 c6
1 1 1te2 dS 12 .i.b3 gS 13 c4!? gxh4?1 14 cxdS
cS 1 5 d6! .i.xd6?! 1 6 lladt gave White tre­
mendous compensation in Crepan-Rezonja,
Bled 2000. However, stronger is 1 3...d4! 1 4
�xd4 gx.h4 when Black has a ck"al' plus.
7 � '11fe7
Or:
a) 7.../8:7?! is very weU met by 8 .i.d3 g6 9
f4, when White has a powerful initiative. One
game continued 9._.i.g7 10 fxeS fxcS 1 1 �gS
c6? 1 2 lLlf5! gxf5 1 3 1Whs+ WfB 1 4 :X f5+ 1 -0 1 3 ltlxg6?!
Saburov-Lutze, correspondence 1 906. Risky. After 13 'ile2 0-0-0 14 f4 .i.cS 1 5
b) 7...g6 8 f4 f5 9 l0f.3 e4 10 lOgs .i.cS+ 1tf2 .i.xe3 1 6 'li'xe3 White i s maybe only
1 1 Wh1 'ti'f6 12 'ti'dS (12 .i.£7+ We7 13 .i.ds slighdy worse.
d6 14 :C1 WEB 1 5 g4 lDe71? 1 6 gxf5 lLlxdS 1 3. . ..i.f7 1 4 11h4 hg& 1 5 •xf& llgB 1 6
17 1txd5 gxf5 is l.'VCO worse, Black having a %lad 1 .i.e7
clear plus in Schlcchter-Marco, Berlin 1 897) Not 1 6...l0e7? 17 l:lxdS! lLlxdS?l 1 8
12...d6 13 lLlxh7 1te7 1 4 ltlg5 lDds and we 1txe5+ .i.e7 1 9 .i.xdS when White wins be­
prefer Black even if it is not entirely clear. cause of the threat of .i.£7+.
1 7 .e6 .i.f7 1 8 11h3 �

8 �f5?!
8 lthS+ Wd8 9 lLlf5 (9 c!l\g6? 1te8 and 1 8...l:td8!? 19 'ilxh7 D.d6 looks good,
Black wins) 9...g6 10 l0xe7 hrxhS 1 1 lLlxc8 White not having <.:nough for the piece. After
l:lxc8 1 2 .i.e3 is better, giving White good 1 8...ltld8 the trend starts to chaf1bre.
compensation fi>r the pawn. 1 9 f4 e4 20 llxd5
8 ......c5 9 .i.b3 d5 1 0 .i.e3 1t'a5 1 1 �h4 20 .i.xdS!? is an interesting option:
Black's position is also more pleasant after 20 ....i.xd5 21 1ibS+ l0£7 22 :XdS ...xa2 23
1 1 l0g3 .i.e6 1 2 1th5+ g6 1 3 1th4 .i.e7. :Cs and White has an auack.

184
Third Mo ve A lternatives for White

20 .. �xd5 21 'lh15+
Game 74
Svetushkin-Miles
Alushta 1999
1 e4 tnf6 2 d3 e5 3 tnf3
Por our purposes, the Petroff move order
would have been 1 e4 e5 2 ltlf3 ltlf6 3 cB.
3 . . .tnc6

21 .•. �8?1
Much stronger is 21 . ..l:g6 22 .ixd5 .lc5
23 ..thl 'ifa6 (23 ...ixe3 24 .if7+ ltlxf7 25
.

11fxa5 leaves White slightly better) 24 l:le I


.ixe3 25 f5 ..f6 26 fxg6 hxg6 27 1i'g4 and
the game is still very unclear.
22 .ixd5 :g7
22.. c6? loses to 23 'irf5+ �g7 24 ..id4+.
.

23 b4 ...86 24 f5! 4 c3
4 g3 is also harmless: 4...d5 5 exd5 ltlxd 5 6
.ig2 .i.cS 7 0-0 0-0 8 :Ct %le8 9 h3?! (9 ltlc3
ltlxc3 I 0 bxc3 .i.g4 would transpose to the
so-called Glek Variation of the Pour Knights
Opening - the chances are equal) 9 .. .i.f5 10
.

ltlh4?! (1 0 ltlbd2 is better, although we pre­


fer Black after 10 .. f6 1 1 ltle4 i..b6) t o...i..e6
.

1 1 ltld2 ltlf4! (this surprising move leaves


Black with a substantial plus) 1 2 gxf4 9xh4
1 3 ltle4 .ixh3! (a nice sacrifice) 14 ltlxc5
i..xg2 1 5 �xg2 exf4 t6 ltle4 ltld4 17 f3 1le6
I 8 ltlf2 ..g3+ 19 �ft Dg6 0-1 Manik-Oral,
Trcncin 1 995.
24. . .tnf7? 4 d5 5 li:lbd2 aS 6 .i.e2 .i.c5
.•.

24...'irf6 25 ..id4 'ifg5 26 .i.xg7+ ��7 27 Black has also tried 6...g6 7 b3 (7 0-0 i..g7
•es was necessary, when White's attack is 8 .l:lcl 0-0 9 .ift J:leR 1 0 b3 b6 I I .la3 .la6
fear.;omc. but the game goes on. 12 'it'c2 'ii'd7 1 3 llad t llad� looks pretty
25 f6! .ixf6 26 b51 1Wd6 27 .ixf7 + b6 standard and pretty level) 7...i..g7 8 .la3
p
27 ...l:lxn s .icS is the end. il)h5 9 0-0 ltlf4 10 J:lc 1 ltlxe2+ I I 'it'xe2
28 .ih6 W87 29 .i.xg7 .txg7 30 ..tb3 :m .ie6 (or 1 1 ...d4!? 1 2 cxd4 exd4 13 .l:lacl ltlb4
31 :n + :xn 32 1Wxf7 + �dB 33 1Wxg7 with equality - Makarychev) 1 2 exd5 'ii'xd5
•d1 + 34 �2 1Wd2 + 35 Wg3 e3 36 13 c4!? tlfd7 14 .i.b2 .ig4 15 d4 0-0-0!? 16
irf6 + lilc8 37 i..e6 + �7 38 1Wf3 + 1 .0 dxeS llhe8 17 'We3 £6 and Black had good

185
The Petroff Defence

compensation in Morozevich-Makarychev, llae8 17 h3 e4!? 18 dxe4 fxe4 1 9 tl'lxe4 .i.f5


Moscow 1992. 20 .i.c4 �h8 21 .i.a3 .i.xhJ! 22 .i.xf8 1Wg4
23 liJM .i.xg2 24 liJg3 .i.e4! (Lcin­
Kupreichik, Hastings 1 982/83). Now after
25 llxe4 'iVxg3+ 26 �h 1 1Wh3+ the game
would have ended with a draw.

0-0 8 a4
7 ()..()
Other moves to consider are:
a) 8 "irc2 a4!? 9 llb 1 lle8 to b4 axb3 1 1
axb3 liJh5 1 2 llet liJf4 1 3 .i.fl wiili equal
chances. 9 exd5
b) 8 b3 lle8 9 a3 .i.a7 (or 9...d4 1 0 .i.b2 White has also played 9 .c2 .i.g4 to lle1
dxc3 1 1 .i.xc3 .i.g4 with equality) to llbt h6 1 1 tl'ln 'iVd7 1 2 .i.e..'� .i.f8 (or 1 2...d4 1 3
dxe4 1 1 dxe4 liJh5 1 2 c!2k4 "irf6 13 b4 axb4 cxd4 .i.xO 14 d5!? .i.xc2 I S dxc6 "irxc6 1 6
1 4 axb4 tl'lf4 1 5 .i.xf4 1fxf4 1 6 b5 with un­ llxe2 .i.xe3 1 7 tl'lxe3 1Wd7 with level pms­
ck'ar play, Gclashvili-Hellsten, Korinthos pects) 1 3 llad1 llad8 1 4 liJg3 d4 1 5 .i.cl
2002. (after 15 cxd4 .i.xf3 16 .i.xf3 liJxd4 17 .i.xd4
"irxd4 1 8 "irxc7 .i.c5 1 9 llf1 b6 Black has
fantastic play for the pawn) 1 5....i.c5 16 liJh4
.i.e6 17 liJhf5 �h7 18 .i.f3 liJg8 1 9 llk2
.i.f8 and Black was a bit better, Vt.>sclovsky­
Scrgeev, Ceske Bude�>Vice 1997.
9 �xd5 1 0 lDe4
•.•

to tl'lc4 .i.f5 1 1 "irh3 "ird7 12 tngs llab8


also gives a fair share of chances to both
players.
1 0...i.f8
Or 1 O....i.b6!? 1 1 tl'lfd2 f5 t 2 l;jg3 .i.e6 1 3
c!ik4 .i.cS with unclear play.
1 1 lle1 h& 1 2 1ib3?!
8. . .1le8 This is the path to destruction. White was
8...h6 9 "irc2 .i.c6 is also a very sensible better off playing 1 2 liJfd2 liJf4 1 3 tl'lc4
way to play. The following continuation ltJxe2+ 14 'irxe2 .i.f5, though Black has easy
looks promising for Black: 1 0 exd5 liJxd5 1 1 equality.
lle1 1Wd7 12 .i.fl f6 1 3 lbc4 .i.a7 14 b3 .i.g4 1 2 . ...te6! 1 3 1i'xb7?!
.

(14 ... 1lad8 1 5 .i.a3 llfe8 is a sound alterna­ 13 'iVbS was better, though after 13 ...'iVc8
tive; we prefer Black) 1 5 .i.e2 f5 1 6 ltJg3 14 .i.ft lla6 White is still struggling.

186
Third Move A lternatives for Wh ite

29 .i.d1 tne6 30 g3 lidS

1 3...li)b6!?
Maybe even stronger is 1 3. ..d71? 14 �5
.. 3 1 f4?1
:Cb8 1 5 li)c5 1t'e8 1 6 li)xe6 l:txb5 17 axb5 The only chance was 31 l:te3 f5 32 llk3
llkb4! (the hidden point) 18 cxb4 ..txb4 with �5 although Black still has all the fun.
a clear plus for Black. 31 ...1lxd3
1 4 1Wxc6 Simpler was 3t .....txc4! 32 l:txe4 l:txd3 33
Forced. After 1 4 ..ie3? ..id71 1 5 ..ixb6 fxe5 li)g5 34 :C1 Wb7 and Black wins.
cxb6 followed by Jla7 Black wins the queen
.. 32 lDf2 lidS 33 fxe5 f5! 34 .*.c1 .*.aS!
for a rook. After this there is no defence.
1 4... -*.d5 1 5 'ttxeS 35 lla2 1Wb7 36 Wf1 1fg2+ 37 ¢>e2
The queen is trapped, and after 1 5 'irb5 c6 .i.f3+ 3S �3 .*.xd1 0-1
it would be sold more cheaply.
1 5 1Wxe8;1'6 .i.d1 c5 1 7 c4
.••

1 7 d4 also leads to a black edge after


1 7... cxd4 1 8 cxd4 ..tb4 19 ..id2 Ld2 20
�xd2 e4.
1 7 ...-*.c6 1 S lDc3 f6 1 9 b3 'tfdS 20 .*.e2
g5!? 21 lDd2 lDcS 22 .i.b2 lDe7 23 I.Ode4
ll'if5 24 lDd5 .*.e7
After the inferior 24... ..ixd5?! 25 cxd5
lDd4 26 ..txd4 cxd4 27 ..tg4 White has excel­
lent countcrplay on the light squares.
25 .*.g4 lDg7 26 .*.f3 g4 27 tl'ixe7+
1txe7 2S .*.e2 h5
It is safe to say dmt White is far from hav­
ing sufficient compensation.

187
The Petroff D e fence

Summary
In practice White's deviations from the main continuations on move three quite often trans­
pose to other openings. The least justifiable is 3 d3, since White gives away his advantage of
the first move. Those who enjoy attacking the king should study 3 .i.c4 l£lxe4 4 l£lc3 l£lxc3 5
dxc3, which is especially recommendable fi>r rapid and blitz tournaments. Unfortunately,
4...l£lf6 cools down White's ardour.
3 l£lc3 .i.b4 leads to l)uite another type of position with a slow, manoeuvring fight and
White having the advantage of two bishops (often in the endgame). Like in many other basic
variations of the Petroff Defence, Black has to neutralise White's pressure with careful play.

1 e4 e5 2 .!Df3 .!Llf6 3 l003


3 .i.c4 (D) - Game 73
3 d3 - Gm11t 74
3 .•• .i.b4 4 c!Dxe5 (D) - Game 71
4 .i.c4 (D) - Game 72

3 -*.c4 4 c!Dxe5 4 .i.c4

188
I INDEX OF COMPLETE GAMES I

Aagaard-I..egky, RHdapest 1996 ................................................................................ 1 54


Alekhine-Rabinovich.A, Moscow 1918..................................................................... 81
Anand-Gelfand, Mo.rcoll' 2004..................................................................................... 4 3
Anand-Hiibner, Dortn11111d 1992 .............................................................................. 126
Anand-Ivanchuk, UtJttrrs 1993 ................................................................................ 1 16
Anand-Mishra, Itulia 1988 ........................................................................................ 1 52
Anand-Shirov, U11arr.r 2000 ........................................................................................ 66
Anand-Shirov, 1\tfoscoliJ'-200 1 ....................................................................................... 12
Anand-Sokolov.1, J)orlnllmd 1999 ............................................................................. 24
Areshchenko-Mista, Ctrppelle Ia Gra11de 200 3 .......................................................... 88
Aronian-Akopian, Ohrid 2001 .................................................................................... 86
Bonch-Osmolovsky - Baranov, Moscon' 1954 ..................................................... 1 73
Burkov-Raetsky, Con"Cspo11dmce 1985........................................................................ 4 7
Cabrera-Collas, Malaga 2003.................................................................................... 169
Dolrnatov-Mamedyarov, MoscoJI' 2002.................................................................. 1 36
Dolrnatov-Raetsky, Podolsk VoroneifJ 1992............................................................. 92
Elizarov-Raetsky, Relorrcbmsk 1989........................................................................ 128
Firman-Bick, .flmllo/1 M01mtaifl 2003 ........................................................................ 59
Grischuk-Adarns, Halkidikj 2002............................................................................ 1 1
Grischuk-Gelfand, Wyk atlfl 'Lee 2002 ..................................................................... 52
Grischuk-Pavasovic, lstaii!JIII 2003 ........................................................................... 23
Gurevich.V-Meijers, GerffiCI'!)' 1999 ...... . ................ . . . . .�
... .. . ... .......... . . .
... .. .............. . . 62
.

Hamdouchi-l.e Roux, Be!Jort 2003 .......................................................................... 60


Herrera- Perez.Ro, f/aradero 2000............................................................................ 72
Howell.J-Makarychev, Fnmze 1989....................................................................... 120
lvanchuk-Shirov, DorlnJIIfld 1998 .............................................................................. 4 1

1 89
The Pet roff Defence

Kasparov-Anand, I i11ares 2000 .................................................................................. 21


Kasparov-Ivanchuk, Debrece11 1992 ........................................................................ 122
Kasparov-Karpov, Moscm11 1981 ................................................................................. 69
Kasparov-Shirov, Wyk aa11 Zee 2001 ........................................................................ 50
Khalifman-Atalik, Ha/kidiki 2002............................................................................. 83
Kholmov-Belousov, Gorf?y 1974 ............................................................................. 101
Kotronias-Motylev, Moscow 2004 ............................................................................. 18
Kovalev-Rozentalis, G/o._�OJI' 200 / .. . . . . . . . . . . . 57
. . . . . .. . . . ........ ...... .... ...... ...... ... ... ... ... .... .... .....

Kupreichik-Yusupov, Mi11sk 1987 . . . . . .. . .. . ...... ..... .


..... . . . . ... 74
.... ....... ......... ............. .. .. .. .

Lasker.Em-Marshall, 51 Petersbu'l!, 1912.................................................................. 85


Lastin-Najer, Elisltt 2000 124
................. . . . . . . . . . . ............................. . . . . . . ............................

Leko-Grischuk, Wijk aa11 Zee 2002 .......................................................................... 14


Leko-Kramnik, World Ch. (GatHe 1), Brissago 2004 .................................................. 8
Leko-Yusupov, JJortflll/1/d 1998.................................................................................. 39
Lev-Alterman, H.atnai-Gan 1992............................................................................... 181
Macieja-Rozentalis, Re11o 1999 ............................................................................... 1 12
Morphy-Barnes, Londo11 1858 .................................................................................. 183
Naiditsch-Timman, Dortt111111d 2002 ........................................................................ 90
Nevednichy-Collas, Montpellier 2003...................................................................... 161
Novozhilov-Raetsky, Comsponde��ce 1983 ............................................................... 98
Nyysti-Rychagov.M, Gothenburg 2003 ..................................................................... 26
Pavlovic-Raetsky, Bie/ 1999 ....................................................................................... 63
Psakhis-Cooper, Port Erin 2003 ................................................................................ 38
Rublevsky-Vallejo, Ohrid 2001 ................................................................................ 107
Sax-Yusupov, Tbessaloniki Ofympiad 1984 ................................................................. 37
Shirov-Gelfand, Aslalla 200 1 ...................................................................................... 28
Shirov-Gelfand, Leon 2001 ........................................................................................ 1 7
Shirov-Kramnik, Belgrade 1997................................................................................. 1 14
Shirov-Piket, Wijk aa11 Zee 2001 ................................................................................. 54
Shirov-Yusupov, Fiuropea11 Tetllll Ch., Bat11mi 1999 ................................................. 35
Shirov-Yusupov, Ter Ape/ 1997 ............................................................................... 147
Smagin-Makarychev, Moscmv 1987 .......................................................................... 94
Srnirin-Alterman, Haija 1995................................................................................... 164
Sorokin-Raetsky, K.ra.r11odar 1984 ............................................................................ 13 3
Stefansson-Yusupov, Eupe11 1994 .......................................................................... 172
Svetushkin-Miles, Alushta 1999 .............................................................................. 185
Svidler-Akopian, IPorld Tet1111 Ch., l_J1ceme 1997 ................................................... 179
Svidler-Ponomariov, Mosco/1' 2001 .......................................................................... 105
Svidler-Yusupov, Bad HonJbHIJ!. 1998 ...................................................................... 109

190
Index o f Complete Games

Tal-Kholrnov,Abna Ata 1968 1 75


..........................................•.......................................

Timofeev-Raetsky, Correspondence 1982 ................................................................. 139


Timoscenko-Yusupov, F"rnn� 1979...................................................................... 150
Tiviakov-Forintos, Porto San Giorgio 1994 ............................................................. 158
Tiviakov-Miles, Unaros 1998 ................................................................................... 131
Topalov-Anand, Wijk aan Zee 2003 .......................................................................... 30
Tukmakov-Bronstein, Mosmw 1971 ........................................................................ 76
Van Der Wiel-MeUado, E/goibar 1998 ................................................................... 14 3
Vitolinsh-Anikaev, Riga 1982 96
....................................................................................

19 1