Attacking the Spanish

By

Sabino Brunello

Quality Chess
www.qualitychess.co.uk
 First English edition 2009 by Quality Chess UK LLP

Copyright 2009 Sabino Brunello

Attacking the Spanish

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, elec
trostatic, magnetic tape, photocopying, recording or otherwise, without prior
permission of the publisher.

ISBN 978- 1 -9065 52- 1 -7 5

All sales or enquiries should be directed to

Quality Chess UK LLP, Suite 45, Central Chambers,
93 Hope St, Glasgow G2 6LD, United Kingdom
e-mail: info@qualitychess.co.uk
website: www. qualitychess.co.uk

Distributed in US and Canada by SCB Distributors, Gardena, California, US

www. scbdistributors.com

Distributed in Rest of the World by Quality Chess UK LLP through

Sunrise Handicrafts, Smyczkowa 4/98, 20-844 Lublin, Poland

Editing: Andrew Greet

Typeset: Andrew Greet and Jacob Aagaard
Proofreading: Colin McNab and John Shaw
Cover design: Peter Woods and Barry Adamson
Cover photos: Jacob Aagaard
Printed in Estonia by Tallinna Raamatutriikikoja LLC
 Contents
Key to symbols used 4
Introduction 5

The Schliemann
1 Schliemann: 4.d3 7
2 Schliemann: 4.ltJc3 37
3 Schliemann: Minor Lines 69

The Gajewski
4 Gajewski 1 1 .d3 and 1 1 .exd5 93
5 Gajewski 1 1 .d4 1 19

The Marshall
6 Anti-Marshall 1 53
7 Marshall: 1 2 .d4 1 83
8 Marshall: 1 2. d3 223
9 Marshall: Modern Ideas 257
 Key to symbols used
;;t; White is slightly better
=+= Black is slightly better
White is better
=+ Black is better
+- White has a decisive advantage
-+ Black has a decisive advantage
= equality
=

00 with compensation
with counterplay
00
unclear

? a weak move
n a blunder
a good move
!! an excellent move
!? a move worth considering
, a move of doubtful value
# mate
(n) nth match game
with an attack
t with an initiative
N new move
Introduction

Hello and thank you for reading this book.

I have been playing the black side of the Spanish, or Ruy Lopez, for a large part of my chess career.
When the opportunity arose for me to write a book on the subject, I jumped at the chance to
study these fascinating positions in greater depth and share my ideas.

For a long time my primary weapon against the Spanish has been the Marshall gambit, which is
certainly the most well known of the 'aggressive' responses to White's opening. At the same time,
I was quite happy when the publisher proposed that I cover not one, but three different systems
for Black. This will enable the reader to develop a more varied repertoire over time. Even if you
decide not to play a particular system in your own games, I hope that you be able to obtain some
useful ideas that can somehow enrich your chess knowledge.

By coincidence, all three of the subject variations are gambits. Each has their own unique
characteristics, advantages and drawbacks, but all are fully playable and have been tested at high
levels of competition. Although the book is written primarily from Black's point of view, I am
confident that the contents will also be of considerable value to those who play the white side.

The first three chapters are devoted to the Schliemann gambit. To be honest, I never paid too
much attention to the black side of this line until Radjabov started to play it consistently against
world-class opposition. I find this variation to be extremely rich with ideas, and it is still far from
clear whether White can achieve a meaningful advantage.

In chapters 4 and 5 we turn our attention to the very trendy Gajewski variation, which came into
fashion a couple of years ago. I studied and experimented with the Gajewski a few times since it
burst onto the theoretical scene. This variation has lost a certain amount of its surprise value since
catching the eye of theoreticians, but it still provides a lot of scope for new ideas. So far White
has not been able to refute the cheeky 1 O . . . d5 , and I expect this line to remain topical for the
foreseeable future.

The final four chapters will address the infamous Marshall gambit, including the various 'anti
Marshall' systems with which White can avoid the most critical variations. The Marshall has
enjoyed a healthy reputation for many years now, and the most recent top-Hight games have done
little to change this assessment. I have been playing the Marshall myself for many years, and have
no hesitation in recommending it as an exciting yet reliable weapon.
6 Attackin g the Spanish

Instead of offering a narrow repertoire comprising a single recommendation against each of

the opponent's options, I decided it would be more useful to cover a few different approaches,
especially in the most critical variations. The reasons for this are many. In the case of the Gajewski,
the theory is still developing quite rapidly, so I felt it especially important to provide an up-to
date picture including the latest games from different variations, along with a lot of my own
analysis. In several of the Marshall lines, there can be more than one equally valid approach to the
position. Thus I hope to provide the reader with the best opportunity to find an approach that
suits his own style, as well as to achieve a better overall understanding of the positions.

This is my first effort as an author, and I have enjoyed working on the project. I would like to
thank Andrew Greet and Jacob Aagaard for their help and support. Finally, I would like to thank
you, the reader, for purchasing this book, and wish you great success in your games.

Sabino Brunello

Rogno, Italy
June 2009
Chapter 1

Schliemann: 4.d3
7
6
5
4
3

a b c d e f g h

l .e4 e5 2.tLlB tLlc6 3 ..ib5 f5 4.d3

Strategic Themes page 8

Theoretical highlights page 9
Theory page 1 0
A) 6.VMd3 page 1 0
B) 6.c3 page 1 1
C) 6.i.g5 page 1 2
D) 6.0-0! i.c5 page 1 5
0 1 ) 7.c3 page 1 5
02) 7.i.xc6 page 1 8
03) 7.VMd3 page 24
8 Attacking the Spanish

1 .e4 eS 2.ti)a ti)c6 3 ..ibS f5 4.d3 S.dxe4 ti)f6

This natural move requires no explanation.
The present position provides our starting
8
position for the present chapter. We will now
7 explore a few of the typical ideas available to
6 both sides, as well as taking a brief snapshot
of some of the most important theoretical
5 variations.
4
Strategic Themes
3
2 White's activity in the centre and queenside
White often tries to install a knight on d5 ,
1
perhaps in conjunction with a bishop on g5 to
a b c d e f g h pin the knight on f6. He may also try to prove
This has always been quite a common that Black's dark-squared bishop developed on
response to the Schliemann, although for a c5 too early by advancing his pawns on the
long time it was considered harmless. I suspect queenside.
that its popularity has partially been due to it
being the default choice of a player who has 8
been surprised and/or intimidated by Black's
7
bold response to the Ruy Lopez.
One of the main selling points of 4.d3 is 6
that White gets to maintain a pawn on the
5
e4-square. Thus he will be in no danger of
being overrun in the centre, unlike the 4. ttk3 4
variation, in which this has been known to 3
happen if White misplays the position.
Since the turn of the millennium, the 2
reputation of 4.d3 has been enhanced by the 1
discovery of some new ideas, which have led
a b c d e f g h
to a re-evaluation of certain critical variations.
Over the past few years it has been tested in In order to save his bishop, Black will have
several games at the highest level. to accept a structural weakening by moving his
a-pawn.
One of the ideas behind White's 4th is that he
is now ready to capture the pawn on fS without Black's kingside attack
having to worry about . . . e4. For this reason I The open f-file makes kingside play an
recommend that Black continues: attractive proposition for Black, especially
in conjunction with the dark-squared
4... fxe4 bishop on the a7-g 1 diagonal. He may also
Black can certainly consider 4 . . . ttJf6 5 .exfS look to transfer his king's knight to the
but overall I consider the text to be the soundest f4-square.
continuation.
 Chapter 1 - Schliemann: 4.d3 9

Sharp tactics
8
7
8
6
7
5
6
4
5
3
4
2
3
1
2
a b c d e f g h
1
From here, he would look to post his queen
a b c d e f g h
on g5 or h4, followed by doubling rooks on
the f-file. White has just played 9.b4!?, instigating
heavy complications. Both players must
Theoretical Highlights demonstrate strong calculating abilities as well
as a good deal of theoretical knowledge. I found

Sacrificing the e5-pawn an important improvement for Black, but was

White may be allowed to exchange bishop unable to claim it as my own novelty. It seems
for knight on c6 in order to capture the pawn that the Indian GM Ganguly independently
on e5. found the same idea, and used it last year to
secure a comfortable draw against Hossain. Full
details can be found in line D3 1 1 ) on page 24.
8
7 The main line
6
8
5
7
4
6
3
5
2
4
1
3
a b c d e f g h
2
We will see that such a sacrifice is par for the
course in the Schliemann. Black has generally 1
scored quite well from here, although he should
a b c d e f g h
definitely come to the board well prepared.
This position is analysed in line D2) on page This has been the most popular variation
1 8. at elite Grandmaster level. White has just
10 Attackin g the Spanish
initiated a forced sequence of exchanges with
10.c3-d5. Play continues:
10 ... xd5 1 1.exd5 d4 12 ..ixd7t Wxd7
At this point White can choose between
capturing a pawn with 13.xe5 and
exchanging knights with 13.xd4 Axd4.
The former leads to great complications in
which Black seems to be holding his own.
This can be seen in the notes to Mamedyarov
- Radjabov in line D3 1 2 1 ) , page 27. The
latter variation is currently considered critical,
and is covered in D3 1 22) on page 3 1 . White
eschews the chance of immediate material gain
and aims for a slight but stable advantage in
a simplified position. The black position is
solid enough, although he still has some minor
long-term problems.
Theory
1.e4 e5 2.f3 c6 3 ..tb5 f5 4.d3 fxe4
5.dxe4 f6
8
7
6
5
4
3
2
1 9.ltJxd4 exd4-+) 9 . . . c6 1 O.d3 b5 l 1 .i.b3 a5
a b c d e f g h
From our starting position we will consider
four main moves. A) 6.Wd3 can transpose
to the main lines, but the independent idea
of 7.c4?! is probably poor. B) 6.c3 looks
natural, but is in fact harmless. C) 6 ..tg5! is
not the most obvious choice, but was recently
tested by Vassily Ivanchuk and thus deserves
to be taken seriously. Last but not least,
the most popular and challenging move is
D) 6.0-0!.
Occasionally White tries to be a bit too clever
by accelerating the thematic queen transfer, as
seen in variation D3) , with:
A) 6.Wd3
6.e2 can also be played with the same
intention.
6 ...i.c5 7.Wc4!
This only leads to problems for White, so he
should prefer 7.0-0 reaching line D3) .
7... We7! 8.0-0
Worse is 8 .lt: k3? ltJd4! (This move is even
stronger than 8 . . . a6 9.,ixc6 bxc6 1 O.i.g5 b8
1 1 .i.xf6 gxf6=t)
a b c d e f g h
Suddenly the bishop on b5 is in trouble. Play
may continue 9.i.a4 (White loses a pawn after
and the threat of . . . a5-a4 means that White
will be forced into the horrible 1 2.a3.
8 ... a6!N
Also playable is 8 . . . ltJd4 9.ltJxd4 ,ixd4 when
Black had no problems in Fox - Freiberger,
Werther 2007.
 Chapter 1 - Schliemann: 4.d3 11

7.Y!Yd3 .bc3t 8.Y!Yxc3!

8
This is the principled continuation, but
7 ultimately it appears to be flawed. Relatively
6 best would have been B .bxc3 0-0 9.0-0
d6, although Black is at least equal here as
5 well.
4
8 xe4 9.Y!Ye3
.

3
No better is 9.'iNd3 d5 1 O.ttJxe5 'iNd6+.
2
9 d5 10.xe5 O-O!
1
.

a b c d e f g h
8
White has no choice but to exchange with:
7
9.,hc6 6
At this point Black can choose between
5
9 dxc6 10.c3 ,ig4i with active piece play,
.

or 9 ... bxc6 10 . .ig5 a5i with a strong centre. 4

In both cases the two bishops provide him 3
with excellent chances.
2
B) 6.c3 1
This is not the most accurate continuation
for reasons that will soon become a b c d e f g h
apparent. Black refuses to waste time defending c6,
6 ,ib4!
.
and instead launches a potent counterattack.

8 1 1 .xc6 bxc6 12 ..ld3

White decides to jettison a pawn and hopes
7 to salvage the opposite bishop position a pawn
6 down. It would certainly take a brave soul to
grab the pawn with 1 2.hc6 i.a6!. Black has
5
a dangerous initiative, as demonstrated by the
4 following variations:
3
a) 1 3.haB E!:xf2 14.i.b7 hb7 1 5.'iNxa7 'iNf8!
2 1 6.i.e3 ( 1 6.'iNxb7 E!:d2! wins after 1 7.'iNa7 d4,
1 or 1 7.'iNaB 'iNxaB 1 B.hd2 'iNa4!) 1 6 . . . 'iNb4t
1 7.@d 1 ( l 7.c3 'iNxb2 1 B .hf2 'iNxal t 1 9.@e2
a b c d e f g h 'iNb2t 20.@f3 d4!-+) 1 7 . . . d4 l B. 'iNxd4 E!:d2t
Black immediately capitalizes on his 1 9.'iNxd2 ttJxd2 20.hd2 'iNxb2 White is
opponent's small mistake. He now threatens losing, as his pieces will not be able to get
to cripple the enemy pawn structure. coordinated.
12 Attacking the Spanish

b) 1 3.d4! is probably the best chance, Radjabov at Linares earlier this year. It looks
although the second player remains on top after like White's main idea is to establish a knight
1 3 . . . liJxf2 1 4 . .ie3 liJxh l 1 5 .0-0-0 gb8 when on d5 as soon as possible. We have already
White does not have enough compensation for seen that the immediate 6.liJc3 is harmless, so
the exchange. instead he waits for Black to commit his dark
squared bishop to a different square before
12 ... tLlxf2 13.0-0 tLlxd3 14Jlxf8t Wixf8 developing the knight.
15.0:d3 .ia6
The following analysis will be based upon the
game Ivanchuk - Radjabov, Linares 2009.
8
7 6...i.c5
6 This was the Azeri GM's choice, but it may
not necessarily be best.
5
4 6 . . . d6 7.liJc3 .ie7 (or the immediate 6 . . . .ie7)
is playable but passive.
3
2 The move I like the most is 6 . . . .ib4t! 7.c3
.ic5
1
a b c d e f g h
Black enjoys a lead in development in
addition to the extra pawn, and in the game
Rausis - Nata, Montpellier 2008, White was
unable to recover. Play continued:

16.i.d2
1 6.e6t looks like a better chance to
save the game, e.g. 1 6 . . . r;!{h8 1 7.xc6 hd3
1 8 . .if4!, or 1 6 . . . f7 1 7.xf7t r;!{xf7 1 8.d4
ge8 1 9 . .id2 ge2 20.gfl t! r;!{g6 2 1 .gf2 when
White should hold.
16 ... Wif5 17JeU! .hd3 18.Wie6t Wixe6
19Jxe6 f8 20.g3 gflt 21 .g2 gbl
22 ..lc3 c5 23.ge1 d4
0-1
C) 6..lg5!?
Under normal circumstances I would not
have viewed this move as one that warranted
special consideration. However, that all
changed when Ivanchuk used it against
a b c d e f g h
This cuts across White's plan nicely by
interfering with the development of the
knight on c3.
8 .d3
After 8 .,ixf6 xf6 9.d5 .ld6! Black is
temporarily disorganized, but will soon force
the enemy pieces back.
In case of 8.0-0 Black can consider 8 . . . h6!?
(Also fine is 8 . . . d6 9.b4 .ib6 1 O . .ic4 e7
I l .liJbd2 a5 1 2.a3 .le6 1 3.b3 a4 1 4.a2
liJd8 White can no longer do anything to
prevent Black from castling.) 9.hf6 xf6
 Chapter 1 - Schliemann: 4.d3 13

1 O.dS .td6! Black will soon organize his 8 ..lxRi

position, for example: l 1 .ttJbd2 a6 1 2 . .td3
.te7 with decent chances.
8 . . . d6 9.a4 a6 1 O . .tc4 e7
1 0 . . . h6! looks more accurate to me, e.g.
1 1 ..txf6 xf6, or 1 1 ..th4 gS 1 2 . .tg3 ttJhS,
with an excellent position in both cases.
l 1 .ttJbd2
l 1 .b4 .tb6 1 2.ttJbd2 as is fine for Black.
1 1 . . .ttJd8 1 2.0-0 ttJe6
8.ttJdS?! allows the typical trick: 8 ... .txf2t
9.<.!ixf2 ttJxe4t 1 O.<.!igl ttJxgS l 1 .ttJxgS xgS .
White is two pawns behind, although he may
be able to salvage half a point with 1 2.h4! (But
not 1 2.ttJxc7t?! <.!id7 1 3.ttJxa8?? e3t 1 4.<.!if1
f8t-+) 1 2 . . . h6 ( 1 2 . . . d8?! 1 3 .hSt g6
1 4.h6 is risky for Black) 1 3 .ttJxc7t <.!id8
1 4.ttJxa8 e3t I S .<.!ih2 f4t 1 6.<.!igl e3t
with a perpetual.

8 8 ...Bxf6 9.d5 Bd8 10.M!?

7 White takes the opportunity to harass the
6
enemy bishop.
5
10 ....ib6 1 1.0-0
4 There was an important alternative in
3 1 1 .a4. If nothing else, this would certainly be
2 consistent with White's last. Black is practically
1
forced to respond with 1 1 . . . a6:
a b c d e f g h

We have been following the game Stepovaia

Dianchenko - Kobalia, Krasnodar 1 997. I see
no reason to regard Black's chances as worse in
any way. The game was eventually drawn.

To summarise, it seems that 6 . . . .tb4t! should

give Black a perfectly acceptable game.
However, we will still take the time to examine
Radjabov's expert handling of the position.
a b c d e f g h

Now White must decide whether it is worth

exchanging his well-placed knight in order to
8 damage the enemy pawn structure.
7
a) 1 2.ttJxb6 cxb6 1 3 . .tc4 .tg4
6 In Garbett - Klinger, Novi Sad 1 990, Black
5 played 1 3 . . . ttJxb4 with complications. The
text is safer.
4 1 4.h3 .thS I S .c3
3 The impulsive I S .g4?! .tg6 certainly does
not help White.
2 I S . . . f6 1 6 ..te2 0-0
1
a b c d e f g h
14 Attackin g the Spanish

The posltlon is balanced, and can be

8
compared with that reached in the main game.
Black's pieces are active enough to compensate 7
his slight queenside weaknesses.
6
b) 1 2.J.c4 J.a7 5
Another move order is 1 2 . . . J.g4 1 3 .h3 4
( l 3.c3 :gfB is similar) 1 3 . . . i.hS 1 4.c3 :gfB
I S .J.e2 J.a7 1 6.0-0 fije7. 3
1 3.0-0 2
After 1 3 .h3 it looks interesting for Black to
try 1 3 . . . J.e6!?, intending to meet 1 4.fijxc7t 1
V1ixc7 I S.he6 with I S . . . fijd4 1 6.fijxd4 a b c d e f g h
V1ic3t 1 7.@f1 hd4.
Also possible is 1 1 . . . 0-0, which is likely to
1 3 . . .J.g4 1 4.c3
transpose after 1 2.a4 J.g4.
8
7
12.a4 0-0 13 ..le2 .txa 14..bf3 a6 15.tLlxb6
cxb6
6
Black's minor weaknesses on d6 and b6 can
5 easily be defended, and his knight is no worse
4 than the enemy bishop.
3
2 16.c3 <it>h8 17 .lg4 f6 18.Wd3 tLle7 19.93

1
a b c d e f g h
White can insert the moves 1 4.h3 J.hS
before playing I S .c3, although this does
not alter the fundamental character of the
position.
1 4 . . . :gfB!N
In Bannik - Romanovsky, Kharkov 1 9S6
Black faltered with 1 4 ... VIid7?!. This could
have let to an unpleasant position after
I S .h3! (in the game the weak I S .bS? spoiled
White's advantage) I S . . . J.hS 1 6.VIid3 It is
hard to suggest a good move for Black here.
I S .J.e2 fije7
White's active bishop has been driven back,
and his knight is about to be exchanged. The
black position is extremely solid, and the king
position does not appear to harm his chances.
1 1 . .ig4
. players.
Wc7
Black has developed harmoniously, and
the chances are approximately balanced. In
the present example the players continue
for another ten moves, before agreeing to a
draw.
20.a5 gafB 21 .We3 bxa5 22.gxa5 Wc4 23.h4
Woo 24.Ae2 c8 25.ga3 cf8 26.<it>g2 d5
27.a Wc7 28.W M6 29.gn d4 30.cxd4
lh-V:z
The position is far from dead, but at the same
time neither side can claim an advantage. It
seems that a sharing of the point was agreeable
to both players.
Overall 6.J.gS can be regarded as a typically
creative idea from Ivanchuk, but ultimately
not as a move that should worry Schliemann
 Chapter 1 - Schliemann : 4.d3 15

D) 6.0-01 1 O.c3 should be met by 10 ... lDe7 l 1 .lDxf6

White exploits the fact that the greedy gxf6 1 2.i.h6 g8
6 . . . lDxe4? allows 7.e2! (this is even stronger
than 7.e I ) 7 . . . 1Df6 8 .,bc6 dxc6 9.lDxe5 e7
(9 . . . i.e7? 1 0.d l ) 1 O.e l when Black faces
severe difficulties.

6 .ic5
.

8
7
6 a b c d e f g h
5 Black has at least equal chances thanks to
4 his central majority and open g-file. Galicki
- Goessling, Dortmund 2004, continued
3 1 3.i.c4 g4 with a good position. It would also
2 have been interesting to consider 1 3 . . . d5!?N
1 4.exd5 1Df5 with promising compensation.
1
a b c d e f g h lO .ig4
..

This is Black's most active and principled
continuation, although the more conservative
6 . . . d6 is also playable.
White must now choose between simple
development with Dl) 7.tilc3, grabbing a
pawn with D2) 7.hc6, and the more subtle
queen manoeuvre D3) 7.Bd3 (or 7.e2)
intending c4.
Dl) 7.tilc3 d6
White has tried several different moves here,
but the only ones that deserve our specific
attention are D 1 1) 8 .ig5 and D12) 8 .ie3 .
1 O . . . i.e6 seems riskier in view of 1 1 .1Dxf6
gxf6 1 2.,be6 fxg5 1 3.c3. Black has some
weaknesses on the light squares.
1 1 .tilxf6
1 1 .c3 is well met by 1 1 . . .,bf3!N 1 2.xf3
(or 1 2.gxf3 d7) 1 2 . . . lDxd5! 1 3.,bd8 xf3
1 4.gxf3 lDb6 1 5 .i.b5 lDxd8 1 6.b4 lDe6
1 7.bxc5 lDxc5 .
8
7
6

5
D1 1) 8 .ig5 0-0 9. til d5

4
9.i.c4t @h8 1 O.lDd5 transposes.
3

9 cj;lh8 lO .ic4
.
2

1 0.lDh4 i.e6 1 1 .i.c4 gives Black no problems 1

after 1 1 . . .,bd5 ( 1 1 . . .lDe7 is also fine) 1 2.,bd5 a b c d e f g h
lDe7, Markovic - Matulovic, Cacak (rapid)
With a superb position for the exchange.
1 998.
16 Attacking the Spanish

1 1 gxf6 12.J.h6
. Black has an excellent position in return for
a very small material investment. He enjoys a
preponderance of pawns in the centre, as well
8
as a strong pair of bishops and an open g-file.
7 Meanwhile it is not at all easy for White to
6 find a useful role for his rooks. Finally, his split
queenside pawns are likely to become weak in
5 the endgame.
4
012) S.,ie3 ,ib6
3
8 . . . he3 9.fxe3 0-0 is also playable, but it is
2 safer for Black to avoid opening the f-file. An
exchange on b6 would be more desirable.
1
a b c d e f g h
8
12 ... lLla5!N
7
Only this move solves Black's problems.
6
1 2 . . J:!:e8 ? is refuted by: 1 3.ttJg5! 1l;Vd7 1 4.1l;Vxg4!
5
1l;Vxg4 1 5 .ttJf7t 'ttg8 1 6.ttJxd6t, Dgebuadze
Held, Schwaebisch Gmuend 1 999, 1 6 . . . 1l;Ve6 4
1 7.he6t xe6 1 8.ttJf5 3

1 2 . . . g8 1 3.hg8 1l;Vxg8 does not give Black 2

enough compensation, although he eventually 1
prevailed in Rigg - Steinbacher, Austria 2000.
a b c d e f g h
13.i.xf8 9.lLld5
1 3 . .id5 ? c6 1 4.hfB cxd5 is better for 9.j,g5 0-0 1 O.ttJd5 reaches the main line.
Black. It would be too risky for Black to try 9 . . . h6?!
1 O.hf6 1l;Vxf6 l 1 .ttJd5 1l;Vf7 1 2.a4 0-0 1 3 .a5
13 lLlxc4 14 ..ih6 lLlxb2 15.Yfe2 lLla4
. j,c5 1 4.a6, as occurred in Benassi - Pitt,
e-mail 1 996.
8
In Amann - Karpatchev, Oberwart 1 997,
7
White played to win a pawn with 9.1l;Vd3 0-0
6 1 O.hb6 axb6 l 1 .hc6 (a more accurate
5
move order would have been 1 1 .1l;Vc4t 'tth 8
1 2.hc6 bxc6 1 3.1l;Vxc6) 1 1 . . .bxc6 1 2.1l;Vc4t
4 'tth 8 (thanks to White's inaccurate move order
3 Black could have played 1 2 . . . d5!) 1 3.1l;Vxc6
j,d7 ( 1 3 . . . a5!?) 1 4.1l;Vc4 ttJh5 Black obtained
2 good compensation.
1
a b c d e f g h
 Chapter 1 - Schliemann: 4.d3 17

9 0-0 IO ..ig5!?

This is a very interesting moment. White

steers the position into something almost
identical to variation D 1 1 ) , except for the
position of the black bishop which is now
on b6 instead of c5 . There are three reasons
why this 'tempo loss' may be of benefit to
him:

1 ) White now has the option of xb6. This

could be especially important should Black a b c d e f g h
wish to play . . . a6, which could now involve We have been following the game Martorelli
a structural concession after xb6 cxb6. (At - Fregonese, Verona 2004. At this point
the same time, the Ivanchuk - Radjabov game the most accurate continuation would have
from line C) has already shown us that this been:
need not be disastrous for Black.) 1 6 . . . h5!
The chances are balanced, as illustrated by
2) White may threaten to trap the bishop with the following sample variation.
a4-a5, perhaps hoping to provoke a structural 1 7.ffl ! h6 I B.h3 hxg5
concession as described above. (This plan Also leading to a draw is I B . . . g3!? 1 9.fe l
would not work against a bishop on c5, as hxg5 20.'Wxg5 'WdB 2 1 .'WxdB axdB 22.hxg4
Black could simply prepare a retreat with . . . a6 xg4 23.xc7 fl t 24.xfl e2t.
or . . . a5 .) 1 9.hxg4 g3 20.'Wxg5 e2t
The game will end in perpetual check.
3) In certain pOSitIOnS Black may wish to
evict the knight from d5 with . . . c6. With the IO ... h8
bishop on b6 this may be less desirable due to
the possibility of xb6 as well as the potential
weakness of the d6-pawn.

The other principled continuation is:

1 O.g5 'it>hB 1 1 .i.c4 i.g4 1 2.'Wd2 'Wd7
I prefer this to 1 2 . . . xd5 1 3.'Wxd5 i.h5
( 1 3 . . . ixe3?? 1 4.'WgBt xgB 1 5 .f7#)
1 4.e6 i.f7 1 5 .xdB ixd5 1 6.exd5 xdB
when White should be slightly better.
1 3.,txb6
1 3.f3 i.h5 led to a balanced position in
Busemann - Neumark, Germany 1 9BB.
It is a similar story after 1 3.xb6 axb6 1 4.f3 a b c d e f g h
i.h5 .
Black proceeds analogously to line D 1 1 ) .
1 3 . . . axb6 1 4.f4
White must maintain his momentum.
I I .a4!?
1 4 . . . exf4 1 5 .xf4 e5 1 6.i.b3
18 Attackin g the Spanish

White plays to harass the bishop on b6. This 1 3 .hB lDe7 1 4.lDxf6 gxf6 I S .J.h6 g8
is one of the ways in which he can attempt to 1 6.aS J.cS 1 7.c3 a6 1 8.J.hS with advantage to
justify the loss of tempo associated with the White.
J.e3-gS manoeuvre.
However, 1 1 . . . J.cS!? is definitely worth
Another approach is: considering. For example, 1 2.c3 can be
l 1 .hf6 gxf6 1 2.lDh4 lDd4 met by 1 2 . . . aS when I see no problems for
1 2 . . Jg8 is also okay. An example is 1 3.c3 Black.
J.g4 1 4.d2 f8 I S .J.c4 gS 1 6.lDxb6 axb6
1 7.f4 hS 1 8.g3 h6 1 9.@h l exf4 20.xf4 12.a5
lDeS with equality, Sznapik - Inkiov, Stara The game Stjazhkina - Semenova, St
Zagora 1 990. Petersburg 2000, resulted in a draw after
The untested 12 ... lDe7!?, challenging 1 2.lDxf6 gxf6 1 3.J.h6 g8 1 4.aS J.cS I S .J.c4
the knight on dS, also deserves serious g6 ( I S . . . dS!? is interesting, e.g. 1 6.exdS lDfS,
consideration. or 1 6.hdS lDxdS 1 7.xdS xdS 1 8.exdS
1 3.J.d3 J.g4, with reasonable compensation in both
1 3.J.c4 c6 1 4.lDe3 fS is comfortable for cases) 1 6.J.e3 he3 1 7.fxe3 f8 1 8.f2 h6
Black, whose central majority may start to 1 9.d2 J.g4 20.afl Y2-Y2 .
come into its own.
1 3 . . . c6 1 4.lDe3 g8 I S .c3 fS 1 6.hS?? 12 .. .tLlexd5 13.exd5 i.c5
After the superior 1 6.lDhxfS lDxfS 1 7.exfS
gS 1 8.B dS Black has a good position, 8
but White should not be in any real danger.
7
6
5
4
3
2
1
a b c d e f g h
a b c d e f g h
We have been following the game Logothetis
We have been following the game Tarumyan - Souleidis, Aghia Pelagia 2004. The chances
- Teemae, Russia 1 99 1 . Unfortunately Black are roughly balanced, and the game eventually
now missed 1 6 . . . xh4! when White can resulted in a draw.
resign.
D2) 7 ..bc6
1 1 .tle7

In this variation White cedes the bishop pair
1 1 . . . J.g4 does not enable Black to solve his in order to win a pawn.
problems after 1 2.J.e2. The game Khalifman
- Inkiov, Moscow 1 989, continued 1 2 . . . hf3 7 bxc6 8.tLlxe5 0-0
.
 Chapter 1 - Schliemann: 4.d3 19

After l 1 .liJd5, Tesinszky - Schenkerik,

8
Hungary 2000, Black should continue
7 1 1 . . .liJxd5 1 2.exd5 .la6.
6 l 1 .b4 xc6 1 2.bxc5 .la6 1 3J!e l xc5 with
decent compensation.
5 1 1 . . .liJxe4 1 2 . .le3
4 1 2.liJxe4 xe4 1 3.c3 e8 1 4.f3 was played
in Mekhitarian - Nezad, Barcelona 2008,
3 and here I suggest 1 4 . . . d5.
2 1 2 . . . liJxc3 1 3.bxc3 g6!?N
13 . . ..lb7 1 4.e l ? g6+ turned out well
1
for Black in Gheng - Held, Schwaebisch
a b c d e f g h Gmuend 1 999, but White could have
For the pawn Black enjoys a lead in improved with 1 4.g4;t.
development, a pair of bishops and active
8
prospects for his pieces. The combined
influence of these three factors should amount 7
to fair compensation. We examine two natural 6
moves: D21) 9.tLlc3 and the more popular 5
D22) 9 .ig5.

3
9.liJd3 has been played several times, but after
the simple 9 . . . liJxe4 1 0.liJxc5 ( l 0.liJd2 liJxd2 2

I I .hd2 .lb6 also gives Black an easy game) 1

1 O . . . liJxc5 1 1 ..le3 liJe6 Black is absolutely a b c d e f g h
fine, one attractive plan being . . . d6 and . . . c5 to
White has retained his extra pawn, but
restrict the enemy bishop.
he will have a hard time obtaining any
meaningful winning chances in view of
D21) 9.tLlc3 d6!
his damaged structure and the opponent's
This secondary offer is an important finesse.
powerful bishops.
Now White must make a difficult decision.
1 4.liJc6
1 4.d3 xd3 1 5 .cxd3 should not be
lO.tLld3
too dangerous. One option is 1 5 . . . .ld7
White can also consider the greedier option
when Black's bishop pair provides ongoing
of:
compensation. He can also consider swapping
1 O.liJxc6
down to an opposite bishop endgame with
This is playable, although Black's initiative is
1 5 . . . hd4 1 6.hd4 ( l 6.cxd4 .le6) 1 6 . . . .lfS
quite potent after:
when a draw is the most likely outcome.
1 O . . . e8
1 4 . . . e8 1 5 .hc5
White's most natural response is:
1 5 .d5t .le6 1 6.e4 .lfS! enables Black to
l 1 .liJd4
equalise after 1 7.liJe7t h8 1 8.h4 he3
l 1 .liJa5, Schumann - Kase, corr. 1 982,
1 9.fxe3 hc2, or 1 7.xe8 fxe8 1 8.hc5
should be met by 1 1 . . .liJg4! with good
dxc5 .
attacking chances.
20 Attacking the Spanish

1 5 . . . xc6 1 6 . .id4 .ifS 1 7.d2 h6 12 We8 13 .bf6 gxf6 14.lL'lg3 Wf7!?

This move was introduced by Magnus

Carlsen. With this modest approach Black
demonstrates great confidence in the
inherent soundness of his position. He is
content to develop his pieces and improve his
position patiently, rather than try to 'prove'
compensation by regaining the pawn or
attempting a direct attack.

Curiously enough, Carlsen had already gained

some experience of this position from the
a b c d e f g h
opposite side of the board. Carlsen - Radjabov,
We have reached a position in which the Monte Carlo 2007, continued: 1 4 . . . i.a6
weakness of the c2-pawn makes it hard for 1 5 .i>h l hd3 1 6.cxd3 xf2 1 7.xf2 hf2
White to claim any real advantage. 1 8.fS

10 .t.d4 1 1.lL'le2 .t.b6

8
7
6
5
4
3 a b c d e f g h

2 At this point the most straightforward

1
a b c d e f
At this point White must decide how best to
deploy his pieces.
12 .t.g5
After 1 2.g3 I propose 1 2 . . . .ia6!?N. Black's
handling of the position resembles that seen
in the Carlsen - Radjabov game, referred to in
the note to Black's 1 4th. Possible continuations
include 1 3.e2 e8 1 4 . .id2 g6 1 5 .i>h l
ae8, and 1 3.i.g5 h6 1 4 . .if4 ( 1 4 . .id2 d7)
1 4 . . . d7 1 5 .d2 ae8 1 6.ae l c5, with
promising compensation in both cases.
response would have been 1 8 . . . g6 1 9.h6t
( 1 9.b3t d5=) 1 9 . . . i>g7 20.d2 i.d4 2 1 .f1
g h e6 22.f7t i>h8 when White has no trace of
an advantage.
Radjabov instead chose 1 8 . . . i.b6 1 9.g4
( 1 9.b3t i>h8 20.c3 d7 2 1 .f1 f8oo)
1 9 . . . f8 20.f1 g6 when Black was okay and
went on to hold the draw.
The fact that the young Norwegian prodigy
subsequently took up the opposite side of
this variation can certainly be seen as an
encouraging sign for Black! We now follow the
game Anand - Carlsen, Bilbao 2008.
 Chapter 1 - Schliemann: 4.d3 21

15.b3 i.e6 16.YlYdl U'8 17J!ael gh6 is slightly behind on material, but his position
is very solid and Anand went on to hold the
draw.
8
7 D22) 9.i.g5
6 This is White's most common continuation,
although the position retains certain similarities
5 with the previous variation.
4
9 ... YlYe8!
3
2
8
1
7
a b c d e f g h
6
18.YlYc3 5
White must be careful - a casual move such as
1 8.c4?! would allow 1 8 . . . 'iBf6! with dangerous 4
attacking possibilities based on . . . 'iBh4. 3

18 .. JU6!? 19J:e2?! 2
Anand was obviously not satisfied with a 1
repetition of moves, but the text could have
a b c d e f g h
led to problems for him.
With this excellent multi-purpose move
The more critical 1 9.'iBxc6 should probably Black unpins the knight while targeting e4 and
be met by 1 9 . . . h5!? (but not 1 9 . . . i.xf2t? facilitating a transfer of the queen to g6 or h5 .
20.tDxf2 xf2 2 1 .tDfS! xf1 t 22.xfl ) with
the possible continuation 20.tDfS .id7 2 1 .'iBc3 10 .ixf6

i.xfS 22.exfS xfS Black has reasonable
compensation based on the active pieces and
pressure against f2.
19 .ig4! 20J:dl?
..
It would have been better to admit the
mistake with: 20Jee I xf2 (20 . . . h5!?) 2 1 .tDxf2
i.xf2t 22.i>h l i.xe I 23.'iBxe l 'iBe7=
20 .ie3!
. destabilise the position of the knight on d3.
Winning an exchange.
21.h3 ixdl 22.YlYxdl .i.c8
The remainder of the game is not so
important for our study of the opening. White
White is forced to 'sacrifice' his second bishop
in order not to forfeit his material advantage.
For instance, 1 O.tDd3?! tDxe4 1 1 .e l ?? tDxf2
1 2.xe8 tDxd l t 1 3.tDxc5 xe8 is winning for
Black.
10 .. Jxf6 1 1.tLld3 i.d4!?
This is an important finesse. Black provokes
the advance of the enemy c-pawn in order to
12.c3
Less challenging is 1 2 .tDd2 .ia6 1 3.b l
( 1 3.c4, Cigan - Srebrnic, Slovenia 1 995, should
have been met by 1 3 . . . c5) 1 3 . . . d6 when the
22
two bishops provide ongoing compensation.
The encounter Anand - Radjabov, Morelial
Linares 2008, proceeded with: 1 4.c4 cS I S .b4
Wl'f7!
a b c d e f g h
With this instructive move Black shows that
he has no fear of the doubled pawns that would
result from an exchange on cS .
The game continued 1 6.@h l ( 1 6.bxcS dxcS
would leave Black extremely active while the
knights lack stability) 1 6 . . Jf8 1 7.f4 hc4
1 8.lLlxc4 Wl'xc4 1 9.bxcS dxcS 20.eS b6
2 1 .c 1 Wl'dS 22.WI'f3 c6 Y2-Y2.
12 ....ib6 13.tLld2 d5!
a b c d e f
Compared with the previous variation,
D2 1 ) , the present position requires a more
confrontational approach.
Attacking the Spanish
Radjabov has tried the more patient 1 3 . . . d6,
but after 1 4.c4 Wl'g6 I S .@h l .ig4 1 6.f3 .ie6
1 7.f4 he was struggling for compensation
and went on to lose in Macieja - Radjabov,
Khanty-Mansiysk 2007.
Black is also not helped by 1 3 . . . .ia6 1 4.c4, as
the critical 1 4 . . . dS I S .WI'a4! dxe4? 1 6.WI'xa6
exd3 1 7.cS saw him lose a piece in Yasakov -
Roldan, e-mail 2000.
14.exd5
1 4.eS is unpromising for White. His kingside
pawns are easily blockaded while the enemy
bishops enjoy tremendous potential.
The game Hernandez - Gomez, Holguin
200 1 , continued 1 4 . . . f8 I S .WI'e2 .ifS with
good compensation, although the more flexible
I S . . . WI' g6! would have been even stronger
as White must now worry about bishop
excursions on g4 and h3, not to mention
. . . ae8 or a doubling of rooks on the f-file.
1 4.WI'e2 .ia6 was also comfortable for Black in
Grott - Leisebein, e-mail 2000.
14...i.h3!N
This new move could certainly come as
an unpleasant surprise to an unsuspecting
opponent! According to my analysis the
position is objectively equal, although from a
psychological perspective it is safe to say that
most of us would prefer to play Black.
The alternative 1 4 . . . cxdS I S .e l gives White
some chances for an advantage, although the
position remains quite complicated, Marez -
Raleus, e-mail 2002.
15.tLlel
g h This is probably White's best, as both the
knight and the g-pawn were vulnerable.
The tactical justification for Black's last can be
seen after I S .gxh3 Wl'g6t 1 6.@h l Wl'xd3.
 Chapter 1
White would be ill advised to venture 1 5 .dxc6?
g6 1 6. e 1 af8 when he will be slaughtered
on the kingside.
Another harmless alternative is 1 5 .c2
g6 1 6.el xc2 1 7.xc2 g6 1 8.el
cxd5, intending ... e8. Black has excellent
compensation and is probably a little better
due to his exceptional piece activity.
15 YlYg6 16.hl
1 6.c2 xc2 transposes to the final line of
the previous note.
16 Jaf8!
a b c d e f g h
Black brings his final piece into the attack.
17.8
1 7.gxh3 leads, after the forcing sequence
1 7 . . . xf2 1 8.xf2 xf2 1 9.g4 xg4 20.hxg4
xd2 2 1 .dxc6 xb2, to an ending in which
only Black can be better.
17 .if5 18.c4
1 8.dxc6 gives Black a choice between two
reasonable lines:
a) 1 8 . . . h6 1 9.c4 ( 1 9.f4 .te3) 1 9 . . . xh2t!
leads to a perpetual after 20.@xh2 h6t
2 1 .@g3 g6t 22.@h2 h6t.
- Schliemann: 4.d3 23
b) 1 8 ... h5!? 1 9.f4 (the threat was ... h6)
1 9 . . . xd l 20.xd l d8
a b c d e f g h
White's pieces remain tangled in spite of
his three (!) pawn advantage. A plausible
continuation would be: 2 1 .ef3 fd6
22.d4!? (otherwise White would have to
worry about . . . .tg4 or . . . .tc2) 22 . . . Ld4
23.c4 .tc5 24.xd6 Ld6. Black stands
better in spite of his nominal material
disadvantage. His bishops are tremendously
powerful and the c6-pawn is likely to fall in
the near future. Meanwhile White is a long
way from creating a passed pawn or achieving
any significant activity for his rooks.
18 YlYh5
Also interesting is 1 8 . . . d8!? 1 9.d6 h5
with compensation.
19.f4
Capturing a pawn with 1 9.dxc6 allows
Black a minimum of a draw with 1 9 . . . xh2t
20.@xh2 h6t 2 1 .@g3 g6t, as 22.@f4??
.td3t 23.@e5 f5 mate, would be inadvisable
to say the least.
19 i.g4 20.8 gxf4 21.xb6 axb6
Black retains sufficient compensation, e.g.
22.dxc6 YlYc5 when the c6-pawn will fall.
Meanwhile Black can compromise the enemy
king's protection by exchanging on f3 at any
moment of his choosing.
24 Attacking the Spanish

D3) 7.%Vd3 approach he prevents castling while also

threatening lDxe5, hence Black's next.
8
8 %Ve7

7 8 . . . i.d7 is generally considered inferior in

6 view of 9.lDc3 i.b6 1 O.lDg5 fie7 I l .lDd5
lDxd5 1 2.exd5 with a strong initiative.
5 At this point White can choose between the
4 extremely direct D3 1 1) 9.b4!? and the more
popular developing move D312) 9.ttlc3.
3
2 D31 1) 9.b4!? .leG!
Without this intermediate move Black
1
would suffer a fatal loss of material.
a b c d e f g h
This is widely accepted as the main line. lo.hOOt bxcG 1 1 .%VaG i.bG
White's main idea is to place his queen on Obviously Black must avoid 1 1 . . . hb4??
c4. From here it prevents Black from castling, 1 2.fib7, which occurred in Fercec - Zelie,
while also menacing his minor pieces on c5 Rabac 2003.
and c6. Before going any further we should
note that 7 . . . 0-0?? is a blunder in view of the 12.a4
fork on c4. This is the point of White's play - he hopes
to embarrass the enemy bishop on b6.
7.fie2 is an equally popular move order, which
usually results in a transposition after 7 . . . d6 12 0-0!
.

8.fic4. However, with the queen on e2 Black

may be able to obtain a slightly improved 8
version of the 7 . . . lDd4!? variation examined
7
later.
6
In a slight break from the usual format, I
5
will now provide coverage of two different
alternatives. D3 1) 7 dG is the traditional

4
main line. It is reliable enough, although 3
the latest theoretical verdict seems to be that
Black may still have some small problems to 2
solve. For this reason we will also investigate 1
the more experimental D32) 7 ttld4!?, an

a b c d e f g h
interesting sideline which has been employed
by Radjabov. This excellent move was introduced by
Ganguly in 2008. Previous games had
D31) 7 dG 8.%Vc4!
seen 1 2 . . . lDxe4 1 3.a5 hf2t 14Jhf2 lDxf2
As mentioned previously, this was the 1 5 .fixc6t @f7 1 6.@xf2 when White has some
idea behind White's 7th. With this concrete chances to be better.
 Chapter 1 - Schliemann : 4.d3 2S

13.a5 This should ultimately lead to an equal

This must be critical, and it should come as
no surprise that the alternatives promise White
nothing:
The timid 1 3.bd2 allows Black to choose
between 1 3 . . . i.d4, 1 3 . . . i.c8 and 1 3 . . . dS!?,
with a good position in each case.
1 3.i.gS is also not dangerous in view of
1 3 .. .'f7 1 4.bd2 i.c8 (This is the most
straightforward move although 1 4 . . . hS!? is
also interesting, e.g. I S .aS i.d4 1 6.xd4 exd4
1 7.'1&xc6 [ 1 7Ja3!?] 1 7 . . . h6 1 8.i.h4 ab8.)
I S .W'd3 as Black's two bishops and central
majority make his position easier to play.
position after:
I S . . . xf3 1 6.gxf3 xh2! 1 7.cxd4 W'f7 1 8.@xh2
W'hSt 1 9.@gl f8!?
Obviously Black can take a draw immediately
with 1 9 . . . W'g6t should he wish it.
20.dxeS dxeS 2 1 .W'e3
White loses after 2 1 .d l ? i.h3 22.W'e3 h6!.
2 1 . . .f6 22.d l i.h3 23.d8t @f7 24.W'b3t
@e7

13 ... i.d4 14.c3 tLlg4!

8
a b c d e f g h
7
6
White's situation looks dire, but he can in
fact force a perpetual with:
5 2S.e8t! @d7
4 2S . . . @xe8?? 26.W'g8t wins, while Black also
has nothing to gain from 2S . . . W'xe8 26.i.gS
3 W'g6 27.W'e3.
2 26.d8t @e7
1
An important alternative is:
a b c d e f g h I S .i.gS W'e8!
Without this move Black would simply be I S ... W'f7? 1 6.cxd4 i.c4 1 7.W'xc6 hfl
losing a piece for very little. Now he obtains 1 8.bd2 is good for White.
quite a potent attack.
8

15.cxd4 7
The following options also deserve our 6
attention. 5
Firstly, let us briefly note that White loses 4
immediately after I S .W'xc6?? xf3 1 6.gxf3
3
W'h4 1 7.W'xa8t @f7 etc.
2

I S .W'e2 1

a b c d e f g h
26 Attacking the Spanish

Now the queen threatens to join in the fun 2 1 .fc l exf4 22.xc6 g4t 23.@xh2 f3
from h5 or g6. 24.xe6t xe6 25 .gl
1 6.cxd4
Worse is: 1 6.h4?! xf3 1 7.cxd4 ( 1 7.gxf3? 8
is refuted by 1 7 . . . 11Jxh2! e.g. I B.@xh2 h5 7
1 9.@g3 f8 20.11Jd2 f6 with mate to 6
follow) 1 7 . . . f4+
5
1 6.e2 is well met by 1 6 . . . h5 . "'" . ... ... ./''''''''.

1 6 . . . xf3 4

Black has a dangerous attack although my 3

analysis indicates that White can hold the 2
balance if he defends precisely. The main 1
line runs:
a b c d e f g h
1 7.gxf3
1 7.h3 1lJxf2 I B.xf2 xf2 1 9.@xf2 g6-+ We are left with a position featuring an
1 7 . . . 11Jxh2 I B.llJd2 unusual material balance that is hard to
I B.a3 h5 1 9.xc6 f8 20.11Jd2 llJxfl assess.
2 1 .@xfl xg5 still leaves White under some
pressure. 15 J'x6!
..

I B . . . h5 1 9.f4
White can hardly benefit from 1 9.xc6 fB, 8
although he can probably still draw.
1 9 . . . f8!? 7
6
5
4
3
2
1
a b c d e f g h

a b c d e f g h
From here the most likely verdict seems to be
equality. Indeed, the game Hossain - Ganguly,
Of course Black can take a draw with Nagpur 200B, soon resulted in a perpetual
1 9 . . . g4t 20.@xh2 h3t, but he may as after:
well continue the attack while keeping the
perpetual in reserve. 16.grl3
20.dxe5 Had White been determined to maintain the
White is not helped by 20.fc l ? exf4. fight he might have considered 1 6.h3!? Play
20 . . . dxe5!? then continues: 1 6 . . . xf2! 1 7.hxg4 (White
Playing for a win. Once again 20 ... g4t is loses after 1 7.xf2? llJxf2 I B.@xf2? h4t
a draw. 1 9.@gl xe4) 1 7 . . . xfl t I B.xfl exd4.
 Chapter 1 - Schliemann: 4.d3 27

The resulting posItlon is rather unclear. lo.lild5

White's extra piece is approximately balanced
by his missing pawns, lack of development and
ragged structure.
16 YHh4 17.h3!
..
The immediate 1 7.fxg4?? would of course
see White get mated after 1 7 . . J&xg4t 1 8.'i!?h l
If White tries to prepare this with 1 O.i.g5 ?!
there follows 1 O . . . a6! 1 1 .hc6 (of course
there is no time for 1 1 .lLld5 ?? as 1 1 . . . axb5
hits the queen) 1 1 . . . bxc6 with an excellent
position.
lo lilxd5 1 1.exd5 lild4

f3t 1 9.'i!?gl i.h3. Black has no real choice, as the alternatives

would leave the knight severely misplaced.
17 YHxh3 18.fxg4 YHxg4t 19.h2 YHh3t

1 9 . . J!f8 20.El:a3 forces Black to take the 12.,hd7t

draw with 20 . . . El:f3 anyway. White can change the move order with
1 2.lLlxd4 i.xd4 1 3.hd7t xd7 leading to
20.gl YHg4t line D3 1 22) .
1/2-1/2

12 YHxd7
.

Based on the evidence of this game, it seems

that after 9.b4!? the theoretical ball is presently
8
in White's court.
7
D312) 9.lilc3 6
5
8
4
7
3
6
2
5
1
4
a b c d e f g h
3
White must now decide whether to grab
2 a pawn with D3 121) 13.lilxe5 or exchange
1 knights with D3 122) 13.lilxd4.
a b c d e f g h
D3 121) 13.lilxe5
This is White's most popular continuation, With this move White wins a pawn but loses
and probably his best try for a theoretical plus. some time.
The knight is headed for d5 at the earliest
opportunity. 13 YHf5 14.lild3 0-0-0
.

Black can regain his pawn with 1 4 . . . b5?!

9 ,ld7
1 5 .El:e l t 'i!?f7 1 6.c3 xd5 , but after 1 7.i.e3
Black unpins his knight in preparation for his position is quite unpleasant. The text is
the following sequence. much more in the spirit of the Schliemann.
2B Attacking the Spanish

1 7 .b 1 xd5+ leaves Black more actively

8
placed.
7
6 1 7.xa7 lLlxa1 1 B.aBt i>d7 1 9.xb7 lLlc2
does not give White any real compensation for
5 the rook, as only his queen is attacking.
4
1 7.i.f4!? xf4 1 B.xc2 xd5 is equal.
3
2 15 .. Jhe8 16.hl!
It is important to position the king away
1
from the checking range of the knight.
a b c d e f g h
15.a4 1 6.lLlxc5 ? dxc5+ 1 7.xc5 xd5 1 B.xa7?
1 5 .i>h 1 b5!? 1 6.c3 xd5 was equal in ( 1 B.c4 was mandatory, although even here
Melia - Shukurova, Kusadasi 2006. Black's 1 B . . . lLlxc2 regains the pawn while keeping a
active pieces compensate the slight weakening much more active position) 1 B . . . lLle2t 1 9.i>h 1
of his king's shelter. a5! trapped the queen i n Meshcheriakova -
Agrest, Stockholm 200B.
1 5 .lLlxc5 dxc5 1 6.xc5 should not be at all
dangerous for Black, as long as he makes the 16 ... g5!
right choice: The slow 1 6 . . . a6 can be met by 1 7.f3 .ta7
1 B.c3 lLle2 1 9.1Llf4;;l;. The text is designed to
In Kozhuharov - Inkiov, Fouesnant 2007, he prevent this possibility.
was successful with 1 6 . . Jhd5? 1 7.xa7 eB?!
1 B.i.e3?? lLle2t 1 9.i>h 1 a5! 0-1 , but White 8
could easily have improved with l B. aBt i>d7
7
1 9.xb7 with a winning position.
6
Therefore Black should prefer 1 6 . . . lLlxc2: 5
4
3
2
1
a b c d e f g h
We now follow the model game Mamedyarov
- Radjabov, Baku 200B.

a b c d e f g h
17..ie3
White has no chance of an advantage, for White gets into trouble after:
example: 1 7.c3? lLlc2 1 B.b 1 e4 1 9.b3 lLle l !
 Chapter I - Schliemann: 4.d3 29

f5 28.al xd5t 29.lDg2 e2 White is

unlikely to survive.
22 . . . hf2 23.f3 f8 24.xf5t xf5 25.f1
25.lDg2 .bgl 26.i>xgl xd5 27 . .te3 xa4
is winning for Black.
25 . . . xe l 26.xe l heFt
Black has excellent winning chances in the
ending.

We must also examine the consequences of the

a b c d e f g h critical 1 7.b4 when I recommend 1 7 . . . e4!.
Black must utilise all eight ranks if he is to (At first I liked the look of 17 ... lDf3!? but
maximise his initiative! the problem turned out to be 1 8.lDxc5 dxc5
20.d l 1 9.a3! when White should maintain some
20.lDxe l ?? xf2! wins. advantage.)
2o.lDxc5? dxc5 2 1 .f3 e2 is not much better
for White.
2o . . . lDxg2!
20 . . . xd5 2 1 .xe l xd3 22 . .bg5 xe l t
23. xe I hf2 is only equal.
2 1 .gl
White loses after 2 1 .i>xg2 h4!, or 2 1 .b4
lDh4!.
2 1 . . .lDe l !
2 1 . . .lDh4 22.xg5 (22.hg5 xd5) 22 . . . f7
gives Black a smaller advantage. a b c d e f g h

The tactics work out well for Black, although

of course the position is very complicated. The
following variations illustrate his possibilities
quite nicely.

a) 1 8.lDb2?! lDc6+

b) 1 8.e l h4 1 9.bxc5 lDf3 20.xh4 lDxh4-+

c) 1 8.c3 lDe2 1 9.g7 ( 1 9.d2 .td4 20.a3

a b c d e f g h xd5 2 1 .xg5 f7+) 1 9 . . . i.d4 20.xg5
The knight's return to this unusual xg5 2 1 .hg5 g8+
destination makes a nice impression.
22.lDxe l d) 1 8.f3 lDxf3 1 9.b3
22.hg5 lDf3 23.hd8 lDxgl-+. Alternatives are no better:
After 22.xg5 h3 23.lDxe 1 hf2 24.lDg2 1 9.a2 e2 20.xf3 g4-+
e l t 25 .xe l he l 26.lDxe l e8 27.gl 1 9.xf3 xf3 20.gxf3 xc4 2 1 .bxc5 dxc5+
30 Attacking the Spanish
1 9 . . . e2 20.tiJf4
20.tiJxc5 tiJd4 2 1 .@gl tiJxb3 22.xf5 tiJxal
23.tiJe6 e8 24.hg5 tiJxc2-+
20.xf3 loses beautifully after 20 . . . g4
2 1 .g3:
a b c d e f g h
2 1 . . .e 1 t!! 22.tiJxe l e2 23.h4 xe l t
24.@h2 gl t 25.@h3 h l t 26.@g4 xh4t
27.@f3 f8t with mate in a maximum of
four more moves.
20 . . . tiJd4 2 1 .h3 gxf4 22.xf5t tiJxf5
23.bxc5 dxc5
Black stands clearly better.
e) 1 8.tiJxc5! seems to be the only way for
White to maintain the balance. My analysis
continues: 1 8 . . . dxc5 1 9.xc5 xd5 20.xa7
tiJxc2 2 1 .b l ! e l 22.xe l tiJxe l 23.hg5!
a b c d e f g White must take care to avoid 20.xc5 ? h4
This is the key move which enables White
to force a draw. 23 . . . xb l 24.a8t @d7
25 .d8t @c6 26.e8t @b6 27.i.e3t c5
28.bxc5t xc5 29.d8t @a6 30.a8t @b6
3 1 .d8t=
17... tLla!
a b c d e f g h
The other option was 1 7 . . . e4 1 8. tiJxc5
dxc5 1 9.xc5 xd5 when Black has some
activity to show for the pawn. Play might
continue 20.c3 tiJe2 2 1 .g7 tiJf4 22.fd l
xd l t 23.xd l b 6 with some, though perhaps
not quite enough, compensation.
Radjabov's choice is more incisive. Black
threatens to transfer a rook to the h-file.
18.tLlxc5
Of course 1 8.gxf3?? loses to 1 8 . . . xf3t
1 9.@gl e4.
18 ... dxc5 19JUdl
1 9.xc5 d6 20.b5 should also lead to a
draw after: 20 . . . e4! 2 1 .gxf3 xf3t 22.@gl
e4 (or immediately 22 ... g4t 23.@h l f3t)
23.fd l xe3 24.fxe3 xe3t 25.@g2
19 Je4 20.Y!ffl
.
2 1 .gxf3 xf3t 22.@gl d6 with a winning
attack.
20 ... tLld4 21 .Y!fd3 Y!fe5
 Chapter 1 - Schliemann: 4.d3 31

03122) 13.tLlxd4 hd4 14.a4!

8
7
6
5
4
3
2
a b c d e f g h
1
Black's active, centralised pieces make a nice
a b c d e f g h
impression.
Tournament practice and analysis have
22.hg5 demonstrated this to be White's most
22.c4!? was possible although 22 . . Je8 promising move. The text gains space on the
would leave Black with enough for the missing queenside while also threatening to trap the
pawn. enemy bishop.

22 .. Je8 23 ..idl gh4 24.h3 ti'xd5 25.c4 14 a6

..

ti'c6 Black prepares a retreat square on a7.

Black's activity provides full compensation
for his small material investment, and I
think that Mamedyarov was quite justified in
repeating the position.
26.ti'g3
The evaluation would be the same after
26Ja3 g8 27.Wf1 lLlfS 28.f3 lLld4 29.c3
lLlf5=
26...tLlS 27.ti'd3 tLld4 28.ti'g3 tLlS 29.ti'd3
tLld4 30.ti'g3 tLlS
Yz-Yz
This was an excellent game. In fact, I would
find it difficult to fault a single move by either
player.
We may conclude that Black is presently
holding his own after 1 3.lLlxe5, although he
should definitely make sure he comes to the
board well prepared.
15.Ae3
Black was intending . . . 0-0 with pressure
against f2. White should therefore exchange
off his opponent's strong bishop while
conveniently opening the f-file and preventing
short castling.
15 ...he3
Obviously it would be far too risky for
Black to venture 1 5 . . . hb2? 1 6.ab l b5 (after
1 6 . . . i.d4 1 7.hd4 exd4 1 8.xb7 White will
emerge with an extra pawn) 1 7.Wa2!? i.d4
1 8.hd4 exd4 1 9.axb5.
16.fxe3 0-0-0
We have reached a posltlon with equal
material, no minor pieces and symmetrical
pawns. Black is close to equality, but White
does control slightly more space as well as the
only open file.
32 Attackin g the Spanish

more advanced square is that the rook might

be able to utilise the third rank, perhaps
switching to g3 or h3 to harass the enemy
pawns. On the other hand, there are certain
times when the white queen might wish to
use the f3-square.
After 1 7.f3 play usually proceeds with:
1 7 . . . df8 1 8.af1 xf3
1 8 . . .'IWe7 1 9.fie4 g6 gives White an extra
tempo compared to Carlsen - Radjabov.
1 9.xf3 @b8

a b c d e f g h
17Jf2
Curiously enough, two of my Quality Chess
associates have made their own significant
contributions to the theory of this line.

In his 2003 book: Starting Out: The Ruy

Lopez, John Shaw recommended 1 4.a4! and
concluded that White stood better in the
diagram position, in view of his attacking a b c d e f g h
prospects against the enemy king.
Some possible continuations include:
This plan was tested in the game Cornejo
- Fernandez, corr. 2003, which continued
a) White can secure control over the f-file with
1 7.b4 hf8 1 8.b5 a5 1 9.b6 xfl t 20.xf1
20.b3 fie7 2 1 .fif1 (2 1 .fie4 g6 leaves White
@b8 2 1 .bxc7t fixc7. Black was fine and a
unable to prevent . . . f8 - a good example of
draw was agreed after a few more moves. We
why White may prefer to keep f3 free for the
must therefore conclude that the attempt to
queen!) when Black has two ideas. He might
bulldoze through the enemy defences brings
try the active 2 1 . . .e4!? or the more patient
White no advantage.
2 1 . . .h5, waiting for White to reveal his
intentions. In both cases Black has better
My editor Andrew Greet arrived at this
chances for equality than in the main
conclusion in 2006 in Play The Ruy Lopez.
line.
Instead of aiming for a direct attack, Greet
advised White to secure control over the f-file
b) 20.fie4 should be met by 20 . . . g6
and play for an endgame advantage.
intending . . . fie7 and . . . f8. (20 . . . g5 ?! does
not work here in view of 2 1 .b3 g4 [or 2 1 . . .h5
He recommended:
22.f6] 22.fif5!) White may still be able to
1 7.f3
obtain a slight edge with 2 1 .1 intending fif3
Since that time, this move has been tested
retaining control over the f-file, although if
at a high level, although the last word now
this is the best he can do then it would seem to
appears to be that the rook will be slightly
suggest that his 1 7th was slightly inaccurate.
better placed on 1. The advantage of the
 Chapter 1 - Schliemann: 4.d3 33

c) The game Naiditsch - Degraeve, Evry 19 g6 20Jf3!

2008, continued 20.'lWh4 when 20 . . . g5!?N was

worth considering. (The game continuation of
20 . . . g6 2 1 .b3 l1:e8 22.e4 a5 23.h3 gave White
the usual slight advantage, although Black was
able to hold the draw.) Play might continue
2 1 .'lWe4 g4 followed by . . . h5 with counterplay.

17 !1df8 18Jafl Vlfe7

After: 1 8 . . . l1:xf2 1 9.11:xf2 @b8 20.b3 'lWe7

2 1 .'lWe4 g6 22.'lWf3! Here we see the usefulness
of keeping the f3-square available. 22 . . . l1:c8
23.c4 c6 24.dxc6 l1:xc6 25.e4 'lWg5 26.h3 White
had a solid advantage in Topalov - Radjabov,
MoreliaiLinares 2008, although Radjabov's a b c d e f g h
superb defence enabled him to cling on for Now White can always recapture with the
half a point. queen, retaining control over the open file.

19.Vlfe4 20 b8 21 .b3 !Ixf3 22.Vlfxf3 Vlfg5 23.h3

.

This was Carlsen's choice, and is probably !1c8 24.h2

the most accurate move. Instead Naiditsch -
Gozzoli, Noyon 2008, continued with the less
convincing 1 9.'lWg4t @b8 20.'lWh5 g6 2 1 .'lWh6
l1:xf2 22.l1:xf2, at which point I propose
22 . . . 'lWd7!N.
Once again Black's position is somewhat
uncomfortable although in Carlsen - Radjabov,
MoreliaiLinares 2008, he eventually escaped
with a draw. I will have to be honest and admit
that I have not been able to find a clear path to
equality against Carlsen's handling of the white
side.

D32) 7 .ti:\d4!?

a b c d e f g h

The queen is eyeing both sides of the board.

Play may continue 23.'lWh4 (23.b3?! 'lWg4 is
awkward) 23 . . . g5! 24.'lWe4 (24.'lWxg5 'lWxa4
25.'lWg7 l1:c8 26.'lWxh7 'lWal t 27.l1:f1 'lWxb2 can
hardly cause Black any problems) 24 . . . g4 25 .b3
'lWg7. Black should be okay here, although as a b c d e f g h
usual he stands marginally worse.
34 Attacking the Spanish
This rare move has not been analysed too
extensively, which makes it seem like a breath
of fresh air in the modern era. I believe that
this will make it an appealing choice for many
readers, especially in view of the small yet
persistent disadvantage suffered by Black in
the main line above.
8.tlxd4
Let us briefly note that B.ttk 3 lDxf3t 9.'Wxf3
0-0 is harmless for Black.
More critical is:
B .lDxeS
This might offer White chances for a slight
plus, although pawn-grabbing may not be
to everyone's taste. Play continues:
B . . . 'We7 9.lDf3 lDxf3t
9 . . . lDxbS is also playable, although White may
be able to keep an edge if he plays precisely.
My analysis runs: 1 O.'WxbS c6 ( l 0 . . . lDxe4
l 1 .lDbd2 c6 1 2.'Wa4 lDxd2 1 3.hd2 0-0
14Jfe l 'Wfl I S .J.e3;;!;) 1 1 .'Wb3 lDxe4
( l 1 . . .d6 1 2.J.e3 lDxe4 1 3.e l J.e6 1 4.'Wd3
dS I S .lDc3 0-0 1 6.lDxe4 dxe4 1 7.hcS
'WxcS I B.xe4;;!;) 1 2.lDc3 dS 1 3 .lDxe4 dxe4
14.J.gS 'Wfl I S .lDd2;;!; White has a very
comfortable position.
1 0. 'Wxf3 0-0
1 O . . . 'Wxe4 1 1 .'Wb3 c6 1 2.lDc3 'We6 1 3.'Wxe6t
dxe6 1 4.J.d3;;!;
l 1 .lDc3 dS 1 2.J.gS c6 1 3.J.d3;;!;
8
7
6 6
5 5
4
3
2
1
a b c d e f g h
White has chances to consolidate his extra
pawn, although Black is very active.
8 ....hd4
The exchange of knights should, in
principle, be helpful to Black. He no longer
needs to worry about hc6 ideas, although
the drawback is that he has lost some time. We
will now follow the game Svidler - Radjabov,
Baku 200B.
9. ttl dl
This looks best.
9.c3 J.b6 1 O.J.c4 d6 l 1 .lDd2 'We7 is
comfortable for Black.
9.lDc3 is also not too dangerous, although
Black must proceed carefully. 9 . . . c6 looks
natural, but does not quite equalise after
1 O.J.c4 d6 ( l O . . . bS 1 1 .J.b3 as 1 2.a3 J.a6
1 3.lDe2;;!;) l 1 .lDe2 J.b6 ( l l . . .J.cS 1 2.J.e3
he3 1 3.'Wxe3 'We7 1 4.ad 1;;!;) 1 2.J.gS;;!; Black
still has some problems to solve.
For this reason, he should prefer 9 . . . 0-0!
1 O.J.e3 he3 l 1 .fxe3 ( l 1 .'Wxe3 d6 1 2.J.c4t
@hB=) 1 1 . . .d6 1 2.J.c4t @hB 1 3 .lDdS J.d7
with equality.
9 ... a6 IO ..tc4 V!le7 n .ttle i.a7
8
7
4
3
2
1
a b c d e f g h
 Chapter 1 - Schliemann: 4.d3 35

12.tLlh4?!
This is a slight inaccuracy. 1 2 ..ig5 d6 1 3.ltJh4
would have been the correct way to reach the
game position.

The creative 12 . .id2 d6 13 . .ia5!? can be met

by 1 3 .. Jf8!? Now Black's king will be forced
to remain in the centre for a while, but all of
his pieces can find good squares.

12 ... d6
Black could have fought for the advantage
with 1 2 . . . b5! 1 3 . .ib3 .ib7, when White is
unable to protect e4. The best he can do is
1 4.ltJf5 he4 1 5 .Wfh3 hfS 1 6.WfxfS .
a b c d e f g h
White has exchanged his opponent's strong
bishop, but no longer enjoys the benefits of a
pair of bishops. His remaining bishop is no
more useful than the black knight, and so the
position is equal. The game continued:

19 ... tLlh7! 20Jadl tLlg5 21 .,id5 c6 22.j,e4

d7 23.g3 tLlxe4 24. Wxe4 ghd8 25. We d5
26.cxd5 xd5 27.gxd5 cxd5 28.f6
White has nothing better than the text, with
which he exchanges down to a drawn ending.

a b c d e 28 gxf6 29.Wxf6 Wxf6 30Jxf6 d4 31.exd4

.

exd4 32. f1 d3 33. e1 ge8t 34. d2

In this very complicated position White has
some compensation for the pawn, but perhaps
not quite enough. One sensible continuation
is 1 6 . . . c6, reinforcing the centre, when Black's
chances look preferable. 1 6 . . . 0-0-0 and
1 6 . . . d5 also deserve consideration.
13.i.g5 i.e6 14.tLlf5 .lxf5 15.exES 0-0-0
Black should be alright here. White's
light-squared bishop is very powerful, but
its opposite number on a7 is no slouch
either!
16.i.e6t b8 17.c4 h6
1 7 . . . g6!? could have been considered.
18.i.e3 he3 19.fxe3
ge2t 35.xd3 gxh2 36.c3 h5 37.ggG h4
38.gxh4 gh3t 39. c2 gh2t 40. c3 h3t
41.c2 gh2t
V2-V2
Overall it seems that7 . . . ltJd4!? gives Blackdecent
chances to equalise. The critical continuation
may well be 8 .ltJxe5, although White will
need to demonstrate a great deal of accuracy
to obtain anything meaningful. Psychological
factors may also come into consideration. If
White wanted to give up the bishop pair in
order to win a pawn, then he could have opted
for variation D2) with 7.hc6 dxc6 8.ltJxe5 on
the previous move. It would seem reasonable
to assume that players who choose to forgo this
opportunity are probably not looking to accept
36 Attackin g the Spanish

a gambit pawn in the early stages of the game.

However, the notes to Svidler - Radjabov seem
to indicate that Black should generally be fine
if White refuses to take the e5-pawn. I believe
that 7 . . . ltJd4!? merits further practical testing,
and it will be interesting to see whether White
will be able to find a convincing reply. In the
end, my advice would be to study both this
and the more traditional 7 . . . d6 before making
your own choice.

Conclusion

Despite its outwardly timid appearance, 4.d3

deserves to be treated with respect. Anyone
doubting this need only remind themselves
of some of the players to have utilised it over
the past few years: Anand, Topalov, Carlsen,
Ivanchuk, Svidler, Leko etc. Despite the best
efforts of these elite Grandmasters, Teimour
Radjabov has almost single-handedly shown
that the Schliemann has every right to exist
at even the highest levels. And we are not just
talking about an occasional surprise weapon
that can be played once before being then
quickly discarded. The young Azeri GM
has continued to obtain good results, even
when opponents know what is coming well
in advance. And this is what defines a truly
respectable opening system.
Chapter 2

Schliemann: 4.c3
7
6
5
4
3

1
a b c d e f g h

l .e4 e5 2.f3 c6 3 . .ib5 f5 4.c3 fxe4 5.xe4 f6

Introduction page 3 8
General Themes page 39
Theoretical highli ghts page 40
Theory page 42
A) 6.d3 page 43
B) 6.e2 d5 page 44
B I ) 7. ttJ xe5 page 45
B2) 7. ttJ eg5 page 46
B3) 7. ttJ g3 page 47
B4) 7. ttJ xf6t page 50
C) 6. ttJ xf6t xf6 page 57
C I ) 7. 0-0 page 57
C2) 7.e2! page 60
38 Attacking the Spanish
1.e4 e5 2.f3 c6 3.i.b5 f5 4.c3
a b c d e f g h
Introduction
4.lDc3 is arguably White's most principled
reaction to the Schliemann, developing
another piece while defending the e-pawn. It
has been established as the main line for many
decades now, although its popularity has waned
slightly over the past few years as new resources
have been discovered for Black (as well as the
realisation that 4.d3!? may not be as harmless
as previously thought - see Chapter 1 for full
coverage of that move) .
One of the main points behind 4.lDc3 is to
facilitate the capture exf5 without allowing a
favourable advance of the enemy e-pawn (as
occurs after 4.exfS?! e4) . This can be illustrated
by a very brief investigation into a few of
Black's main responses.
a) 4 . . . lDf6 is a natural move, although White
should be able to maintain some advantage
after 5 .exfS as 5 . . . e4 can now be met by
6.lDh4!. The knight is not at all badly placed,
and Black will find it difficult to regain his
sacrificed pawn. Another option is 5 . . . lDd4!?,
although here 6.lDxe5 leaves Black struggling
to demonstrate full compensation.
b) 4 . . . lDd4!? is another interesting move
which can also be met by 5.exf5 . Play usually
continues 5 . . . c6 (5 . . . lDf6 6.lDxe5 reaches the
above note) when 6 . .ia4 gives good chances
for an advantage, while 6.lDxe5!? is also quite
dangerous.
In view of the above, it makes sense for Black
to capture on e4 himself
4... fxe4
This has become firmly established as Black's
main response.
5.xe4
This move requires no explanation. Black
no longer has to worry about exfS ideas,
and he also benefits by exchanging one of
the opponent's prized central pawns for his
own f-pawn, thus securing a valuable central
majority as well as an open file that can be used
for attacking purposes after short castling. On
the negative side, the white knight is actively
placed in the centre of the board, while Black
has still only managed to develop a solitary
piece.
At this point Black must choose between two
principled alternatives. The aggressive 5 . . . d5!?
used to be considered the main line, and has
certainly not been refuted. Unfortunately most
of the critical variations have been analysed
quite extensively, and it seems that White can
always maintain an edge if he knows what
he is doing. For this reason I have decided to
recommend the alternative:
5 ... f6
This move appears to have superseded 5 ... d5
as the experts' choice. Most notably, it has been
used on a regular basis by Radjabov, which is
about as high a recommendation as one could
ask for. Black develops another piece and
challenges the opponent's centralised knight.
He also retains the option of occupying the
centre with . . . d5 , depending on how White
responds.
 Chapter 2 - Schliemann: 4.lik3 39

General Themes In the above diagram we see Black struggling

for counterplay while White is consolidating
White's speedy development smoothly. Black may have a pair of bishops,
One of the main points behind White's 4th but he lacks a target. The pawn on f2 is securely
move is to develop quickly and exert pressure defended, so it is hard to imagine Black making
against the opponent's pawn centre and/or anything of the open f-file.
king position. The following diagram shows
the type of scenario that Black should avoid. Weak squares
Another scenario might be that White makes
use of his lead in development to install his
8
pieces in the key squares weakened by Black's
7 pawn moves.
6
5 8

4 7

3 6

2 5

1 4

a b c d e f g h 3

Black lags in development and his king is 2

vulnerable. 1
a b c d e f g h
The e5-pawn
Over the course of the present chapter we Here we see that Black's pawn structure has
will encounter several lines in which Black been compromised on both sides of the board.
'sacrifices' (or White 'wins', depending on your The biggest danger is the kingside, where the
perspective!) the e5-pawn. f6-pawn severely restricts the bishop on g7.
If White can install a knight on f5 then he
will enjoy positional domination in addition
8
to excellent attacking chances. Black might
7 prevent this with a timely . . . f5, although
6 this might restrict his other bishop, as well
as weakening the dark squares. White might
5 exploit the latter with the manoeuvre .tf4-e5 .
4
We have seen that there are numerous pitfalls
3
that Black must avoid. Indeed, I would say
2 that the main virtue of the 4.lLlc3 variation is
that White's rapid development enables him
1
to keep a variety of options open. He might
a b c d e f g h equally aim for an early confrontation or a
40 Attacking the Spanish

long-term strategical approach, depending on

8
how the opponent plays.
7
Fortunately it is not all doom and gloom for 6
Black! His position contains plenty of resources
and ifWhite plays inaccurately he can easily get 5
into trouble. We will now turn our attention to 4
some of the more positive aspects of the black
side of the Schliemann. 3
2
Central Control
1
One drawback of White's 4th move is the
loss of a pawn presence in the centre of the a b c d e f g h
board. In an ideal world Black might end up Possible attacking ideas include . . . lLlg4
with a powerful pawn centre resembling the to target f2, or . . . i.g4 to develop with gain
following: of tempo. White must also be wary of . . . h6
followed by .. Je6 when his queen would be
8 embarrassed.
7
Theoretical highlights
6 Here are a few of the most noteworthy
5 theoretical variations contained within the
present chapter. Our first example shows how
4 even a seemingly placid variation can contain
3 hidden dangers.

2 l.e4 e5 2.ttla ttlc6 3.Ab5 f5 4.ttlc3 fxe4

1 5.ttlxe4 ttlf6 6.d3 d5 7.ttlxf6t Vxf6 S.Ag5
Vf5
a b c d e f g h
Black's extra central pawn exerts a huge
8
influence over that area of the board, while
the open g-file may be used for attacking 7
purposes. 6

Gambit Play 5
We have already seen that White may aim 4
to capture the e5-pawn. Whilst this approach
can work well in certain positions, it can also 3
carry great risks, especially when his material 2
gains come at the expense of development.
1
In the following diagram Black's initiative is
becoming very dangerous indeed. a b c d e f g h
 Chapter 2 Schliemann: 4.ttk3
- 41

White's 6th move may appear excessively

8
timid, and has generally not been considered
at all threatening for Black. However, in the 7
diagram position I found a strong novelty 6
for White: 9.c4!, after which Black's position
can become dangerous if he does not react 5
precisely. Full details can be found in line A) 4
on page 43.
3
Here is a more popular variation, which Black 2
should definitely study:
1
1.e4 e5 2.tLlf3 tLlc6 3.J.b5 f5 4.tLlc3 fxe4 a b c d e f g h
5.tLlxe4 tLlf6 6.'ife2 d5 7.tLlxf6t gxf6 8.d4 This position should be okay for Black,
Ag7 9.dxe5 0-0 although my research revealed that he has
often failed to handle it in the most accurate
8 way. Play continues 10.'ife2 (After 1 O.1Mfh5t
I suggest a subtle yet significant improvement
7
which completely solves Black's problems)
6 10 c6 1 1 .J.d3 d5 when Black should be fine
.

5 as long as he plays precisely. Full details can be

found in line C l ) on page 57.
4
3 Finally we come to the main line, which is
covered in variation C2) on page 60.
2
1 l.e4 e5 2.tLlf3 tLlc6 3.i.b5 f5 4.tLlc3 fxe4
5.tLlxe4 tLlf6 6.tLlxf6t 'ifxf6 7.'ife2 Ae7
a b c d e f g h
8 ..bc6 dxc6 9.tLlxe5 J.f5 10.0-0 0-0 1 l .d4
This is considered to be one of the critical
lines. It is dangerous for White to try to
8
keep the extra pawn so he invariably plays
e5-e6, either with or without a preliminary 7
exchange on c6. Line B4) on page 50 shows 6
why Black has nothing to fear from either
approach. 5
4
The following variation is also seen quite
frequently. 3
2
1.e4 e5 2.tLlf3 tLlc6 3.Ab5 f5 4.tLlc3 fxe4
1
5.tLlxe4 tLlf6 6.tilxf6t 'ifxf6 7.0-0 tLld4
8.tLlxd4 exd4 9J:e1 t i.e7 a b c d e f g h
42 Attackin g the Spanish
White has an extra pawn and keeps some
advantage, but I do not believe it to be
anything serious. As you will see in the notes
to Kramnik - Radjabov, my recommendation
involves simplifying into an endgame a pawn
down which I expect Black to hold quite
easily.
Theory
1.e4 e5 2.ttla ttlc6 3 .ib5 f5 4.ttlc3 fxe4
5.ttlxe4
It should come as no surprise that White
gains nothing from:
5 .hc6 bxc6!
Instead 5 . . . dxc6 6.ttJxe4 ttJf6 would
transpose to the note to Black's 5th in line
C of Chapter 3. Usually in the Schliemann
Black would prefer to recapture with the
d-pawn, but in the present position the text is
stronger.
6.ttJxe4 d5 7.ttJg3 e4
Black must of course avoid 7 . . . i.d6? 8 .ttJxe5
he5 9.h5t.
8 .ttJe5 ttJf6!?N
With this untested move Black gives up a
pawn in order to fight for the initiative.
9.d4
Or 9.0-0 i.d6 1 O.ttJxc6 d7 1 1 .ttJd4
0-0 with the bishop pair and a lead in
space and development to compensate the
pawn.
9 . . . i.d6 1 O.ttJxc6 d7 1 1 .ttJe5 e6
a b c d e f g h
Black has a promising position. Indeed, if we
look a little deeper we can see how easy it is
for White to fall into difficulties:
1 2.c4!?
This may be his best chance.
After 1 2.0-0 he5 1 3.dxe5 xe5 1 4.i.e3
0-0+ Black stands better thanks to his
powerful central pawn majority.
1 2.i.f4 i.a6 1 3.c4 ( 1 3.ttJe2 0-0+) 1 3 . . . 0-0
1 4.0-0 hc4 1 5 .ttJxc4 hf4 1 6.ttJe3 ab8
also leaves Black with an obvious advantage.
1 2 . . . 0-0
1 2 . . . he5?! leaves Black with some structural
defects after: 1 3.dxe5 xe5 1 4.cxd5 0-0
1 5 .0-0 d8 1 6.b3 xd5 1 7.i.d2;!;
1 3.cxd5 xd5 1 4.b3 xb3 1 5 .axb3 i.e6
With two powerful bishops and a lead in
development, Black maintains excellent
compensation notwithstanding the queen
exchange.
8
7
6
5
4
3
2
1
a b c d e f g h
5 ... ttlf6
5 ... d5 is the other principled continuation,
but the present work will focus exclusively on
the text. Black is not in a hurry to occupy the
centre and will instead be happy to continue his
kingside development while also challenging
the strong enemy knight.
We will consider the quiet response A) 6.d3
before moving on to the much more common
 Chapter 2 Schliemann: 4.tik3
- 43

B) 6.YlYe2 and C) 6.tnxf6t.
The position after 6.hc6 dxc6 is analysed in
line C) of Chapter 3 (note to Black's 5th move
on page 77) .
A) 6.d3
This timid move should not be too
dangerous, although we will see on move 9
that the white position still contains some
interesting resources.
6 ... d5!
There is no reason for Black to refrain from
this natural move.
7.tnxf6t
White is simply worse after the passive
7.lDg3 .td6 8.0-0 0-0+, for example 9.c3 (after
9Je 1 lDd4!? 1 O.lDxd4 exd4 1 1 .h3 c6 1 2 ..ta4
8.d4 e4 9.lDe5 i.b4t 1 O.c3 0-0 1 1 .0-0 .td6+
was Todorovic - Dinie, Sokobanja 1 989.
8.0-0 .tg4 9Je 1 (9.c4 was tried in Nakamura
- Chernoff, Dos Hermanas 2003, when
Black could have obtained the advantage with
the natural 9 . . . hf3 1 O.xf3 x8 1 1 .gx8
0-0-0+) 9 . . . .td6 1 O.h3?! ( l 0 . .ta4 would have
left Black with only a slight edge) 1 O . . . h8+
Black went on to win in Vamos - Jakubowski,
Szombathely 1 993.
8.e2 .td6 9.c4 (9 . .tg5 f5 is analysed in
the note to White's 9th move below) 9 . . . 0-0
1 O.cxd5 lDd4 1 1 .lDxd4 exd4+ was once again
better for Black in Tchimino - Belmonte,
Santiago 1 994.
8 YlYf5 9.c4!N
.

c7+ Black eventually prevailed in Schmaus
Maier, Munieh 1 993) 9 . . . lDe7 1 O . .ta4 c6, as in
Kochyev - Kuzmin, St Petersburg 1 992.
7 YlYxf6
This untested idea was mentioned in the
introduction. I believe this to be only move
that can cause Black any significant problems.
A few examples:

9. d2 .td6+ was Bernal Caamano - Arbakov,

Linares 1 995.
8
7 9.0-0 .td6 1 O.c4 looks fairly sensible, e.g.
6 1 0 . . . 0-0 1 1 .cxd5 lDd4 1 2.lDxd4 exd4 with
equality in Toledano Llinares - Ferron Garcia,
5 Vilanova 1 993.
4
After 9. e2 .td6 White can try many
3
moves, none of which enables him to
2 equalise, e.g. 1 O ..th4 (Neither 1 0.0-0 0-0+,
Diez-Schlesinger, Bad Homburg 2005, nor
1
1 0.0-0-0 0-0 1 1 ..te3 .te6+1+, Huellen
a b c d e f g h - Rieke, Enger Spenge 1 998, offer any
Black has no problems, and it is White improvement.) 1 0 . . . 0-0 ( l 0 . . . h5!? is also very
who has to tread carefully to avoid becoming strong) 1 1 .hc6 bxc6 1 2.0-0?! h5 1 3 . .tg3
worse. .tg4 Black was already winning in Yap Choow
Tun - Handoko, Kuala Lumpur 200 1 .
8.i.g5
The alternatives are no better: 9 ... <it>f7!
44 Attacking the Spanish

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
a b c d e f g h a b c d e f g h
This surprising move is the one which best In this position Black clearly has long-term
meets the demands of the position! Black's compensation for the pawn. White suffers
number one priority is to facilitate a knight from some structural defects, although his
jump to d4. weaknesses are not so easy to attack. Overall I
In case of 9 . . . .!b4t 1 O . .!d2 0-0 1 1 .hc6 consider the chances to be roughly equal after
hd2t 1 2.V;Vxd2 bxc6 1 3.0-0 dxc4 1 4.dxc4 the simple 14 Ad7 or 14....if5.
.

e4 1 5 .tt)d4 V;Vg6 Black's structural weaknesses On the other hand the tempting 14 b5!? is ..

render his position strategically quite interesting but probably not best in my opinion.
dangerous. Play continues 15.,ixb5 b8 16.Ac4 gxb2
17.0-0 when White has some advantage, e.g.
10.cxd5 17 Ah3 18.gAl1 gxb1 t 19Jxb1 a5 20.d4
.

The surprising 1 0.g4!? should be met by followed by a bishop transfer to e4.

1 0 . . . i.b4t ( l 0 . . . V;Vxg4?? loses to 1 1 .hc6)
1 1 .@f1 ( l 1 ..!d2 hd2t 1 2.tt)xd2 enables B) 6.'llYe2
Black to pick up a pawn on either d3 or g4.)
1 1 . . .V;Vxg4 1 2.hc6 V;Vh3t 1 3.@gl ( l 3.@e2?
8
bxc6 1 4.tt)xe5t @g8+) when Black can either
take a perpetual with 1 3 . . . V;Vg4t or continue 7
the fight with 1 3 . . . bxc6. 6

lo ... lLld4 1 1.Ac4 5

1 1 .tt)xd4 exd4 leads to an equal position 4
after 1 2.V;Vd2 .!d6 1 3.0-0 a6 1 4 ..!c4 b5
3
1 5 . .!b3 .!b7= .
2
1 1 . .. Ad6 12.Ae3 lLlxf3t 1
1 2 . . . b5?! 1 3.tt)g5t @g6 14 . .!b3 is dangerous
for Black, e.g. 1 4 . . . a5 1 5 .h4 when the white a b c d e f g h
pieces coordinate very well. This is a more challenging continuation than
6.d3. The queen sets up an x-ray attack against
13.'llYxf3 'llYxf3 14.gxf3
 Chapter 2 Schliemann: 4.ltk 3
- 45

the enemy monarch while also menacing the

e-pawn.

G d5!

Once again Black should not shy away

from the most aggressive and principled
continuation. At this point we will begin by
considering the unusual Bl) 7.llxe5, which
should not cause Black too many problems.
Next will come the two knight moves B2)
7.lleg5 and B3) 7.llg3. Both ensure him
of an extra pawn, but the good news is that
Black's rapid development and active pieces a b c d e f g h
should provide excellent compensation. Black's extra space and fluid development
Finally we will address the main line of offer promising compensation for the pawn.
B4) 7.llxfGt.
13.llb3
Bl) 7.llxe5 dxe4 Feeble is 1 3.hd5 ?! cxd5+, Kudryavtsev -
This resembles the line 5 . . . d5 (instead of our Duriez, e-mail 2000.
preferred 5 . . . ltJf6) 6.ltJxe5 dxe4, although the
extra moves e2 and . . . ltJf6 will create some 1 3.ltJe2 also should not give Black too much
unique opportunities for both sides. to worry about. Bossuyt - Van Beers, Antwerp
1 992 continued 1 3 . . . 0-0 (alternatives include
8.YlYc4!? 1 3 . . . f8 1 4.0-0 o-o-o and 1 3 . . . b5 14.i.b3
This looks like the most principled way for f8 1 5 .0-0 O-O-O) 1 4.0-0.
White to utilise the position of his queen.
At this point the game continued: 1 4 . . . @hB?!
Instead the unimaginative B.ltJxc6 bxc6 1 5 .ltJg3 aeB 1 6.d3 ( 1 6.b3;!;) 16 ... exd3
9.hc6t i.d7 1 O.haB xaB leads to a 1 7.hd3 ltJf4 ( 1 7 . . . ltJb4 1 B . .ie3) 1 B.ltJe4
position in which Black can claim at least equal ltJxd3 1 9.cxd3 i.b6 20.i.e3;!;
chances despite his nominal material deficit,
for example 1 1 .d3 @f7 ( 1 1 . .. c6!? could be Instead Black should have preferred 1 4 . . . b5!?N
considered, after which 1 2.f3 is the critical 1 5 .i.b3 aeB.
test) 1 2.dxe4 xe4 1 3.xe4 ltJxe4 1 4.i.e3
i.b4t!? ( 1 4 . . . a6 1 5 .0-0-0 i.d6 may have been
better, keeping an unbalanced position in
which I slightly prefer Black's chances) 1 5 .c3
ltJxc3= led to an equalization of the material in
Rutherford - Cook, Sydney 1 992.

After the text there follows a forced sequence.

8 YlYd5 9.YlYxd5 llxd5 lo.llxc6 Ad7

1 l .lld4 cG 12.Ac4 Ac5

a b c d e f g h
46 Attacking the Spanish

Black enjoys full compensation here. His

8
pieces are very actively placed, and it is not easy
for White to develop his queenside pieces. 7
6
13 i.b6 14.d4 exd3 15.cxd3
.

8 4

7 3

6 2

5 1
4 a b c d e f g h

3 Offering a second pawn. Black is building a

useful lead in development, and is well placed
2 to take advantage of the open f-file.
1
9.tLlxc6
a b c d e f g h
9 . .ixc6 can be met by 9 . . . bxc6 as in the note
We have been following the game Raty - to White's 8th, as well as 9 . . . .ixe5!? 1 O.xe5
Reich, corr. 1 983, in which Black failed to bxc6t, which may be even better.
equalise after 1 5 . . . 0-0 1 6.0-0 ae8 1 7 . .id2
<tt> h 8 1 8.d4;;!;, and subsequently blundered 9 bxc6 10.i.xc6 gb8 1 1.d3
.

with 1 8 . . . lLlf6? 1 9 . .ib4 1 -0. The alternatives are no more appealing.

Instead I propose 15 ....lf5N 16.0-0 0-0-0;;
with reasonable chances. Black is very active
and White's d-pawn is weak. The primary
threat is . . . lLlb4.
B2) 7.tLlegS .id6
Black should have no qualms about offering
the e-pawn.
After I I .c4 b6 1 2 ..ia4 b4+ Black regains a
pawn without dampening his initiative.
I I .d4 b6 1 2 . .ia4 was played in Malmstroem
- Kudryavtsev, e-mail 200 1 . At this point the
strongest continuation would have been 1 2 . . . h6
1 3.lLlf3 .ia6 with a very powerful attack.
1 l Jlb6 12.i.a4 i.c5
.

8.tLlxe5 1 2 . . . lLlg4!? was another interesting possibility.

8 . .ixc6t bxc6 9.lLlxe5 0-0 also gives Black
plenty of play. Westerinen - Pokern, Germany
1 982, continued 1 O.d4 c5! 1 1 .c3 cxd4 1 2.cxd4,
and now 1 2 . . . c5! would have been consistent
and strong. We will encounter this method of
undermining the white centre more than once
over the coming pages.
8 ... 0-0!
The position after the text can be found in the
introduction, in which it was remarked that
White was in some danger. In case the reader
was sceptical of this assessment, the evidence
can be found in the remainder of our analysis.
13.0-0
After 1 3.c3 lLlg4 Black's attack is extremely
strong, as demonstrated by the game Morais
 Chapter 2 - Schliemann: 4.ttJc3 47

- Beaumont, e-mail 2000, which continued: It is not often that one can achieve such
1 4.f3 .tf2t (perhaps even stronger may have an attack after just 1 4 moves with the black
been 1 4 . . . h6!? I S .fxg4 hxgS intending . . Je6) pieces! In the only practical encounter White
I S .i>d l .th4 1 6.h3 e6 1 7.c2 e7 1 8.b3 failed to find the correct response.
e2 1 9.93 i>h8 20.b8 xh2 2 1 .xh2 xh2
22.gxh4 xf3 23.f4 .tg4 24.xa7 el t 15Jxf2?!
2S.i>d2 The only satisfactory continuation was
I S .xf2! xf2 1 6.dxcS b4 1 7 . .tb3 e4
8 1 8.xe4 xe4 1 9 . .tf4 when White has enough
7 for the queen.
6
15 .ixd4 16 .ib3 .ixf2t 17.mhl .ic5+

5
In Grazinys - Raleus, e-mail 1 999, Black had
4 restored the material balance while retaining a
3 very strong attack.
2

1
B3) 7.tLlg3
This move carries similar intentions to the
a b c d e f g h
previous variation, namely to capture the eS
2S . . . d4! 0-1 . An elegant finish to a fine pawn. The difference is that White hopes his
attacking game. knight will be more secure and less tactically
vulnerable on g3.
13 ... tLlg4 14.d4
In the game Stock - O'Hara, e-mail 1 998, 7 ...,id6
White was swiftly demolished after: 1 4.f3? Black's response is identical.
e6! I S . .te3 (this is hopeless, although
I S .d l can also be refuted by I S . . . xf2! 8.tLlxe5
1 6.xf2 hf2t 1 7.i>xf2 h4t 1 8.i>gl xf3!) 8 .hc6t bxc6 9.xeS 0-0 reaches the main
I S . . . xe3 1 6.fxe3 xe3 0-1 line.

14 .. Jlxf2! 8 ... 0-0!

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
a b c d e f g h a b c d e f g h
48 Attacking the Spanish

Once again Black should not hesitate to

offer a second pawn. Now White must decide
which minor piece to exchange for the knight
on c6.

9 .hc6

Probably best, although relinquishing the

bishop pair is never a decision to be taken
lightly.

The drawbacks of 9. ttJxc6 bxc6 are twofold: a b c d e f g h

White loses a valuable blockader of the open
e-file, and his bishop is forced to spend a tempo.
The loss of time is important, notwithstanding
the fact that a second enemy pawn is about to
fall. Play continues 1 O ..bc6 b8 1 1 .0-0 b6
1 2 . .ta4 ttJg4
At this point the game Riepe - Leisebein,
e-mail 2002, was agreed drawn, presumably
in view of 20 . .bd5 xb2 2 1 .ae 1 g6, when
White has nothing better than 22 . .tc6 e6
23 . .td5 e8 repeating.
d) 1 3.d4 should be met by 1 3 . . . h4!? if Black
is playing for a win. A logical continuation
is 1 4.h3 .ta6 1 5 .e6t h8 1 6.xg4 .bg3
1 7.xh4 .txh4 1 8.f3 .bfl 1 9.xfl e6
20 ..td2 c6.

a b c d e f g h

The following variations illustrate Black's

chances:

a) After 1 3 .d3? i.c5 White was already busted

in Gomez Trigo - Antunes, Seville 1 993. a b c d e f g h

The position is quite balanced. White has

b) No better was 1 3 .h3? .bg3 1 4.fxg3 two pawns for the exchange, but he is not
( 1 4.hxg4 e6!-+) 1 4 . . . e6 when White was completely safe as Black can try to invade on
losing his queen for insufficient compensation e2.
in Jackson - Lyell, West Bromwich 2003.
Alternatively, should he wish it, Black can more
c) White can probably maintain equality or less force a draw with: 1 3 . . . ttJxh2 14.xh2
with: 1 3 . .tb3 .tc5 1 4.d4! .bd4 1 5 . .te3 ttJxe3 h4t 1 5 .gl .bg3 ( 1 5 . . . .ta6?! 1 6.c4 .bc4
1 6.fxe3 e8 1 7.f3 .be3t 1 8.h 1 h8 1 7.e 1 is less convincing)
1 9.xd5 xd5
 Chapter 2 Schliemann: 4.ltJc3
- 49

12.'1'8
White does not appear to have a satisfactory
way to stabilise his centre.

1 2.vgbS, Roiz Baztan - Menendez Rey, Gijon

2003, can be met by 12 . . . cxd4 1 3.lLlc6 .td7!
with advantage to Black.

1 2.c3 cxd4 1 3.cxd4 cS!+ continued Black's

undermining policy in Beveridge - Curnow,
a b c d e f g h Cheltenham 2004. The doubled c-pawns
Play continues 1 6.fxg3 xf1 t 1 7.vgxf1 (risky certainly proved useful in this instance.
is 1 7.@xf1 ?! vgh 1 t 1 8.@f2 f6t 1 9 . .tf4 vgxa 1 )
1 7 . . . vgxd4t 1 8.vgf2 vgd 1 t with a perpetual. 1 2 . .tgS!? is an interesting suggestion of
Postny. I think that Black should play 1 2 . . . h6
9 ... hxc6 ( 1 2 . . . cxd4 1 3.f4 b8 1 4.lLlhS is dangerous)
1 3 . .txf6 vgxf6 (Postny analyses only a long
and complicated variation starting with
8 1 3 . . . gxf6 but in my opinion it's not necessary
7 to weaken the kingside) 1 4.c3 cxd4 l S .cxd4
cS .
6
Black's bishop pair and active pieces ensure
5 him of a full share of the chances.
4
We now return to the main line (after
3 1 2.vgf3) .
2
1 8
a b c d e f g h 7
10.0-0 6
1 0. lLlxc6?? vgd7 leaves White facing 5
embarrassment along the e-file.
4
1 O.d4 cS 1 1 .0-0 is another try, after which 3
I like Postny's suggestion of 1 1 . . . cxd4.
2
Surprisingly, this move has never been played.
Play continues 1 2.lLlc6 vgd7 1 3 .lLlxd4 e8 1
with sufficient compensation according to the
a b c d e f g h
Israeli Grandmaster.
This position was reached in the game
10 . Je8 I I .d4 c5!
.
Vitiugov - Ponkratov, Orsk 2002, which
This undermining move is key, as Black continued: 1 2 . . . .txeS 1 3.dxeS lLlg4 ( 1 3 . . . xeS
needs to open lines for his bishops. 1 4 . .tf4;;!;)
50 Attacking the Spanish

8
B4) 7.xf6t
This is White's most promlSlng move,
7
eschewing the gain of a pawn but avoiding the
6
loss of time.
5

4 7 .. gxf6
.

2 8
1 7
a b c d e f g h
6
The game resulted in a draw after: 1 4.c4 c6
5
( I 4 . . . tDxe5 1 5 .'Wxd5t 'Wxd5 1 6.cxd5 tDd3!?
is okay for Black) 1 5 .i.f4 ( I 5 .cxd5 tDxe5 4
[ 1 5 . . . cxd5 1 6..!! d l ] 1 6.'Wc3 'Wxd5 1 7.i.e3t) 3
1 5 .. ..!! f8 1 6.e6 he6 1 7.h3 tDf6 1 8Jae l
'Wd7 1 9.'We3 ae8 20.'Wxc5 dxc4 2 1 .i.e5t 2
Y2-Y2 1

Instead White could have secured a slight a b c d e f g h

advantage with 1 4.'Wa3! 'Wh4 1 5 .h3 tDxe5
1 6.'Wxc5 .
In view of the above, 12 cxd4 looks more .
promising, when play may continue: 13.c6
%Yd7 14.xd4 gb8 15.df5 Ae5
Naturally, Black recaptures with the pawn in
order to avoid the loss of the e-pawn. From
a purely strategic point of view Black's extra
central pawn is a big plus. The downside is
that, in the short term at least, his king will be
slightly exposed.

8 8.d4
White had better stake a claim in the centre
7 rather than being tempted by the lure of a check
6 on h5, as shown by the following variations.
5 8.tDh4 can be met by 8 . . . 'We7 9.0-0 i.d7+ when
4 White had given up the centre for nothing in
Lanin - Smirnov, St Petersburg 2007.
3
No better is 9.hc6t bxc6 1 0.0-0 g8!?+
2 when Black's pawn centre makes a powerful
1 impression, especially when backed up by a
pair of bishops.
a b c d e f g h
Postny concludes that Black has good 8.tDd4?! may appear attractive, but after the
compensation, and I agree. Black controls the calm response 8 . . . i.d7 9.'Wh5t @e7 Black's
centre, his bishops have great potential and his magnificent centre gives him the advantage
remaining pieces are all very active. once again.
 Chapter 2 Schliemann: 4.lik 3
- 51

8 Ag7
. b) 1 O.h3 i.xf3 1 1 .V;Vxf3 0-0 1 2.cxd5 ? (after
Black should have no qualms about 1 2.i.xc6 bxc6 only Black is playing for
sacrificing a pawn, especially considering that the advantage, but this was the lesser evil)
8 . . . e4?! 9.ltJh4 would leave him under some 1 2 . . . ltJxd4 1 3.V;Vd3
pressure. Compared with the previous note the
d4-pawn secures White some additional space, 8
while Black's own centre can be undermined 7
by a timely f2-f3.
6
5
9.dxe5
From time to time White tries 9.c4?! but this 4

is unlikely to bring him anything good after 3

9 . . . i.g4!, maintaining the option of castling on 2
either side. White has tried three moves here: 1

a b c d e f g h
a) 1 0.dxe5 0-0 I I .e6?! (White should be able
to equalise with l 1 .cxd5 ltJd4 1 2.V;Vd3 i.xf3 The players agreed a draw here in De Vreugt
1 3.V;Vxd4 fxe5 1 4.V;Vd3 i.xd5 1 5 .i.c4 c6 1 6.0-0 - Hendriks, Wijk aan Zee 1 997, but I am sure
f7=) 1 1 . . .ltJe5 1 2.i.d7 dxc4 Black would have continued had he spotted
1 3 . . . c6! 1 4.i.c4 ( 1 4.dxc6? V;Va5t picks up the
8 bishop) 1 4 . . . cxd5 1 5 .i.b3 f5 when his extra
7 pawn and powerful centre should add up to a
decisive advantage.
6

5
c) 1 0.cxd5 has been seen at a high level, but
4 after 1 O . . .V;Vxd5 l 1 .dxe5 0-0-0 1 2.i.xc6
3 V;Vxc6
2
8
1
7
a b c d e f g h
6
The pos1tion appears messy, but is
5
favourable to Black who enjoys a useful lead in
development. For example: 4

3
a l ) After 1 3.e7 ltJd3t 14.@f1 V;Vxd7 2
1 5 .exf8=V;Vt xf8 Black went on to win quickly 1
in Acosta Ruiz - Rybak, e-mail 1 999.
a b c d e f g h

a2) 1 3.h3 may be a better try, but after it was dear that Black had won the opening
1 3 . . . i.xf3 1 4.gxf3 V;Ve7 1 5 .i.e3 ltJxd7 1 6.exd7 battle in Morozevich - Aronian, Monte Carlo
V;Vxd7 1 7.V;Vxc4t V;Vf7 1 8.V;Vxf7t @xf7 Black's 2006.
superior structure give him the advantage in
the endgame. 9 0-0
..
52
a b c d e f g h
This position was featured in the introduction.
It should be noted that Black gains nothing
with 9 . . . fxe5 ?! 1 O.ttJxe5. At this point White
usually chooses between B41) 10 .ixc6 and
B42) 10.e6.
The attempt to hang on to the extra pawn
with 1 0.exf6 V;;Vxf6 leaves Black with plenty of
compensation, e.g. I l .c3 J.g4 1 2.hc6 V;;Vxc6
1 3.ttJe5.
a b c d e f g h
In this position Black enjoys the luxury of
more than one route to an advantage.
In Westerinen - Binham, Helsinki 1 986, he
opted for 1 3 . . .he2 1 4.ttJxc6 ae8 1 5 .J.e3
bxc6 1 6.xe2 c5 1 7.d3 d4 1 8.cxd4
cxd4 1 9 .J.d2 xf2 and retained an edge,
although White eventually held on for a
draw.
Attacking the Spanish
An interesting deviation to the above game
would have been: 1 5 . . . J.h5!? 1 6.ttJb4
( I 6.ttJa5 c5 1 7.0-0 d4 1 8.cxd4 cxd4 1 9.J.d2
e2 20.ad l c8+) 1 6 . . . d4 1 7.cxd4 hd4
(or 1 7 . . . c5 1 8.dxc5 hb2 1 9.0-0 hal
20.xa I + when White will offer a tough
resistance in the endgame) 1 8.0-0 hb2+
Although both of the aforementioned options
are quite agreeable, Black can also consider the
more straightforward 1 3 . . . V;;Vxc3t!? 1 4.bxc3
he2 1 5 .xe2 he5+.
B41) 10 .ixc6 bxc6 1 1 .e6
White opts for a strategic handling of the
position. Instead of opening for the game for
the enemy bishops, he returns the extra pawn in
the hope of exploiting his opponent's structural
weaknesses on both sides of the board. One
could certainly make the argument that the
f6-pawn is more of a liability than an asset for
Black, so this approach is eminently sensible.
1 1 ... e8
The rook no longer has a future on the f-file,
but it will be happy enough opposite the white
queen.
a b c d e f g h
12.0-0
Queenside castling is riskier, as Black's heavy
pieces enjoy much easier access to the b-file
 Chapter 2 - Schliemann: 4.ltJc3 S3

than the g-file. Nevertheless a few games have 12 gxe6 13.i.e3

proceeded with: After 1 3.d3 it looks quite interesting for

Black to try 1 3 . . . aS!? (the more popular option
1 2.i.e3 he6 1 3.0-0-0 of 1 3 . . . cS is also fine) .
1 3.0-0 i.g4 sets up an unpleasant pin, so
the safest continuation is probably 1 3 .'1Wd2!?
d6 1 4.0-0 with equality.
The alternative 1 3.lDd4 i.d7 1 4.0-0-0
looks unpromising after 1 4 . . Jb8 I S .lDb3
as 1 6.d2, Anka - Micic, Balatonbereny
1 993, when Black could have secured a big
advantage with 1 6 . . . a4 1 7.lDcS a3.
1 3 . . . d6
Black's central control combined with the
two bishops and open b-file more than make
a b c d e f g h
up for his structural deficiencies.
1 4.a6 Possible continuations include:
White is not helped by 1 4.d2 cS I S .i.h6
d4 1 6.hg7 @xg7+, Luther - Caruso, Verona a) 1 4.lDd4 e4 I S .d l (or I S .f3 i.a6 1 6.c3
200S , with good attacking chances. xd4 1 7.xd4 hfl 1 8.@xfl =) I S . . . f5 1 6.c3
It was a similar story after 14.lDd4 cS e8 1 7.i.d2 g6 was equal in Severiukhina
I S .lDxe6 xe6+ in Hanisch - Karpatchev, Grebenshikov, Dubna 2007.
Neuhausen 2004.
b) 1 4.e l is playable although 14 . . . xe l t
8 I S .lDxe l left the knight in a passive position
7
in Lenz - Chulis, Austria 1 996.
6
c) In Fisher - Koons, e-mail 2000, White soon
5
lost after 1 4.d l i.a6 1 S .f5 e8 1 6.f4? i.e2
4 1 7 .e I ? e4 0-1 . A perfect result for Black,
3 although White could easily have improved by
2 means of 1 6.i.d2 with equality.
1
It is for this reason that I propose the more
a b c d e f g h
ambitious I S . . . e4!?N. This is a multipurpose
The diagram position was reached in move, whose numerous merits are revealed in
McShane - Nordfjord, Copenhagen 2003. At the following sample variations:
this point Black should change plans, forgoing
the attack in favour of 1 4 . . . cS! I S .xd6 cxd6 c 1 ) The first point worth mentioning is that
when the modification of the pawn structure 1 6.i.d2?? can be refuted instantly by 1 6 . . . i.c8!,
helps him considerably. The doubled pawns trapping the queen. Please note, however, that
are not at all weak, and the d6-cS-dS trio the same idea cannot be used against other
will ensure his domination over the central moves, as White could simply capture the rook
squares. on e4 thanks to the pin along the d-file.
54 Attacking the Spanish

c2) 1 6.1.e3 should be met by 1 6 . . . ;Ve8!, 1 8 .ixg7 .ixf3 19.9xf3 cttxg7

threatening to trap the queen with . . . 1.c8.

Following 1 7.lLld2 b4 1 8 .b3 ;Ve5 Black is
8
starting to take control over the centre. After
1 9.;Vd7 ;Vd6 White must either retreat or 7
allow a useful improvement in the enemy 6
pawn structure.
5
c3) 1 6.d4 may be White's best. Here I would 4
suggest 1 6 . . . ;Vc8!? 1 7.;Vxc8t Axc8 when
3
the powerful bishops make up for Black's
damaged structure. 2
1
13 c5..

a b c d e f g h
8 We have been following the game Klundt
- Bock, Forchheim 2002. The chances can
7
be evaluated as approximately level, as the
6 mutual pawn weaknesses are of roughly equal
significance.
5
4 B42) 10.e6
3
8
2
7
1
6
a b c d e f g h
5
Postny correctly recommended this move as
the best. 4
3
Less good is 1 3 . . . e8 1 4.;Vd3 1.g4 1 5 .1.c5;!;
when Black's position is somewhat unpleasant, 2
albeit defensible. (Instead after 1 5 .lLlh4?! c5! 1
Black obtained counterplay and triumphed in
Shirov - Radjabov, Odessa 2007.) a b c d e f g h
With this move White aims for a roughly
14.%Yd2 d4 15 ..ih6 similar type of position, but without ceding
It looks natural for White to exchange one the bishop pair. Strategically speaking, he
half of his opponent's bishop pair, although hopes to show that Black's split kingside
there is nothing wrong with 1 5 .1.f4 1.b7 pawns will be a liability. In an ideal world,
1 6.fe 1 b6 1 7.b3 Axf3 1 8.gxf3= (Postny) . he dreams of securing the fS-square as a
base of operations from which to launch a
1 5 Ab7 16JUe1 %Yd7 17.gxe6 %Yxe6
..
devastating attack. I hope to show you how to
 Chapter 2 - Schliemann: 4.ltJc3 55

prevent such a nightmare from ever becoming
a reality in your games.
1 0 e5
Black begins by posting his knight on an
active, central location. At the same time he
prepares to capture the doomed e6-pawn as
well as to bolster his centre with . . . c6. Please
also note the misplacement of the bishop on
b5.
1 1 .0-0
There does not seem to be anything better
than this natural move.
just as in the analogous posltlon after
1 3.i!fe l . The position of the second white
rook does not alter the evaluation.
1 3. d4 also achieves very little for White
after 1 3 . . . f7 or 1 3 . . . b6!?
13 ... c6 1 4.i.d3 xd3 1 5 .xd3 g4 1 6.d2
i.d7 1 7. i!e7 i!f7 1 8.i!ae 1 i.f8 1 9.i!xf7 cj;>xf7
The position is equal, although I would say
it is White who may have to be slightly more
careful. The bishop pair combined with the
extra central pawn gives Black some chances
to play for the advantage, and in Ovetchkin -
Zvjaginsev, Krasnoyarsk 2007, he eventually
managed to win.

1 1 .i.f4 Occasionally White castles long, such as

With this move White tries to misplace the after:
enemy queen. 1 1 .i.e3 c6 1 2.i.d3 Le6 1 3.0-0-0!?
1 1 . . .d6 1 3.d4 can be met by 13 ... i.g4 1 4.f3
Black must avoid 1 1 . . .c6? 1 2.ltJxe5 fxe5 i.d7 1 5 . b3 b6! ? ( 1 5 . . . i!e8 was a sensible
1 3.Le5 when he will not have time to alternative) . Black was fine and eventually
capture on b5 due to the impending e6-e7 won in Tseshkovsky - Yilmaz, Kusadasi
fork. 1 990.
Compared with variation B4 1 ) above,
8 queenside castling is less risky here as the
7 b-file is closed.
6
8
5
7
4
6
3
5
2
4
1
3
a b c d e f g h
2
1 2.0-0 xe6
1
Compared with the main line, Black has
a b c d e f g h
been obliged to capture the e-pawn with his
queen rather than bishop, but this does not Nevertheless, Black can still obtain a
appear to harm his chances. comfortable game with:
1 3.i!fe l 1 3 . . . i.g4 1 4.h3 xd3t 1 5 .xd3 i.h5 =
Nothing is really changed by 1 3.i!ae l c6 Now i n Hracek - Werle, Germany 2008,
1 4.i.d3 ltJxd3 1 5 .xd3, Hellers - Antunes, White faltered with:
Novi Sad 1 990, when I suggest 1 5 . . . g4, 1 6.g4?! i.g6 1 7.d2?!
56 Attacking the Spanish

After 1 7.b3 .te4 1 8 . .td4 a5 1 9Jhe l 14.f4 c6! 15.fxe5

hf3 20.xf3 xa2 2 1 .f5 White should 1 5 . .ta4 is well met by 1 5 . . . b6! 1 6.c3 .tg4
be able to hold the position. when Black has everything in order.
1 7 . . . .te4=t
White had some problems although he 15 fxe5!
..

eventually salvaged half a point. Of course Black had no intention of taking

1 1. ,he6 12.lLld4
The game Mills - Botsari, Manila 1 992,
continued 1 2.h3 c5 ? 1 3.c3? ( 1 3 . .tf4 d6
1 4Jfe l ) 1 3 . . . b6 with equality. Instead
on b5 immediately. Now two enemy pieces
are under fire, while the g7-bishop is suddenly
looking a lot happier!
16Jxf8t %Yxf8 17 .id3 e4!

I propose the improvement 1 2 . . . .tf7!?N, Black must of course avoid 1 7 . . . exd4??

intending . . . .th5 with a good position. 1 8.h5 with a decisive attack.

12 ...i.g4 18 .be4

Black should take the opportunity to provoke 1 8 . .te3 exd3 1 9.xd3 f6 ( 1 9 . . . .td7 20.f1
the f-pawn into moving. e7 is another way) 20.f1 g6 left White
struggling to equalise in Cisler - Rybak, Czech
13.a Republic 1 998.

8
18 .bd4t 19.Ae3
.

7
8
6
7
5
6
4
5
3
4
2
3
1
2
a b c d e f g h
1
13 .ic8!
..

a b c d e f g h
I like this move, although I suppose that
there is nothing particularly wrong with 19 ....be3t
1 3 . . . .td7 1 4.hd7 xd7 1 5 .f4 ltJc6 with This leads to a forced draw, so let us see if
balanced chances. The point of the text is there is anything better.
that the white bishop is misplaced on b5,
and will soon be forced back by . . . c6 in any There is nothing to be gained from 1 9 . . . c5
case. Meanwhile Black's extra central pawn 20.hh7t h8 2 1 .hd4t xd4t 22.h l
gives him a solid space advantage, so it almost xh7 23.f1 .tg4 24.e7t when White
seems too generous to offer a straightforward can force a draw at any moment of his
exchange of bishops. choosing.
 Chapter 2 Schliemann: 4.lik3
- 57

The only genuine winning attempt seems to

be 1 9 . . . i.xb2!? Play then continues 20.c3
(20Jf1 ?! e7! [20 . . . g7 2 1 .h5 i.e6
22.i.d3] 2 1 .i.d3 i.d7 leaves the onus on
White to prove compensation) 20 . . . .ixc3
(20 . . . hal 2 1 ..ixh7t @xh7 22.h5t=)
2 1 .i.xh7t @xh7 22J:!f1 e8 23.c2t g6
24.xc3 i.h3 25 JH2 e8

a b c d e f g
Now White chooses between the natural
Cl) 7.0-0 and the immediate attack on the
e-pawn with C2) 7.'iNe2.

Cl) 7.0-0 d4!?

a b c d e f g h With this ambitious move Black com
promises his pawn structure in order to gain
Objectively Black cannot claim any real
space. The similarity to the Bird defence
advantage here, but the game goes on.
against the Ruy Lopez (3 . . . ft:)d4) is almost too
obvious to warrant mentioning, although I
After 19 Axe3t the game Polgar - Radjabov,
would tend to regard the present variation as
.

Wijk aan Zee 2008, was agreed drawn after:

offering better equalising chances.
20.'iNxe3 dxe4 21 .'iNg5t 'iNg7 22.'iNd8t 'iNf8
23.'iNg5t 'iNg7 112-1f2 . 8.xd4
8.el i.e7 9.ft:)xd4 exd4 transposes to the
It seems fair to conclude that Black is in
main line.
excellent theoretical shape after 6.e2 d5 . We
will now move on to what I believe to be the
8 exd4
most challenging line at White's disposal.
..

Upon first impressions, it looks as though the

opening has gone badly for Black. His pawn
C) 6.xf6t 'iNxf6 structure has been damaged (if we removed all
Compared with line B4) , the alternative
other pieces except the kings then White would
6 . . . gxf6?! 7.d4 would be too risky here. The
easily be winning) , he lags in development and
white queen has not wasted any time moving
his uncastled king stands on an open file.
to e2, which makes the prospect of a check on
Fortunately the position also contains a few
h5 all the more dangerous. For instance, after
redeeming features. Most obviously, the pawn
7 . . . d6 Black would have to reckon with such
on d4 severely cramps the White position, and
violent attacking ideas as 8.ft:)g5!. Though I
makes it hard for him to find a useful role for
am all in favour of playing ambitiously with
his bishops. After the subsequent . . . c6 and . . . d5
the black pieces, one has to draw the line
Black will be able to develop his own bishops
somewhere.
on good, active squares.
58 Attacking the Spanish

Ultimately we can summarise the situation Black should have no real problems in the
by saying that Black is investing in quality of ending.
development at the expense of both time and
pawn structure. IO c6 1 1.i.d3
..

The light-squared bishop has found a decent

9Jel t i.e7 IO.YlYe2
It looks obvious to keep the enemy king
pinned down in the centre. Nevertheless,
White occasionally prefers:
1 O."lNh5t g6 1 1 ."lNh6
1 1 ."lNe5 "lNxe5 1 2J:!xe5 c6 is equal, but easier
home. Unfortunately for the first player, the
future of its brother on c 1 - not to mention
the rook on al - is an altogether different
proposition.
1 t. dS

to play for Black.

1 1 . . .c6 1 2.i.f1
8
Black should be alright here, although care is
required. I suggest: 7
1 2 . . . @f7! 6
This is an important finesse. Instead after
1 2 . . . d5 1 3 .d3 @f7 White can make a 5
powerful exchange sacrifice with 1 4J:!xe7t! 4
"lNxe7 1 5 .i.d2 i.e6 1 6Je 1 .
3
2
1
a b c d e f g h
12.b3
In Zeleic - Lalic, Rijeka 2009, White instead
tried:
1 2.f4 @d8
12 . . . @f7!? 1 3.c4 dxc3 1 4.dxc3 i.d6 also
a b c d e f g h
looks fine for the second player.
Black's position is quite precarious. In
Kupreichik - Jonkman, Groningen 1 995, he
slowly lost his pawns after 1 6 .. J:!ae8 1 7.i.f4
"lNfB 1 8."lNh4 "lNe7 1 9.i.g5 "lNd6 20."lNxd4
i.f5 2 1 .i.f4 "lNd7 22.i.e5 hfB 23."lNxa7.
1 3.d3
Compared with the previous note, the point
behind delaying the advance of the d-pawn
is seen after 1 3.xe7t "lNxe7 1 4.d3 when
1 4 . . . d6! allows Black to cover his dark
squares more easily. a b c d e f g h
1 3 . . . i.fB 1 4."lNf4 "lNxf4 1 5 .i.xf4 i.c5 1 3.c4
 Chapter 2 Schliemann: 4.ltJc3
- S9

1 3.b3 also fails to create any problems after Black is able to leave his bishop en prise
1 3 . . . i.d7 1 4.i.b2 i.cs 1 S .'lWf2 f8 1 6.g3 thanks to some specific tactical resources.
as . Black easily manages to develop his Players who do not wish to give their opponent
pieces. the option of forcing the draw (as in the note
1 3 . . . dxc4 1 4 . .txc4 i.d6 1 S .d3 rtic7 1 6.i.d2 to White's 1 Sth in the main line below) may
i.d7 1 7.'lWf2 ae8 1 8.e4 cS 1 9.ae 1 hf8 wish to investigate 1 2 . . . rtid8!? intending to
complete development by . . . i.d6, . . . rtic7 etc.
8 On the other hand, 1 2 . . . rtif7?! 1 3.i.b2 i.d6
7 is well met by 1 4.c4!.
6
13.%Yxe7
5
If White does not take up the challenge
4 then Black will obtain an easy game, as can be
3 seen after 1 3.i.b2 i.cS intending . . . i.d7 and
2 . . . ae8.
1
13 ...%Yxf2t 14.hl i.h3!
a b c d e f g h
Black tosses another log onto the fire.
White went on to win the game, but at this
stage the position is balanced. 15Jgl
A draw could be agreed after 1 S .gxh3 'IWf3t.
It is also worth mentioning that 1 2.c4 can be
met convincingly by 1 2 . . . 0-0!: 15 Jae8

8
8
7
7
6
5 6
4 5
3 4
''''''''''" , , , , , ,

2
3
1
2
a b c d e f g h

1 3.cxdS ( 1 3.'lWxe7? 'lWxf2t 1 4.rtih 1 i.d7! 1

leaves White in trouble, e.g. 1 S .'lWe2 'lWxe2! a b c d e f g h
1 6 . .txe2 ae8 winning) 1 3 . . . cxdS 1 4.b3
16.%Yxf8t
( 1 4.'lWxe7? 'lWxf2t 1 S .rtih 1 i.d7! is similar to
1 6.i.a3?? xe7 1 7 . .txe7 was played in
the above note) 1 4 . . . i.cS 1 S .i.b2 i.f5 Black's
Virovlansky - Skrivanek, St Petersburg 1 998,
activity more than makes up for the doubled
at which point Black missed 17 ... f7! 1 8.i.gS
d-pawns.
h6 1 9.af1 .txg2t winning.
12 0-0!
16 %Yxf8 17JU1 .bg2t 18.xg2 %Yd6
.

.
60 Attacking the Spanish

Black's slight material deficit is balanced 23 c4 24.i.f5 g6 2S ..tg4 flYeS

.

by the exposed position of the enemy king. 25 . . . h5 26 . .tf3 'lWe5 was a more forcing route
Provided he does not allow White to coordinate to the same position.
his army in a harmonious way, he should be
able to maintain the balance. Before moving 26..la hS 27.ge2 flYgst 28.hl flYf5
on I will briefly mention that 1 8 . . . 'lWe7!? looks Black creates a threat on every turn.
like a playable alternative.
29.g2
19 ..ih2 cS 20Jlfl U'8!
It makes a lot of sense to exchange a pair of
8
rooks before White can double on the f-file.
20 . . . b5 2 1 .i.xb5 'lWg6t 22.c;t>h l 'lWe4t 23Jg2 7
is unclear according to Tseitlin, but it looks to 6
me as though White is almost winning.
5
21 JlafJ. gxflt 22.gxfl 4
Thanks to the rook exchange White's
3
attacking potential has been reduced, while
his own king will have a harder time escaping 2
checks.
1
a b c d e f g h
8
29 ...flYgSt
7
Black cannot continue the game with
6 Tseitlin's suggestion of 29 . . . g5 on account of
30J:e8t c;t>h7 3 1 .i.xd4 g4 32.i.d l .
5
4 30. f1 flYf5=
3 Now the 'correct' conclusion of the game
would be 3 1 .c;t>g2 'lWg5t followed by a swift
2 handshake. Instead, in Kalegin - Tseitlin,
1 USSR 1 986, White blundered with 31.gfl??
flYxc2 and went on to lose.
a b c d e f g h
22 ... hS! C2) 7.flYe2!
Black makes use of a tactical opportunity I believe this to be the most critical test of
to gain additional space on the queenside. It 5 . . . lLlf6. White simply intends to capture on
is important that White is not given time to c6 followed by e5.
coordinate his pieces effectively.
7...ll.e7
23.h4 The gambit approach is firmly established
White is obviously not helped by 23.i.xb5? as Black's best. It is pointless to defend with
'lWg6t followed by 24 . . . 'lWxc2. 7 . . . d6? as 8.d4 leaves Black struggling.
 Chapter 2 - Schliemann: 4.ltJc3 61

8 .bc6
. 1 2 . . . VNg6!
S.O-O allows Black to choose between This is the clever point behind Black's
transposing to variation C l ) with S . . . ltJd4!? previous move. At this point the game Volaks
9.lDxd4 exd4 l OJ!e l c6, or to the main line - Auzins, corr. 1 994, continued:
below with S . . . O-O 9.hc6 dxc6. 1 3.VNg3
Perhaps White can improve with 1 3.cxb4,
8 dxc6
... although 1 3 . . . VNxg2 1 4J!gl VNxf3 1 5 .i.e3
S . . . bxc6!? is an interesting alternative that eS looks like decent value for a pawn.
deserves attention. White can capture the 1 3 . . . VNxd3 1 4.lDe5 VNe4t 1 5.i>dl
e5-pawn in one of two ways: Although White went on to win, Black could
have obtained a fine position with:
a) 9.VNxe5 has been the most popular move 1 5 . . . d6!N 1 6.e l VNd5t 1 7 . .id2 .ia5
historically, although it has been eschewed by White's extra pawn looks less important than
the world's top players over the past couple of his awful king position.
years. I do not know if there is a specific reason
for this, so will merely present what I believe to b) 9.lDxe5 has been much more popular over
be one of Black's better responses. the past couple of years.
Play proceeds with:
9 . . . VNf7 1 O.VNxc7 9 . . . 0-0 1 0.0-0 VNe6 1 1 .e l
Capturing the second pawn is the most White prepares for a queen exchange.
principled response. 1 1 . . . .ic5
1 0 . . . 0-0 l 1 .d3 .ib4t! In Karjakin - Zvjaginsev, Dagomys 200S,
Black was unsuccessful after 1 1 . . . .ih4
8 1 2.lDf3 VNxe2 1 3.xe2 .if6 1 4.d3 d6 1 5 .h3
7 c5 1 6.b l a5 1 7.a4 .id7 1 S.b3 abS I 9.lDd2
6
feS 20.xeSt xeS 2 1 .i>f1 d5 22 . .ib2
i.xb2 23.xb2. White went on to convert
5
his extra pawn smoothly.
4 1 2.lDf3 VNxe2 1 3.xe2 d6
3

a b c d e f g h

This looks better than 1 1 . . . .if6, 1 1 . . . .idS

and 1 1 . . .VNe6t, none of which seem to be
sufficient for equality.
1 2.c3
Perhaps White could consider 1 2. i>f1 ,
although in that case it will clearly take him
a b c d e f g h
a long time to catch up on development.
If Black is not happy with this position This is an important position. The question
then he should be playing a different is whether Black's two bishops and better
opening. mobility can provide enough for a pawn.
62 Attackin g the Spanish

White has two choices: 10.0-0!

a) Leko - Radjabov, Monte Carlo 2007
proceeded with 1 4.d3?! i.g4. White has no
chance of an advantage, and an equal ending
was soon reached after: 1 5 .i.e3 i.b4 1 6.a3 i.a5
1 7.b4 i.b6 1 8.hb6 axb6 1 9.e7 hf3 20.gxf3
f7 2 1 .xf7 @xf7
b) If White wants to play ambitiously then he
should prefer 1 4.d4 i.b6, and now:
b I ) 1 5 .e7 i.g4 1 6.i.e3 hB 1 7.gxf3 xB
1 8.c4 @f8 1 9.e4 f5 was okay for Black in
Kharlov - Vuilleumier, Dresden 2007.
b2) 1 5 .c3 i.g4 1 6.e3 c5 was played in Chan
Wei Xuan - Lyell, Beijing 2008. Here White
could have played 1 7.dxc5 hc5 1 8.ttJd4;!;; .
b3) 1 5 .a4!? looks promising. Solomon -
Lyell, Queenstown 2009, continued 1 5 . . . i.g4
1 6.e4 hf3 1 7.gxf3 a5 1 8.a3;!;; . White went
on to convert his advantage.
This is White's most accurate move order,
keeping the d-pawn's options open until the
opponent's intentions have been revealed.
1 0.d3 0-0 1 1 .0-0 reaches the note to White's
1 1 th below.
Instead 1 O.d4 gives Black an additional
opportunity in:
1 0 . . . 0-0-0!?
1 0 . . . 0-0 1 1 .0-0 reaches the main line.
1 1 .i.e3 c5
Also possible is 1 1 . . .i.d6 1 2.f4 he5 1 3.dxe5
0 3.fxe5 g6) 1 3 . . . g6 1 4.c1 h5 1 5 .0-0
with good chances of equality, as in the
game Wessels - Freiberger, Germany 2005.
Compared with the main line, the black
monarch has to feel safer on the queenside.
1 2.0-0-0 cxd4 1 3.hd4 g5t 1 4.@b 1 xg2
1 5 .hg1 d5

8 . . . bxc6!? is certainly playable. However, at

the present time it appears that Black still has
some problems to solve in line b) above.

a b c d e f g h
8
Black was okay in Solodovnichenko -
7 Ponkratov Naberezhnye, Chelny 2008.
6
10 ... 0-0
5 The tactical justification for White's last is
4 revealed after 1 O . . . hc2? l 1 .d3 0-0-0 1 2.e 1
i.d6 1 3.g4t @b8 1 4.i.g5, winning material
3
for insufficient compensation.
2
1 l 1.d4
1 1 .d3 is slightly less accurate in view of
a b c d e f g h 1 1 . . . ae8.
 Chapter 2 Schliemann: 4.ltJc3
- 63
a b c d e f g h
With the following possibilities:
1 2.lDc4 b5!? 1 3.lDe3 i.d6 is unconvincing for
White.
1 2.i.f4 (Nijboer - Degraeve, France 2007)
1 2 . . . i.c5!N leads after 1 3Jae 1 i.d4 1 4.c3
he5 1 5 .he5 g6 1 6Jd 1 i.h3 1 7.f4 i.g4
to a position in which White can probably
maintain equality but nothing more.
After 1 2.f4 i.d6 White has to waste time with
1 3.d4, admitting that his 1 1 th was a mistake,
although unfortunately the present position is
not one in which a single tempo is likely to
alter the evaluation in a pivotal way.
1 1. ..Ad6
Black clears the e-file while giving himself
the option of swapping down to an opposite
coloured bishop position at a moment's notice.
12.c3
An important alternative is:
1 2.f4
Black should respond with:
1 2 . . . i.xe5
1 2 . . . c5 is conveniently met by 1 3 .i.e3.
1 3.dxe5 g6
I must emphasise that if Black is going to
exchange on e5, then it is crucial that he
chooses the right moment to do so. This
will usually mean waiting for white pawns
to appear on both d4 and f4. This will
ensure that, regardless of which way White
recaptures, his pawns will be securely
blockaded on dark squares where they will
restrict his own bishop.
If White is given time to develop his bishop
on f4 (and subsequently g3) , exchanging
on e5 will become a lot less palatable as the
white bishop would replace its colleague on
a superb square.
Finally, in the event that White recaptures
on e5 with the d-pawn, he will obtain a
kingside pawn majority of four versus two.
In this case it is essential that Black be in
a suitable position to prevent his opponent
from mobilising the kingside pawns. We will
now witness a demonstration of how he can
frustrate his opponent's ambitions.
1 4Jf2 ad8 1 5 .i.e3 d5
a b c d e f g h
From this position Black has good chances to
establish a successful light square blockade.
Here we can see another reason why the
second player should wait for a white pawn
to appear on f4 before exchanging on e5. If
the pawn were one square further back then
it would be easy for him to prepare the g4
advance, crushing Black's dreams of a light
square blockade. By contrast, in the diagram
position Black will, as a rule, almost always
answer h3 with . . . h5. An example is Oral
- Jonkman, Pardubice 1 996. That game
continued:
64 Attacking the Spanish
1 6Jk 1 fd8 1 7.ff1 h5 1 8.f2 a6 1 9.h3 f8
20.h2 h7 2 1 .fd 1 fd8 22.xd5 xd5
a b c d e f g h
The precise moves here are not too important.
What matters is that Black constantly
maintains the all-important blockade.
23.e l ! ? g6!
Black avoids 23 . . . hc2?! 24.e6 when his
defence would be more difficult. Instead
Jonkman sticks to the plan, and is never in
any real danger. For the sake of completeness
I will leave you with the remainder of
the game, which provides an excellent
demonstration of how to hold such a position
with the black pieces.
24.c3 .te6 25 . .td4 d8 26.a3 f8 27.h4
f5 28.fI .td5 29.f2 f7 30.d8t h7
3 1 .h4 g8 32.b4 b6 33.g3 d3t 34.h2
f5 35 .g3 e7 36.e3 e6 37.e2 g6
38.c2 .te4 39.a2t h7 40.e6 xe6 4 1 . .ie5
g6 42.d2 .td5 43.c2 .te4 44.d2 .td5
45.d4 b5 46.hc7 e6 47 . .te5 g6 48 . .ic7
e6 49.d2 g8 50.e2 xe2 5 1 .xe2 e4
52.xe4 he4 53.g4 hxg4 54.hxg4 g6 5 5 .g3
f7 56.h4 f6 57 . .te5t e6 58.g5 .tc2
59 . .td4 .td3 60 . .ta7 .tc2 6 1 .h4 i.d3 62 . .tb8
.tc2 63.g5 .tb 1 64.h6 f7 65.h7 .tc2
66 . .tc7 .tb 1 67.h6 .tc2 68.f5 gxf5 69.g5 f4
70.hf4 .tg6 7 1 . .te3 .tc2 72 . .td4 .tg6 73.i.h8
.tc2 74.h5 .tg6t 75.g4 e6 lI2-lI2 .
Let us now return to the main line. One
advantage of 1 2.c3 is that White's bishop
enjoys greater freedom.
12 .ie6!?
...
It is important for Black to prevent the
manoeuvre .tf4-g3. If White were allowed to
accomplish this then he could continue to
consolidate his extra pawn while Black would
be starved of any meaningful counterplay.
Meanwhile the exchange of bishop for knight
on e5 would be undesirable for reasons
described previously. There follows a summary
of the alternatives. I do not consider any of
them to represent an improvement for Black
over the main line. However, I do believe that
by playing through these lines you will improve
your understanding of what both sides are
trying to achieve.
a) 1 2 . . . c5
This is well met by:
1 3 . .tf4
1 3.f4 cxd4 1 4.cxd4 e6 is not convincing
for White, and 1 3.e 1 .te6 1 4.f4 ae8 gives
a balanced position.
1 3 . . . ae8 14.fe 1 cxd4 1 5 .cxd4
Despite the isolated d-pawn and Black's pair
of bishops, the second player does not have
enough compensation. A couple of games
have continued:
1 5 . . . .te6 1 6 ..tg3 c6
a b c d e f g h
Curiously the Greek GM Vasilios Kotronias
has encountered this position twice. In
Stefansson - Kotronias, Gausdal 1 990, White
obtained a clear plus after 1 7.d2 d8 1 8.c3
 Chapter 2 Schliemann: 4.llJc3
- 65

i.c7 1 9.a4 .tb6 20.llJf3! i.g4 2 1 .i.h4. Kotronias
obviously learned his lesson, because two years
later he switched to the opposite side!
The game Kotronias - Vouldis, Greece 1 992,
saw him temporarily return the pawn but
retain the advantage after 1 7.c2 i.b4 1 8Je3
d8 1 9.1lJd3 xd4 20 . .te5 h4 2 1 .a4 a5
22.a3 b5 23.c2.
b) After 12 . . Jae8 1 3.i.f4 there appears to be
nothing better than 1 3 . . . c5 reaching the above
note.
c 1 ) 1 3 . . . f7 1 4.g3 ae8 1 5.i.f4 e6
( 1 5 . . . e2 1 6.b3 does not change the assessment
of the position) 1 6 . .te5 g6 1 7.e3 with
good winning chances, as in the game Galkin
- Bezgodov, Russia 1 998, although Black was
eventually able to salvage half a point.
c2) 1 3 . . . xe5 1 4.dxe5 ad8 1 5 .e 1 @f7
1 6.f3

c) The immediate 1 2 . . . he5?! looks premature

in view of 1 3.xe5!'

This looks best, although White can also

maintain an edge with 1 3.dxe5, e.g. 1 3 . . . g6
1 4Jd 1 ( 1 4Je 1 is well met by 1 4 . . Jad8;
White should not allow Black to dominate
a b c d e f g h
the d-file) 1 4 . . . i.g4 1 5 .c4t @h8 1 6Jd4
( 1 6Je 1 .tf3 is obviously satisfactory for Black) Black has serious problems. He is practically
1 6 . . . .tf3 1 7.fl f5 two pawns down, and his only chance to save
the game is associated with the presence of
8 opposite-coloured bishops. It is crucial that
7 the f-pawn has not yet advanced to f4, as
now White will have no problem mobilising
6
his kingside pawns with g4 (which can be
5
supported with h3, should Black attempt to
4 prevent it with . . . h5) .
3

2 c3) 1 3 . . . ae8 1 4.xf6 ( 1 4.g3 e2 is less

1
a b c d e f g h
1 8.@h l ! (without this Black obtains a
comfortable position, e.g. 1 8 . .te3 xe5 and
1 8.f4 af8 1 9.xf5 xf5) 1 8 . . . c5 1 9.h4!
( 1 9.f4 .tc6 20.xf5 xf5) 1 9 . . ..tc6 20 . .tf4
d8 2 1 .i.g3 d2 22.b4 cxb4 23.xb4 f7
White keeps a slight advantage, although Black
should be able to hold.
good. Note that White would be ill-advised to
continue: 1 5 .xc7? .td3 1 6.f3 xg2t 1 7.@xg2
hfl t 1 8.@gl xf3+) 1 4 . . . xf6 1 5 . .te3
@f7t Black faces an unenviable defensive task,
although I believe that it should be possible for
him to hold the draw.
13.f4
1 3 . .td2 c5 1 4.f4 cxd4 1 5.cxd4 f5!? offers
Black decent play on the light squares.

After 1 3.xe5! Black may try:

66 Attacking the Spanish

8 8
7
7
6
6
5
5 4

4 3

3 2

1
2
a b c d e f g h
1
23.f2 xg4 t 24. xg4 hg4
a b c d e f g h Black had equalised the material, and the
So far we have been following the game game was eventually drawn after 68 moves.
Kramnik - Radjabov, Monte Carlo 2007,
which continued: In view of the improvements available to White
1 3 . . . ae8 1 4 . .te3 f5 1 5.b3 on moves 1 9 and 20, I think that the correct
path for Black would have been:
8
7
13 .be5! 14.fxe5
.

1 4.dxe5 g6 1 5 . .te3 ad8 should be

6
compared with the note to White's 1 2th. Once
5 again Black has good chances of achieving a
4 successful blockade. The inclusion of the move
3 c2-c3 is less than helpful for White, as the d3-
2
square may require protection.
1
14 ...VHxfl t 15.VHxfl gxfl t 16.cj;lxfl gmt
a b c d e f g h 17.cj;lgl
Black has failed to equalise from the opening.
His position remains quite solid, but it is not 8
easy to suggest a useful plan for him.
1 5 . . . a5 1 6.ae 1 a4 1 7.c4 axb3 1 8.axb3 h5 7
1 9.h3 6
This may have been the right time to
5
improve the bishop's position with 1 9 . .td2!?,
e.g. 1 9 . . . c2 ( 1 9 . . . h4 20 . .tc3) 20.e3 h4 4
2 1 .h3!? when the h4-pawn becomes a target.
3
1 9 . . . h7 20.g4?
20 . .td2 was still possible, as suggested by 2
Postny. 1
20 . . . hxg4 2 1 .hxg4 he5 22.dxe5 g6
a b c d e f g h
 Chapter 2 Schliemann: 4.ltk3
- 67

We have been following the analysis ofPostny

who concludes by stating: "Black's chances for
a draw are higher than White's chances to win".
I would tend to go even further and say that
White's winning chances are very low indeed,
and Black should have little trouble securing
the half point.

Conclusion

Virtually all of the variations considered within

the present chapter have featured plenty of
rich and dynamically unbalanced positions.
By following the recommended repertoire
you will, at the very least, be forcing your
opponents to confront unusual problems that
they would not typically expect to encounter
in the Ruy Lopez.
In variations B4) and C l ) the material
balance remained equal, but Black was still
able to play ambitiously by seizing additional
space in return for a slight compromising of
his pawn structure.
In variations B2) , B3) , and C2) we saw
how Black could fight for the initiative by
sacrificing his e5-pawn (and perhaps, in the
first two cases, an additional queenside pawn
to boot) . Our analysis showed that, especially
in lines B2) and B3) , White's position could
quickly become quite precarious.
The critical test of our repertoire came in
variation C2) after 7JWe2! i.e7 8.hc6 dxc6
9.ttJxe5. Black must certainly demonstrate
some precise knowledge here, but I believe his
position to be quite defensible. It is important
for the second player to understand when to
swap down to an opposite coloured bishop
position, and I hope that after absorbing the
contents of the present chapter the reader will
feel equipped to make that decision. When
carried out correctly, this should lead to a
position that the second player ought to be
able to draw without too many problems.
Chapter 3

Schliemann: Minor Lines

7
6
5
4
3

1
a b c d e f g h

l .e4 e5 2.tLla tLlc6 3.i.b5 f5

Introduction page 70
Strategic Themes page 70
Theoretical highlights page 70
Theory page 7 1
A) 4.exf5 ?! page 7 1
B) 4.Wfe2 page 72
B 1 ) 1 0 .d3 page 74
B2) 1 0. e 1 page 75
C) 4 . .hc6 page 77
D) 4.d4 fxe4 page 8 1
D 1 ) 5 . .hc6 page 8 1
D2) 5 . ttJ xe5 page 83
70 Attacking the Spanish

l.e4 e5 2.ttla ttlc6 3 ..tb5 f5 5.ttlc3 ttlf6 6.e2 .id6 7.d4!?

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
a b c d e f g h a b c d e f g h
Introduction The above diagram shows a critical position
The present chapter is concerned with from variation C) , coverage of which begins on
White's alternatives to the two main lines of page 77. Fortunately I found a strong novelty
4.ltJc3 and 4.d3. These range from the meek on move 1 1 , which ensures Black of at least
(4.exf5 ?!) , to the aggressive (4.d4) , to the equal chances.
strategic (4.,ixc6) , to the highly unorthodox
(4J&e2!?) . None of the options presented here A dangerous piece sacrifice
should worry Black unduly, although each 1 .e4 e5 2. ttl a ttlc6 3 .ib5 f5 4.d4 fxe4 5. ttlxe5

has the potential to cause problems against an ttlxe5 6.dxe5 c6 7.ttlc3!? cxb5 8.ttlxe4
unprepared opponent.
8
Strategic Themes
Over the course of this chapter we will be 7
investigating four very different variations, 6
each with their own unique characteristics. For
this reason it makes little sense to attempt to 5
offer generic advice. However, I believe that the 4
reader will be able to learn everything he needs
3
to know about each line from the theoretical
section. 2
1
Theoretical Highlights
There follows a brief preview of some of the a b c d e f g h
most interesting theoretical variations that we Usually in the Schliemann it is Black who
will encounter over the course of the present attempts to seize the initiative through material
chapter. sacrifices. In the present position White is
trying to turn the tables! Fortunately Black
Tactics in the centre should be able to emerge with the advantage
1 .e4 e5 2.ttla ttlc6 3 . .ib5 f5 4.hc6 dxc6
 Chapter 3 Schliemann: Minor Lines
- 71

if he responds correctly. Most o f the correct Theory

defensive methods have been known for some
time, although I was still able to find a number 1.e4 e5 2.fila filc6 3 .lb5 f5

of minor improvements in order to maximise We now consider the following four moves:
Black's chances. A) 4.exf5?!, B) 4.We2!?, C) 4.hc6, and
D) 4.d4.
Bishop pair versus pawn structure
1 .e4 e5 2.fila filc6 3.i.b5 f5 4.d4 fxe4 The exceedingly rare 4.0-0?! fxe4 5 .hc6 dxc6
5.hc6 dxc6 6.filxe5 6.lDxe5 can be strongly met by 6 . . :d4!. Rafid
Kifl - Bruchmann, Cologne 1 993, continued
8 7.Wfh5t (Black is also better after 7.lDg4 h5
8.lDe3 i.e6 intending ... 0-0-0) 7 ... g6 8 .lDxg6?!
7 hxg6 9.Wfxg6t (the queen on d4 defends h8)
6 9 ... @d8 when White had very little to show
for the piece.
5
4 A) 4.exf5?!
3 This feeble move gives Black good chances
to seize the initiative. The missing f-pawn can
2 usually be regained in due course.
1
4 ... e4 5 ..bc6
a b c d e f g h White is more or less forced to make this
In this variation Black's bishop pair should exchange in order to avoid what would be a
provide plenty of compensation for his slightly humiliating knight retreat. 5 .Wfe2 merely leads
compromised pawn structure. In most games to a transposition after 5 . . . Wfe7 6.hc6 dxc6.
he has developed the king's bishop on e7 in
order to take the sting out of the i.g5 pin. 5 dxc6 6.We2

Instead in line D 1) I suggest the more active, No better is 6.lDe5 hf5=t.

yet seldom played, development of the bishop
on d6, which seems to offer Black good 6 We7 7.fild4 filh6!
.

chances.
8
Generally speaking, the variations presented in
this chapter are theoretically fine for Black. At 7
the same time it is important to know how best 6
to handle them. After familiarising yourself
with the contents of the following pages, you 5
will have every reason to feel confident should 4
you encounter any of these lines over the
3
board.
2
1
a b c d e f g h
72 Attackin g the Spanish

It is not often that the knight visits this square
in the Schliemann. In the present position the
text is the perfect way to regain the pawn.
8.0-0
8.tDc3 tDxfS 9.tDb3 a5 1 O.a4 .!e6 1 1 .0-0
,ixb3 1 2.cxb3 0-0-0 was uninspiring for
White in Vetter - Leisebein, e-mail 200 1 .
8.h5t is also unpromising after 8 . . . tD f7 (there
is nothing wrong with 8 . . . f7 although Black's
initiative has the potential to become more
potent with the queens on the board) 9.tDc3
(9.g4 g6 did not help White in Cherner -
1 1 . . .f6 also gave Black a comfortable game
in Rosek - Leisebein, e-mail 2000. The point
of the text is that, by supporting the knight on
f5, Black turns the idea of . . . ,ixb3 into quite a
serious positional threat. I would rate Black's
position as preferable, and offer the following
plausible continuation.
12.tLld4
After 1 2 .d3 ,ixb3 1 3.cxb3 xe4 1 4.dxe4
tDd4 1 5 .tDc3 tDxb3 Black's sounder queenside
structure gives him the advantage.
12 tLlxd4 13.'i'xd4 .ig7 14.'i'e3 0-0
..

Jasinski, e-mail 2000) 9 . . . e5 1 O.tDce2 .!c5
1 1 .tDb3 .!b6 when the white position is a
sorry sight.
8 ... tLlxf5 9.tLlb3
It is difficult to say whether White should
prefer the text over 9.tDxfS .!xf5 1 O.tDc3 0-0-0
when Black also enjoys an excellent position,
e.g. 1 1 .e 1 f7!? ( 1 l . . .e8 is fine, but the
text may be even better) intending to meet
1 2 .tDxe4 ,ixe4 1 3.xe4 with 1 3 . . . .!c5 when
the f2-pawn will fall.
9 a5! 10.a4 .ieG 1 1 .'i'xe4
..
1 1 .tDc3 transposes to 8.tDc3 (note to White's
8th above) after 1 1 . . .,ixb3 1 2.cxb3 0-0-0.
1 1 g6!?N
. plan, and quite frankly he should not even try,
8 well enough.
7
Black could even consider queenside castling,
but the text is more straightforward.
15.tLlc3 gae8
Black has excellent compensation for the
pawn, with a lead in development and superbly
placed pieces.
B) 4.'i'e2
What's this - the Worrall Attack versus the
Schliemann? It appears terribly unnatural for
White to allow his queen to be drawn into the
centre, and it is hardly surprising that the text
has seldom been played. At the same time, this
move does have one clear purpose: White is
simply and unashamedly playing to win the
e5-pawn. It is not easy for Black to prevent this
as the typical gambit approach seems to work
4 fxe4
.

6 4 . . . tDf6?! is not such a good idea as after

5 5 .exf5 e4?! White can simply play 6.d3 with
advantage.
4
3 5.'i'xe4
Black has no problems after:
2 5 .,ixc6 dxc6 6.xe4 .!d6 7.tDxe5 tDf6 8.e2
1 0-0
a b c d e f g h
 Chapter 3 Schliemann : Minor Lines
- 73

8
6 . . . i.d6!? is also possible. This move would
definitely turn out well should White elect to
7
capture on c6 and eS, as the bishop would be
6 much more active on d6 than e7 in the resulting
5 position. Indeed, 7.,bc6 dxc6 B .lDxeS 0-0
4 would transpose directly to the note to White's
3 Sth above. However, if White refrains from
this show of gluttony then it will not be easy
2
to develop the cB-bishop.
1 L-L-

a b c d e f g h After the text White must decide whether or not

One does not have to be Mikhail Tal to see
that Black has excellent compensation for a
pawn here. Play may continue:
9.d4 i!eB 1 0.0-0
l O.lDc3?? is refuted by 1 O . . . cS!.
1O ... cS!
1O . . . ,beS 1 1 .dxeS d4 is also possible, e.g.
1 2.i!e l lDg4 regaining the pawn. The text is
more ambitious.
1 1 .i.e3 cxd4 1 2.,bd4 cS!?
1 2 . . . ,beS 1 3.heS lDg4 1 4.f4 lDxeS I S .fxeS
d4t 1 6.f2 xf2t is close to equal, with
a slightly edge to Black on account of his
superior minor piece.
1 3.i.c3 bS!
Black has a strong initiative for the pawn.
5 f6 6.%Ye2
to exchange on c6 before taking the e-pawn.
7 ..bc6
With this move White cedes the bishop pair
while also helping to develop the cB-bishop.
7.lDxeS was also less than an ideal solution.
Following 7 . . . lDxeS B.xeS 0-0 White's
queen becomes a target while his kingside is
bereft of defenders. The game Bruzon Bautista
- Gomez, Matanzas 1 997, continued with:
9.0-0 dS 1 0.d4 lDg4 1 1 .e l i.d6 1 2.h3
( 1 2.f4 c6 1 3.i.d3 f6 followed by . . . i.f5
gives Black more than enough compensation)
1 2 . . . lDh2! Black was winning the exchange,
which he subsequently converted to a full
point.

7... dxc6 8.xe5 0-0 9.0-0 .if5

8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
1
a b c d e f g h
a b c d e f g h
6...i.e7
74 Attacking the Spanish

This pOSitIOn has seldom been reached,

although there is an obvious resemblance to
variation C2) of Chapter 2 (page 60) .

In principle, the presence of an additional pair

of knights should favour Black, whose forces
may benefit from a 'non-linear' fighting unit
whereas White's second knight may struggle to
find a useful role. This does not mean that Black
stands better. However, compared with the
aforementioned line, he should enjoy a higher a b c d e f g h
probability of creating practical problems for Black has full compensation for the pawn.
his opponent. Even if White finds a satisfactory way to
complete his development (which is easier said
We will now analyse two alternatives: Bl) than done) , the presence of opposite-coloured
IO.d3 and B2) 10Jlel . bishops would mean that he could scarcely
hope for any more than a draw.
It should also be pointed out that after
l O.lt:k3?! Black can get away with l O . . . hc2!' Bl) IO.d3 ge8
For example: The rook wastes no time in occupying the
crucial e-file. It is also possible for Black to aim
8 for a set-up with rooks on e8 and f8:
7

6 In Petrushin - Shabanov, Smolensk 1 992,

Black failed to equalise with: l O . . . d4?!
5
1 1 .e3 (I would prefer l 1 .lLJd2 ae8 1 2.lLJdf3
4
d6 1 3.d4;!;;) 1 1 . . .d6 ( I 1 . . .h4 1 2.lLJc3 i.d6
3 1 3.lLJf3;!;;) 1 2.lLJc3 ae8 1 3.i.d2;!;;
2

1 A more promising approach seems to be

1 0 . . . d6!?N l 1 .lLJd2 ae8 1 2.lLJdc4 dS,
a b c d e f g h
with reasonable compensation.
I I .d3 does not trap the bishop because of
1 1 . . . i.d6!. I I .Ae3
l 1 .lLJd2 brings White no advantage
I I .d4 i.fS is simply better for Black thanks after 1 1 . . .i.d6 (also possible is 1 1 . . .dS
to his bishop pair and the opponent's weak 1 2.c4 d6) 1 2.f4 ( I 2.d4 is well met by
d-pawn. 1 2 . . . cS!) 1 2 . . .heS 1 3.fxeS d4t 14.f2
xf2t I S Jxf2 xeS 1 6.lLJf3 ee8 with an
Relatively best is 1 1 J!e 1 i.fS 1 2.lLJxc6 bxc6 equal position.
1 3.xe7 xe7 1 4J!xe7 fe8 I S .xe8t
( I S .e3 would be met by I S . . . lLJg4) I S . . . xe8 1 1 Ad6
..

1 6.f3 lLJdS : 1 1 . . .lLJdS was a good alternative, after which

a plausible continuation is 1 2.lLJd2 i.f6 1 3.f4
 Chapter 3 Schliemann : Minor Lines
- 75

he5 1 4.fxe5 xe5 1 5 .lLlc4 lLlxe3 1 6.lLlxe5

8
lLlxfl 1 7.xfl 'lWd4t 1 8.2 f8 1 9.c3 'lWd6
with equality. 7
6
12.ttlc4?
Better would have been 1 2.f4 lLld5 1 3.d4 5
(or 1 3 .'IW2 'lWf6 1 4.lLlc3 lLlxe3 1 5.'lWxe3 he5 4
1 6.fxe5 xe5=) 1 3 . . . c5 1 4.lLld2 lLlxe3 1 5 .'lWxe3
cxd4 1 6.'lWb3t i.e6 1 7.'lWd3 with roughly equal 3
chances. 2
1
8
a b c d e f g h
7
This is a logical developing move, and perhaps
6 a more theoretically challenging option than
the previous line. The drawback is that the 2-
5
pawn could become vulnerable.
4
3 10 ...,ic5!
Guarding against the threat of lLlxc6 while
2 immediately highlighting the drawback of
1 White's previous move.
a b c d e f g h
1 O . . . e8?! l 1 .lLlc3;;!; would be less accurate.
The text was played in Ficet - Harnois, White's f-pawn is a genuine target, so on this
Romans 1 999. Black missed the opportunity occasion it is the queen's rook that should head
to play: for e8.

12 .txh2t!
.. l 1.ttlc3
After the game continuation of 1 2 . . . i.f4? This enables White to defend 2 with lLldl
1 3.'lWd2 he had no real compensation for the and perhaps lLle3 to block the a7-g1 diagonal,
pawn. as well as the e-file.

13. c;t>xh2 ttl g4t 14. c;t>g3 1 1 ...d6 12.c4t

This is forced, as after 1 4.@gl ?? 'lWh4 White
can resign.
14.. J;e6
Black has a strong attack, e.g. 15.3 ttlxe3
16.ttlxe3 g5t 17.c;t>n gae8, regaining the
piece while maintaining the initiative.
B2) 10Jtel
Another possibility is:
1 2.lLlc4 'lWd7 1 3.d3 ae8 1 4.i.e3 lLlg4
1 4 . . . b5!? can also be considered, although I
prefer the text.
1 5 .'lWd2
A quiet move such as 1 5 .a3 can be met by
1 5 . . . i.g6!?, increasing the pressure against
2.
After 1 5 .h3 lLlxe3 1 6.lLlxe3 ( 1 6.fxe3? b5)
16 ... e6 Black's bishop pair and piece
76 Attacking the Spanish

activity are more significant than a pawn. 1 7.e7 is well met by 1 7 . . . tiJg8! 1 8.d7
1 5 . . . tiJxe3 1 6.tiJxe3 J.g@ b4 1 9.tiJe3 e l t 20.tiJf1 J.g6 2 1 .f3 J.d3
Black's pieces are developed very actively and 22.J.d2 xal 23.xd3 i.xb2 24.xc7 J.d4
harmoniously. 25.xc6 f8 when Black's position is slightly
preferable.
12 ... h8
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
1
a b c d e f g h
a b c d e f g h After the text, Black must decide where to
13.lDdl place his queen.
The greedy 1 3 .tiJf7t? xf7 1 4.xf7 is easily
punished by 1 4 . . . f8 1 5 .c4 tiJg4 with a 17 V!id5
..

crushing attack. This is the option that sets White the most
difficult problems. In addition to the simple
13 b5!
.. idea of capturing the knight on d 1 , Black also
Forcing White to execute his 'threat'. threatens to win material with . . . J.g6, as the
reply e7 can be met by . . . e8.
14.lDf7t
14.b3?! i.e4 is certainly not an 1 7 . . . b4 is also playable, when the main line
improvement for White. of my analysis runs as follows: 1 8.J.g5 J.g6
1 9.xc7 tiJe4 20.h4 tDxf2 2 1 .tiJxf2 hf2t
14 gxf7 15.V!ixf7 Axe2!
.. 22.h2 g4 23.xc6 f8
Black takes the time to capture an important
pawn. 1 5 . . . f8 is met by 1 6.b3 when Black
certainly has compensation, but no knockout
blow that I can see.

16.d4
White needs to catch up on development.
The d-pawn was isolated and would probably
have to be sacrificed anyway at some point.

16 .bd4 17Jle6
.

a b c d e f g h
 Chapter 3 - Schliemann: Minor Lines
White's extra material is balanced by his
unsafe king, e.g. 24Jk l .Lh4 25 ..Lh4 'iNxh4t
26.g l 'iNf2t with perpetual check.
The text leads to greater complications,
although the final outcome should remain the
same after correct play by both sides.
18 .ih6!
White must combine attack and defence.
There is no other satisfactory way of dealing
with the aforementioned pair of threats.
18 Jg8! 19.e3 ixe3 20.ixe3!
. gain some stability in the centre, specifically
White can get away with this move thanks
to a tactical nuance. The alternative 20.fxe3
ttJh5+ only helps Black.
20 .i.g6 21.Be7 ge8
. As usual Black should recapture with the
It looks as though White must either lose his
queen or give up the rook on e6. Unfortunately
he can save himself with the following mini
combination.
22.gxf6! gxe7 23.gf8t Bg8 24.gxg8t Cit>xg8
25 ..ba7
After numerous sacrifices and counter
sacrifices, the game has petered out to an equal
ending.
25 .. J!e2
8
7
6
5
4
3
2 Black can also consider:
1 5 . . . fxe4 6.ttJxe4 ttJf6
a b c d e f g h
77
The reduced material and opposite-coloured
bishops make the draw a virtual certainty. In
the event that White positions his b-pawn on
b3 and targets the vulnerable c7-pawn with his
bishop, Black can always obtain counterplay
by advancing the other c-pawn.
C) 4.ixc6
This move should not be too dangerous, but
it deserves to be treated with respect and has
been used by some strong players. Capturing
on c6 without provocation is undoubtedly a
concession of sorts. In return, White hopes to
for a knight on e4, which will no longer be
prone to harassment from the enemy d-pawn.
4 ... dxc6
d-pawn in order to facilitate the development
of the light squared bishop.
5.c3
Feeble is 5 .ttJxe5?! 'iNd4 6.ttJf3 'iNxe4t 7.'iNe2
ttJf6 8.ttJc3 'iNxe2t 9.xe2 .id6 when the
bishop pair gives Black the better chances.
8
7
6
5
4
3
2
1
a b c d e f g h
5 ... f6
This should enable him to equalise with
78 Attackin g the Spanish

careful play, although compared with the
main line the reduced central tension will
make the white position easier to handle. I
will offer some brief coverage here, as the
position could also be reached via the move
order 4.lDc3 fxe4 5 .lDxe4 lDf6 6.hc6 dxc6,
as mentioned in the note to White's 6th in
Chapter 2 on page 43.
Play continues with:
7.V!ie2 J.g4 8.h3
8.d3 should be met by 8 . . . .ib4t! 9.J.d2 he
when Black had a comfortable position in
Bojkovic - Micic, Moscow 1 994.
8 . . . he
This is the safest move. Instead 8 . . . J.h5 is
riskier in view of 9.d3 V!id5 (Compared with
the previous note 9 . . . J.b4 t works less well
here, as after 1 O.c3 J.d6 l 1 .lDg3 Black must
acquiesce to an unfavourable exchange on f3
if he is to avoid losing a pawn) 1 O.g4! J.f7.
Now in Milos - Van Riemsdijk, Sao Paulo
1 993, White obtained a clear advantage with
1 1 .c4! (In the high level encounter Vallejo
Pons - Aronian, Monte Carlo 2007, he
instead played l 1 .lDc3 and went on to lose.)
1 1 . . ..ib4t 1 2.J.d2 hd2t 1 3 .lDexd2 V!ie6
1 4.lDxe5 .
9.V!ixf3 lDxe4 1 0.V!ixe4 V!id5 l 1 .d3 J.c5
8 7
7 6
6
5
4
the draw. A glance at the database reveals a
positive score for White, but I think this has
more to do with a higher average rating than
the superiority of his position.
We may conclude by saying that 5 . . . fxe4 gives
Black good chances to equalise, although I
would definitely regard 5 . . . lDf6 as the more
principled move which sets the opponent more
difficult problems.
6.Y!Ye2
White is unlikely to benefit from 6.lDxe5
fxe4 as the opening of the position will tend
to favour the black bishops. Play may continue
7.d4 (7.lDxe4? turns out badly for White after
7 . . . lDxe4 [7 . . . V!ie7!?] 8 .V!ih5t g6 9.lDxg6 hxg6
1 0.V!ixh8 i>f7=t) 7 . . . exd3 8.0-0 J.d6 9.i!e 1 0-0
1 O.V!ixd3. White can perhaps claim equality
here, but certainly nothing more.
6 ....!.d6
6 . . . fxe4 7.lDxe4 reaches the note to Black's
5th.
7.d4!
8
5
4

3
3
2 2
1 1
a b c d e f g h
a b c d e f g h
This position has been reached in several This move is extremely double edged. Its
games, the earliest being Bellin - Wessels, appeal is obvious: White hopes to exploit the
Guernsey 1 990. As long as he plays carefully tactical vulnerability of the minor pieces on d6
Black should have little difficulty holding and f6, which can be forked by a pawn on e5,
 Chapter 3 - Schliemann: Minor Lines 79

as well as the position of the enemy king on 8

the e-file. At the same time, one should always
7
be wary of opening too many lines when the
6
opponent enjoys the advantage of the bishop
pair. 5

4
In any event, the quieter alternatives promise 3
White very little. 7.d3 0-0 leads to a balanced
2
position, while 7.exfS 0-0 B.d3 hfS already . ,=./' , , '.

1
looks somewhat more comfortable for Black.
His active piece play should easily offset the a b c d e f g h
minor weakness of the eS-pawn. Black enjoys an enduring initiative in return
for the piece. The rook on h I will not be
After the text, it is clear that concrete tactical entering the game any time soon, and
motifs will take precedence over longer term l S .E!:f1 ?
strategic considerations. Fortunately the black is refuted instantly by
position is quite robust, and we will see that l S . . . bS!
with the help of some accurate moves he can
obtain a fine position. 8 dxc3 9.exf6t

Weaker is 9.exd6t e4 1 0.dxc7 xc7, when

7 exd4!
Black must meet the challenge head on!
Anything else would lead to an inferior
position.
8.e5
This is the consistent follow-up.
B.exfSt i.e7 9.e4, as in the game Lauer -
Saathoff, Bavaria 2006, is refuted by the simple
Black will simply develop and enjoy a better
structure after white recaptures the pawn on
c3.
9 ... <t>f8 10.fxg7t
After 1 O.0-0?! xf6+ Black's posmon is
perfectly safe, while he enjoys the bishop pair
and a material advantage.
10 <t>xg7 1 1.0-0 f4!N

9 . . . 0-0+.

B.xd4
This was played in Sedina - Scetinin, Biel
1 997. Black's best would have been:
B . . . xe4! 9.f3
Less critical are 9.xe4 fxe4 1 0.xe4t e7+,
and 9.0-0 0-0 1 0.xe4 fxe4 1 1 .xe4+.
In both of these cases Black's bishop pair
gives him the advantage, irrespective of
whether or not the queens remain on the
board.
9 . . . h4t 1 O.c;t>f1 O-O! 1 1 .fxe4 fxe4t 1 2.c;t>gl
i.g4 1 3.c4t c;t>hB 1 4.i.e3 E!:aeB a b c d e f g h
80 Attacking the Spanish

This important improvement secures Black a
good game. Just look at what a difference this
small move makes for the respective bishops
on cl and c8!
Instead the game Kozhuharov - Toth, Bar
2006, saw 1 1 . . .f6 1 2 . .ig5 g6 1 3.e3 e8
1 4.xc3t with advantage to White.
12.bxc3 EleS 13.%\'d3
This looks like White's best chance to
maintain equality. The alternative is 1 3.d 1
f6 1 4.b 1 m g8 when Black stands better. His
pieces are more active, and the light-squared
bishop in particular has fantastic potential.
True, his king may be slightly exposed but
the open g-file may also become useful for
attacking purposes.
13 %\'6 14.c4!
playing for a win he may wish to consider a
different approach.
1 4 . . . i.f5 might also be playable, but after
1 5 .b3 the impending i.b2 will gain time
for White, and a subsequent ae 1 will see his
piece coordination improving.
1 4 . . . mh6!? would be my own choice if I was
playing for a full point. Play may continue
1 5 .b 1 b6 1 6.i.b2 g6, with a highly
unclear position. Neither king is particularly
safe; Black's is far from ideally placed, while
his opposite number must worry about the
open g-file as well as the long diagonal after a
subsequent . . . c5 and . . . i.b7. Finally, it should
be pointed out that endgames should tend
to favour Black in view of White's appalling
queenside structure.

Only this move will do for White! The slower 15.m3

1 4.b 1 is inadequate after 1 4 . . . i.f5 ( l 4 . . . g6!?
also looks strong) 1 5 .d2 i.e4:
8
7
6
5
4
3
2
1
a b c d e f g h
a b c d e f g h
Black's pieces are much better coordinated. The following sequence now appears forced:
It is too risky for White to try 1 6.xb7?! hf3
1 7.gxf3 g8 1 8.mh 1 mf7, as Black's kingside 15 %\'6
. 16 .lb2 .le5 17.tlxe5 gxe5

attack is extremely dangerous. IS.%\'c3

White will regain the rook to reach a level
14 %\'xal
.
endgame.
This seems to be the safest move, although it
only leads to an equal endgame after accurate Is ... cbg6
play from both sides. If Black is intent on 1 8 . . . mf7 1 9.xe5 xe5 20.he5 f3=
 Chapter 3 Schliemann: Minor Lines
- 81
19.'ilYxe5 'ilYxe5 20.he5 .ie6
The chances are balanced, and the opposite
coloured bishops will only increase the
likelihood of a draw.
0) 4.d4
a b c d e f g h
This is a very aggressive move. It is interesting
to compare the Vienna game after the opening
moves l .e4 e5 2.tlJc3 tlJf6 3.f4, when Black's
best response is widely recognised as 3 . . . d5 .
A naive observer might be tempted to remark
that in the present variation White is playing
the same position with colours reversed, except
that he must surely be helped by the inclusion
of the extra move .ib5 . Unfortunately for the
first player, we will see that this 'free' move can
be a mixed blessing.
4... fxe4
We now consider the somewhat prosaic
01) 5.hc6 along with the more principled
02) 5.tlxe5.
01) 5.,hc6
White plays it safe. He does not wish to
commit himself to a material sacrifice as in
line D2) below, but he can hardly hope to
fight for the advantage by ceding the bishop
pair voluntarily.
5 ... dxc6
This is the soundest move. However, if Black
wishes to put the opponent under immediate
pressure then 5 . . . bxc6 6.tlJxe5 h4!? could be
considered.
8
7
6
5
1
....:=::::!l_
a b c d e f g h
The idea is quite enterprising, although it
may not quite stand up to the highest scrutiny.
Possible continuations include:
a) 7.tlJc3 .ib4 8.e2 tlJf6 9.0-0 ixc3 l O.bxc3
0-0 is comfortable for Black.
b) 7.e2!
This is the critical test.
7 . . . d5
7 . . . tlJf6 8.tlJc3 d5 9.h3! is an important
idea, threatening to trap the queen with g3.
Following 7 . . . d6 8.tlJxc6 tlJe7 9.tlJxe7 he7
1 0.0-0 0-0 1 1 .c4t @h8 1 2.tlJc3 Black
does not have enough compensation for the
missing pawn.
8.c4!?
8.tlJxc6 is playable, when Black should
continue 8 . . . tlJf6 with reasonable
compensation.
8 . . . .ia6
After 8 . . . .id6 White can win material
with 9.cxd5 tlJf6 (Black is not helped by
9 . . . cxd5? 1 0.b5t) l O.dxc6, intending
to meet 1 O . . . a5 1 1 .tlJc3 .ia6 with 1 2.e3
followed by g3 or g5 .
9.c2 .ib4t 1 O . .id2 hd2t 1 1 .tlJxd2
82 Attacking the Spanish
a b c d e f g h
Black stands worse, e.g. 1 1 . . .e3 1 2.ltJd3
xf2t 1 3.xf2 exf2t 1 4.@xf2, or 1 1 . . .ltJe7
1 2.0-0 0-0 1 3.a4 f4 1 4.xa6 xd2
l S .ltJxc6 ltJxc6 1 6.xc6. White has won a
pawn, although Black may have chances to
hold the major piece endgame.
6.ttlxe5 ttlf6
6 . . . h4!? can be considered, although once
again Black will have to proceed with caution
as a subsequent . . . ltJf6 could be met by h3
threatening g3. The text is more reliable.
7 .ig5!?
.
This looks like the only way for White to put
the opponent under any pressure.
The obvious move is 7.0-0, after which 7 . . . i.d6
8.i.f4 0-O 9.i.g3 e7! gives Black an excellent
game.
a b c d e f g h
The idea is to strike at the centre with . . . cS ,
undermining the knight on eS . A possible
continuation is 1 O.ltJd2 (after 1 O.c3 cS White
has nothing better than 1 1 .ltJd2 transposing)
1 O . . . cS 1 1 .c3 cxd4 1 2.cxd4 cS when the game
is opening up nicely for the black bishops.
Before we return to the main line, please note
that the immediate 7.ltJc3 gives Black the extra
opportunity of 7 . . . i.b4.
The point of 7 . .igS!? is to wait for Black to
commit this bishop to d6 or e7 before bringing
the knight to c3 to attack the e-pawn.
7 ..Ad6!?
.
This is the most active choice. 7 . . .i.e7 is also
quite playable, when play may continue 8.ltJc3
0-0 9.0-0 i.f5 . Black is not worse here, e.g.
1 O.ltJe2 h6 1 1 .ltJg3 .ill7 .
8. ttl c3 0-0
8
7
6
5
4
3
2
1
a b c d e f g h
9.0-0
The principled 9. e2 should be met by
9 . . . e8! 1 O.hf6!? ( l 0.ltJxe4 ltJxe4 1 1 .xe4
i.f5 gives Black typical compensation for
the Schliemann) 1 0 . . . gxf6 1 1 .ltJc4 i.b4 with
roughly equal chances.
White must of course avoid 9.ltJxe4?? heS
when he loses a piece.
 Chapter 3 Schliemann: Minor Lines
- 83

9 .if5 10.e2 e8 1 1..bf6 gxf6 12.ttlc4

This is the key move. The point is that
g6 bishop retreats can be met by . . . 'I&a5t winning
the e-pawn. We now analyse 021) 7.i.e2 and
the piece sacrifice 022) 7.ttlc3. The latter is
8
undoubtedly the more principled response.
7
6 7.i.c4? has been played but after 7 . . . 'I&a5t
8.c3 'l&xe5 White simply has an inferior
5 version of line D2 1 ) , as a subsequent . . . d5 will
4 gain time by hitting the bishop.
3
021) 7.j,e2
2 If White intends to gambit the e-pawn, then
this is the only sensible option.
1
a b c d e f g h 7 a5t 8.ttlc3

The position is roughly equal, although if I White can also play:

had to choose a side I would tend to favour
Black's bishops over White's slightly sounder
pawn structure.
02) 5.ttlxe5 ttlxe5
5 . . . f6 would be the standard response
in the Vienna (i.e. in the same position with
colours reversed and the bishop on f1 instead
of b5), but the text enables Black to call into
question the position of the bishop.
8.i.d2
This enables him to force the enemy king to
move.
8 . . :xe5 9.i.c3
9.i.h5t will usually transpose after 9 . . . @d8
I O.i.c3 ( 1 0.0-0 f6 1 1 .i.e2 d5 is excellent
for Black) I O . . :g5 1 1 .0-0 f6 1 2.i.e2
d5.
9 . . :g5 I O.i.h5t @d8 1 1 .0-0 f6

6.dxe5 c6!

8
7
6
5
4 a b c d e f g h

3 Can White profit from the position of the

2 king on d8? In my opinion Black has every
reason to feel confident, as his central pawns
1 will form a powerful barrier against the
a b c d e f g h enemy pieces.
1 2.i.e2
84 Attacking the Spanish

The game Cunha Pereira - Machado, 1 3 . . . exf3 1 4.hf3 .id6

Sao Paulo 1 983, instead saw: 1 2.f4 i.cst Black stands clearly better. His extra pawn
1 3.@h l f5 1 4.g4 ( I 4.i.e2 is well met is an important asset, and his king will be
by 1 4 . . . dSFF) 1 4 . . . e6 I S .f5 e7 1 6.gS perfectly safe on c7.
xhS 1 7.f6 e6 1 8.xhS g6 1 9.e2
( I 9.h6 e3 wins) 8 'Bxe5 9.J.e3
.

After 9.0-0 f6 1 O.g3 dS 1 1 ..if4 e7+

Black will develop easily.

8
7
6
5
a b c d e f g h
4
Now instead of the game continuation of
1 9 . . . e3, Black could have obtained a large 3
advantage with 1 9 . . . dS!. 2
12 . . . dS
1
8 a b c d e f g h
7 10.'Bd2
6 White maintains the option of long castling.
5 This move was tested in Kuzmin - Stjazhkina,
4
St Petersburg 2000.
In Buenjer - Leisebein, e-mail 200 1 , White
3
failed to demonstrate compensation after:
2 1 O.i.d4 gS 1 1 .0-0 (not surprisingly, a queen
L , = J'="',

1 exchange is of no help to White: 1 1 .d2

a b c d e f g h xd2t 1 2.@xd2 dS 1 3.f3 exf3 1 4.hf3 i.d6
I S J!ae l t @f7 1 6J!hfl !!f8+) 1 1 . . .dS 1 2.c 1
White has tried a few different moves here,
xc 1 1 3.!!axc1 i.d6 1 4.f3 exf3 I S .!!xf3 0-0
without coming close to demonstrating
Black was simply a pawn up for nothing.
adequate compensation.
1 3.f4
10 Ac5!N
1 3.i.d2 f5 1 4.i.e3 ( I 4.c4? d4-+) 1 4 . . . i.d6
.

I believe this to be the most accurate move.

I S .c4 @c7 1 6.c3 dxc4+ was Pedersen -
Grynfeld, Helsinki 1 9S2.
1 1 .0-0-0
The feeble 1 3.@h l is hopeless after 13 ... i.d6
No better is 1 1 .0-0 he3 1 2.xe3 dS, or
1 4.d2 @c7, as first shown in Hoenlinger
1 2.fxe3 0-0 followed by . . . dS , with a clear
Steiner, Vienna 1 9S 1 .
advantage to Black in both cases.
 Chapter 3 Schliemann: Minor Lines
- 85

1 1. ..0-0 12 ..bc5 'lWxc5 13.'lWd6 9 ... f6

This attempt to obstruct Black's queenside
development is easily rebuffed.
13 'lWxd6 14Jxd6 e8 15Jd2 d5+
..
There is no reason why Black should not go
on to convert his extra pawn.
D22) 7.c3
This is the only critical move. The previous
variation saw White gambit a pawn for meagre
compensation, so he may as well up the stakes
in order to make a real fight of the game.
7 cxb5 8.xe4
..
Suddenly things are looking a lot rosier for
Black. His knight has entered the game, while
the c8-bishop is free to do the same. On the
other hand his king is still some way from
castling, and the pawn on d6 has the potential
to become a nuisance.
At this point White must decide how to
meet the attack on his knight. Exchanging on
f6 would be pointless, as Black would simply
recapture with the queen and then gobble the
d6-pawn.
Instead White normally chooses between
D221) 10.'lWd4 and D222) 10 ..lgS.

Occasionally White tries the cheeky move:

1 0.0-0?!
8
In this case Black should not be afraid to
7 accept the gift.
6 1 O . . . tDxe4 1 1 .h5t
No better is: l 1 .:!:!e l .if5 1 2.d5 d7
5 1 3.:!:!xe4t .!xe4 1 4.xe4t @f7-+
4 1 1 . . .g6 1 2.e5t
After 1 2.xb5t @f7 1 3 .d5t @g7 1 4.xe4
3
i.xd6 Black should convert his extra material
2 without too much trouble.
1 2 . . . @f7 1 3.xh8 tDf6
1
a b c d e f g h
White has only a single pawn for the piece,
but his initiative should not be underestimated.
His immediate plan is simply to castle and
centralise his pieces. Meanwhile it is far from
easy to suggest a convenient way for Black to
develop his pieces. That is, until you notice the
following move.

8 ... d5! a b c d e f g h
If it were not for this pawn sacrifice then
Black would be in considerable danger. White has regained a portion of the sacrificed
material, but his initiative has evaporated
9.exd6 and he will soon have to defend an inferior
White has no real choice but to accept the ending. Possible continuations include:
offer. 1 4 . .ig5 .if5 1 5 .:!:!adl .ig7 1 6.xd8 :!:!xd8+,
86 Attacking the Spanish

Frolov - Kuzmin, Cappelle la Grande 1 994, The alternative 1 2.VNxe4t VNe6 1 3.VNd3
and 1 4.i.h6 .ie6 1 5 .,ixfB VNxfB 1 6.VNxfBt ,ixd6 1 4.d l ( 1 4 . .if4 .ixf4 1 5 .fe l .ie5
xf8+, Martjukhin - Efendiyev, e-mail 1 999. 1 6.VNxb5t .id7-+) is well met by 1 4 . . . .ie7
1 5 . .ig5 ( 1 5 .VNxb5t f7+) 1 5 . . . ,ixg5
D221) 10.ti'd4 ti'd7! 1 6.e l 0-0 1 7.xe6 ,ixe6. The white queen
is outgunned by the opposing bishop pair
and rook.
8
Also playable is 1 4 . . . .ifB!?
7
6
5
4
3
2
1
a b c d e f g h
a b c d e f g h
This is not the most obvious move, but it has
been recognised as Black's strongest since well White has some compensation here, but it is
before the days of Fritz and Rybka. very hard for him to create concrete threats.
1 2 . . . VNxd6 1 3.VNxe4t
It is worth mentioning that the less After 1 3.xe4t f7 14.f4t g8 White
sophisticated 1 O . . . ltJxe4 is also far from bad, cannot force a draw with 1 5 JhfBt VNxfB
and after I 1 .VNxe4t f7 1 2 . .if4 VNe8 1 3 . .ie5 1 6.VNd5t because of 1 6 . . . .ie6! 1 7.VNxe6t
VNc6 Black is doing fine. This could be a useful VNf7+.
secondary option in case you are ever surprised 1 3 . . . f7 1 4.VNf3t
over the board and find yourself struggling to 14.VNe8t g8+ is safe for Black.
recall the details of the main line. 1 4 . . . VNf6!N
This is a significant improvement. The
l 1 .tLlxf6t previously played 14 . . . g6 is much less clear
This is not the only move to have been after 1 5 .e8!N.
tested.

1 1 ..if4? ltJxe4 1 2.VNxe4t VNe6 forces the queens

off, after which Black should have no difficulty
converting his material advantage.

A much more respectable alternative is:

1 1 .0-0
Black can maintain his advantage, but will
need to play accuracy. There follows:
1 1 . . . ltJxe4 1 2J!e l
 Chapter 3 Schliemann: Minor Lines
- 87

It seems that the most White can hope for

here is to regain one or two more pawns. The
trouble is that, even if he restores theoretical
material equality of three pawns for a
piece, Black's bishop pair will always be too
strong.
1 5 .VNh5t VNg6
1 5 . . . g6 1 6.VNd5t i.e6 1 7.VNxb5 is less clear.
1 6.VNd5t
1 6.VNxb5 VNc6-+
1 6 . . . i.e6 1 7.VNxb7t i.e7
Black is ready to complete development
and should be the strong favourite to win the a b c d e f g h
game. The game is not yet over, but Black should
win with careful play.
1 1. gxf6 12.,if4

White must support the pawn on d6. He can 0222) 10.Ag5

hardly consider 1 2.0-0? VNxd6-+ or 1 2.VNxf6?! This is the main line.
VNe6t 1 3.VNxe6t he6+.
10 VNa5t!

12 VNe6t!

a b c d e f g h
a b c d e f g h
This is a very useful move. Black not only
This disruptive check ensures that White's escapes the pin while forcing one of the enemy
king will have its own problems. pieces to retreat, but also prepares to activate
the queen along the fifth rank. We will see
13.'it>d2 'it>f7 14.hel several instances over the coming pages where
1 4.b3? hd6! wins immediately as in Schlick Her Majesty plays a key role in the defence.
- Hermann, Bad Neuenahr 1 984. With apologies for the clumsy labelling, we
now reach a final division between 0222 1)
14 VNc4 15.VNxc4t bxc4+

1 1 .i.d2 and 02222) 1 1 .c3.
88 Attacking the Spanish

D2221) I 1.W b4 12.tLlxf6t

The alternatives are no better.

1 2.0-0 lDxe4 1 3 Je 1 VBe5 1 4.f3 i.fS 1 5 .d7t

d8 is winning for Black.

1 2.VBe2 VBe5 1 3.lDxf6t gxf6 was seen in Barry

- Marshall, Cambridge Springs 1 904. Again,
White has no compensation whatsoever.

1 2.VBf3 a b c d e f g h
This can be met convincingly by: This seems to be clearest, although 1 4 . . . i.f8
1 2 . . . VBfS! also brought Black the advantage in
Already we see an important example of the Kubacsny - Seitaj, St Ingbert 1 993.
queen's mobility. 1 5 Je l t
1 3.lDxf6t After 1 5 .VBd3 f7 I see nothing better for
1 3.0-0-0 proved hopeless after 1 3 . . . lDxe4 White than 1 6Jfe 1 VBc5 transposing to the
1 4Jhe 1 VBxf3 1 5 .gxf3 f7 1 6Jxe4 hd6, main line.
Terwey - Kunisch, Ruhrgebiet 2003. 1 5 .i.g7 g8 1 6.hf6 VBfS 1 7.e 1 t ( 1 7.i.h4
1 3 . . . gxf6 1 4.i.f4 VBg4!? 1 5 .VBe3t VBe6+ allows Black to end the game in style
with 1 7 . . . xg2t!) 1 7 . . . i.e6 is winning for
Black.
1 5 . . . f7 1 6.VBd3
1 6.VBd4 i.e5 1 7.VBh4 i.fS-+
1 6 . . . VBc5!

8
7

6
5
a b c d e f g h
4
Black should convert his advantage without 3
difficulty.
2

12 gxf6 13 .if4

1

White has also experimented with the idea a b c d e f g h

of sacrificing the d-pawn: Preventing a check on c4.
1 3.0-0 i.xd6 1 4.i.h6 1 7.ad 1
The point of White's play is that the opening 1 7.i.e3 VBd6 1 8.VBc4t i.e6 1 9.VBh4 ag8-+
of an additional central file may make it 1 7 . . . i.e6!
harder for the black king to find shelter. I This is the most accurate move. White has
suggest the following new move: practically nothing to show for the sacrificed
1 4 . . . .ic7!N piece.
 Chapter 3 Schliemann: Minor Lines
- 89

13 ...Ae6 14... f7 15.0-0

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
a b c d e f g h a b c d e f g h
Black should be better, provided he avoids We have been following the game Jonasson -
any careless mistakes. Van Beers, Eupen 1 994. At this point I would
suggest:
14.Wd4
1 4.VNe2 @f7 1 5 .0-0-0 VNxa2 1 6J!he l VNc4 15 .. J:c8
is winning for Black. I see very little compensation for White
here.
1 4.0-0 can be met by 1 4 . . . @f7!.
D2222) 1 1 .tl)c3
8 This looks like White's best chance to obtain
7 an acceptable position.
6
1 1 . .. h4
5

4
8
3
7
2
1 6
a b c d e f g h 5
The king is quite safe here. Note the role of 4
the queen on as, preventing any checks on
3
h5 .
A logical continuation would be 1 5 J!e l 2
VNd5 1 6.VNe2 .!! g8 1 7 . .tg3 .tg4 1 8 . .!! ad l VNc6 1
1 9.f3 .tf5 . Black maintains his advantage and
will gradually continue bringing his pieces into a b c d e f g h
the game, starting with the queen's rook at the Once again Black continues to harass the
earliest opportunity. enemy pieces, while at the same time activating
90 Attacking the Spanish

his queen. The text uncovers an attack against 15 ... @d7 16.0-0
the bishop on g5, so White's next is forced. 1 6.ltJxe6? We5t! wins for Black.

12.i.xf6 gx:f6 13.tLld5 b3t! In Kotov - Stjazhkina, St Petersburg 2000,

I cannot overemphasise the importance
of this move! The alternative is much less
convincing:
1 3 . . . .ie6 14.Wh5t @d8 1 5 .0-0-0
White's initiative is starting to become
dangerous. he 1 is threatened, and Black is
practically forced to play:
1 5 . . . b3!
Instead 1 5 . . . Wxa2? 1 5 .he l .id7 1 6.ltJxf6
should win for White.
White tried 1 6.ltJxa8 .ixd6 1 7.axb3, at which
point 1 7 . . . We5t! would have left Black with an
obvious advantage.
16 ....hd6 17. tLlxe6
1 7. ltJxa8 was seen in Parkanyi - Semenova,
Tapolca 1 997, and now the cool 1 7 . . . bxa2!
would have been excellent for Black.
17 Ve5
.

8
7
6
5
4
3
2
a b c d e f g h
1
Now 1 6.axb3 might result in a repetition
after 1 6 . . . Wal t 1 7.@d2 Wa5t 1 8.@c l etc. a b c d e f g h
White can also try 1 6.cxb3!? c8t (after 18J e l!
1 6 . . . Wc5t 1 7.@b l .ixd6 1 8.Wf3 Black is This looks like White's best chance, although
in some danger) 1 7.@b l .ixd6 1 8.Wf3 f8 it would require strong nerves to allow the
1 9.b4 Wb5 20.he l with reasonable practical queen to capture on h2 with check.
compensation.
Instead the timid 1 8.g3 failed to put Black
14.c3 .ie6 under any pressure in Wolny - Mallee, corr.
Compared with the previous note, the 1 986, which continued 1 8 . . . Wxe6 1 9.e l Wfl
inclusion of the moves . . . b3 and c3 makes a 20.axb3, and now 20 . . . @c7 would have been
significant difference. clearly better for Black.
15.tLlc7t 18 Vxh2t 19.fl Vhl t 20.e2
..

Thanks to Black's accurate 1 3th move, it Now both kings are out in the open.
would now be hopeless for White to continue
1 5 .Wh5t? @d8 1 6.0-0-0? bxa2 1 7.@c2 a b;;W 20 Vxg2
..

1 8.xal Wxd5 winning easily. 20 . . . Wh5t!? is also at least equal for Black.
 Chapter 3 Schliemann: Minor Lines
- 91

We have been following the game Pavlovic -

Garbarino, Internet 2002. Black's accurate play
has left him with slightly better chances. His
8
king is safer, while White must always worry
7 about reaching an ending in which his knight
6 will struggle against the enemy h-pawn.

5 Conclusion
4
In the present chapter we have considered
3
four very different answers to the Schliemann
2 variation. Line A) with 4.exfS ?! (page 7 1 ) is
hardly worth mentioning as a serious option
1
for White. Variation B) with 4.'We2!? is
a b c d e f g h actually rather interesting, despite breaking
This seems to be White's only chance. one of the fundamental principles of opening
play by exposing the queen to an early attack.
22 ... c.!?c8 It is hard to see this variation becoming
22 . . . 'Wf3t looks tempting, but should only fashionable for White, but in any case it seems
lead to a draw. My analysis continues 23Je3 that Black can sacrifice the e5-pawn for enough
'WfSt 24.@e2 !he6 (24 . . . b5?! 25.d4 gave compensation.
White the advantage in Pavasovic - Srebrnic,
Ljubljana 1 992) 25.'Wxb7t @e8 26.'Wc8t @f7 The positional option of 4.hc6 dxc6 5 .c3
27.'Wxh8 'Wb5t 28.@e 1 !!xe3t 29.fxe3 .tg3t should also not worry Black. The most
30.@d2 'Wxb2t 3 1 .@d3 'Wxa1 32.'Wxh7t @f8 important thing for him to remember after
33.'Wh8t @f7 34.'Wh7t Black is unable to 5 . . . f6 6.'We2 is that he should maintain
avoid the perpetual. the tension with 6 . . . .td6, not fearing the
complications that might arise after 7.d4!?
23.c4t c.!?b8 exd4!. Following the more or less forced
sequence of 8.e5 dxc3 9.exf6t @f8 1 O.fxg7t
@xg7 1 1 .0-0, my new suggestion of 1 1 . . . f4!
8 should ensure an excellent game for Black.
7
Finally, the active D) 4.d4 (page 8 1 ) is a move
6
for which Schliemann players should definitely
5 be prepared. After the normal 4 . . . fxe4, D 1 )
5 . .txc6 (page 8 1 ) is not too threatening,
4
and I rather like the idea of developing the
3 bishop actively on d6. D2) 5 .xe5 (page
2 83) is a different matter entirely - White
elects to fight for the initiative, but more or
1 less commits himself to a piece sacrifice (the
a b c d e f g h feeble D2 1 ) 7 . .te2, as analysed on page 83, is
clearly insufficient) . Fortunately the correct
92 Attacking the Spanish

defensive methods have been tried and tested,

and my analysis has reinforced the widely held
view that Black can obtain an advantage with
accurate play.
Chapter 4

Gajewski: 1 1 .d3 and 1 1 .exd5

8
7
6
5
4
3

a b c d e f g h

l .e4 e5 2.f3 c6 3.b5 a6 4.a4 f6 5.0-0 e7

6Jel b5 7.b3 0-0 8.c3 d6 9.h3 a5 1 0.c2 d5!?

Introduction page 94
Kuznetsov - Gajewski page 9 5
Wang Hao - Gawain Jones page 9 6
Theory page 99
A) 1 1 .d3 page 99
B) 1 1 .exd5 e4 page 1 0 1
B 1 ) 1 2 .ltJg5 ? ! page 1 0 1
B2) 1 2 .i.xe4 page 1 04
94 Attacking the Spanish

1.e4 e5 2.ltla ltlc6 3 ..ib5 a6 4..ia4 ltlf6 8

5.0-0 .ie7 6J:el b5 7 .ib3 0-0 8.c3 d6 9.h3

ltla5 10 ..ic2 d5!? 7

6
8 5
7 4
6 3
5 2
4 1
3 a b c d e f g h
2 A dream position for him might look
1 something like the above. White has retained
his light-squared bishop and can develop his
a b c d e f g h queenside pieces smoothly. Meanwhile Black
Introduction has no counterplay in sight.
This aggressive move is named after the
young Polish GM Grzegorz Gajewski, who Black's kingside attack
began playing it in 2007. Black's 1 0th was not Black will sometimes advance his e-pawn with
completely new, but it had never been taken . . . e4. He would love to bring his f-pawn into a
remotely seriously at a high level. supporting role, as seen in the diagram below.

It soon transpired that Gajewski's approach 8

enables Black to fight for the initiative and
create a variety of problems for his opponent at 7
an early stage. Not surprisingly, this variation 6
immediately attracted many adherents and
quickly became a subject of intense theoretical 5
debate. Two years later we are in a better 4
position to draw conclusions, although the
3
theory is still developing. This makes it an
ideal weapon for players who like to analyse 2
at horne. 1

Strategic Themes a b c d e f g h
We will now look at a few of the general Black has the makings of a powerful
ideas for both sides. initiative.

White's extra pawn Two important games

By taking on d5 White intends to win Before we look at any theoretical lines, it is
Black's e-pawn and play his own pawn to d4, worth familiarising ourselves with the following
consolidating the material advantage. two games. Between them they illustrate many
 Chapter 4 Gajewski: I I .d3 and I I .exd5
- 95

important attacking ideas, as well as helping to In line B 1 ) we will see that 1 7 . . . g6! is even
explain how the present variation became so stronger.
popular in a short time.
We will begin with the game in which 18.,ie4 Ab7?!
Gajewski unveiled his new weapon. Strictly speaking this is not quite sound,
although the idea is a very attractive one.
Kuznetsov Gajewski
- 1 8 . . . c6 and 1 8 . . .i.e6 were objectively better.

Pardubice 2007 19.tLl5 IS 20 ..bf5 gm 21 J;e6

After 2 1 .Wfg4 Wff6 22.i.e6t h8 23.f3 tDc4
1 .e4 e5 2.tLla tLlc6 3.Ab5 a6 4 .b4 tLlf6 Black's compensation is obvious.
5.0-0 Ae7 6Jel b5 7.Ab3 d6 8.c3 0-0 9.h3
tLla5 10.,ic2 d5!? 1 l .exd5 21 gIS 22.gxd6 hd6
..

Confronted by an unexpected move, White

reacts in the most natural way. The other main
8
line is I I .d4, which will be discussed in the
following chapter. 7
6
1 1. e4
.

8 4

7 3

6 2

5 1

4 a b c d e f g h

3 Black has just two pieces for the queen, but

he is well ahead on development and the g2-
2 pawn is a real threat. The idea is wonderfully
1 imaginative, even if it is objectively not quite
sound.
a b c d e f g h
12.tLlg5?! 23.a4
White plays to preserve his bishop pair, but 23.Wfe2! looks like the best way to begin
we will see in variation B 1) that this approach the consolidation process. After the game
is flawed. continuation Black is able to whip up some
dangerous counterplay.
12 tLlxd5 13.tLlxe4 5!
..

Gajewski uses the f-pawn as a battering ram 23 ....ig3! 24.a

to break open the enemy kingside. After 24.fxg3 fl t 25.Wfxfl gxfl =Wft
26. xfl the sting in the tail can be found in
14.tLlg3 4 15.tLle4 a 16.d4 fxg2 17.tLlg3 26 . . . tDb3!.
YHd6
96 Attacking the Spanish

24...J.f4 Now Black finishes the game in style:

Black achieves little with: 24 . . . lDc4 25.Wfe2
lDde3 26.i.xe3 xf3 27 . .ic 1
25.axb5 bel 26Jxa5?
White could have retained the advantage
with 26.Wfxc1 :
26 . . . xf3 27.bxa6! .ia8 28.Wfg5! f1 t
(28 . . . lDc4 29.Wfxg2 lDf4 30.Wfxf3+-) 29.i>xg2
lDe3t 30.i>h2 h l t 3 1 .i>g3 gl t 32.i>f4
xg5 33. i>xg5 is winning for White.
26 . . . lDf4 27.Wfxf4 (after 27.Wfe3?! lDc4
28.Wfe8t f8 Black's initiative is too strong)
27 . . . xf4 28.bxa6 .ia8 29.xa5 xf3 30.i>xg2
xc3t 3 1 .i>f2 The dangerous passed a-pawn
gives White a clear advantage in this ending.
26 ... f4!
Now Black's initiative is very dangerous.
30 ... g1=t! 31.ftxgl gxa5
0-1
White resigned, as he will be forced to give
up his queen to avoid mate. Gajewski's novel
treatment of the opening would probably have
attracted attention irrespective of the outcome
of the game. However, the fact that he won so
spectacularly must have made his new system
all the more seductive to Grandmasters and
amateurs alike.
In our next game, which took place a couple
of months after the above, we see the young
English GM Gawain Jones using I O . . . d5!? to
surprise the formidable Chinese GM Wang
Hao.
Wang Hao - Gawain Jones

27.fteH Liverpool (UK - China) 2007

White cracks under pressure. 27.lDd2 was
the last chance. 1.e4 e5 2.f3 c6 3 ..t.b5 a6 4..h4 f6
5.0-0 .t.e7 6J:el b5 7.Ab3 d6 8.c3 0-0 9.h3
27 ....ixf3 a5 10.i.c2 d5 I l.exd5 e4 12.g5 xd5
Now there is no defence. 13.xe4 f5 14.g3 f4 15.e4 f3 16.d4 fxg2
17.g3 ftd6
28.bxa6 xh3t 29.cti?h2 .if4t 30.cti?xh3
8
8 7
7 6
6 5
5 4
4 3
3 2
2 1
1 a b c d e f g h
a b c d e f g h 18.ftd3
 Chapter 4 Gajewski: 1 1 .d3 and 1 1 .exd5
- 97

Up to this point we had been following the @xf7 3 1 .bxa6 hf2t 32.@xg2 hd4t
previous game, in which 1 8.Ae4 was played. With a winning ending.

18 ...g6 19.e4 Be6 20 ..ih6 f4 26.xe4 ,be4 27.Be3 ,bel 28.bxa6 Ae4
2D . . . f5 was a good alternative, e.g. 2 1 .ttJbd2 29.a7 gm 30.f4
h5 with good attacking chances. The position is approximately equal, and
was eventually drawn. For the record, the
21 .i.xf4 xf4 22.bd2 Bf7 23.a4 remaining moves were:
Another possibility was 23.g3 Ah4
24.xg2 Ab7 25.e2, although Black would 30 ...AdS 31.ge2 BfS 32Jlal AgS 33.BeS
obviously retain fine compensation here as .bf4 34.Bxf5 gxf5 3S.c4 .!.f3 36Jlxg2t
well. i.xg2 37.xg2 gaS 38.f3 Ah6 39.b4 f7
40.dS Ad2 41.a4 f4 42.cS Ae3 43.a6 e8
23 ...Ah4 24.axbS Ab7! 2S.xaS? 44.h4 hS 4S.e4 d8 46.d6 cxd6 47.cxd6
Too greedy! Necessary was 25.e2 axb5 c8 48. f3 .ba7 49.gaS b7 SO. xf4
26.xb5 with good chances to defend. A gh8 S I.gS i.b6 S2.gfS c6 S3.g6 gg8t
possible continuation is 26 . . . f5 27.xb7!? S4. f7 gg4 SS.xhS xb4 S6.gh6 d7
(27.d3) 27 ... ttJxb7 28.xa8t @g7 29.Ab3 S7.hS h4 S8.g6 ghl S9.f5 Ae3 60.gh8
d7 3D.@xg2 when White may be slightly i.d4 61.gh6 i.e3 62.gh8 xd6 63.g4
for choice, although the position remains very e7 64.a8 ggl t 6S. f3 i.gS 66.ga7t
unclear. m 67.ga6 i.e7 68.gaS g7 69.f2 gel
70J:la7 f6 71. f3 gcS 72.h6 ghS 73. g4
8
ghl 74Ja6t f7 7S. fS gfI t 76. g4 i.f6
77.ga3 g6 78.gh3 h7 79.gh2 ggl t
7 80.f3 gg6 81.e4 AgS 82.gg2 xh6
6 83Jh2t g7 84.g2 f6 8S.ga2 f7
86Jg2 ge6t 87.dS i.f6 88.ga2 geSt
5 89.c4 e6 90.ga6t fS 91 .b6 ge8
4 92. dS gel 93. c4 i.eS 94. d3 gdl t
9S.e2 gd4 96.gb3 gh4 97.d3 gh8
3
98.gbl d8t 99.c4 gd2 100.fIt e6
2 101.gbl d6 102.gb6t c7 103.g6 .id6
1
104.gg8 d7
112-1J2
a b c d e f g h Regardless of the final result, Gajewski's
2S .. Jlxe4? system once again proved its worth in securing
Black could have crowned his energetic play a winning position against a high calibre
with: opponent.
25 . . . he4! 26.ttJxe4
26.xe4 g4!! is a brilliant point. Theoretical Highlights
26 . . . xe4 27.e2 Before moving on to the main theoretical
27.d2 xe 1 t 28.xe 1 e8 29.d2 f3 section, we will briefly note some of the key
. . variations.
Is wmmng.
.

27 . . . xe2 28 .xe2 e8 29.Ab3 xe2 3D.hf7t

98 Attackin g the Spanish

White refuses the bait White wants to take the gambit pawn without
Accepting a gambit pawn is not to everyone's relinquishing the bishop pair. Unfortunately for
taste, so the following is seen from time to him this approach j ust seems to be too ambitious.
time. In variation B l ) we will see that it is Black who
should be fighting for the advantage here.
1.e4 eS 2.tLla tLlc6 3.AbS a6 4.J.a4 tLlf6
S.O-O Ae7 6Jel bS 7.i.b3 0-0 8.c3 d6 9.h3 Extra pawn versus bishop pair
tLlaS 10.Ac2 dS 1 l .d3 If White wants to accept the gambit pawn
then he should do so in the following way.
8
l.e4 eS 2.tLla tLlc6 3.AbS a6 4.J.a4 tLlf6
7 S.O-O J.e7 6J3el bS 7.J.b3 0-0 8.c3 d6 9.h3
6 tLlaS 10 ..ic2 dS 1 l .exdS e4 12 ..be4 tLlxe4
13J3xe4 Ab7 14.d4 ge8
5
4
8
3
7
2
6
1
5
a b c d e f g h
4
This is a reasonable choice for players who
3
yearn for a solid game with little theory.
However, I hardly need state that it is not a 2
serious try for a theoretical advantage. 1
Having his cake and eating it a b c d e f g h
We have already encountered the following White has tried a few different moves here.
approach in the two illustrative games: He has scored quite well with:
1.e4 eS 2.tLla tLlc6 3.AbS a6 4..ia4 tLlf6 IS.gel xdS 16.Af4
S.O-O Ae7 6Jel bS 7.i.b3 0-0 8.c3 d6 9.h3 However, in line B22) I found a new move
tLlaS 10 ..ic2 dS 1 l .exdS e4 12.tLlgS which I believe to be the answer to Black's
problems. The other main line is:
8
7 IS ..i4
This is perhaps the most theoretically
6
challenging move. Virtually every game has
5 continued with:

4
1 5 ... ti)c4 16.b3 ti)b6 17.c4
3 Based on the evidence presented in B23 1 2) ,
i t does not appear that Black can equalise here.
2
1
a b c d e f g h
 Chapter 4 Gajewski: 1 1 .d3 and 1 1 .exd5
- 99

However, in variation B232) I reveal a simple

8
novelty on move 15 which offers reasonable
compensation. 7
6
At the present time, I believe that Black is holding
his own in all of the variations contained within 5
the present chapter. This may explain why the
4
active I I .d4! ? has become an increasingly popular
way for White to meet the Gajewski. This move 3
will be discussed in its own dedicated chapter.
2
Theory 1
a b c d e f g h
l.e4 e5 2.f3 c6 3 ..ib5 a6 4..b4 f6
5.0-0 .ie7 6J:el b5 7 ..ib3 d6 8.c3 0-0 9.h3 This is a safe choice for White. The ensuing
a5 10.i.c2 d5!? queenless position should be approximately
equal, but at the same time there is plenty
This will be our starting position for the present of scope for either player to outplay the
chapter. opponent. In the following section we will see
some examples of how Black should handle
We will now consider the solid A) 1 l .d3, the position.
followed by the more principled B) 1 1.exd5.
The only other sensible moves are: 1 1 ... dxe4 12.dxe4 Y9xdl 13J:xdl .ib7
Black can also consider delaying the
1 1 .d4!? will be examined in Chapter 5 . development of this bishop with 1 3 . . . i.d6
1 4.lLlbd2 lLlb7!? ( 1 4 . . . i.e6 and 1 4 . . . i.d7 are
l l .lLlxe5 is met by 1 1 . . .dxe4, when two moves more natural, but the text is also interesting) :
have been tried:

a) 1 2.b4 should be met by the zwischenzug

1 2 . . . i.d6! 1 3.d4 exd3 1 4.hd3 ( 1 4.lLlxd3 lLlc4
is also fine for Black) 1 4 . . . he5 1 5 Jhe5 lLlc4
1 6Je 1 'IWd5 . Black is doing well and eventually
won in Leon Hoyos - Miton, Lubbock 2007.

b) 1 2.d4 is also harmless after 12 . . . exd3

1 3.'lWxd3 'lWxd3 1 4.lLlxd3 eB, when Black has
completely equalised and can perhaps claim
a small advantage. S.Pedersen - Schandorff,
Denmark 2007, was soon agreed a draw: 1 5 .a4
i.b7 1 6.axb5 axb5 1 7.i.g5 i.d6 I B.xeBt
lLlxeB 1 9.1Lld2 lLlc6 Y2-Y2 .
A) 1 1 .d3
a b c d e f g h
Najer - Safin, Moscow 200B, continued
1 5 .lLlb3 eB (Black continues to play flexibly)
1 6.i.e3 h6 1 7.lLlbd2 i.d7 I B.lLlfl i.c6
1 9.1Ll3d2 lLlc5 with chances for both sides.
Black went on to lose this game, but his play
up to this point was fine.
1 00 Attacking the Spanish

14.tLlbd2 Ud8 Even 17 . . . g6!? would have been a more

There is no reason to avoid this very natural
move.
In Shirov - Leko, Moscow (blitz) 2007, Black
preferred 1 4 . . . i.d6 1 5 .a4 (after 1 5 .b4 ttJc6
1 6.a4 ttJe7 Black's pieces are well coordinated)
1 5 . . . ttJc6?! ( 1 5 . . . c5 was correct, with equality)
1 6.ttJb3! ttJe7 1 7J!e 1 abB ( 1 7 . . . fbB seems
preferable) 1 B .axb5 axb5 1 9.i.e3 and White
eventually won. Of course, with more time
available Leko would surely have played more
accurately.
useful waiting move than the text.
1 B .ttJf1 bxa4
1 B . . . ttJb6 1 9.a5 ttJc4 20.ttJe3 ttJxe3 2 1 .he3
looks pleasant for White.
1 9.ha4 ttJcbB
This is very passive. A better try would
have been 1 9 . . . a5!? 20.hc6 hc6 2 1 .bxa5
i.d6 when the strong light-squared bishop
provides some compensation for the pawn.
20.ttJg3
White had a very pleasant position and went
on to convert his advantage.

15.gel 16.b4 tLlc6 17.a4 tLld7

1 5 .ttJxe5 ttJxe4 brings White nothing. 1 7 . . . ttJe7 1 B.ttJb3 allows White a useful grip
over the queenside.
More logical is 1 5 .b4 ttJc6 1 6.a4. In this case
1 6 . . . g6!? looks like a sensible choice. Both 18.axb5 axb5 19.uS guS 20 ..id3 .b6!
sides should aim to improve their positions
slowly but surely. Overall I would evaluate the
8
chances as approximately equal.
7
8 6
7 5
6 4
5 3
4 2
3 1
2 a b c d e f g h
1 This looks clumsy, but Black will soon be
able to harmonise his pieces. As soon as he
a b c d e f g h plays . . . c6 White's temporary advantage will
15 ... Ad6! disappear.
This is where the bishop needs to be. In
Lupulescu - Baratosi, Bucharest 200B, Black 21.tLlbl tLld8 22.tLla3 gb8 23 ..ie3 f8
failed to equalise after: 24.gal c6 25.i.e2 .ib7 26Jdl e7
Black has fully equalised, and the game
1 5 . . . ttJd7 1 6.b4 ttJc6 1 7.a4 h6?! Konguvel - Sriram, Mangalore 200B, was
It was not too late for 17 . . . i.d6!, transposing eventually drawn.
to the main game.
 Chapter 4 - Gajewski: 1 1 .d3 and 1 1 . exd5 101

B) I l.exd5 Instead, the superior 1 7 . . . i.d6 would have

This is where the real fun begins! kept an advantage. For instance, after 1 8 . .!d2
f6 1 9.ttJd3 ttJc4 20.ad 1 c5 2 1 .i.cl ad8
1 1 e4
. 22.xe8t xe8 23.@f1 g5! White is still some
way short of equality.
8
BI) 12.g5?!
7 This was played in several games shortly
6 after the Gajewski carne into fashion. It is
now considered to be too risky, as Black soon
5 develops a formidable attack.
4
12 xd5 13.xe4
..

3
Instead 1 3.ttJxh7?! @xh7 1 4.he4t @g8
2 1 5 .hd5 might look like a clever combination,
but White can easily find himself in trouble
1
after: 1 5 . . . xd5 1 6.xe7 i.xh3!
a b c d e f g h
One of the advantages of the Gajewski over 8

the Marshall is that Black can advance this 7

pawn without being troubled by a hanging 6
knight on c6. We now consider BI) 12.g5?! 5
and B2) 12.,he4.
4

3
1 2.ttJe5?! is unpromising in view of 1 2 . . . xd5
1 3.d4 exd3 1 4.xd3 i.b7 when Black's 2
position is already preferable. Haznedaroglu 1

- Brunello, Crete 2007, continued: 1 5 .xd5 a b c d e f g h
ttJxd5 1 6.ttJd2 fe8 1 7.ttJdf3
1 7.f3 ( 1 7.gxh3? g5t picks up the rook)
1 7 . . . xf3 1 8.gxf3 ae8 1 9.xe8 (no better is
1 9.xc7 e 1 t 20.@h2 xc l 2 1 .@xh3 ttJc4)
1 9 . . . xe8 20.ttJa3 e 1 t 2 1 .@h2 .!f5 White
must struggle to equalize in this difficult
endgame.

13 5!
..

This pawn can force its way to f3 with

gain of tempo. We have already encountered
this important motif in the two introductory
a b c d e f g h games.
At this point the game continued: 1 7 . . . b4?!
1 8.cxb4 hb4 1 9 . .!d2 hd2 20.ttJxd2 ttJf4 14.g3 f4 15.e4
2 1 .ttJef3 White was okay and eventually drew. 1 5 .ttJf1 f3 1 6.d4 fxg2 1 7.ttJg3 transposes.
1 02 Attackin g the Spanish

15 ... S

8
7
6
5
4
3 a b c d e f g h
2 Here White could have defended better,
1 by playing 20.lLlbd2 or 20 . .td3. Black still
has the initiative, but the exact evaluation
a b c d e f g h is not too important to us. If he had played
One way or another, White's kingside is 1 9 . . . ,txf2t he would have won as in the game,
going to be damaged. His best reaction is to which concluded swiftly with 20.lLlxf2? .txf2t
develop his queenside pieces as quickly as 2 1 .@xf2 h4t 22.@gl f8 0-1 .
possible.
17 g6!
.

16.d4 fxg2 17.ttlg3

From this square the knight helps to shield
the king, while also unblocking the sights of
the rook on e1 and bishop on c2. It is doubtful
that White has anything better:

1 7. @xg2 .tfS gives Black a healthy initiative.

In Guliyev - Fressinet, Ajaccio 2007, White

lost quickly after 1 7.hS g6 1 8.h6 fS
1 9.1Llg3 .tfB 20.d2 f7+ 2 1 .b3? lLlf4 when
the game was as good as over.
Although Black's play made for a positive
impression, he could have played even more
strongly with 20 . . . lLlc4! 2 1 .e2 f7, when
White's kingside is under heavy pressure.
1 7.a4 looks like a luxury that White can
ill afford. In Brinck-Claussen - O. Larsen,
Denmark 2007, Black won in fine style after
1 7 . . . .txh3 1 8.axbS .th4! 1 9.xaS xf2? Black
plays the correct combination incorrectly. It
was best to take with the bishop first, to make
the h4-square available for the queen.
a b c d e f g h
With this excellent move Black gains control
over two key squares: fS and hS. The former
is important because White can use it to
exchange pieces in some variations. The latter
might have been utilised by the white queen.
Finally, the knight on g3 and bishop on c2 are
both restricted. I probably was the first to play
this move, albeit only in an unpublished rapid
game, in 2007.
 Chapter 4 Gajewski: 1 1 .d3 and 1 1 .exd5
- 1 03

1 7 . . . "lNd6 is a reasonable move, and was used in 20 YNd6 21.d2 .tm 22 ..tg5 Ag7

both of the introductory games at the start of
the chapter. This move would have been good
enough to recommend, were it not for the even
stronger main line.
18 ..te4
White has to transfer some pieces to cover
his kingside. We now follow the game Dabo
Peranic - Stevie, Bizovac 2008.
18 .bh3 19.i.h6
.
The patient 22 . . . c6!? deserved attention,
stabilising the position of the knight on d5 . In
this case 23.l2Jh5 could be met by the simple
23 . . . .tg4, forcing the knight to retreat. I would
definitely prefer Black here in view of his safer
king.
23.a4
23.l2Jh5! would have enabled White to ease
his defence by exchanging a few pieces.

1 9.hd5t "lNxd5 20Jhe7 is refuted brilliantly 23 ... h4 24 ..bg2?

by 20 . . Jhf2!: This loses instantly, although the refutation
is easy to overlook. 24.l2Jh5 once again looks
8 like the best try.
7

6 24 .lxg2 25. Wxg2 YNf4!

..

4 8
3 7
2
6
1
:i<....!:::: 5
a b c d e f g h
4
2 1 .@xf2 f8t 22.@gl fl t 23.l2Jxfl gxfl ="lNt
24."lNxfl hfl 25.@xfl "lNh l t and wins. 3
2
19 .. Jf7 20.YNd3
After 20.@h2 Black can calmly play 1
20 . . . g1 ="lNt! 2 1 .@xg l c6+ with ideas of . . . .td6 a b c d e f g h
and . . . "lNh4.
A beautiful move! It is not often that one sees
a bishop 'trapped' in such an ingenious way.
8
7 26.ge5
6 26.l2Jde4 runs in to 26 . . . "lNxg5!.

5 26 .be5 27.dxe5 YNxg5 28.YNxd5 gd8

.

4 29.YNxa5
29.l2Jde4 "lNxg3t will leave Black a whole
3 rook up.
2
29 ... gxd2 30JUl h5 0-1
1
a b c d e f g h
1 04 Attacking the Spanish
This was not a perfect game, but it
highlighted once again the kinds of tactical
pitfalls into which White may so easily fall.
Ultimately it seems that 1 2.tLlg5?! is too risky
to be an effective practical choice.
B2) 12 .be4 tll xe4 13.Eixe4
eventually proved to be too much for White
a b c d e f g h
If White wishes to accept the gambit
pawn, then this is the sensible way to do it.
The position will resemble a Marshall gambit
after Black captures on d5 . Both players can
claim certain advantages compared to that
opening. White has been forced to part with
his light squared bishop in return for an
enemy knight. However, the remaining black
knight is somewhat misplaced on the edge of
the board. White is less likely to succumb to
a direct attack as he still possesses a knight on
f3 to protect the kingside. On the other hand,
Black is not so reliant on a mating attack, as
his bishop pair will often provide a degree of
positional compensation.
13 .t.b7
..
This is the best place for the bishop.
14.d4 ge8
This is the best and most popular move.
Black keeps his options open and waits for
White to reveal his intentions.
Instead 14 . . . ,ixd5?! would be premature after
1 5 .!!e l :
a) In Onischuk - Drozdovskij, Kharkov 2007,
Black was successful with 1 5 . . . Wid6 1 6.b3 tLlc6
1 7.i.e3?! f5 with good counterplay, which
to handle. However, 1 7.i.a3 is an obvious
improvement, after which White stands better.
b) Zherebukh - Swiercz, Kirishi 200S, instead
saw 1 5 . . . i.d6 1 6.tLlbd2 eS 1 7.xeSt WixeS
I S.tLlfl when White was starting to get
coordinated. Following I S . . . tLlc4 1 9.b3 tLlb6
20.tLle3 he enjoyed a stable advantage.
Another idea is 1 4 . . . i.f6, preparing to recapture
on d5 with the queen. The drawback is that he
commits the dark-squared bishop to what may
turn out to be an unfavourable square. Play
continues 1 5 .i.f4 and now:
a) 1 5 . . . tLlc4 is well met by 1 6.b3 Wixd5
( I 6 . . . tLlb6 1 7.i.e5;!;;) 1 7.el Wif5 ( I 7 . . . tLld6
is safer but the position is already unpleasant.
Both I S.tLlbd2 and I S.i.e5 give White a stable
advantage.) I s.hc7 tLlb2 1 9.Wid2 In Prasad
- Ivanov, Selestat 200S, it turned out that
Black's active play had brought him nothing
but problems. After the further 1 9 . . . hf3
20.Wixb2 Wig6 2 1 .i.g3 i.b7 22.tLld2 White had
consolidated and went on to win.
b) The game Degraeve - Karpatchev, Saint
Chely d'Aubrac 200S, resulted in a quick
draw after 1 5 . . . Wixd5 1 6.e l tLlc4 1 7.Wic2
( I 7.b3!? transposes to Prasad - Ivanov above)
1 7 . . . c5 Y2-Y2 . However, if White had been
feeling more ambitious he could have obtained
a pleasant advantage with: I S. b3 tLld6 1 9.i.e5
cxd4 20.cxd4;!;;
We may conclude that Black does best to follow
the path of the main line.
 Chapter 4 Gajewski: 1 1 .d3 and 1 1 .exd5
- 1 05

After 14 .. .11e8 we reach an important 16 Jad8

position in which White has tried three main Black has a couple of reasonable alternatives
moves: B21) 15.ttlbd2, B22) 15J:e1, and in 1 6 . . . i.d6!? and 1 6 . . . c5!? 1 7.dxc5 xc5
B23) 15.Af4. 1 8.lLle4 b6.

B21) 15.ttlbd2 17.ttlfl c5 18.ttle3

This is not the most theoretically challenging
move. The bishop on c 1 is now blocked, so
8
White will have to move the knight for a second
time to make room for it. In the meantime 7
Black will be able to complete development 6
and open the centre with . . . c5 .
5
15 ...'ilYxd5 16Jle1 4
1 6.e2 d7 1 7Je3 ( 1 7.lLle5?! e6 1 8.e3
3
i.g5 1 9 .g3 f6) should be met by 1 7 . . . c5
(Kosten) with good play for Black. 2
1
In Tuncer - Kosten, Cappelle la Grande 2008,
White preferred 1 6.e5 d7 1 7.a4. a b c d e f g h
18 'ilYh5
.

The game Hoffmann - Nyback, Germany

2008, showed another promising path for
Black in 1 8 . . . e6!? 1 9.i.d2 i.e4 20.b3 lLlc6
2 1 .d5 hd5 22.lLlxd5 xd5 Black has a very
slight edge, although the game was eventually
drawn.

a b c d e f g h
The game continued 1 7 . . . i.d6?! 1 8 .xe8t
xe8 1 9.b3?! lLlc6 20.axb5 axb5 with
approximate equality and an eventual draw.
Unfortunately 1 9.axb5 would have been an
obvious improvement, and after 1 9 . . . xb5
20.a4 White stands clearly better.
Therefore Black should have preferred 1 7 . . . f6
1 8.e2 i.d6 ( 1 8 . . . i.d5 also deserves
consideration) 1 9.xe8t xe8 (Kosten) when
the two bishops should give Black enough
compensation.
19.d5 ttlc4!?
1 9 . . . hd5 would have been equal. The text
is more ambitious, although is probably does
not change the final evaluation.
20.ttlxc4 bxc4 21 .ge5 'ilYg6 22.'ilYe2 cJ?f8
23.'ilYxc4 f6
We have been following Kryvoruchko -
Nyback, Plovdiv 2008. The position is rather
unclear, with mutual chances. White has
retained his extra pawn, but Black enjoys a pair
of bishops and a healthily centralised army. The
game was eventually drawn.
B22) 15.ge1
The rook will soon be chased away regardless,
1 06 Attacking the Spanish

so White reasons that by moving it immediately
he will keep all of his options open regarding
the placement of the other pieces. The
drawback is that he loses some flexibility with
the rook which might, in certain positions, be
.ih4 is examined after 1 9.'lNe2!) 20 . . . i.g5
2 1 .:gxeBt :gxeB 22.'lNxb2 :ge l t 23.i>g2 i.el
(23 . . . i.f4 also wins) 24.lDd2:
8

better off moving to e2 or e5. 7

6
IS ...BxdS 16 ..if4 5
1 6.lDbd2 transposes to line B2 1 ) above.
4

3
8
2
7 1

6 a b c d e f g h

5 Now the most efficient route to victory is

24 . . . :gh l ! 25.i>xh 1 hb2.
4
1 9 . . . hf3
3 The alternatives are no better:
2 1 9 . . . i.f8?! 20.'lNxb2 hf3 2 1 .:gxeB :gxeB
22.gxf3 :ge l t 23.i>g2 'lNd3 24.lDd2 :gxa1
1 25 .'lNxa1 'lNxd2 26.c4
a b c d e f g h 19 ... lDd3 20.:gd 1 i.a3 (after 20 ... :gacB
2 1 .'lNxd3 i.e4 22.'lNe3 :gxc7 23.lDbd2
16 ... cS! Black does not have enough for two pawns)
After a detailed examination of the 2 1 .'lNxd3 'lNxd3 22.:gxd3 i.b2 23.d5 hal
alternatives, I came to the conclusion that this 24.c4 White's connected passed pawns are
was Black's most promising move. very dangerous, e.g. : 24 . . . i.cB 25 .c5 i.fS
26.:gd 1 i.f6 27.lDbd2 :gacB 2B.d6
In Zhigalko - Teterev, Minsk 2009, Black 20.'lNxf3
failed to equalise with 1 6 . . . lDc4?! 1 7.b3 lDd6 After 20.'lNxb2 i.h4 2 1 .lDa3 i.c6 Black has
1 B.'lNc2!? i.cB 1 9.:ge5 'lNc6 20.a4 (promising excellent prospects, despite being two pawns
alternatives include 20.lDbd2, or 20.d5!? down. The bishop pair is very strong, White's
'lNd7 2 1 .:ge 1 ) 20 . . . f6 2 1 .axb5 axb5 22.:gc5. kingside is under pressure and the knight on
White was clearly better although the game a3 is a mere spectator.
was eventually drawn. 20 . . .'xf3 2 1 .gxf3
1 6 . . . 'lNf5!?N 8
This is more interesting. For a while I 7
believed that this move would ensure Black
6
of an excellent game. Unfortunately I then
discovered that White can obtain a slightly 5

better ending with precise play. 4

1 7.hc7 lDc4 1 B.b3 lDb2 1 9.'lNe2! 3
Hopeless is 1 9.'lNd2 hf3 20.gxf3? (20.'lNxb2 2

1
L-
a b c d e f g h
 Chapter 4 - Gaj ewski: 1 1 .d3 and 1 1 .exd5 1 07

2 1 . . JacB force White to compromise his structure by

This looks better than: 2 1 .. . .ig5 22.a3 capturing with the c-pawn.
d3 23.xeBt xeB 24.d l b2 25.b l
d3 26.cj;>f1 cB 27.i.a5 .id2 2B.cj;>e2 17 'ifxc5 18.bd2 c4!?N
. .

.ixc3 29.cj;>xd3 (29 . .ixc3 xc3 30.cj;>d2 b4=) In Mekhitarian - Perdomo, Sao Paulo 200B,
29 . . . .ixa5 30.b4 .idB 3 1 .e l .ih4 32.e2;;!;; . the continuation was IB ... .if6 19 . .ie3 %Vd5
White has quite good chances to convert his 20 ..id4 .ig5 2 1 .f1 .if4:
extra pawn; the centralised king is a big help
to him. 8
22 . .ig3 7

6
8
5
7
4
6
3
5
2
4 L,=-/'="',

1
3
a b c d e f g h
2

1 22 . .ie3?! %V5 ? (the simple 22 . . . .ic7 leaves Black

with no problems) 23 . .ixf4 %Vxf4 24.%Vd7;;!;;
a b c d e f g h
(24.%Vd4!) . Black eventually managed to
22 . . . b4! draw although the results of his opening were
22 . . . 5 23.e2 dl 24.a4 unimpressive.
23.c4 From the diagram position, White should
After 23.cxb4 .ixb4 24.xeBt xeB it is have preferred 22.e3 %Vh5 23 . .ib6 .ixe3
hard for White to develop. 24 . .ixa5 .ixf3 25.%Vxf3 %Vxf3 26.gxf3 with
23 . . . .if6 24.d2 quite good winning chances in the ending.
24 . .ie5 ? d3
24.xeBt xeB 25 .i.e5 d3 is also fine for 19.xc4 'ifxc4 20.'ifd4 'ifc6
Black.
24 . . ..ixd4 25 .ab l d3 26.xeBt xeB
27.cj;>f1 5! 8
Threatening to trap the bishop. 7
2B.f4
6
Black probably has enough positional
compensation to draw this position, but he 5
can hardly hope for anything more. 4
17.dxc5 3
This was White's choice in the only practical 2
encounter I was able to find. It certainly seems
like a natural choice, as after 1 7.bd2 cxd4 1
the pressure on the long diagonal would a b c d e f g h
l OB Attacking the Spanish

Black's powerful light-squared bishop will B231) 15 tl)c4

.

provide ongoing compensation for the pawn.
We will see that similar positions can arise in
certain variations of the Marshall Gambit. It
is not easy for White to make his extra pawn
count, even if he manages to exchange queens.
B23) 15.i.f4
8
7 hd5 1 7.ttJe5 .ig5 1 B . .ig3 ttJxe5 1 9.xe5
6 xe5 20.he5 . Black still had some problems
5 draw after a long battle.
4
3
2
1
a b c d e f g h
This is the main line. White develops the
bishop on its most natural square. He intends
to follow with ttJbd2, completing development
with good chances to consolidate his extra
pawn. We will now analyse two moves. B231)
15 ... tl)c4 has been the most common choice,
although I am not convinced that it can suffice
for full equality. However, my new suggestion of
B232) 15 .ixd5!?N may be more promising.
.
This is a very logical move. The first point is
that Black forces his opponent to spend time
dealing with the threat against the b2-pawn.
At the same time he improves the misplaced
knight, which can proceed to a better central
location via either . . . ttJd6, or . . . ttJb6xd5.
16.b3
In Adams - Carlsen, Khanty-Mansiysk 2007,
White obtained a slight edge with 1 6.e2
to solve, although Carlsen duly held on for a
The improvement 1 6 . . . xd5 has been
suggested by Kosten who goes on to analyse:
a) 1 7.ttJbd2 ttJxb2 1 B.c2 .ia3=
b) 1 7.b3 f5 with two choices for White:
b 1 ) 1 B.hc7?! hf3 1 9.9xf3 .id6 20.hd6
ttJxd6 is good for Black, as pointed out by
Kosten. White's two pawn advantage is
outweighed by his shattered kingside.
b2) 1 B.ttJbd2 looks better, when there follows
1 B . . . ttJxd2 1 9.xd2 hf3 20.gxf3 xh3
2 1 .ae l d7.

Before moving on, let us briefly note that

1 5 . . . xd5 does not equalise after 1 6Je5! d7
1 7.ttJbd2. Yagupov - Zubarev, Tula 200B,
continued: 17 ... f6 (After 17 ... .id6 1 BJxeBt
xeB 1 9.hd6 cxd6 20.fl !? Black's strong
bishop gives him some compensation, but he
has no real targets to attack and White has
reasonable chances to convert his extra pawn.)
1 B.e 1 i.d5 (Also insufficient is 1 B . . . ttJc4
1 9.ttJxc4 bxc4 20.ttJd2 intending b3) 1 9.b3 a b c d e f g h
acB 20.c2 ttJb7 2 1 .c4 .if? 22.e2;!;
White may be able to claim a very slight
 Chapter 4 Gajewski: I I .d3 and I I .exd5
- 1 09
edge thanks to his control over the e-file, but
overall I agree with Kosten's assessment that
Black should not be in any real danger.
16 tLlb6
8
7
6 B231 1) 17.i.g3
5
4
3
2
1
a b c d e f g h
White now faces a choice of his own.
B231 1) 17.i.g3 should not be too dangerous
for Black, although the second player must still
react with some precision. The strongest move
is B23 12) 17.c4!.
Less challenging is 1 7. d3 ttJxd5 when
White has tried two moves:
a) I S . .ie5 f6 1 9 . .ih2 c5 was fine for Black in
Nyysti - Nyback, Mantta 200S.
b) The game Ivkina - Ikonomopoulou, Plovdiv
200S, instead saw I S . .ig3 cS 1 9.b4?! ttJb6
( 1 9 . . . f5 20.e I f4 2 1 ..ih2 .id6 22.ttJbd2 is
promising for White) 20.e2 c5 2 I .bxc5 xc5 .
Black obtained good compensation, although
he eventually lost in the game. However,
White's play can easily be improved. 1 9.ttJbd2
is simple and good, as can be seen after
1 9 . . . c5 20.dxc5 hc5 2 1 .xeSt xeS 22.e I
c6 23.ttJe4. In this position White's well
placed knights give him excellent chances.
Therefore Black should probably have preferred
one of the alternatives on the previous move:
b l ) I S . . . f5!? 1 9.e l f4 has been suggested by
Kosten. Following 20 . .ih2 c5 the poorly placed
bishop on h2 ensures Black of a reasonable
game.
b2) I S . . . c5 is another logical move. Play may
continue 1 9.ttJbd2 cxd4 20.xd4 i.c5 2 1 .e4
b6 22.b4 .if8 with fair compensation.
8
7
6
5
4
3
2
1
a b c d e f g h
17 tLlxd5
This is my recommendation. Black has a
second satisfactory continuation in:
17 . . . hd5!? I S.e I c5
Black does not obtain equality with: I S . . . .id6
1 9.xeSt xeS 20.hd6 cxd6 2 1 .ttJbd2
c6 22.c2 eS 23.e l xe l t 24.ttJxe l t
1 9.ttJbd2 cxd4 20.ttJxd4
20.cxd4 cS also looks fine for Black.
20 . . . cS
Black has decent compensation. Possible
continuations include 2 1 .ttJf5 .if8 22.xeS
xeS 23.cl d7, and 2 1 .ttJe4 .if8 22.d3
he4 23.xe4 xe4 24.xe4 ttJd5, regaining
the pawn to reach an equal position.
18Je1
Black should not be worried by I S.d3 f5
1 9.e2 f4 20 . .ih2 c5 2 1 .ttJbd2 cxd4 22.cxd4
(Kosten) 22 . . . cS with sufficient compensation
1 10 Attacking the Spanish

based on the bishop pair and well centralised
pieces, not to mention the poorly placed
bishop on h2.
18 ... c5 19.a4 cxd4 20.axb5 axb5
Also possible was 20 . . . dxc3 2 1 .bxa6 :gxa6
22.:gxa6 c2 (22 . . . ha6?! 23.ttJxc3 ttJxc3
24.YMxd8 :gxd8 25.:gxe7 gives White good
winning chances) 23.YMxc2 ha6 24.ttJbd2:
8 decent compensation.
7
With the following possibilities:
a) After 23.c4 ttJc3 24.ttJxc3 bxc3 the c3-pawn
is very powerful. It is White who is more likely
to fall into danger here.
b) 23.YMd2 .tc5 (also playable is 23 . . . .tf6
24.:gxe8t YMxe8 25.f3 YMa8) 24.ttJf5 :gxe l t
25 .YMxe l looks a little better for White, so
Black should probably prefer 25 . . . :ge6!? with

6 c) 23.ttJf5 bxc3 leads to no more than equality

5
for White, e.g. 24.ttJxc3 ttJxc3 25 .YMg4 (25 .YMd7
ttJd5 26.ttJxe7t ttJxe7 27.:gxe7 :gxe7 28.YMxe7
4
hg2+ should lead to a draw after 29 . .tc7
3 .txh3 30.YMd8t YMxd8 3 1 ..ixd8) 25 . . . .tf6
2 26.ttJh6t c;t>h8 27.ttJxf7t c;t>g8 28.ttJh6t with
1 a perpetual.
a b c d e f g h
d) 23.YMd3 .tf6 24.:gxe8t YMxe8 should be fine
According to Kosten Black does not have for Black. It is important that 25 .c4 can be met
quite enough compensation for the pawn by: 25 . . . ttJc3!
here. My own view is that he should not face
too many difficulties after 24 . . . ttJb4 25 .YMc3
(25 .YMc7 YMxc7 26.hc7 ttJc2 27.:ge5 f6
28.:ge6 .td3 is very close to equal) 25 . . . ttJd3
26.:ge4 ttJc5 27.:ge3 ttJe6, when the bishop
pair provides ongoing compensation.

21.xa8 xa8 22.tLlxd4 .lc5!?

Kosten enthusiastically awards this move not
one but two exclamation marks. The idea is
certainly interesting, although there was a fully
satisfactory alternative in 22 . . . b4: a b c d e f g h

26.ttJxc3 YMe l t (26 . . . bxc3 27.ttJe2 YMe4

28.YMxe4 he4 should also suffice for a draw)
27.YMfl YMxc3 28.YMe2 h6 29.YMe8t c;t>h7
30.YMxf7 hd4 3 1 .YMxb7 YMxg3 32.YMe4t c;t>h8
33.YMxd4 YMxb3 34.c5 YMc3 when White will
have to take a perpetual.

23.tLlxb5 gxe1 t 24.xe1 tLlf6!

a b c d e f g h
 Chapter 4 - Gajewski: 1 1 .d3 and 1 1 . exdS 111

25 Val
.

8
2S . . . %Vd8 looks slightly less accurate in
7 view of 26.b4 .ib6 27.tDbS %VdS 28.%Vfl tDe4
(Kosten) 29.c4 tDxg3 30.cxdS tDxfl 3 1 .i>xfl
6
.ixdS, although Black should probably be
5 able to hold the draw.
4
26.b4 Ab6 27.Vdl
3
2 8
1 7
a b c d e f g h 6
The knight vacates the long diagonal while 5
also preparing a possible . . . tDe4. Another
important point is that the checking square 4
on e8 is now protected, thus freeing the black 3
queen from defensive duties.
2
25.c7 1
2S.b4 .ixg2 26.bxcS (the computer suggests
a b c d e f g h
26.i>h2 .ifB 27.tDd2, although White's
exposed king is likely to cause him problems Kosten has suggested 27.tDbS %Va8 28.%Vfl
for a long time to come) 26 . . ..ixh3 27.f3 %Vxf3 tDe4 29.i>h2 as a possible improvement for
28.%Vf2 %Vd l t 29.%Ve l %Vf3 This is a draw, as White. I would tend to agree, although I still
pointed out by Kosten. think that Black has enough compensation.
His bishops are superbly placed and his overall
However, it seems that Black can do even piece coordination is excellent.
better with 28 . . . %Vd3!
27 h6
..

27 . . . hS!? is more ambitious.

28.ctf?h2 Ae4 29.Vd8t ctf?h7 30.d2 Vxc3

31.Vd6
At this point the game Durarbeyli - Kosten,
Cappelle la Grande 2008, was agreed drawn.
Overall we have seen that 1 7 . .ig3 presents
Black with multiple routes to an acceptable
position.

a b c d e f g h
B2312) 17.c4!
Black regains the piece to obtain a clear This looks like White's best chance for a
advantage, as 29.tD l a3 can be met by plus.
29 . . . tDe4.
1 12 Attacking the Spanish

17... bxc4 18.bxc4 xc4 19.c3 20.fib3 llb8

It is important to defend the bishop.
20 . . . fLlxd5 runs into 2 1 .hc7! (2 1 .xb7 fLlxc3
8
should enable Black to equalise) 2 1 . . . xc7
7 22.fLlxd5 hd5 23.xd5 with a solid extra
6 pawn.

5 21Je5!
4 This looks like the most accurate move:
3
a) 2 1 .ae I ?! fLlxd5 can only be better for
2 Black.
1
b) After 2 1 .fLlg5 hd5 22.fLlxd5 fLlxd5
a b c d e f g h 23.hc7 xc7 24.xd5 f8! White is unable
This is a more difficult line for Black to to consolidate his extra pawn, e.g. 25.a4 ( . . . b5
meet. Compared with the previous variation was a serious threat) 25 . . . bd8 26.e5 xe5
White is playing much more dynamically. We 27.xe5 f6 (27 . . . hg5 28.xg5 xd4=)
will see that he has a lot of tactical resources at 28.a5 xd4 29.fLlf3 d6. The position is
his disposal. equal, with perhaps a miniscule edge to Black
in view of his better minor piece.
19 ... b6
In Kaplan - Ragger, Yerevan 2007, Black c) 2 1 .e2 fLlxd5 22.fLlxd5 hd5 23.hc7 gives
was able to equalise with 1 9 . . . fLld6 20Je l ?! Black a choice:
fLlb5! (after this move Black is fine) 2 1 .cl
c8 22.a4 d6 (an even faster route to c l ) After 23 . . . xb3 24.hd8 b7 25 .he7
equality would have been: 22 . . . fLlxc3 23J!xc3 bxe7 26.xe7 xe7 White must be a little
d6 24.xe8t xe8 25.xe8t xe8 26.hd6 better although the most likely result should
cxd6=) 23.xe8t xe8 24.e l d7 25 .fLle5 be a draw.
he5 26.dxe5 d8. White was unable to keep
his extra pawn, and an equal endgame ensued c2) It also looks interesting to try 23 . . . d7!?
after 27.b3 hd5 28.fLlxd5 xd5 29.xd5 24.fLle5 (24.xb8 xb8 25 .hb8 hf3 26.gxf3
xd5 The game was subsequently drawn. h5 is balanced) 24 . . . e6 25 .xb8 (25 .c2
bc8 26.ae l b4 27.fLlf3 he l 28.xe6 fxe6
However, White could have secured a plus 29.fLlxe l e7 30.fLld3 exc7 should be alright
with 20.hd6!N. Play continues 20 . . . cxd6 for Black) 25 . . . xb8 26.hb8:
(20 . . . hd6 2 1 .xe8t xe8 22.b3 b8
23.e l ) 2 1 .b3 c8;t. From a purely
materialistic point ofview, White's extra pawn is
not too important. However, from a positional
perspective the doubled d-pawns are extremely
useful in restricting the freedom of the enemy
bishops. We may therefore conclude that Black
has failed to equalise from the opening.

a b c d e f g h
 Chapter 4 Gajewski: 1 1 .d3 and 1 1 . exd5
- 1 13

The position is approximately equal. White

has a slight material advantage, but Black's
queen and bishop pair work together very well.
However, it is worth mentioning that he must
avoid: 26 . . .f6? 27 Jae 1

2t. ..W
The alternative is 2 1 . . .lLld7:

8
7

4 a b c d e f g h
3
This energetic move targets the weak f7-
2 pawn.
1

a b c d e f g h 22 .. JU'8
22 . . . lLld7 does not equalise after 23.c2:
When White can choose between:
a) 23 . . . lLlf6 24.ae l ! (weaker is 24.e3 h6
22Je2 lLlb6 23.ae l (after 23.b l hd5 25.lLlf3 ,bd5= or 24.lLlce4 lLlxe4 25.xe4
24.lLlxd5 lLlxd5 25 .xbB xbB 26.xbB hg5 26.hg5 xe5 27.xe5 xd5=) 24 . . . h6
xbB 27.hc7 b l t 2B.i>h2 f6 Black should 25.lLlge4 lLlxd5 26.lLlxd5 hd5 27.lLlc5 b6
be okay, as his pieces are extremely active) 2B.e2
23 . . . lLlxd5 24.lLlxd5 hd5 25 .hc7 hb3
26.hdB exdB 27.axb3 i.f6 Black should be b) 23 . . . g6 24.e2 lLlb6
able to draw this position. After 24 . . . lLlf6 25.ae l i.b4 26.xeBt lLlxeB
27.lLlge4 hd5 2B.lLlc5 hc3 29.xc3 b6
22.ee l Black is surviving, although he is still under
This looks like the best try for an advantage. pressure after 30.e3!?
Play continues: 25 .ae l d7
22 . . . lLlb6 23.ab l
23.lLle5 lLlxd5 24.ab l i.b4=
23 . . .i.aB
23 . . . hd5 24.,bc7 xc7 25.lLlxd5 lLlxd5
26.xd5 Ieaves Black struggling.
24.a4 lLld7 25.c4 xb l 26.xb l lLlb6
27.xa6 lLlxd5 2B.lLlxd5 hd5
If Black defends well than he may be able to
hold the draw. At the same time, this is not the
type of position that one generally wishes to
obtain from the opening.
a b c d e g
1 14 Attacking the Spanish

26.lDge4! This was White's choice in Carlsson - Stevic,

After 26.lDf3 f6 Black's kingside has been Dresden 2008. The move is not at all bad, but
compromised slightly, but he will soon I do not believe it to be the strongest. We will
regain his pawn and his bishop pair should return to this position in due course, after first
ensure a reasonable game. Play may continue: reviewing the course of the game.
27.b3 @g7 28.lDd2 lDxd5 29.lDxd5 xb3
30.xe7t xe7 3 1 .xe7t xe7 32.lDxe7 26 5
.

b4 Black should be fine in this ending.
26 . . . bc8
26 . . . lDxd5 27.lDxd5 xd5 28.xc7 bc8
29.e5
27.d3 lDxd5 28.lDxd5 hd5 29.xa6
Black is struggling. Not only is he a pawn
down, but the dark squares around his king
are also weak.
This is the only satisfactory response to the
twin threats of xc4 and iog5 .
27.Wxc4 fxe4 28.d6t Wfl 29.Wxf7t gxf7
30.dxc7 gc8 31 .Ag3
3 1 .lDe2 may have been more precise,
although White has probably lost the greater
portion of his advantage already.

23.Wc2 g6 24.tLlge4 31 gfxc7 32 .ixc7 gxc7

24.b l ? d7 25.ee l lDxd5 leads nowhere

for White, who might already stand slightly
8
worse.
7
24 tLlc4
..
6
No better is: 24 . . . lDxd5 25 .lDxd5 hd5
26.ioh6 f6 (26 . . Je8 27.e l ) 27.hf8 fxe5 5
28.he7 xe7 29.lDc3 4
3
25Jxe7! Wxe7
2
8 1
7 a b c d e f g h
6 Black has reasonable chances to draw,
although his task is not easy.
5
4 33Jcl .td5 34.tLle2 gxcl t 35.tLlxcl <it>fl
3 36.<it>f1 Ac4t 37.<it>e1 <it>e6 38.<it>d2 <it>d5
39.<it>e3 g5
2 39 . . . a5 40.lDb3 a4 4 1 .lDc5 a3 42.lDxe4 iof1
1 43.lDd2 hg2 44.h4 is difficult for Black.
a b c d e f g h 40.g3
We have reached a critical position. The last winning chance was: 40.lDb3 iof1
4 1 .g3 i.xh3 42.lDc5 iof5 43.lDa4 @c4 44.lDb6t
26.We2!? @b5 45.lDd5 h5 46.lDc3t @b4 47.lDxe4 h4
 Chapter 4 Gajewski: 1 1 .d3 and 1 1 .exd5
- 1 15

4B.c!tJxg5 h3 49.c!tJxh3 hh3 50.@e4 It is still 30 ..bh6 gxf6 31 .tDxf6 'ilYxf6 32.'ilYxc7 gc8
far from trivial for Black to hold this ending. 33.i.g7t 'ilYxg7 34. 'ilYxd6

40 aS

Now the draw is simple.

41.h4 gxh4 42.gxh4 a4 43.hS a3 44.h6 AbS

4S.tDb3 Ac4 46.tDc1
1/2-1/2

Returning to move 26, I believe that White has

a strong improvement available:

26.i.gS!N

a b c d e f g h
8
White has four pawns for a bishop, as well
7
as a strong attack against the enemy king. His
6 advantage ought to be enough to decide the
game.
5
4 B232) lS i.xdS!?N
.

3
8
2
7
1
6
a b c d e f g h
5
26 f6
.

This looks more or less forced. In case of 4

26 . . . Wla3 White can choose between 27.c!tJf6t
3
@g7 2B.c!tJd7, and 27J%e l !? with a very strong
initiative. 2
1
27.tDxf6t <i!.>h8
27 .. J%Xf6 2B.c!tJe4 is virtually winning for a b c d e f g h
White, e.g. 2B .. J%Xf2 29.@xf2 c!tJa3 3o.Wld3 Amazingly this natural move has yet to be
Wlb4 3 1 .Ekl c!tJb5 32.Wle3 when the black tested, even though it may be Black's best
kingside is just too weak. chance to equalise.

28.tDce4 tDd6 29Jel h6 16.gel .Ad6

29 . . . c!tJxe4 30J%xe4 Wlg7 3 1 .c!tJg4 is winning According to the database this position has
for White. been reached just once, via a different move
order.
1 16 Attacking the Spanish

17..bd6 I B . . . c7 is also possible, keeping the . . . ltJc4

1 7J3xeBt xeB I B.hd6 cxd6 1 9.1tJbd2 move in reserve.
is similar to the main line. Once again my
suggestion is 1 9 . . . ltJc4!? with reasonable In either case, Black has reasonable
compensation. compensation thanks to the superbly placed
bishop on dS , which is currently the best
17 ... cxd6 performing piece on the board.
Also possible is 1 7 . . J3xe l t I B.ltJxe l
( I B.xe l cxd6 1 9.1tJbd2 i s similar to the main
line) I B . . . xd6 1 9.1tJd2 eB:

8
7

a b c d e f g h

Black has sufficient compensation. The

game Rusjan - Marinc, Ljubljana 2007,
continued 20.ltJd3 hg2 2 1 .g4?! .ib7 when
Black stood better. Instead after the superior
2 1 .i>xg2 g6t 22.g4 xd3 23.e l f8! the
chances are approximately equal.

18.bd2 c4!?

8
7
6
5
4
3
2
1
a b c d e f g h
 Chapter 4 - Gajewski: 1 1 .d3 and 1 1 .exd5
Conclusion
The Gajewski variation has only been topical
for a couple of years at the time of writing.
Although the theory is still developing, we
have reached the stage where we can begin to
draw at least some tentative conclusions about
certain variations.
If White decides to play solidly with A) 1 1 .d3
(page 99) , then he can hardly hope for an
opening advantage. At the same time, the
move should not be ignored completely. From
a psychological perspective, it is not always
easy to switch from offering a sharp gambit
to manoeuvring without queens. I hope that
the guidance offered in section A) will help
the reader to react appropriately should he
encounter this variation.
Aside from the central counter-strike 1 1 .d4,
which will be covered in the following chapter,
the critical test of Black's idea has to be the
acceptance of the gambit pawn with B) 1 1 .exd5
(page 1 0 1 ) . Following the standard response of
1 1 .. .e4, it seems too risky for White to try B 1 )
1 17
1 2. g5 ?! (page 1 0 1 ) , so he should instead prefer
B2) 1 2.he4 xe4 1 3Jhe4 (page 1 04) . Then
after the logical 1 3 . . . .!b7 1 4.d4 e8 we analysed
three main continuations for White. Generally
speaking, it seems to me that Black has enough
compensation in most lines, although further
practical testing is clearly needed before any
firm conclusions can be drawn. For instance,
after B23) 1 5 . .!f4 (page 1 08), the response
of 1 5 . . . hd5!?N (line B232, page 1 1 5) ,
which may well b e Black's most promising
continuation, has not even been played
yet!
Overall I consider the Gajewski to be a good
practical weapon against the Spanish, although
the downside to its newfound respectability
is the diminishing of its surprise value. As I
have already mentioned, the theory is still
developing in many lines, so I would regard
it as an especially attractive option for creative
players who like to search for their own original
ideas.
But let us take a look at the move that has
quickly developed to be the main option:
Chapter 5

Gajewski: 1 1 .d4
7
6
5
4
3

1
a b c d e f g h

1 .e4 e5 2.tLla tLlc6 3.Ab5 a6 4.Aa4 tLlf6 5.0-0 Ae7

6.gel b5 7.Ab3 0-0 8.c3 d6 9.h3 tLla5 1 0.Ac2 d5!?
1 1 .d4

Introduction page 1 20
Theoretical Highlights page 1 22
Theory page 1 23
A) 1 1 . . . dxe4 1 2. ttJ xeS cS page 1 24
A I ) 1 3 .dxcS page 1 24
A2) 1 3 . ttJ d2 page 1 26
A3) 1 3 . .igS page 1 28
A4) 1 3 . .ie3 page 1 32
B) 1 1 . . . ttJ xe4 page 1 38
B l ) 1 2. dxeS page 1 38
B2) 1 2. ttJ xeS page 1 39
1 20 Attacking the Spanish

1.e4 e5 2.tLla tLlc6 3.Ab5 a6 4.i.a4 tLlf6 Strategic themes

5.0-0 j,e7 6Jel b5 7.j,b3 0-0 8.h3 d6 9.c3 There are a number of recurring themes in
tLla5 lO.i.c2 d5 1 l .d4 the I I .d4 variation.

8
Attacking the e4-pawn
In the 1 1 . . . dxe4 variation, It is quite common
7 for White to mount an assault against the pawn
6 on e4.

5
8
4
7
3
6
2
5
1
4
a b c d e f g h
3
This has become the most popular reaction
to the Gajewski amongst top players. White 2
eschews the gain of a pawn and instead fights 1
for the initiative.
a b c d e f g h
Introduction White can increase the pressure with Wb 1
Despite White's refusal to accept the pawn and lDg4 to exchange the defensive knight on
sacrifice, the text should not be viewed as in f6. Alternatively, he might develop his dark
any way less principled than I I .exd5 . White's squared bishop on a square like g5 . Then the
1 1 th move signifies a willingness to fight for bishop can remove the defensive knight, while
the central squares, in accordance with classical the rook on e 1 would assist in attacking the
chess principles. Both players have developed e-pawn.
the same number of pieces, although advocates
of the white side may point to the less than Black has very little chance of defending
ideal placement of the knight on a5 . the pawn by conventional means, so he will
Obviously the situation in the centre of the usually need to rely on tactical counterplay.
board will need to be clarified over the next You can find plenty of examples throughout
few moves. Black will almost always capture our coverage of line A) , beginning on page
the e4-pawn on the next turn, with either the 1 24.
knight or the d-pawn. In both cases, White
will respond by capturing the e5-pawn with White's positional trumps
his knight. The evaluation of the resultant The following type of position can occur in
positions will then hinge on a variety of tactical the 1 1 . . . lDxe4 variation. Black's pieces are quite
and strategic motifs. Both of Black's options active, but he may experience problems if he
will be explored in detail in the theoretical allows White to carry out the plan described
section. below.
 Chapter 5 - Gaj ewski: I I .d4 121

Black begins by moving his bishop out of

8
harm's way. He will follow up with a rapid
7 advance of the f-pawn, conveniently gaining
6 time by targeting the enemy knight. In
variation A3), note to White's 1 4th move on
5 page 1 29, you can find an example in which a
4 top class player was crushed by this attacking
plan.
3
2 Voluntary acceptance of doubling pawns
The following situation might arise in
1
variation B) . It is quite possible that White
a b c d e f g h will try to exchange his own 'bad' bishop for
White can arrange his knights in an ideal his opponent's 'good' one, in order to increase
way with ttJd3 and ttJfeS. In doing so he eyes his positional advantage. Whenever this
a second potential outpost square on cS . More happens, Black should always keep in mind
importantly, he frees his f-pawn which will be the possibility of recapturing with the c-pawn.
able to evict Black's best placed piece. Finally,
he will often insert the move a2-a4 at a suitable 8
moment. This helps to soften up the enemy
7
queenside, while also activating what would
otherwise be a passive rook on al . 6
5
Black's kingside attack
In some positions White may be tempted 4
to compromise his position in order to gain 3
material. In the following case, Black is unable
to regain the cS-pawn due to the possibility of 2
b2-b4, forking queen and knight. However, he 1
can instead obtain fantastic compensation as
a b c d e f g h
indicated below:
The structural transformation helps Black
in a big way, primarily by guarding the weak
8
squares on cS and eS. The open c-file is an
7 additional benefit. The doubled pawns are not
6 really weak, as White has no way of attacking
them.
5
4 Playing on the light squares
Sometimes White will exchange the knight
3 on e4 and install his own knight on cS . In
2 response, Black can consider exchanging this
piece for his dark-squared bishop. The resulting
1
position may look something like this:
a b c d e f g h
1 22 Attackin g the Spanish

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
a b c d e f g h a b c d e f g h
This type of position can occur with or This is the main tabiya for the 1 1 . . . dxe4
without queens. Generally speaking, Black's variation. White has tried many different
control over the light squares combined with his moves, as seen at the branching point for line
better pawn structure should provide sufficient A) on page 1 24.
compensation for the loss of the bishop pair.
In order for this idea to work properly, it is An interesting piece sacrifice
crucial that White must already have moved In the following line I found a novelty on
his a-pawn. If this pawn were back on a2 then move 1 4, followed by a promising sacrifice.
White could simply play b3 to bolster his light
squares, after which the knight on a5 would l.e4 e5 2.tLla tLlc6 3.i.b5 a6 4.L4 tLlf6
look very silly indeed. This idea can be found 5.0-0 JJ.e7 6J3e1 b5 7.JJ.b3 0-0 8.h3 d6
in line B22 1 ) on page 145, as well as the note 9.c3 tLla5 10.Ac2 d5 1 1.d4 dxe4 12.tLlxe5 c5
to White's 1 9th in line B2222) on page 1 49. 13.dxc5 VIIc7 14.tLlg4 i.xg4!? 15.hxg4

Theoretical Highlights 8
A starting point 7
Virtually all games involving 1 1 . . .dxe4 reach 6
the following position.
5
1 .e4 e5 2.tLla tLlc6 3.Ab5 a6 4.h4 tLlf6 4
5.0-0 Ae7 6Je1 b5 7.Ab3 0-0 8.h3 d6 9.c3
3
tLla5 10.i.c2 d5 1 1 .d4 dxe4 12.tLlxe5 c5
2
1
a b c d e f g h
Black can now launch a sacrificial attack
with:
 Chapter 5 - Gajewski: I I .d4 1 23

15 ... hc5! 16.b4 .ixflt 17.<if?xf2 ti'hl 13.ttld3 i.d6 14.ttld2 e8 15.ti'h5!
According to my analysis this should be
enough for at least a draw. Full details can be
found in line A I ) , beginning on page 1 24.

The acid test

According to the latest games and analysis,
the following looks like the most critical test of
the 1 1 . . . dxe4 variation.

l.e4 e5 2.ttla ttlc6 3.Ab5 a6 4.i.a4 ttlf6

5.0-0 i.e7 6J: e 1 b5 7.i.b3 0-0 8.h3 d6
9.c3 ttla5 10.Ac2 d5 1 1.d4 dxe4 12.ttlxe5 c5
13.i.e3 i.b7 14.ttld2 ti'c7 15.ti'bl!
a b c d e f g h
8 This is line B222) , coverage of which
begins on page 1 45 . My own choice with
7
Black was 1 5 . . . .ib7 (line B222 1 ) , page 1 46) .
6 Although I suffered an unpleasant reversal
against Zoltan Almasi, it is clear that there is
5
a lot more for both sides to discover in this
4 variation. Alternatively, in line B2222) on page
3 1 49, I provide coverage of the as yet untested
1 5 . . . c6!?N. According to my analysis Black
2 should be alright here, although we will need
1 to wait for some practical tests before drawing
any firm conclusions.
a b c d e f g h
White targets the e4-pawn directly. In line Theory
A42) , beginning on page 1 36, you can see how
Black has attempted to combat his opponent's 1.e4 e5 2.ttla ttlc6 3.i.b5 a6 4.h4 ttlf6
plan. So far he has not been able to demonstrate 5.0-0 Ae7 6Jel b5 7.Ab3 0-0 8.h3 d6 9.c3
a convincing route to equality. ttla5 10.i.c2 d5 1 l .d4
The present chapter will be organised slightly
The ultimate main line differently from most others. The Gajewski has
Due to the current status of the previous only been around for the past couple of years,
variation, Black's primary equalising attempt and fresh ideas are still being tested on a regular
is currently connected with 1 1 . ..tt:)xe4 rather basis. Instead of recommending a specific
than 1 1 . . . dxe4. At the time of writing the repertoire, I will instead conduct a slightly
following appears to be the most critical test. wider theoretical survey, while at the same time
providing some specific recommendations. The
1.e4 e5 2.ttla ttlc6 3.i.b5 a6 4.h4 ttlf6 intention is to offer the best possible coverage
5.0-0 i.e7 6Je1 b5 7.i.b3 0-0 8.c3 d6 9.h3 of the entire 1 1 .d4 variation, according to the
ttla5 10.i.c2 d5 1 1 .d4 ttlxe4 12.ttlxe5 f6 information available at the present time.
1 24 Attackin g the Spanish

The chapter will be divided between the 14 .ixg4!?N

.

two primary options: A) 1 1 dxe4 and B). This is a new idea. Black gives up the bishop
1 l ... lLlxe4. pair in order to accelerate his development. A
previous game had seen:
Instead 1 1 . . . exd4 is unlikely to equalise after 1 4 . . . lLlxg4 1 5 .hxg4 fS
1 2.e5, e.g. 1 2 . . . lLle4 1 3.cxd4 c5 1 4.lLlbd2 with 1 5 . . . .ib7 is not very logical. After 1 6.ixe4
a nice position for White. i!ad8 1 7. f3 ixe4 1 8.i!xe4 White has
improved his coordination while taking a
A) 1 1 . dxe4 12.lLlxe5 c5
second pawn.
1 6.d5t @h8 1 7.xa8 ixc5
Inferior is 1 7 . . . .ib7? 1 8.xfSt ixfS 1 9.9xfS .
8
White has a lot of material for the queen,
7 while Black has no real attack.
6
8
5
7
4 6
3 5

2 4

3
1
2
a b c d e f g h
1
This is the universal choice. Black needs to a b c d e f g h
challenge the enemy centre while also dealing
with the threat of b4. At this point White faces This position was reached in Apicella -
an important choice between four alternatives: Hamdouchi, Clichy 2008, which soon
AI) 12.dxc5, A2) 12.lLld2, A3) 12.i.g5 and resulted in a draw after:
A4) 12.Ae3. 1 8.d5 ? .ib7 1 9.e6
1 9.d l ?! i.xf2t 20.@xf2 h2 is only
AI) 13.dxc5 dangerous for White.
This simplistic move releases the central 1 9 . . . .ic8 20.d5 .ib7 2 1 .e6 .ic8 22.d5
Y2-Y2 .
tension. Logic tells us that this should not
worry Black, although I needed to find a new
move in order to demonstrate it. However, White could have refuted his
opponent's play with:
13 lYc7

1 8.b4! ixf2t
1 3 . . . ixc5 is also a playable option, although 1 8 . . . .ib7 1 9.xfSt ixfS 20.bxa5 is winning
the text would be my own choice. for White.
1 9.@xf2
14.lLlg4! With an extra rook and bishop, White has
This is the most challenging move. White an overwhelming material advantage and
intends to win the e4-pawn, and it is up to should win easily provided he avoids being
Black to demonstrate compensation. mated.
 Chapter 5 Gajewski: 1 1 .d4
- 1 25

1 9 . . . h2 15.hxg4 .bc5!
1 9 . . . fxg4t? 20.c;!?gl defends easily.
2o.lLld2
8
7
6
5
4
3
2
1
a b c d e f g h
a b c d e f g
Black's attack should not be sufficient: In this position the sacrifice appears to be
playable.
a) 20 . . . h4t? is feeble: 2 1 .c;!?e2 xg4t 22.c;!?f1
White escapes the checks and claims the full 16.b4
point easily. Without this move Black would obtain
an easy game, e.g. 1 6.he4 ae8 1 7 . .if3
b) After 20 . . . fxg4t 2 1 .c;!?e2 xg2t 22.c;!?d l (or 1 7.f3 lLlc4! with a great position)
lLlc4 23. xe4 White also wins. 1 7 . . . xe l t 1 8.xe l e8 1 9.f1 lLlc4! White
has great difficulty developing, e.g. 2o.lLld2?
c) 2o . . . lLlc4! lLle3! 2 1 .fxe3 he3t 22.c;!?h l g3 and
This looks like the best chance, although wins.
White should still win after:
2 1 .d5! e3t 22.c;!?e2 .ie6 16 .bflt 17.<bxn Yfh2

22 . . . exd2 23.hd2 only helps White.
23.lLlf1 ! h4 24.d4 f4
24 . . . f2t is also insufficient after 25.c;!?dl
(25.c;!?d3?? xe l ) 25 . . . f4 26 . .tfS! hf5
27.gxfS e8 28 . .td2!.
25.lLlxe3! fxe3 26 . .if5! f2t 27.c;!?dl hfS
28.gxfS e8 29.d7! f8 30.d3 xg2
3 1 .e2
The king will escape to c2. It would be
difficult for most players to find so many
accurate defensive moves over the board,
so adventurous souls might still consider
rolling the dice with this sacrificial approach.
Unfortunately, it simply does not cut the
mustard as a main recommendation.
Also deserving attention is 1 7 . . . lLlc4!? 1 8 . .ig5
ad8 1 9.e2 e5 2o.hf6 xf6t 2 1 .c;!?gl
b6t 22.f2 e3 23.f4 fe8. Black certainly
has some compensation here, although overall
I consider this approach to be slightly riskier
than the main line.
18Jxe4!?
This is more ambitious than 1 8.he4 h4t
1 9.c;!?gl lLlxg4 20 . .id2 (White is not helped
by 20 . .ie3? lLlc4 2 1 ..ic5 fe8) . At this point
Black can force a draw immediately with
20 . . . f2t 2 1 .c;!?h l h4t should he wish it.
Alternatively he can continue the fight with
20 . . . ad8!?, although this brings no advantage
against accurate play:
1 26 Attacking the Spanish

20 Jad8 21 .'ilYe2 Ue8 22.ttld2 ttlc4

.

23.ttlxc4 xe4 24.ttle3 'ilYf6

Black's initiative fully compensates his small
material deficit.

A2) 13.ttld2
This leads by force to a position where Black
needs to defend carefully for a long time in
order to achieve a draw.

a b c d e f g h 13 ... cxd4 14.ttlxe4 dxc3 15.'ilYf3! .ib7

2 1 .bxa5 :gd6!? (2 1 . ..'lNf2t once again draws) 1 5 . . . cxb2 would be too greedy. After 1 6.i.xb2
22.'lNf3 'lNh2t 23.@f1 :gf6 24.i.f5 'lNh5 i.b7 1 7.:gad 1 White is tremendously active.
25 .@gl (White can also consider playing
on with 25 .i.xg4!? 'lNxg4 26.'lNxf6 gxf6 8
27.i.h6 with a highly unclear position.) At
this point Black really has no choice but to 7
repeat the position with: 25 . . . 'lNh2t 26.@f1 6
'lNh5
5
18 ... ttlxe4t 19 ..be4 'ilYh4t 4
3
8
2
7
1
6
a b c d e f g h
5
16.b4!
4 Without this move White would struggle to
3 justify his play.

2 16 ....be4
1 An equally valid move order is 1 6 . . . ttJxe4
1 7.i.xe4 i.xe4 1 8.'lNxe4. However, Black must
a b c d e f g h avoid 1 6 . . . i.xb4? 1 7.i.g5 .
20.<bgl
The alternative is: 20.g3 'lNh2t 2 1 .@f1 17..be4 ttlxe4
:gad8 22.'lNe2 'lNh3t 23.i.g2 'lNxg3 24.'lNf3 1 7 . . . i.xb4? is refuted by 1 8.i.g5!, e.g.
'lNxf3t 25 .i.xf3 ttJc4 Black should be alright 1 8 . . . i.e7 1 9.:gad 1 'lNc8 20.i.xf6! i.xf6 2 1 .ttJd7!
in this ending. His pieces are more active and winning.
he has good chances to advance his kingside
pawns. The white king is also less than 18.'ilYxe4 f5
comfortable. Otherwise Black loses a piece.
 Chapter 5 - Gajewski: I I .d4 1 27

White stands slightly better. Black has three

8
vulnerable pawns on a6, c4 and f5, although
7 the c-pawn could also prove to be an asset in
6 some endings. The advance of the f-pawn has
also weakened his kingside. Whether or not
5 these factors amount to a significant advantage
4 for White is open to debate.
3 23 .if6 24.adl!
..

2 24Jac 1 is less precise. The game Caruana

- Aveskulov, Plovdiv 2008, soon resulted in a
1
draw after 24 . . . 1&d3 25.1&xd3 cxd3 26Jxc8
a b c d e f g h xc8 27.e3 c2 28.xd3 xa2 29.g3 Yz-Yz.
19.1!
This i s White's only try for an advantage. In 24 ...'ilYd3
Hillarp Persson - S.B.Hansen, Denmark 2007, It is hard to determine Black's most accurate
he got nowhere with 1 9.1&f4 llJc4 20.11Jxc4 move here.
bxc4 2 1 .1&xc4t @h8 22.i.e3 Y2-Yz. 24 . . . 1&b6
This is a legitimate alternative to the main
19 c2!
..
line.
Forced, as knight moves would allow a fork 25 .i.c3 hc3 26.1&xc3 fd8 27.xd8t 1&xd8
on c6. 28.e6

20.2
White has no choice, as the queen must
defend the b-pawn.

20 c4 21.xc4 bxc4 22.'ilYxc2 c8

..

23.i.d2

8
7 a b c d e f g h
6 28 . . . 1&d3
5 28 . . . 1&d l t 29.@h2 1&d3 30.1&e5! 1&e4
3 1 .1&d6! 1&b7 32.a3 1&b8 (32 . . . c3 might
4
lead to a similar ending after 33.e7 1&b8
3 34.1&xb8 xb8 35.c7 a5 36.xc3 axb4
37.b3) 33.1&xb8 xb8 34.c6 a5 35.bxa5
2
(35 .xc4 axb4 36.axb4 is also unpleasant)
1 35 . . . b3 36.xc4 xa3 37.c5 Black faces
a b c d e f g h a long and difficult defence.
29.1&e l !
1 28 Attackin g the Spanish

This is a very precise move. Instead 29.V;Ve5
V;Vb 1 t 30.@h2 V;Vxb4 3 1 .e7 V;Vb8 32.xg7t
@f8 33.V;Vxb8 xb8 34Jhh7 b2 should be
a draw.
29 . . . @f8
After 29 . . . h5 30.xa6 V;Ve4 3 1 .V;Vc3 V;Vd3
32.V;Vxd3 cxd3 33.d6 c l t 34.@h2 b l
35J!d4 White has good winning chances.
30.V;Ve5 V;Vb l t 3 1 .@h2 V;Vxb4 32.V;Vxf5t @g8
33.V;Vd5;!;
8 to offer maximal resistance after:
7
After 28.d6 xa2 29.xd3 b8 it will be
difficult for White to make progress.
28 .. Jxa8 29 ..le3 ga3
Also deserving consideration is 29 . . Jha2
30.xd3 b2 3 1 .d7t (3 1 .i.c5!? ) 3 1 . . . @e6
32.b7 h5!? with good defensive chances.
30.b5
We have been following the game Kravtsiv -
Brkic, Gaziantep 2008, in which Black failed

6 30 ...i.e5 31 .g3 e6 32.b6

5
White had very good winning chances and
eventually triumphed. However, Black could
4
have improved with:
3

2 30 .. Jxa2N 31.xd3 gb2 32.d7t e6

1 33.gb7 g5!
a b c d e f g h

White maintains slight pressure, but I doubt 8

that it is enough to win. 7
33 ... @h8 34Jk6 V;Vb8t 35.g3 xc6 36.V;Vxc6
V;Vg8 37.V;Vxa6 h6 38.V;Vc6 c3 39.V;Vxc3 V;Vxa2 6
Black should hold this endgame without any 5
real problems.
4
25.YHxd3 cxd3 26Je6 f7 27Jxa6 a8 3
2
8
1
7
a b c d e f g h
6
Active defence is the key. I believe that the
5 position is closer to draw than to a white win.
4 For instance:
3
34.g3 e5 35. g2 f4 36.gd4t gd4 37.Aa7
2 e4 38.gxh7 at 39. g3 gxb5 40.i.e3 gb8
1 41.g4 g8t 42.h5 f5
Black draws easily.
a b c d e f g h
28.xa8 A3) 13.i.g5 i.b7

a b c d e f
Here White faces an important choice
between A31) 14.tLld2 and A32) 14.a4.
After the feeble 1 4.dxcS ?! c7! Black's
position might already be preferable. The
game Volokitin - Wojtaszek, Crete 2007, soon
turned into a disaster for White after:
I s .Lf6?! ad8! 1 6.e2
1 6.1L1d7 xd7 1 7.i.d4 Lcs 1 8 ..he4 Le4
1 9.xe4 1L1c6+
1 6.g4 Lf6 1 7.c6 lLlxc6 1 8.1L1xc6 Lc6
1 9 .he4 fe8+
16 . . . Lf6 1 7.1L1g4 i.gS
17 . . . i.h4! would have been even stronger
thanks to the additional pressure against n.
1 8.1L1e3 fS! 1 9.b4 f4
This pawn will help to destroy the opposing
king's protective cover.
20.1L1fl
No better is 20.bxaS fxe3 2 1 .fxe3 i.h4.
a b c d e f g h
Chapter S Gaj ewski: I I .d4
- 1 29
20 . . . 3! 2 1 .gxf3 e3! 22.fxe3
The last chance may have been 22.1L1xe3
although 22 . . . Lf3 23.fl still leaves White
in trouble after 23 . . . 1L1c6, or even 23 . . . 1L1c4!?
24.1L1xc4 bxc4 2S .xc4t h8.
22 . . . xf3
White's kingside is in ruins, and he soon
lost after:
23.e4 df8 24.g2 i.h4 2S.1L1bd2 xc3
26.ec 1 i.c8! 27.bxaS xcst 28.h l n
29.i.b3t h8 30.xc3 xc3 3 1 .xn Ln
32.d l i.xh3 33.1L1h2 g3 0-1
g h
A31) 14.tLld2 cxd4 15.cxd4 ge8
Black gains nothing from I S . . . xd4?!
1 6.1L1xe4.
Also unsatisfactory is I S . . . c8 1 6.i.xf6 hf6
1 7.1L1xe4 Le4 1 8.he4 lLlc4 1 9.b3 ( 1 9 . .ifS
c7 20.b3) 1 9 . . . 1L1xeS 20.dxeS xd l
2 1 .gaxd l heS 22.Lh7t xh7 23.xeS.
White has good winning chances, although in
Pierrot - Liascovich, Buenos Aires 2008, Black
eventually salvaged a draw.
16.tLlb3
In Charbonneau - Benjamin, Internet 2007,
Black equalised easily after 1 6.Lf6 i.xf6
1 7.1L1xe4 LeS 1 8 .dxeS xeS.
a b c d e f g h
So far we have been following the game
1 30 Attacking the Spanish

Dominguez - Carlsen, Khanty-Mansiysk

8
2007. At this point the young Norwegian
prodigy erred with: 7
6
16 xb3?! 17 .hb3 d5
..

1 7 . . . i.d5 1 8Jk 1 h6 1 9.i.h4;t;; also leaves 5

Black under pressure. 4

18.he7 gxe7 19Jcl f4 3

After 1 9 . . .'IWd6 correct is 20.c2!;t;; (Marin) . 2
1
Black also fails to solve his problems with
1 9 . . . a5 20.a3 a4 2 1 .i.a2;t;; . a b c d e f g h
This position appears balanced. For
20.%Yg4 g6 instance:
20 . . . tDd3 can be met by 2 1 .f5!, at which
point Black should play 2 1 . . . tDxe5 22.dxe5 19J1cl gc8 20 ..ibl h6 21 ..th4 .td5
transposing to the game. He only risks falling Black has a full share of the chances.
into even greater difficulties with: 2 1 . . .g6
22.f6 tDxe5 (other moves lead to a fatal loss A32) 14.a4
of material) 23.dxe5 d7 24. xd8t dxd8
25 .e6 fxe6 26.i.xe6t @h8 27.c7
8
21 .%Yf5 xe5 22.dxe5 7
White's pieces are much more active,
although the game was eventually drawn. 6
5
Fortunately Black has a significant improvement
4
available.
3
16 c4!N
.
2
This was suggested by Marin in his
annotations to the aforementioned game. It 1
appears to be Black's only route to equality. a b c d e f g h

17.xc4 White plays to soften up the enemy

Harmless is 1 7.i.xf6 i.xf6 1 8.tDc5 i.xe5
1 9.dxe5 xd 1 20.axd 1 tDxb2 2 1 .tDxb7
(2 1 .b 1 does not win a piece in view of
2 1 . .. ac8) 2 1 . .. tDxd 1 22.xd 1 xe5 when
Black is certainly not worse.
queenside. He does not mind postponing the
development of the knight on b 1 , as this piece
may be able to find a better home on c3.
14... cxd4
In Luther G. Jones, Maastricht 2008,
-

17 bxc4 18.c5 .*.c6

..
Black tried a different approach:
1 8 . . . i.d5 ? 1 9.i.a4! would be very annoying. 1 4 . . . e8!?
 Chapter 5 - Gajewski: 1 1 .d4 131

This is riskier than the main line, but may

nonetheless be playable.
1 5 .axb5
After 1 5 .dxc5 V!ffc7 1 6.hf6 hf6 1 7.ti:)g4
i.g5 Black has reasonable compensation,
e.g. 1 8.axb5 axb5 1 9.V!ffd6 V!ffc 8 20.ti:)a3 a6
2 1 .V!ffd l d8 22.V!ffb l f5 23.ti:)e3 i.h4 24.b4
f4 25.ti:)g4 h5 26.ti:)e5 e3 with a strong
initiative for the pawn.
1 5 . . . axb5 1 6.ti:)d2
1 6.dxc5 V!ffc7 1 7.hf6 hf6 1 8.ti:)g4 i.g5
reaches the above note.
1 6 . . . cxd4 1 7.cxd4 V!ffxd4 1 8.ti:)xe4 V!ffxd l a b c d e f g h
1 9.exd 1 At this point, instead of the game's 1 8 . . . he4?
1 9.xe4, Black could have equalised with
1 8 . . . V!ffxd4 1 9.hb7 ti:)xb7.

However, White also missed an earlier

opportunity to improve with:
1 7.ti:)b3!N

8
7

6
a b c d e f
5
1 9 . . . ac8 4
1 9 . . .he4 20.he4 ti:)xe4 2 1 .he7 f6 is
3
equal.
2
20.hf6?!
The superior 20.ti:)c3 ti:)c6 would have led to 1
a balanced position. L-
a b g

20 ... gxf6 2 1 .ti:)g4 @g7 22.xa5 xc2+ Instead of targeting the e4-pawn, White plays
Black's bishop pair and activity were more to activate his pieces. Play may continue:
significant than the pawn weaknesses on the 1 7 . . . i.d5
kingside. 1 7 . . . ti:)xb3 1 8.hb3 ti:)d5 ( l 8 . . . i.d5 1 9.c 1 )
1 9.he7 xe7 20.c1 leaves Black under
15.cxd4 pressure.
We are following the game Kamsky - Miton, 1 8.ti:)xa5 V!ffxa5 1 9.hf6 hf6 20 ..be4 he5
Dagomys 2008, which continued: 2 1 .dxe5 he4 22.xe4 xe5 23.f4;!;;
Black's position is deceptively difficult, and
15 ... h4 16.ttld2 e8 17.Lf6 .bf6 18.,be4 he faces a long fight for a draw. His best
1 8.ti:)xe4 he5 1 9.dxe5 xe5 is equal. course of action may be to activate his pieces
at the cost of a pawn with:
1 32 Attacking the Spanish

23 .. J!ae8 24.b3 c7 25.xb4 g6 23.gxel .lc6 24.xe4

Black has reasonable chances to hold the
draw, although the ending will not be much 8
fun for him.
7
In view ofthe above, I would like to recommend 6
an improvement on move 1 5 . Curiously
enough, the same idea was also suggested in 5
line A3 1 ) above. Not surprisingly it involves 4
improving the position of the a5-knight.
3
15 ... e4N! 2
1
a b c d e f g h
White has won a pawn, but the bishop pair
will be very powerful in this open position
featuring pawns on both sides. I consider the
most likely result to be a draw after 24.. Je8
or 24 ... gc8!?

A4) 13 ..le3

8
a b c d e f g h 7
This is a useful improvement of Black's worst 6
placed minor piece.
5
16.xe4 4
1 6.b3 ttJd6 gives Black no problems.
3
16 ... bxe4 17.e3 2
This is dearly the best square for the knight. 1

Instead, should play continue with 1 7.ttJd2?! a b c d e f g h

c8, Black would be fine. This has emerged as White's main try for an
advantage.
17.. Je8 18.Wd2 h6 19 .ixf6 .ixf6 20.xe4
.

20.he4 leads to equality after 20 . . . he4 13 ....lb7 14.d2 We7

2 1 .xe4 xe4 22.ttJxe4 xd4 23.ttJxf6t This seems to be best.
xf6.
In Kotronias - Mastrovasilis, Igoumenitsa
20...Wxd4 21.Wxd4 .ixd4 22.d6 gxelt 2007, Black's position quickly went downhill
 Chapter S Gajewski: 1 1 .d4
- 1 33

after 1 4 . . . dS ?! I S .tDg4! cxd4 (or I S . . . tDxg4 A couple of other moves have also been tried.
1 6.hxg4 fS 1 7.gxfS xfS I B.f3, intending to
meet I B . . . .ih4 with 1 9.tDxe4! he I 20.tDgS +-) I S Jk l .B:adB 1 6.a3 is hardly the most critical
1 6.hd4 e3? 1 7.tDxe3 and White won easily. approach, although White is still not worse.
Spraggett - Zubarev, Cappelle la Grande 200B,
A more interesting alternative is: continued 16 ... tDc6 1 7.tDxc6 xc6 I B.e2
1 4 . . JkB .B:feB with a balanced position.
This was played in Macak - M.lvanov,
Prague 200B. This would be a decent move In Arakhamia - Wells, Liverpool 200B, White
were it not for the strong reply: obtained an edge with: I S .a4 .B:fdB 1 6.axbS
I S .b l !N axbS 1 7.tDg4 tDxg4 ( 1 7 . . . tDeB I B.tDxe4 fS
Unfortunately this accurate move spoils [ I B . . . cxd4 1 9.hd4 fS 20 . .B:xaS! .B:xaS 2 1 ..ib3t
Black's idea. @hB 22.tDh6+-] 1 9.tDxcS hcs 20.hfS
I S . . . cxd4 1 6.hd4 .ics 1 7.tDxe4 hd4 White has three pawns and an attack) I B.hxg4
I B.cxd4 .ixe4 cxd4 1 9.hd4 tDc6 20 . .B:xaB haB 2 1 .he4
I B . . Jxc2 1 9.xc2 tDxe4 20J!xe4 he4 tDxd4 22.cxd4;1;;
2 1 . xe4 leaves White a safe pawn up.
1 9.he4 xd4 However, I think that Black could have
Or 1 9 . . . tDxe4 20.xe4 improved with l S . . . b4!N leading to the
20.b4! tDc4 2 1 .tDc6 following possibilities:
White wins the exchange as 2 1 . . .d6 can be
met by 22.tDe7t anyway. a) 1 6.cxb4 cxd4 1 7.hd4 hb4 looks at least
equal for Black.
8
b) 1 6 . .B:c 1 bxc3 1 7.bxc3 .B:adB reaches a tense
7 position with chances for both sides.
6
c) 1 6.e2 tDdS!
5
4
3
2
1
a b c d e f g h
We have reached a final branching point for
the 1 1 . . . dxe4 variation. One critical idea is
A4I) I5.i.f4, losing a tempo with the bishop in a b c d e f g h
order to embarrass the enemy queen. This can
lead to great complications, with some chances White has two ways to capture the e-pawn:
of a white advantage if he plays very precisely.
However, I believe that the greatest threat to c 1 ) 1 7.tDxe4 fS I B.cxb4 tDxb4 1 9.tDxcS hcs
Black's survival comes from A42) I5.Bbl!. 20 . .B:ac 1 d6 2 1 .hfS tDb3
1 34 Attackin g the Spanish

The complications turn out in Black's Black is not helped by 22 . . . xd4?! 23.d3.
favour. 23.bxc3
23.xc3 xd4 24.d3 b6 should not
c2) 1 7.i.xe4 bxc3 1 B.bxc3 tDxc3 1 9.i.xh7t worry Black.
1 9.d3? tDxe4 20.tDxe4 c4 2 1 .c2 f6 23 . . . i.xe5
22.tDf3 acB is clearly better for Black. 23 . . . f6 may not give quite enough
1 9 . . . @xh7 20.d3t 5 2 1 .xc3 f4 compensation after 24.tDd3 c6 25.f1
White's situation appears difficult, but he can i.xg3 26.fxg3.
secure half a point with a tactical sequence. 24.i.xe5
22.c2t @gB 23.i.xf4! xf4 24.tDb3! tDxb3 24.dxe5 c6
There is nothing better. 24 . . . .td5
25 .c4t! Black has reasonable compensation, and the
25 .xb3t?? c4 wins. opposite-coloured bishops will enable him to
25 . . . @h7 26.d3t hold many endings even a pawn down.
26.c2t?? .te4! 27.xe4 tDxa1 hits the
queen. 16 ... d5
26 . . . @gB 27.c4t
The game ends with a repetition of moves. 8
7
A41) 15.M4 cxd4 16.xe4
This is the most ambitious move. 6
5
1 6.tDg6 is unpromising after 1 6 . . . .td6
1 7.i.xd6 xd6 1 B.tDxfB d3 1 9.tDxh7 @xh7, 4
when Black's powerful pawns make up for his 3
small material deficit.
2
1 6.cxd4 tDd5 1 7.xe4 .td6 1 B . .tg3 acB 1
1 9.c1 tDb4
a b c d e f g h
8 17..lg3
7
1 7.h5!? demands an accurate response:
1 7 . . . 5! 1 B.tDg5 i.xg5 1 9.i.xg5 dxc3 20.bxc3
6
g6!? Black's slightly airy kingside is compensated
5 by his opponent's weaknesses on the opposite
4 Hank.
3
17... f5!
2
Once again Black should not shy away from
1 complications.
a b c d e f g h
18.g6 Yfc6 19.xf8 xf8 20.Yfxd4
This looks to be winning a piece, but White
After 20.cxd4 fxe4 2 1 .i.xe4 .tb4 (2 1 . . .tDc4??
can rescue himself with:
22.h5+-) 22.e2 g6 the position is
20.e3! tDxc2 2 1 .c3 b6 22. l xc2 xc3
balanced.
 Chapter 5 - Gaj ewski: I I .d4 1 35

2o ... lLlc4 a pawn versus three minor pieces. White seems

After 20 . . . fxe4 2 1 .xe4 g6 22J!ad a Black to be better, e.g. 27 . . . lLlb5 (27 . . . f6? 28.e4!
faces a difficult defence. is almost winning for White) 28.d3.

8
24 ..i.f3 xf3
White was threatening to exchange on d5
7 followed by e5 winning material.
6
24 . . . c8? is no good in view of 25 .ad l .
5
4 25.ftxf3
25.gxf3 lLlxc3 26.xc6 Lc6 (Kosten)
3
reaches a very unclear ending.
2
1
25 lLlc2 26.c4 bxc4 27.bxc4 lLldb4 28.ftxc6
.

hc6 29.e5 .txf2t 3o ..ixfl lLlxal 31 ..ic5!?

a b c d e f g h lLld3 32.f5
We have reached a critical position. The game
z.Almasi - Stevie, Plovdiv 2008, continued: 8
7
21.a4? lLlf6+
White was losing material for insufficient 6
compensation, and was unable to salvage the
5
game.
4
Instead, the critical continuation would have 3
been:
2
21 .b3! 1
The position is extremely complicated but it
seems that White can retain an advantage. a b c d e f g h
In this position White's rook and pawn
21. lLla3 22.i.dl! fxe4 23.ftxe4
. should be more effective than the opponent's
Positions featuring rook versus two minor two minor pieces. The passed c-pawn is a
pieces can be hard to evaluate. This one would threat, and the knight on al is misplaced. A
be fine for Black, were it not for the fact possible continuation is:
that his pieces are rather unstable and badly
coordinated. The immediate threat is .if3 32 h6 33J:U'8t h7 34.gc8 .tb7 35Jc7
.

which might be followed by ad l . lLlxc5 36.gxc5 lLlc2 37.f2

Black faces a difficult defence.
23 ...Ac5
Another possibility is 23 . . . .if6 24 . .if3 Lc3 It appears that White may be able to obtain an
25 .ac 1 b4 26.ed l xf3 27.xf3, leading to advantage with 1 5 . .if4, although he will need
an unusual material balance of two rooks and to find a lot of accurate moves to do so.
1 36 Attacking the Spanish

A42) 15.Vbl! 8
7
8 6
7 5

6 4

3
5
2
4
1
3 a b c d e f g h
2 Black has tried a few different approaches
1 here. The position is rather complicated, but
ultimately it seems that White can always
a b c d e f g h
maintain an edge.
This has become White's principal weapon
in the 1 1 . . . dxe4 variation. The e4-pawn is now a) 2 1 . . . lDc6 gives White a choice:
under heavy fire, and Black will have to stir up
some complications. a 1 ) In Kotronias - Estremera Panos, Liverpool
2008, Black obtained a good position after
15 ... cx:d4 16.hd4 22.i.e3 he3 23.xe3 lDe5 24.Wtd 1 (24.Wte 1
1 6.cxd4?! makes no sense here. Kreisl - lDg4 25 .lDf6t lDxf6 26.gxf6 Wtc6 27.f3 xe3
Freitag, St Veit 2008, continued 1 6 . . J!ac8 28.Wtxe3 Wtxf6 29.d 1 g6 is close to equal
( 1 6 . . . i.b4!? is also possible) 1 7 Jcl ( 1 7.he4 for Black - Korneev) 24 . . . lDc4 25.e2?!
lDxe4 1 8.lDxe4 f6 1 9.1Df3 lDc4 gives Black lDxb2 26.Wte 1 e6+. Therefore White should
more than enough for a pawn) 1 7 . . . Wtd6 with investigate one of the alternatives:
a good position for Black.
a2) 22.i.f6!? is a promising idea of Kosten, who
16 ...,id6 offers the following sample variation: 22 . . . lDe5
Otherwise Black will simply lose the e4- 23.Wtd 1 h6 24.Wth5 e6 25 .gxh6 gxf6 26.Wtf5
pawn for no compensation. i.xh6 27.lDg3, with a very dangerous attack.

17.tLg4 xg4 18.hxg4 rut 19.mhl 5
This does not appear to equalise, although it
is doubtful that Black has anything better.
1 9 . . . Wtf4?? 20.lDfl was a disaster for Black in
Baramidze - Djuraev, Gaziantep 2008.
19 . . . i.f4 20.lDxe4 fe8 2 1 .g5
This has been played in a couple of games,
although 2 1 .Wtd l ! ? also deserves attention.
a3) 22.i.c5 also looks strong. Korneev analyses
22 . . . lDe5 23.Wtd 1 ad8 24.Wth5 he4 25.xe4
Wtxc5 26.xf4 lDg6 27.f3 d2 28.b4 Wtb6
29.i.b3 e7 when Black does not have enough
for the pawn.
b) Korneev - De Jong, Porto Mannu Palau
2008, continued: 2 1 . . .e6 22.lDc5 (22.lDf6t!?
gxf6 23.i.xh7t also looks dangerous)
22 . . . xe 1 t 23.Wtxe 1
 Chapter 5 Gaj ewski: I I .d4
- 1 37

8
Black is not helped by: 20 . . . !!xfS 2 1 ..ixe4
!!h5 22 . .ixb7 lDxb7 (no better is 22 . . . .if4t
7
23.i>gl lDxb7 24.YNd l ) 23.YNd l !!h6 24.YNb3t
6 i>hB 25 .!!e6
5

4 So far we have been following the game

3
2
1 The following sequence now looks more or
a b c d e f g h
b l ) 23 . . . .idS 24.YNe2 is dearly better for White
according to Korneev.
b2) 23 . . . .ic6 24.YNe2! is another strong
suggestion from the Russian Grandmaster.
The idea is that 24 . . . .ixgS can be refuted
by 25.lDe6! YNe7 (25 . . . fxe6 26.YNxe6t i>hB
27.YNfS+-) 26.YNh5 h6 27.lDxg7+- (Korneev) .
b3) In the game Black avoided these
tactical problems with 23 . . . .ixg5, but after
24.lDxb7 lDxb7 he had lost the bishop pair.
Following 25 .YNe4 g6 26.!!e l White had an
indisputable advantage, and went on to win
convincingly.
20.gxf5
Kononenko - Gajewski, Pardubice 200B,
which continued:
2 1 .lDf1
less forced:
2 1 . . .!!xfS 22 . .ixe4 !!h5t 23.i>gl .ih2t
24.lDxh2 YNxh2t 25.i>f1 YNf4 26 . .ixb7 lDxb7
27.YNd3
White kept an extra pawn as well as the
superior minor piece, but her king was still
not completely happy.
27 . . . lDd6 2B.YNf3?
Better was 2B . .ie3 !!h l t 29.i>e2 YNg4t
30.i>d2 !!dB 3 1 .YNd5t! lDf7 32 . .id4! h6
33. YNf3 YNxf3 34.gxf3, when White retains
an edge.
2B . . . !!h l t 29.i>e2 !!eBt 30 . .ie3 YNc4t 3 1 .i>d2
!!xe l 32.i>xe l
Black enjoyed excellent compensation.
Despite the improvement noted at move 2B,
White could have obtained an even better
position with:

8 21.lLlxe4! gxf5 22.VHdl gaf8 23.VHg4 lLlc4

7 24.lLlg3 gg5
It is hard to suggest any improvements for
6
Black.
5
4 25.VHe6t <i!.>h8
Perhaps a better try, relatively speaking,
3 would be 25 . . . YNf7 26.YNxf7t i>xf7. The
2 trouble is that even if Black regains his pawn
by taking on g3, he will still stand dearly worse
I
thanks to the opponent's strong bishops.
a b c d e f g h
26.VHh3 h5 27.lLle4
20 ....i.f4 We have been following the analysis of
1 3B Attacking the Spanish

Korneev. Both sides have positioned their
pieces on active squares, but White stands
clearly better. Aside from his extra pawn, he
also dominates the centre thanks to his securely
anchored pieces.
1 2 . . . c5 1 3.ge2 ttJc4 1 4.b3 ttJb6 1 5 .b4
e7 1 6.ttJd4 d7 1 7.f4 c5 1 B.bxc5 hc5
1 9.he4 dxe4 20.gxe4 gcB 2 1 .iVf3 ttJd5
22.ge 1 e6 when Black had more than enough
compensation for the missing pawn.

Overall, it seems to me that 1 5 .iVb 1 is quite a 13.exf6 hf6 14.tLlbdl

solid bet for a white advantage. The position
can become quite complicated, but after
correct play from both sides I do not see how
Black can equalise. For this reason, at the
present time I consider the following variation
to be Black's main hope against 1 1 .d4.
B) 1 1 . .. tLlxe4
14.f4 b7 1 5 .h2?! ( 1 5 .ttJbd2 hc3
1 6.bxc3 gxf4 1 7.ttJxe4 dxe4 1 B.ttJd2=)
1 5 . . . ttJc4! 1 6.b3 ttJcd6 gave Black a solid
advantage in Salgado Lopez - Chirila, Szeged
200B.
14... tLlxdl 15.Yfxdl tLlc4 16.Yfd3 g6
17.Yfdl

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
a b c d e f g h a b c d e f g h
Occasionally White plays Bl) 12.dxe5 here, 17 c6
..

but the better and more popular move is B2)
12.tLlxe5.
Bl) 12.dxe5
This should not worry Black at all. The
position resembles the Open variation
(5 . . . ttJxe4 6.d4 b5 7.b3 d5 B.dxe5 e6)
except for the placement of a few pieces. The
differences seem to favour Black, who can
obtain an equal position without difficulty.
There is nothing wrong with this, although
1 7 . . . b7!? was also fine. It is worth adding
that the attempt to invade on e6 with 1 B.ttJd4
would be strongly countered by 1 B . . . geB!=t.
18.b3 tLld6
1 B . . . hc3 would be risky in view of 1 9.g5
ttJb2 20.iVc1 when the knight may have
trouble escaping from b2, although the
position remains complicated.

12 5
. 19.tLld4 Yfb6 20 ..ih6 a:f7
This is not the only reasonable move. We have been following the game Kotronias
Lukasevicius - Egan, corr. 200B, continued - Lie, Malmo 200B. Black has an extra
 Chapter 5 - Gajewski: I I .d4 1 39

central pawn, he has developed his pieces B21 1) 14.a4

harmoniously, and can look to the future with We are following the game T. Hansen - Lie,
confidence. Tonsberg 200B, which continued:

B2) 12.xe5 14 h4
.

This gives White more than one route to an

advantage. Instead Black could have obtained a
satisfactory position with 1 4 . . . h4!N. We will
return to this move after we finish analysing
the game continuation.

15.h3
This is a good move, although there was a
second promising line in 1 5 .f3 lLlf6 1 6.lLlb3
lLlxb3 1 7 . .txb3;1;; .

a b c d e f g h
This looks more logical than 1 2.dxe5 . By
maintaining a pawn on d4 he keeps the c5-
square as a potential outpost. The knight is
well placed on e5, while the rook on el also
enjoys greater freedom.
At this point Black has tried B21) 12 .td6 ..

and B22) 12 f6, the latter of which seems to

.

be more promising.

B21) 12 ....id6 13.d2 f5

In an ideal world Black would prefer not a b c d e f g h
to play this committal move at such an early
stage. Unfortunately, in the present position it In this position White could have obtained a
is the only really consistent choice. distinct advantage with:
Now White usually chooses between the
commencement of a queenside offensive with 16.f3!N
B21 1) 14.a4 and reinforcing his grip over e5 Instead the game continued:
with B212) 14.d8. 1 6.lLlxc4 dxc4 1 7.lLla5
1 7 . .txe4 fxe4 I B.lLld2 allows I B . . Jhf2!,
Also deserving attention is 14.f3!?N lLlxd2 (after which should lead to a draw after 1 9. @xf2
1 4 . . . lLlg3 1 5 .b4 lLlc4 1 6.lLldxc4 dxc4 1 7.a4 h4t 20.@gl .ixh3 2 1 .xe4 g3 22.f1
.tb7 I B . .tf4 lLlh5 1 9 . .td2;1;; White maintains h2t 23.@f2 f8t 24.lLlf3 xf3t 25.@xf3
a slight positional advantage) 1 5 .xd2 lLlc4 g3t 26.@e2 d3t 27.@f2 g3t etc.
1 6.f2. White maintains a slight edge without 1 7 . . . h4 I B.f3 .te6 1 9.1Llb7
taking any risks. Now:
1 40 Attacking the Spanish

1 9 . . . bxc3 20.bxc3 .id5 2 1 .ttJxd6 cxd6 16.b3

This would have been fine for Black.
16 ... bxc3
Black wins a pawn, but only temporarily.
17.bxc3 xc3 18.Wd3 xe5 19.dxe5 .ib4
20 ..ld2
Also strong is 20.d4 b8 2 1 .e3 c5
22.ttJxc5 hc5 23.xc5 d4. Here White has
a pleasant choice between 24.c4t <bh8
25 .e l with a nice middlegame position, and
24.e6 b6 25.xb6 xb6 26.e7 with a lasting
advantage in the endgame.
20 ...c5 21 ..bc3 c4 22.Wd2 .bc3 23.Wxc3
cxb3 24..bb3i
Throughout the above varlatIOn we saw
numerous different ways in which White
could have secured an advantage. Fortunately,
I found an important improvement which was
mentioned at move 14. This should enable
Black to achieve a satisfactory game, as shown
in the following analysis.
1 6.ttJdf3 f6 1 7 . .if4 b8 is comfortable for
Black. White has an outpost on e5, but Black's
knight on e4 is also excellent. The second
player also enjoys good prospects on both
Hanks. His queens ide pieces are already well
placed, and he may be able to consider a timely
. . . g5 to gain space and perhaps even start an
attack.
16 ... xb3 17 ..bb3 .le6 18.8 .be5 19.dxe5
c5 20 ..lc2 d4!?
This is the most dynamic move, although
that does not necessarily make it the best.
20 . . . ab8 2 1 ..ie3 e7 is a calmer alternative,
leaving White just marginally better.
21.cxd4!
White must react precisely. He would be ill
advised to accept the sacrifice with 2 1 .cxb4?!
d3!.
8
7
6

5
14...Wh4!N 15.We2
This is clearly the most natural way to defend 4

the pawn. 3

2
15 ... b4 1

a b c d e f g h
8
Here we have:
7 After 22.hd3?? ttJxd3 White cannot
6 recapture as the rook on e 1 hangs
22.f2 xb4 is better for Black.
5 22.e3 xb4 (also playable is 22 . . . dxc2!?
4 23.bxc5 ad8) 23 . .id2 xb2 24.ac 1 ttJe4!
leaves White in difficulties.
3 22.d2 dxc2 23.bxc5 ad8 Black has more
2 than enough play for a pawn.
1
21. ..Wxd4t 22 ..le3 Wxe5
a b c d e f g h
 Chapter 5 - Gajewski: I I .d4 141

Black has gone from offering a pawn It is worth remembering this method of
sacrifice to accepting one. White certainly arranging the knights. Usually it is seen
has compensation, although it is hard to say in the Stonewall Dutch. Instead 1 9.tLld2
whether it is enough for an advantage. A tLlcxd2 20.i.xd2 b4 gives Black a reasonable
possible continuation might be: position.
1 9 . . . i.xe5
23.Vfl tLld7 24 .tg5 Vd6 25Jladl Vc6
More passive is 1 9 . . . g6 20 . .if4 e8 2 1 .YNc 1 ! ?
26 .te7 Ue8 27.hb4 tLlb6
with a slight edge. The queen move defends
White has regained his pawn, but Black has b2 while also eyeing Black's weak squares on
reasonable prospects on the queens ide. the kingside.
20.dxe5 YNb6
B212) 14.tLlcJa

8
7
6
5
4
3
a b c d e f g h
2
2 1 .YNe2
1 After 2 1 .<it>h2? tLlxf2 22.tLlxf2 YNxf2 23.f1
a b c d e f g h YNb6 24.i.xfS g6 White has no advantage.
2 1 . . . e8 22.i.e3 tLlxe3 23.YNxe3 YNxe3
This is probably White's best. The only 24.xe3
drawback is that he will not be able to challenge Black faces a difficult defence in this
the strong enemy knight with f2-f3 for a few endgame. His light-squared bishop is passive
more moves. We will now follow the game and the white knight may jump to c5 later.
Hou Yifan - D. Howell, Gaziantep 2008.
18.Ve2
14... tLlc4 15.a4!
This is an important move for White. The
8
difference between this position and the
previous variation is that Black does not have 7
the . . . YNh4 idea at his disposal. 6

15 ....tb7 16.axb5 axb5 17J::lxa8 has 5

The other possibility is: 4
17 . . . YNxa8
This is playable enough, although White can 3
still obtain an advantage with precise play. 2
1 8.tLld3! YNa7 1 9.tLlfeS
1
a b c d e f g h
1 42 Attacking the Spanish

Instead 1 8.d3!?N deserved serious 13.tLld3

attention. The idea is to continue with feS 1 3.f3 is less logical. The f3-square should
followed by a timely f3. be kept free for the queen's knight, or perhaps
for a pawn to drive away the intruder on e4.
18 c5!?
Now Black has two playable options:
Black takes the opportunity to free his
position. a) The untested 1 3 . . . f5!? warrants consider
ation. Compared with line B2 1 ) above, it is
19.tLld3 likely that White will have to lose a tempo with
A good move, threatening to take on cS and f3-eS at some point. Following 1 4.bd2
exchange one of Black's better minor pieces. Black can choose between 1 4 . . . c6, 1 4 . . . i.b7
and 1 4 . . . i.e6, with good chances in each case.
19 cxd4 20.tLlxd4 Y!Yb8

White retains a very slight advantage thanks b) 1 3 . . . c4 has been played in two games:
to her strong knight on d4. On the other hand
Black remains quite active, and went on to b 1 ) In Pavasovic - Stevic, Sibenik 2008, White
hold a draw without any real problems. decided to snatch a pawn with 1 4.he4 dxe4
l S Jhe4. There followed l S . . . .ib7 1 6J!e 1
We may conclude that 12 . . .i.d6 is a viable e8:
option. At the same time, it seems that both
of the main lines examined offer White some 8
chances of an advantage. Black may also 7
need to worry about the promising but as yet 6
untested idea of 1 4.f3!?
5

B22) 12 f6
4

8 2
1
7
a b c d e f g h
6
Black had achieved what looks like a very
5 comfortable version of line B2) in Chapter 4.
4 The game proceeded with 1 7.bd2 b6 1 8.b3
'!MId7 1 9.c4! ? (White returns the pawn in order
3 to simplify to an equal position) 1 9 . . . M3
2 20.xf3 bxc4 2 1 .bxc4 ttJxc4 22.'!MIb3 '!MIbS
23.i.f4 lh-lh.
1
a b c d e f g h b2) 1 4.bd2 cxd2 l s .hd2 .ib7 1 6.i.f4
This is Black's most ambitious and principled i.d6 1 7.hd6 '!MIxd6 1 8.d2 ae8 1 9.b3
continuation, forcing the well-placed knight
to retreat.
 Chapter 5 Gajewski: 1 1 .d4
- 1 43

square on cS as well as opening the c-file. The

game was soon agreed drawn after 1 8.tDf4 f5
1 9.a4 bxa4 -.

In Bosiocic - Stevie, Rijeka 2009, White

gained no advantage with: 1 4.'lWhS i.e6!

8
7

6
a b c d e f g h
5
White obtained a minimal edge in Smirin -
4
Greenfeld, Beer Sheva 2007. He will gain a niee
outpost on cS, but Black has enough prospects 3

for counterplay on the opposite flank. 2

13 Ad6 14.lDdl
..
This is White's most natural and strongest
move. However, it is worth paying attention to
the following two alternatives, as they illustrate
some important motifs that can also arise in
the main variation.
1 4.i.f4 should not cause Black any problems
after 1 4 . . . ll'k4. The game So - D. Howell,
Wijk aan Zee 2009, continued: I S .b3 tDb6
1 6.'lWc 1 e8 1 7.,hd6 cxd6!
1
-.J
a b c d e f g h
The bishop transfer to f7 enables Black to
cover all of his weaknesses, while conveniently
gaining time. This game also resulted in an
early draw after I S .i.f4 ( 1 S .tDd2 i.f7 1 6.'lWh4
tDxd2 1 7.,hd2 i.g6 1 8.b3 'lWd7 gives Black no
problems) l S . . . i.f7 1 6.'lWf3 tDc4 1 7.b3 tDb6
1 8.,hd6 'lWxd6 1 9.'lWf4 Y2-Y2 .
Let us return to the main line after 14.lDdl:

8
7

6
5

1
L.J
a b c d e f g h

This excellent move secures Black a good

game. The doubled pawns are not at all weak, a b c d e f g h
as White has no realistic way of attacking 14 Je8
..

them. On the contrary, the structural change This is almost certainly Black's best option.
helps Black considerably by covering the weak
1 44 Attacking the Spanish
14 . . . .tb7 does not equalise after I S .lLlb3 lLlc4
( 1 S . . . lLlxb3 1 6.axb3 is also undesirable for
Black) 1 6.f3 lLlg3 1 7.lLlbcS . The outpost on
cS guarantees White a stable advantage, e.g.
1 7 . . . i.cB I B.a4 or I B.lLlb4!?
14 . . . f5
This is positionally risky.
I S .lLlf3
The two white knights work in perfect
harmony. Play might continue:
I S . . . lLlc4 1 6.lLlfeS h4 1 7.f3
8 Another sensible-looking alternative is:
7 I S .lLlfl c6!?
6
5
4
3
2
1
a b c d e f g h
1 7 . . . cS!?
After 1 7 . . . lLlgS I BJ!e2!? is a very annoying
move. White's kingside is protected and
he may even plan e l , completely killing
Black's hopes of kings ide counterplay.
I B.a4!
Refusing to be drawn into complications.
I B . . . cxd4 1 9.cxd4 lLlf6 2o.lLlf2;!;
White has a very pleasant position.
14 . . . i.f5
This is also unsatisfactory in view of:
I S .lLlcS!
I S .lLlxe4 Axe4 1 6.f3 ( 1 6.lLlf4 eB 1 7.Axe4
dxe4 I B.hS c6=) 1 6 . . . i.g6 1 7.lLlf4 Axc2
I B.xc2 eB is equal.
I s . . . Axcs 1 6.dxcS eB 1 7.lLlfl !
Black's pieces will be pushed back, and the
pair of bishops will soon come into their
own.
1 7 . . . d7 I B.lLle3 i.g6
No better is: I B . . . i.e6 1 9.b4 lLlc4 2o.lLlxdS
lLlxcS 2 1 .lLlxf6t gxf6 22.bxcS i.fS 23.i.b3
1 9.xdSt xdS 2o.lLlxdS lLlxcS 2 1 .xeBt
xeB 22.Axg6 e l t 23.@h2 hxg6 24.b4 lLlab3
2S.axb3 lLlxb3 26J!xa6 xc 1 27Jk6
White wins a pawn, giving him good
winning chances.
Returning to the main line of 14 J;e8, we
will now examine B221) 15.tLlxe4 followed
by the main line of B222) 15.Yfh5.
8
7
6
5
4
3
2
1
L-
a b c d e f g h
This is often a useful prophylactic move to
stabilise the centre and queenside, as well as
making way for the queen's rook to swing to
the centre from a7.
1 6.i.f4
1 6.f3 brings White no advantage after:
1 6 . . . lLlg3 1 7.i.f4 lLlxfl I B.xeBt ( 1 B.@xfl
Axf4 1 9.1Llxf4 lLlc4 looks fine for Black)
I B . . . xeB 1 9.hd6 lLle3 20.e2 i.fS 2 1 .e l
lLlac4 22.b3 f7!=
1 6 . . . hf4 1 7 .lLlxf4 a7
After 1 7 . . . d6 I B.f3 f5 1 9.1Lle3 lLlc4
2o.lLlxc4 bxc4 2 1 .b3 i.d7 Black's passive
bishop renders his position slightly worse,
albeit still defensible.
I B.a4 ae7
Black has decent chances.
 Chapter 5 - Gajewski: I I .d4 1 45

B221) 15.tlxe4 dxe4 16.tl:k5 f5 B222) 15.Wh5!

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
a b c d e f g h a b c d e f g h
17.a4 This is the most challenging option at White's
This looks like a principled attempt to disposal. Black's usual response has been
fight for the initiative. It is difficult to suggest B2221) 15 ...Ab7, but I do not believe this to
another way to put Black under any pressure. be sufficient for equality. However, I do think
that the untested B2222) 15 ... e6!?N should
17 .be5!
. provide a satisfactory game. Before addressing
Black should not be afraid to relinquish the these options, it is worth mentioning that the
bishop pair here. following two alternatives can be dismissed
relatively easily.
18.dxe5 Wxdl 19.,bdl Ae6
Black's control over the light squares 1 5 . . . i.e6?! just loses a pawn after 1 6.tDxe4 dxe4
compensates for the loss of the dark-squared 1 7.xe4 i.f7 1 8.Wff3 xe4 1 9.Wfxe4, when
bishop. Black has no real compensation.

20.B 1 5 . . . f5
White tries to open the position. This is usually unpromising when White is
in a position to place his knights on d3 and
20 b3!
.. f3, and the present position is no exception.
This pawn sacrifice virtually forces the draw. 1 6.tDf3
White gets nowhere with 1 6.tDxe4 dxe4,
21 .,hb3 ,hb3 22.axb5 axb5 23.gIaS gu8 e.g. 1 7.tDe5 he5 1 8 .dxe5 tDc4 1 9.a4 i.e6
24.fxe4 fxe4 25.gxe4 20.f3 g6! 2 1 .Wfg5 exf3 22.b3 tDb6. White
25.i.f4 c6 26J'!:xe4 a4 27.xa4 ha4 is a is in danger of suffering in a 'bad bishop'
trivial draw. endgame. Black should maintain the
blockade of the e5-pawn while gradually
25 gal 26.ge1 .id5 27.c;t>fl c;t>f7 28.g3 e6
.. improving his pieces.
It is highly unlikely that White will achieve 1 6 . . . g6
anything in this ending. 1 6 . . . tDc4?! 1 7.tDg5!
1 7.Wfh6;!;
1 46 Attacking the Spanish

Black has a difficult position and must try

to neutralize his opponent's initiative. If
he attempts to play too actively then his
problems can quickly escalate, for example:
1 7 . . . ltJc4 1 B .ltJg5 1tJf6!? 1 9J!xeBt xeB 20.b3
1.8 2 1 .h4 e2?
2 1 . . .ltJd6 looks like the lesser evil, although
after 22.1tJf3 White's grip over the e5- and c5-
squares ensures him of a stable advantage.
22.bxc4 xc2
a b c d e f g h

Black can react in a variety of ways, although

it is not clear whether he can equalize.

a) 1 B . . . ltJc4 is well met by 1 9.b3 ltJb6

20.a4! bxa4 2 1 .bxa4 when Black's position
is quite unpleasant. He cannot seek relief
through a queen exchange with 2 1 . . .d6??
in view of 22.xd6 cxd6 23.b 1 winning a
piece.
a b c d e f g h
b) The attempt to blockade the light squares
Now White can break through in stunning with 1 B . . . i.d5 1 9.i.f4 c6 also fails to solve
fashion. Black's problems after 20.b3 ltJb7 2 1 .a3;!;; .
23.1tJf7!! 1.g7 White's bishop pair is a real asset. Furthermore,
23 . . . ltJe4 24.dB! <tt> xf7 25.ltJe5t gives Black's bishop is not very stable on d5 , with
White a decisive attack. c3-c4 ideas being a constant worry.
After 23 . . . ltJd7 24.ltJh6t <tt> g7 25.1.f4!
xd3 26J!e l ltJf6 the move 27J:!eB!! wins c) I think that the strongest response is
beautifully. 1 B . . . d6!
24.ltJh6t <tt> h B 25 .ltJe5 Initially I thought that White could obtain a
Threatening mate. safe edge with:
25 . . . i.e6 26.1.g5 1tJh5 27.e 1 xc3 2B.ltJxg6t 1 9.xd6
hxg6 29.xe6 Obviously this is not forced, and White
Black is unlikely to survive for long. may be better off investigating alternatives.
However, Black will certainly benefit from
B2221) 15 .tb7 ..
the removal of the annoying queen.
I tested this move myself in the game 1 9 . . . cxd6
Z. Almasi - Brunello, Dresden 200B, which This small change in the pawn structure
we will now follow. should improve Black's chances considerably.
The d-pawn can help to support the centre
16.tifl while the open c-file is a potential source of
White may have been able to improve with: activity.
1 6.ltJxe4!?N dxe4 1 7.ltJc5 hc5 1 B .xc5 20.a4
 Chapter 5 Gaj ewski: 1 1 .d4
- 1 47

20.i.f4 d5 looks playable enough for Black. a) 16 . . . ltJc4 1 7.f3 ltJg3

My first reaction was that White should
automatically be better after the text move.
However, a closer examination revealed that
Black is not without chances after:
20 . . . d5!
The attempt to activate the bishop with
20 . . . i.d5?! runs into 2 1 .i.f4.
Instead it is more important to stabilize the
centre. Now Black's queenside prospects
should compensate the opponent's bishop
pair.
2 1 .i.d l
1 7 . . . g6 does not equalise after 1 B.h4 ltJg5
1 9JxeBt xeB 2o.hg5 fxg5 2 1 .xgS
e2 22.,!:k1 f8. At first glance this appears
quite tempting for Black, but it seems that
White can maintain the advantage with
either 23.h4!, intending f2, or 23.i.b l !,
threatening e 1 . In both cases Black has no
real compensation for the pawn.
I B.ltJxg3 hg3
This enables White to whip up a dangerous
attack.
1 9.1tJeS!

8 8

7 7

6 6

5 5

4 4

3 3

2 2

1 1

a b c d e f g h a b c d e f g h

This piece was doing nothing on c2, so it
makes sense for it to relocate.
2 1 . . .b4!?
Black could of course consider a more
patient approach with 2 1 .. .@f7. In this case
White's bishop pair should be enough for a
slight plus, although Black should not be in
any great danger.
22.cxb4
Otherwise White will be left with a queenside
wealrn.ess.
22 . . . ltJc6 23.i.d2 ltJxd4
In my opinion this position is dynamically
balanced. White's bishop pair is a definite
asset, but the same can now be said of Black's
powerful central pawns.
16 ti'd7
...
Two other moves also deserve our attention.
When the queen comes to hS Black should
always be on the lookout for this move.
1 9 . . . fxeS 20.xh7t @f8 2 1 .hBt @f7
22.hSt @f8 23.i.gS
White can force a draw with 23.hBt, but
he should be playing for more.
23 . . . d6 24.i.g6 e6 2S .b3 ltJd6
2S . . . ltJb6 26.e3 intending f4 gives White a
huge attack.
26.e3 gB
26 . . . ltJfS? 27.i.xfS xfS 2B.i.e7t wins the
queen. And 26 . . . ltJf7 is also unsatisfactory in
view of 27.f4 ltJxgS 2B.xeS!.
27.g4 e6
After 27 . . . e6, the simple 2B.i.c2 leaves
Black without a satisfactory response.
2B.heB xg4 29.fxg4 i.f4 30.f1 @gB
3 1 .hf4 exf4 32.e7 ltJxeB 33.xf4
White should win this ending.
1 48 Attackin g the Spanish

b) Fortunately Black has a stronger The only differences are that the black queen
alternative: is on d7 rather than d8, while the rooks
1 6 . . . g6!N 1 7.h4 on e l and a8 have been removed from the
1 7.h6 lDc4 1 8.f3 lDg3 is equal. board. Unfortunately these factors do very
1 7 . . . f5 1 8.xd8 axd8 little to ease Black's defence:
20 . . . g6!?
8 The best chance. Instead 20 . . . fxe5 2 1 . xh7t
7 @f8 leads to defeat after some precise
6
moves: 22.i.f5! exd4! 23.i.g5! (23.hd7??
e l #) 23 ... f7 24.i.g6 e6 25 .h8t
5
g8 26.xg8t @xg8 27.he8. Compared
4 with the comparable variation noted above,
3 please note the enormous significance of the
2 undefended rook on e8 and the absence of
1
the rook on e 1 .
2 1 .hg6! xe5!
a b c d e f g h
The only chance, but it is still not good
In this position Black's active pieces should enough.
make up for the slightly weak squares on e5 22.dxe5 hxg6 23.xg6t g7 24.e8t f8
and c5 . 25 .e6t f7 26.g4t g7 27.exf6 xg4
28.hxg4
If this evaluation stands up to scrutiny (and I White is the strong favourite to win this
see no reason why it should not) then White ending.
should focus all his efforts on the earlier
suggestion, 1 6.lDxe4!?, if he intends to fight
for an opening advantage.
8
17..lf4
White could have obtained a strong attack 7
using the same idea as in variation a) in the 6
note to Black's 1 6th move:
1 7.f3! lDg3 1 8.xe8t xe8 1 9.1Dxg3 hg3 5
20.lDe5! 4
8 3
7 2
6
1
5
a b c d e f g h
4

3
18 ..bd6 Yfxd6
1''''''''.,,, ,,.'/'
1 8 . . . cxd6!? was a worthy alternative, just
2
as in the So - D. Howell game from page
1 1 43. Black should still be a little worse in the
a b c d e f g h present position, but would retain reasonable
 Chapter 5 Gajewski: I I .d4
- 1 49

defensive chances after something like 1 9.f3 31.3 d6 32.tLlfe6 gb8 33.fxe4 .be6
g6 20.'lWg4 'lWxg4 2 1 .hxg4 g5;t. 34.tLlxe6 fxe4 35Jlxe4 bxc4 36.bxc4

19.tl:k5 .too 20.b3 tLla3 21..he4 dxe4 8

22Jac1 g6 23.YHh6
7
8 6
7 5
6 4
5 3
4 2
3 1
2 a b c d e f g h
1 White has won a pawn, and his pieces are
much more active than their black counterparts.
a b c d e f g h Almasi finished the game smoothly.
23 ... a5?!
23 . . . f5! would have prevented the manoeuvre 36 gb2 37.c5t xd5 38.tLlxc7t xe4
..

that shortly follows. In this complex position 39.tLlxe8 b8 40.c6 tLlb5 41.c7 gc8 42.a4
both players need to play accurately. Black 1-0
should play . . . e7 and . . . f8, supporting his Despite this reversal, there is clearly much
kingside pawns. The position is marginally more to discover in this line. I hope that the
better for White, mainly due to the offside suggestions and accompanying analysis will
knight on a3. A plausible continuation might provide a useful starting point for future
be 24.'lWg5 (24.g3 e3! solves all of Black's practical tests.
problems) 24 . . . e7 25.g3 f8 26.h4 a5
27.h5 'lWf6 28.'lWxf6 xf6. Black should be B2222) 15 ... c6!?N
alright here. In case of 29.c4 he can play a
strong regrouping move: 29 . . . .ie8!, with equal
8
chances.
7
24.tLlg3! 5 25.tLle2! 6
Having successfully completed this
manoeuvre, White can be very happy about 5
his position. Unfortunately I failed to defend 4
optimally and my position soon deteriorated.
3
25 .. Jad8 26.tLlf4.td5 27J:e3 YHf8 28.YHxf8t 2
xf8 29.c4 j.f7 30.d5 e7?! 1
30 . . . @g8 looks like the lesser evil, although
Black is in trouble regardless. a b c d e f g h
1 50 Attackin g the Spanish

This is a new idea which, in my opinion,
may well be the best move. Compared with
the previous variation, Black's b5- and d5-
pawns are now a lot more stable, and the light
squared bishop may fine a useful role on the
c8-h3 diagonal. Finally, the text also allows a
useful transfer of the queen's rook via a7.
White gains nothing from avoiding the
queen exchange. For instance, after 20. 'IWh6
exf3 2 1 .gxf3 c4 Black stands well.
20 ...9xg5
Less promising is 20 . . . .ie7 2 1 .'IWg3 exf3
22.'IWxf3 hc5 23.dxc5 .ie6:

16.tl)xe4 8
The idea of the rook transfer is well 7
illustrated by the following variation: 1 6.f4 6
f5 1 7.xe4 fxe4 1 8.f3 a7! 1 9.fxe4 g6 20.'IWf3
5
dxe4 2 1 .xe4 f7
4

8 3

7 2

6 1

5 a b c d e f g h
4
White's position is preferable thanks to
3 the strong dark-squared bishop. The main
2 difference between this position and the
1 note to White's 1 9th is that White has not
weakened his queenside with a4, so there
a b c d e f g h
is no weak square on b3. Black may be able
White finds himself in an awkward pin along to defend any immediate threats, but White
the f-file. Black's activity is worth much more should be able to increase the pressure
than a pawn. gradually.

16 .. Jxe4!? 21.i.xg5 exf3 22.gxf3

1 6 . . . dxe4 1 7.c5 g6 ( I 7 . . . f5 1 8 . .ig5 also
looks pleasant for White) 1 8.'IWh4 is slightly
awkward, as Black will have to weaken his 8
kingside. 7
6
17Jxe4 dxe4 18.tl)c5 f5 19.3
After 1 9.a4!? hc5 20.dxc5 .ie6 2 1 .axb5 5
axb5 the position is about equal. White 4
has a powerful dark-squared bishop, but
Black has one more pawn on the kingside 3
and the knight on a5 will come into play 2
soon.
1
19 ... g6 20.9g5 a b c d e f g h
 Chapter 5 - G ajewski : I I .d4 lSI

White's kingside pawns have been prevents the knight from coming to dS (and
compromised, but on the other hand he still subsequently f4) .
has an extra central pawn.
24 g7
.

22 lLle4
.
The king should vacate the diagonal.
Black should begin by improving this piece.
If he tries a move like 22 . . . @f7 then 23.a4! 24 . . . ltJd7 2S.ltJd3 maintains a slight edge for
might prove slightly awkward, e.g. 23 . . . ltJc4 White.
24.axbS cxbS when White's central domination
increases. The immediate pawn exchange 24 . . . bxc4
2S.bxc4 only improves the scope of the white
23.b3 lLlb6 bishop, e.g. 2S . . . @g7 26 . .tb3!? (26 . .td3
The alternative is no better: reaches the main line) 26 . . . aS 27.ltJa4 ltJxa4
23 . . . ltJeS 28.ha4 cS 29.dS when Black faces a difficult
This only loses time. defence.
24.@f2 ltJd7
25.i.d3
It looks logical for Black to attempt
2S . .td8 causes Black no problems: 2S . . Jb8
to exchange or improve this knight.
26 . .td3 bxc4 27.bxc4 hcs 28.dxcS ltJd7
Unfortunately White can maintain his
29 . .tc7 (29Je l ltJxcS+) 29 . . Jb2 30 . .td6
advantage with the resolute response:
@f7 3 1 .e l xa2 32.e7t @f6 with equality.
2S .b4! ltJb6
White has active pieces, but Black has an extra
Black has no time for 2S . . . aS ? in view of
passed a-pawn and his king protected by a solid
26 . .tb3t @g7 27.ltJe6t.
structure. I can see nothing better for White
26 . .tb3t @g7
than taking a draw with 33.e8 .tb7 34.e7
26 . . . ltJdS 27.a4 maintains the pressure.
forcing 34 . . . .tc8.
27Je l
White enjoys an ongoing initiative. 25 bxe4 26.bxe4 f7 27.lLlb7!?
.

27.b l hcs 28.dxcS ltJd7 29 ..te3 ltJeS is

24.e4 equal.

8 27....hb7 28Jbl ga7 29Jxb6 ll.e7 30Jb2

i.e8
7
6 8
5 7
4 6
3 5
2 4
1 3
a b c d e f g h 2
This looks like the critical move. White 1
a b c d e f g h
1 52 Attacking the Spanish

Black will follow with .. Jb7, after which

he should be able to hold the double bishop
ending without too many problems.

Conclusion
We have reached the end of what has been
a fascinating journey into the still relatively
unknown Gajewski variation. In the present
chapter we have seen that the plan with 1 1 .d4
must be respected as a serious try for a white
advantage. Throughout the chapter I have
combined a thorough survey of the latest games
with my own original analysis, to provide a
clear theoretical picture.

At the present moment I do not see a way for

Black to equalise with 1 1 . . . dxe4. According to
my analysis White has more than one route
to an advantage, but the most straightforward
seems to be 1 2.ttJxe5 c5 1 3.i.e3 i.b7 1 4.ttJd2
Vffc7 1 5 .Vff b l !, which was discussed in line
A42) on page 1 36. Unless Black can find a
major improvement here, I would recommend
that he avoids this path.

At the present time, Black's best hope of

equalising seems to be 1 1 . . .ttJxe4, followed
by 1 2.ttJxe5 f6 1 3.ttJd3 i.d6. This has been
less popular than 1 1 . . . dxe4, probably because
most Gajewski players have tended to prefer
an unbalancing approach over a position with
symmetrical pawns. Logic tells us that White's
position ought to be slightly preferable, thanks
primarily to the potential outpost square
on c5 . At the same time Black has an active
position with plenty of resources. At the time
of writing there have still been very few games
played, and I was able to find a variety of new
and interesting possibilities for both sides. The
Gajewski remains a fertile area for fresh ideas,
so I would encourage the reader to examine
the latest games and look for improvements of
his own. Happy hunting!
Chapter 6

Anti-Marshall
7
6
5
4
3

1
a b c d e f g h

l .e4 e5 2.tLlf3 tLlc6 3.i.b5 a6 4.i.a4 tLlf6

5.0-0 i.e7 6J1el b5 7.i.b3 0-0
Introduction page 1 54
Theoretical Highlights page 1 54
Theory page 1 57
A) 8 . c3 d5 9.d4 page 1 57
B) 8 . d4 page 1 58
B I ) 8 . . . d6 page 1 59
B2) 8 . . . lLJxd4! ? page 1 65
C) 8.a4 b4 page 171
C I ) 9 . d4 page 171
C2) 9.d3 page 1 72
0) 8 .h3 i.b7 9 . d3 d5 ! ? 1 0.exd5 lLJxd5 page 1 76
0 1 ) I l .a4 page 1 77
02) I l .c3 page 1 78
03) I l .lLJxe5 paeg 1 79
1 54 Attacking the Spanish

1.e4 e5 2.f3 c6 3.i.b5 a6 4..la4 f6 Theoretical Highlights

5.0-0 il.e7 6J:te1 b5 7.il.b3 0-0
White strikes back in the centre
8
l .e4 e5 2.f3 c6 3 ..lb5 a6 4.L4 f6
5.0-0 i.e7 6Jlel b5 7.i.b3 0-0 8.c3 d5
7
6 8
5 7
4 6
3 5
2 4
1 3
a b c d e f g h 2
Introduction 1
With his last move Black prepares to meet
a b c d e f g h
B.c3 with B o o .d5 , the notorious Marshall
Attack, which will be covered in Chapters 7-9. 9.d4
Before moving on to this subject we must first White retaliates with his own central strike.
address the numerous ways in which White The idea is somewhat reminiscent of the
may sidestep this formidable weapon. previous chapter, but less effective for White.

The present chapter is an extremely important 9 exd4 10.e5 e4 1 1 .cxd4 .iS

..

one. The Marshall Attack has enjoyed a

fearsome reputation for several decades, and
8
there are a great many Ruy Lopez players
who prefer to avoid it entirely. Even the 7
best prepared player in chess history, Garry 6
Kasparov, would invariably opt for one of the
anti-Marshall systems found in the present 5
chapter. 4
3
Strategic themes
Just as in Chapter 3, we are going to be dealing 2
with quite a varied selection of different lines.
1
Each has its own individual characteristics,
so I will once again refer the reader to the a b c d e f g h
theoretical section for an in-depth discussion The bishop is ideally placed on the h7 -b 1
of each variation. diagonal. Play may continue 12.c3 xc3
13.bxc3 ftd7 when Black is doing fine, as seen
in variation A) on page 1 57.
 Chapter 6 - Anti-Marshall 1 55

The same idea, a move earlier

8
If 9.d4 is playable for White, then he might
very well consider the same idea on the previous 7
move.
6
1.e4 e5 2.tLla tLlc6 3 .tb5 a6 4 .ta4 tLlf6

5
5.0-0 ILe7 6Jel b5 7.i.b3 0-0 8.d4 4
3
8
2
7
1
6
a b c d e f g h
5
Here White's most natural and strongest
4 continuation has not been tested. 16.tLlc3!N
3 (page 1 68) is a good move, but Black can still
obtain full chances if he plays energetically.
2 The pawn sacrifice with 1 O.c3 leads to
1 interesting positions, but if both sides play
accurately then the usual result will be a
a b c d e f g h
draw. White seems to have enough activity to
I analysed two responses to this move: regain his pawn, but not enough to claim an
advantage. The following position was reached
a) 8 d6 is the most solid option. After 9.c3
.
in Dominguez - Aronian, Wijk aan Zee 2009,
i.g4 we reach a position that has been played as given in the note to White's 1 3th move in
thousands of times. The latest games indicate line B222) , page 1 69.
that Black is still doing fine here, as seen in line
B 1 ) starting on page 1 59.
8
b) 8 tLlxd4!? is a more combative move which
. . 7
aims to unbalance the game. White can react 6
in a few different ways.
5
bl) After 9 .ixf7t gxf7 10.tLlxe5 U8 1 1 .Vxd4

4
c5 12.Ve4 tLlc7 Black has plenty of activity for
the pawn, as seen in line B2 1 ) on page 1 65 . 3
2
b2) White usually exchanges knights with
1
9.tLlxd4 exd4, followed either by recapturing
the pawn (line B22 1 , page 1 66) or aiming for a b c d e f g h
a lead in development by means of a gambit Black could have gained the advantage with
(line B222, page 1 69) . 16 i.a8!N 17 .ta7 c5 when the bishop is
.

In the former case, the following position trapped on a7.

might be reached:
1 56 Attacking the Spanish

The traditional anti-Marshall
1.e4 e5 2.a c6 3 ..lb5 a6 4.L4 f6
5.0-0 Ae7 6.!1el b5 7.Ab3 0-0
For the last few decades of the 20th century,
White's most principled way of avoiding the
gambit was considered to be:
8.a4
To this day, this move remains a popular
choice at all levels. In line C) I recommend the
response:
8 ... h4.
Compared with the main lines of the closed
Ruy Lopez, the inclusion of these two pawn
moves introduces certain nuances. One
particularly striking example was the game
Kulaots - Bartel, Gjovik 2009:
During the past decade the following system
became extremely popular:
8.h3!? .ib7 9.d3
White is hoping to discourage the gambit
approach with . . . d5. By comparison with the
standard Marshall with B.c3 d5, he hopes that
the moves h3 and d3 will prove more useful
in the event that Black gives up a pawn. For
a long time, Black responded with 9 . . . d6,
at which point White would respond with
1 0.a3 . With this approach White avoided the
potential weakening of his queenside as seen
in the previous example. The resultant semi
closed positions were tested in many games
at the highest level. Although Black probably
has enough defensive resources, I prefer to
recommend a more active approach:
9 ... d5!?

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1

a b c d e f g h a b c d e f g h
White's premature central advance has Despite the earlier comments, it turns out
allowed Black to gain a lot of space on the that the text is playable nonetheless.
queenside. Moreover, White's queens ide pieces
are completely shut in. Details can be found in 1 0.exd5 xd5
the note to Black's 1 1 th in line C2) , page 1 74. Full analysis of this variation can be found in
line D) , beginning on page 1 76. Black's main
The fashionable anti-Marshall idea is revealed after:

1.e4 e5 2.ttla c6 3 ..!.b5 a6 4.L4 ttlf6 1 l .ttlxe5 ttld4!

5.0-0 Ae7 6.!1el b5 7.Ab3 0-0 Unlike the main line of the Marshall, Black
 Chapter 6 - Anti-Marshall I S7

is in no position to play for a direct attack.
Instead he prepares to eliminate one of the
opponent's strong bishops, after which he
should have enough positional compensation
to ensure a satisfactory game.
Overall, it seems to me that Black is holding
his own in all of the above lines. We will now
conduct a more detailed examination.
9 ... dxe4 1 O.tDxeS also leaves Black some way
short of equality.
10.e5
Nothing else can threaten Black, e.g.
1 O.exdS tDxdS ( I O . . . tDaS!? also looks playable)
I l .tDxd4 tDxd4 1 2.cxd4 ( I 2.xd4?! .ib7+)
1 2 . . . i.e6 1 3.tDc3 eB with equality.

Theory 10 ttle4 1 1.cxd4 i.f5

..

1.e4 e5 2.ttlf3 ttlc6 3.i.b5 a6 4.i.a4 ttlf6 8

5.0-0 i.e7 6Jel b5 7..lb3 0-0
7
8 6
7 5
6 4
5 3
4 2
3 1
2 a b c d e f g h

1 12.ttlc3
1 2.tDbd2 tDb4 1 3.tDfI cS 1 4.tDe3 i.e6
a b c d e f g h
l S .dxcS hcs 1 6.a3 tDc6 gave Black no
We examine four options: A) 8.c3 d5 9.d4, problems in Stellwagen - Vescovi, Wijk aan
B) 8.d4, C) 8.a4, and D) 8.h3. Zee 2006.

A) 8.c3 d5 9.d4 In Sulskis - Brunello, Arvier 2007, I fell into

This is nowhere near as effective as the difficulties after 1 2.h3 d7 1 3 .tDbd2 tDb4
comparable anti-Gajewski variation: 7 . . . d6 1 4.tDfI cS l S .a3 tDc6 1 6.tDe3:
B.c3 0-0 9.h3 tDaS 1 O.i.c2 dS I l .d4, which
we encountered in the previous chapter. In the 8
present position Black has avoided moving his 7
knight on c6 away from the centre. He can 6
also make use of the fact that his opponent's
5
bishop is on b3 rather than c2.
4

9 ... exd4! 3
1,.,.,.!' c=!I&>
9 . . . tDxe4 1 O.dxeS .ie6 transposes to an 2
unusual line of the Open variation. White may 1
have some chances for an edge with I l .tDd4!.
a b c d e f g h
1 58 Attacking the Spanish

1 6 . . . J.e6? 1 7.tDxdS! hdS 1 8.dxcS ad8 Now the game comes down to a battle
1 9.'lWxdS 'lWxdS 20.hdS xdS 2 1 .xe4 Black between pawn majorities. Black must be careful
had no real compensation for the pawn. not to allow the white pawns to advance too far,
as this could easily lead to a crushing attack.
However, Black could have obtained the It turns out that Jakovenko has evaluated the
better chances with 1 6 . . . c4! 1 7.tDxfS ( I 7.J.c2 position correctly, as demonstrated by the next
J.g6+) 1 7 . . . 'lWxfS 1 8.J.c2 ad8! ' This excellent few moves.
prophylactic move pre-empts any exchange
of the knight on e4. Now Black has good 22.g4 Bd7 23.f4 f5!
chances to seize the initiative with . . .f6. It is imperative that Black prevents the f4-
Instead, the immediate 1 8 . . .f6 could have been fS advance. The text allows him to neutralise
met forcefully by 1 9.94! (rather than 1 9.J.e3 White's pawn majority.
ad8+) .
24.exf6 .txf6i
12 .ti)xc3 13.bxc3 Bd7 14..lc2
. White's pawn advances have done little
After 1 4.J.gS J.a3!? l S .'lWe2 ( I S .J.cl J.e7=) except weaken his kingside pawns. Black has
l S . . . h6 1 6.J.f4 tDaS 1 7.J.c2 tDc4 1 8.ad 1 the initiative, as well as the sounder structure
J.e7 1 9.h3 a s Black had good counterplay and superior bishop.
and eventually triumphed in Bacrot - Anand,
Bastia 2004. We may conclude that 9.d4 is of no danger
to practitioners of the Marshall Attack. Let
14 ...h6 15.i.e3 ttla5 us now see whether White can do better by
employing a similar idea a move earlier.
8
B) 8.d4
7
6 8
5 7
4 6
3 5
2 4
1 3
a b c d e f g h 2
The knight heads for the outpost on c4 while 1
also making way for the c-pawn to advance.
a b c d e f g h
We will follow the game: Nijboer - Jakovenko,
Sabadell 2008. Now we consider two moves: Bl) 8 d6

and B2) 8 ... ttlxd4!? The former is perhaps

more reliable, while the second is intended to
16.ttld2 gab8 17.i.xf5 Bx5 18.ttlfl ttlc4
unbalance the game.
19 .lc1 c5 20.ttle3 ttlxe3 21.he3 c4

 Chapter 6 - Anti-Marshall 1 59

Bl) S d6
.

8
This is Black's most solid answer. Now he
threatens . . . ttJxd4, winning a pawn without 7
allowing the troublesome e4-e5 advance. 6
(White would be unable to regain the d4-pawn
because of . . . c5 and . . . c4, trapping the bishop 5
on b3) . For this reason White's next move is 4
almost universally played.
3
9.e3 2
Alternatives are not at all challenging: 1
9.d5?! is undesirable for White, as after a b c d e f g h
9 . . . ttJa5 his valuable light-squared bishop will 12.h3 .leS
be exchanged for a mere knight. It is worth knowing that 1 2 . . . i.d7? should
be avoided due to 1 3.ttJxe5! dxe5 1 4.d6
9.dxe5 relinquishes White's space advantage in when White's bishop pair will give him the
the centre, leading to equality after 9 . . . dxe5, or advantage.
9 . . . ttJxe5!?
13.dxe6 Y!Je7
9 ....ig4 1 3 "l...l
J!':\ XC6 IS
. less accurate. Black should
The present variation is a major branch arrange to recapture with the queen in order to
of the Ruy Lopez, and has been tested in improve his control over the d5-square.
thousands of games. To analyse it in full detail
would take a great deal of space, which would 14.tLlbd2
hardly be appropriate for what is essentially This is White's most common and almost
a mere sideline in the context of the present certainly best move.
book. Instead I will focus on the most popular
and critical lines. Before moving on, it is worth 1 4.i.g5 ttJc4 1 5 .b3 ttJb6 is comfortable for
noting that this section may also be of use to Black. It is useful to remember this method
Gajewski players, who might encounter this of improving the wayward knight, especially
variation via the move order 7 . . . d6 (instead of when it can be achieved with gain of tempo.
7 . . . 0-0) 8.c3 0-0 9.d4 (9.h3 ttJa5 1 O.i.c2 d5
reaches the Gajewski) 9 . . . i.g4. 1 4.a4 100ks slightly premature. Black can react
in a few different ways, including 1 4 . . . Wfxc6
At this point the pressure against f3 and d4 1 5 .ttJbd2 i.e6 with equal chances. However, I
generally encourages White to react in one think that the option most in the spirit of the
of two ways: Bl l) 10.d5 and B12) 10 ..!.e3. present book is 1 4 . . . b4!?, temporarily giving
A number of sidelines have also been tested, up a pawn in order to fight for the initiative.
but we will focus on the main lines for reasons Malaniuk - Smyslov, USSR 1 988, continued
explained earlier. 1 5 .cxb4 ttJxc6 1 6.i.d2 Wfb7 with good play for
Black.
Bl l) 10.d5 tLla5 1 1 .'!'e2 c6
14 Y!Jxc6 15.tLlfl
.
1 60 Attacking the Spanish

1 5.a4 should be met by 1 5 . . . J.e6, e.g. 1 6.tLlg5 1 7. tLlg5 i.d7 I B.a4 J.dB!? (I B . . . h6 followed by
J.d7 1 7.tLlf1 h6 I B.tLlf3 i.e6 with decent . . . J.e6 would have been simplest) 1 9.e3 tLlc4
chances. 20.d3:

I5 ...i.e6 I6.tlg3 Ue8

8
7
6
5
4
3 a b c d e f g h

2 20 . . . i.e7 (once again 20 . . . h6 looks natural,

although Black may have been unsure about
1
the complications ensuing after 2 1 .b3!?) 2 1 .b3
a b c d e f g h tLla5 22.tLlh5! tLlxh5 23.xh5 hg5 24.xg5
White has no advantage. Here we will h6 25 .d2 e6 26.d5 Black's creative
explore a few of the more instructive examples opening play only brought him problems in
from recent years. Kamsky - Jakovenko, Sochi 200B, although he
eventually managed to draw.
I7.tlh4
With this interesting move White hopes to I7... d5!
provoke a kingside weakness. Here are a few I like this dynamic approach very much,
other possibilities: although the more mundane 1 7 . . . g6 also looks
perfectly playable.
1 7.a4 ecB I B.e2 J.f8 1 9.J.d2 b4!:
I8.exd5 .bd5 I9Jxe5 c4 20J;te2 i.c5
8
7 8
6 7
5
6
4
5
3
4
2

1 3
a b c d e f g h 2
20.cxb4 xc2 2 1 .bxa5 xd l t 22.xd l J.b3 1
Black's bishop pair gives him the advantage,
a b c d e f g h
Balogh - Wang Hao, Taiyuan 2007.
 Chapter 6 - Anti-Marshall 161

Black's beautiful pieces provide full 12 . .lc2 c5 (or 12 . . . ttJc4 13 . .lc 1 c5, but my
compensation for the pawn. We have been own subjective preference is for the text) .
following the game Kamsky - Grischuk,
Khanty-Mansiysk 2005, which was eventually 12.e5
drawn. Black has nothing to fear from 1 2.exd5
ttJxd5 1 3.ttJc3 ttJxe3 1 4.fxe3. The game Liwak
B12) 10.,ie3 - B. Socko, Lubniewice 2003, continued
1 4 . . . ttJa5 1 5 . .lc2 g6 1 6.h3 hf3 1 7.xf3
ttJc4 1 8Je2 c5 1 9.ttJd5 cxd4 20.exd4
8
i.g5 . Black was doing well and went on to
7 win.
6
12 .tle4

5
4 8
3 7
2 6
1 5
a b c d e f g h 4
This time White supports his centre and 3
maintains the tension. If he is allowed to
complete development with ttJ bd2 then he will 2
stand very well, so Black needs to take action. 1
Before we move on, please note that capturing
a b c d e f g h
the e4-pawn here would allow a fork on d5 .
Some players have nonetheless experimented 13.ti)c3
with 1 O . . . ttJxe4!?, the idea being l 1..i d 5 d7! 1 3.h3 .ih5 is likely to transpose to the main
1 2.he4 d5 with compensation. I would line.
encourage interested readers to conduct their
own investigation into this line, although 1 3. ttJ bd2 is not too threatening after 1 3 . . . ttJxd2
ultimately I suspect that White can maintain 1 4.xd2 hf3 1 5 .gxf3 .lb4 1 6.c2 ttJa5
an advantage. For this reason we will instead ( 1 6 . . . he l 1 7.xc6 .la5 1 8.hd5 b8
focus on a tried and tested continuation. 1 9 . .le4 gave White reasonable compensation
in Pulkkinen - Lindstedt, Helsinki 1 990)
10 exd4
.. 1 7.ec 1 ttJc4 with equality in T. Horvath -
Usually it would be a mistake to concede Zupe, Austria 1 995.
the centre voluntarily, but Black has a specific
follow-up in mind. 13 ti)xc3 14.bxc3 ti)a5 15.,ic2 ti)c4

The knight settles on its optimal square.

l 1 .cxd4 d5 Black can now start thinking about . . . a5
Black must stake a claim for the central and . . . b4 to open a queenside file and begin
squares. The other main line is 1 1 . . . ttJa5 undermining his opponent's strong centre.
1 62 Attackin g the Spanish

16.h3 underestimated. One important idea, which

It is clear that White's chances lie on the is easy to overlook, is that he may be able to
kingside. His next job will be to develop an consider exchanging the bishop on f5 at a
initiative on that flank, and the text seems suitable time. Though it may appear almost
like a sensible starting point. The following suicidal to permit the opening of the g-file,
alternative does not appear too threatening: we will see that there are certain times when
the idea can work well. Of key importance is
1 6.d3 g6 1 7 ..ih6 e8 1 8.e2 the idea of meeting gxfS with the blockading
White must waste time in order to avoid an move . . . f6!. This prevents the fS-f6 advance
unfavourable exchange of bishops. while fixing the fS-pawn as a potential target.
1 8 . . . a5 1 9.h3 .ie6 If Black is able to combine these moves with
Black should certainly not exchange this . . . e8-h5, then he might even be able to
piece for an ineffective knight. take over the initiative on that side of the
20 . .id3 c6 2 1 .lLlh2 board.
White needs to find a useful role for the
knight, but in the meantime Black can fight 18 ... a5!
back on the queenside. Black must strive for queenside counterplay
2 1 . . .b4! 22.cxb4 axb4 23.lLlg4 a3 before he gets steamrollered on the opposite
Black had achieved good counterplay and flank.
eventually prevailed in Li Chao - Tomashevsky,
Ningpo 2008. 19 ..if4
It is hard to determine White's most precise
16 ....ih5 17.g4 method of handling the position. In any case, I
1 7 . .if4 a5 1 8 .g4 .ig6 1 9 . .ifS b4 transposes to think that the following examples will illustrate
the game. Black's chances quite well. I suggest that the
reader pays particular attention to the timing
17....tg6 18.i.f5 of the bishop exchange on f5 in lines b) and
c) .
8
a) 1 9.@h 1
7 This is a reasonable move. It is clear that
6 White is anticipating an exchange on fS,
and hopes to utilise the g-file for his rooks.
5 Therefore, in this particular case Black
4 should try to avoid the bishop exchange for
as long as possible. In Jakovenko - Shomoev,
3
Internet 2004, he obtained an excellent
2 position after:
1 9 . . . b4 20.d3 bxc3 2 1 .gl b8
1
2 1 . . . .ib4 keeps the pawn but Black's position
a b c d e f g h looks dangerous after 22.h4.
Although White possesses a somewhat However, the defensive move 21 ... f6!? looks
unpleasant bind over the kingside, reasonable, e.g. 22.hg6 hxg6 with an
Black's defensive resources should not be unclear but roughly balanced position.
 Chapter 6 - Anti-Marshall
22Jg2 @h8 23J!ag1
White is really banking on the subsequent
opening of the g-file. Fortunately for us, this
dream does not have to become a reality.
24 . . . 'We8 24.'Wxc3
24.h4 f6!
24 . . . f6 25 .hg6 'Wxg6+
White's strategy has failed, and his rooks are
starting to look ridiculous on the g-file.
b) In Sznapik - Sydor, Poznan 1 984, White
began with a different king move.
1 9.@g2 b4 20.cxb4 axb4 2 1 ..if4 i!a3 22.h4
hfS!
Excellent timing. Black has seen that he
can force the enemy queen to an awkward
position.
23.gxfS 'Wd7 24.'Wb 1
After 24.'Wc2? i!c3 25 .'Wb 1 ttJa3 Black picks
up the fS-pawn.
24 . . . i!fa8
a b c d e f g h
The white pawns on e5 and f5 may look
impressive, but he is unable to mount an
attack as his pieces are completely lacking in
harmony. The a-pawn is a serious weakness,
and . . . b3 is already a serious threat.
c) In Svidler - Delchev, Plovdiv 2003, the
Russian Super-Grandmaster based his play
around the e- rather than g-file.
1 9.'Wd3 b4 20.i!e2:
1 63
a b c d e f g h
We reach an interesting branching point:
c 1 ) The game continued 20 . . . i!a6 2 1 .@g2
i!b6 22.cxb4 axb4 23.i!ael . The position is
extremely complex. Black should probably be
alright, although Svidler eventually won in
thematic fashion with a timely advance of the
e-pawn.
c2) I would instead propose the following quite
thematic solution: 20 . . . hfS!? 2 1 .gxfS f6! This
looks very logical for two reasons. First, with a
white rook already committed to e2 it will take
a long time for him to transfer his rooks to the
g-file. Second, it is difficult for White to prevent
the enemy queen from reaching her ideal square
on h5. A few sample continuations:
c2 l ) The latter of the above points even holds
true mer 23.i!ae 1 'We8! exf6 i.xf6. White has
no useful discovered attacks. The queen will
come to h5, with good chances for Black.
c22) If White wishes to prevent the queen
transfer, he should try 23 . .if4. Now it would
be inadvisable for the queen to come to e8, as
the opening of the e-file will be too dangerous.
However, Black can change plans with
23 . . . bxc3. White can hardly recapture this
pawn, as his fS-pawn would be lost after
. . . fxe5 . He may instead try 24.e6. This prevents
. . . fxe5, but the downside is that White
relinquishes the ability to open the e-file. Thus
1 64 Attacking the Spanish

Black can return to his original plan with 23.exf6

24 . . .'eB. (24 . . . !b4!? was possible, but the text
is simple and strong) 25.YMxc3 YMh5 I prefer
Black's position, as the fS- and e6-pawns could
easily become weak.
Evidently White was not happy with the
position after 23.e6 YMeB, when the queen
comes to the ideal h5-square.
23 .ixf6

19 b4
. Black's position seems to be slightly
We now follow the game Ronchetti -
Brunello, Fiuggi 2007.
20.h4 .bf5
As we have already seen, it can be difficult
to decide on the correct timing of this move.
In general, unless there is a pressing reason I
would say that Black should avoid revealing
his intentions for as long as possible. For this
reason, today I would prefer the more flexible
20 . . . bxc3.
21.gxf5 Wlc8
This looks like the most consistent follow
up, although 2 1 . . .bxc3 was also possible.
22.Wld3
favourable, thanks primarily to his sounder
pawn structure. At the same time, the position
is rather complicated and White is also not
without his chances.
24Je6 bxc3 25Jael Wld7 26.Wlxc3 Wlf7
The queen will be ideally placed on h5, as is
so often the case in this variation. Nevertheless,
26 .. JabB also deserved serious consideration.
27.Wld3 gad8
27 . . JabB! looks best, e.g. 2B.Ek6 (perhaps
White should play cautiously with 2BJH e2)
2B . . . YMh5 29.hc7 bcB 30.c5 lLJb2!
8
7

6
8 5

7 4

3
6
2
5
1
4
a b c d e f g h
3 White is unable to maintain the defence of
2 both f3 and fS .
1
28. <tt> g2 Wlh5?!
a b c d e f g h This pawn sacrifice was a bit too optimistic.
22 f6!
29 .bc7 Wlg4t 3o.Ag3 <tt>h8 31.g1e2

Black fortifies the kingside in the usual way.

The computer wants to play 22 . . . bxc3.
However, after the natural 23. <i!?h2 the
open g-file will always ensure some practical
compensation for White.
Vl-Vl
By now the tide has turned in White's favour,
so I was happy to accept a draw. It seems to me
that Black has no real problems in the B . . . d6
 Chapter 6 - Anti-Marshall 1 65

variation. At the same time, the reader may also 1 5 .exd5 ttJxd5 1 6.he7 ttJxe7 1 7.ttJg5 Wl'f4
wish to investigate a less common continuation 1 B.WI'd2 ttJg6 1 9.WI'xf4 ttJxf4. Black had full
offering good chances to unbalance the game. compensation and the game was eventually
It is to this possibility that we will now turn drawn.
our attention.
b) In Ivanchuk - Leko, Morelia/Linares 200B,
B2) 8 ti)xd4!?
. White preferred 1 3.lLlg4 and achieved a slight
edge after: 1 3 . . . ttJxg4 1 4.WI'xg4 d5 1 5 .WI'h5
dxe4 1 6.WI'd5t i>hB 1 7.WI'xaB i.b7 1 B .WI'a7
8
aB:
7
6
5
4
3
2
1

a b c d e f g h
a b c d e f g h
This leads to more unusual and dynamic
positions than B . . . d6. It is a good option for 1 9.i.f4! Wl'c6 20.WI'xaBt haB 2 1 .i.e3
players who are happy to take some risks in The position is close to equal, although
order to play for a win with the black pieces. We White retains a small advantage and in the
examine: B21) 9 ..lrl'7t and B22) 9.ti)xd4. present game Ivanchuk eventually managed to
grind out a win.
B21) 9.i.xf7t gxf7 10.ti)xe5 gf8
If Black retreats the knight from d4 then
ttJxf7 followed by e5 will be problematic for
him. 8
7
1 l .'ifxd4
White has emerged with an extra pawn, 6
but Black's bishop pair and open f-file should 5
provide full compensation.
4
1 l ...c5 12.'ifdl .lb7 3
This would be my own preference, although
1 2 . . . WI'c7 is also playable. Now White can try 2
two moves: 1

a b c d e f g h
a) In Timofeev - Nyback, Plovdiv 200B, the
continuation was 1 3.ttJf3 i.b7 1 4.i.g5 d5! 14... ti)xg4
1 66 Attacking the Spanish

Also playable is: 1 4 .. Jae8 1 5 .1L1xf6t ( 1 5 .i.g5 Black had regained his pawn while keeping
lLlxg4 1 6.'W'xg4 i.d6 might transpose to the his positional trumps.
main line after 1 7.g3) 1 5 . . . bf6 1 6.1L1d5 'W'e5
1 7.1L1xf6t xf6 1 8.'W'xd7 i.c6 1 9.'W'd2 d6 B22) 9.tilxd4 exd4
20.'W'f4

8
7

1
--
a b c d e f g h

20 . . . 'W'xe4! 2 1 .'W'xe4 xe4 22.i.e3 g4 23.g3 a b c d e f g h

c4 White's extra pawn does not count for
much, and the most likely result is a draw.
IS.Wxg4 .id6
Black's lead in development and powerful
pair of bishops ensure full compensation for
the pawn. We will follow the game Sivanandan
- Negi, Chennai 2008.
16.g3 gae8
Also possible was 1 6 . . . i.e5.
17..igS .ieS 18.gadl d6 19.tildS Wf7
20 ..tof4 .bb2i
IO.eS
White must of course avoid the so-called
"Noah's Ark Trap": 1 0.'W'xd4?? c5 followed by
. . . c4, trapping the bishop.
IO ... tile8
. Now White can regain his pawn with B221)
I l .Wxd4 or adopt a gambit approach with
B222) I l .c3.
B221) I l .Wxd4 cS!?
This has not been the most popular choice,
but I believe it to be a fully viable option. More
common is:
1 1 . . .i.b7 1 2.c4 bxc4 1 3.'W'xc4 d5 1 4.exd6
lLlxd6 1 5 . 'W'g4

a b c d e f g h a b c d e f g h
 Chapter 6 - Anti-Marshall 1 67

White's sounder pawn structure gives him
some chances for an edge.
1 5 . . . c!Llb5
Heading for d4. Nijboer - Sokolov,
Netherlands 1 996, instead saw: 1 5 . . . .tf6
1 6. c!Llc3 bB 1 7 . .tf4;!;
1 6.c!Llc3!
White relinquishes his structural advantage,
instead pinning his hopes on the greater
activity of his pieces.
1 6 . . . c!Llxc3
1 6 . . . .tf6 1 7.c!Lle4 he4 I B."lWxe4 .td4 1 9."lWf4
c5 20 . .te3 he3 2 1 .fxe3 "lWe7 22.aca,
Balogh - lenni, Dresden 2007.
1 6 . . . c!Lld4 also fails to equalize after 17 . .th6
( I 7.xe7? "lWxe7 threatens mate on e l )
1 7 . . ..tf6 I B.ad 1 .
1 7. bxc3 .td6
The position looks close to equal, but in
Ivanchuk - Aronian, MorelialLinares 200B,
the brilliant Ukrainian was able to turn
The alternative 1 2 . . . bB is also fully playable.
Kholmov - Smyslov, Sochi 1 974, continued:
1 3.c4 .tb7 1 4."lWe2 b4 (also very interesting
is the pawn sacrifice 1 4 . . . bxc4 1 5 .hc4 d5!?
1 6.ha6 ha6 1 7."lWxa6 c!Llc7 with decent
compensation) 1 5 .c!Lld2 c!Llc7 1 6.c!Lle4 ( I 6 . .tc2
f5 1 7.exf6 hf6 I B.c!Llb3 d6 is equal) 1 6 . . . c!Lle6
1 7."lWd3 f5!. Black obtained interesting
counterplay.
13.c4 bxc4
1 3 . . . bB is possible, although if Black wanted
to place the rook here then it would have made
more sense to do so on the previous move.
14 ..ic2
After 1 4.hc4 d5 1 5 .exd6 hd6 Black's fluid
development makes up for the slight weakness
of his structure.
14 g6
..

his slight initiative into something more

tangible.
8
I B . .tf4 "lWf6 1 9.hd6 cxd6 20.ad l adB
2 1 ."lWb4 .taB 22.e3 g6 23."lWb6 "lWg5 24.g3 7
"lWb5 25.xd6 xd6 26."lWxd6 6
White had won a pawn.
5
12.%'fe4 4
3
8
2
7
1
6
a b c d e f g h
5
Black has enough activity to offset his slightly
4 loose pawn structure. Statistically White's
3 results have been extremely poor from this
position, although objectively I would tend to
2 evaluate the chances as approximately equal.
1
15.%'fxc4
a b c d e f g h Two other moves have been tried.
12 ... tLlc7
1 68 Attacking the Spanish
a) l S.lbc3 dS
This is the simplest response, although
l S . . J!b8 is also playable. Van der Wiel -
Ki.Georgiev, Wijk aan Zee 1 988, continued
1 6.i.a4?! at which point Black could have
obtained an edge with 1 6 . . . dS!N 1 7.exd6
hd6 1 8 . .ih6 .ifS 1 9.c6 ( 1 9.xc4?? !!b4
20.e2 h4-+) 1 9 . . . !!xb2+.
1 6.exd6 hd6
White has no advantage, and the game Sikora
Gizynska - Maskova, Naleczow 1 987, was
agreed drawn after:
1 7.xc4 .ie6 1 8.a4 Ad7 1 9.c4 .ie6
20.a4 lh-lh.
b) l S . .ih6 !!e8 1 6.lbc3 !!b8 1 7.xc4 !!b4
This is an important method of activating
the rook. We will follow the game: Brooks -
Kaidanov, New York 1 990.
1 8.d3 !!h4!?
Black is playing quite creatively, although
1 8 . . . .ib7 1 9.!!ad 1 !!d4 20.g3 lbe6 was
also fine.
1 9 ..ie3 dS 20.exd6 hd6 2 1 .h3 .ixh3! 22.gxh3
!!xh3 23.f1 !!h2 24.c4
a b c d e f g h
24 . . . !!xe3! 2S.!!xe3 VNgst 26.!!g3 Axg3
27.lbe4 h4 28.fxg3 !!h 1 t 29.@f2 h2t
30.@f3 !!xa1
Black has emerged with a decisive advantage,
and soon won after:
3 1 .lbf2 lbe6 32 . .ib3 lbd4t 33.@e3 xg3t
34.@e4 !!e 1 t 3S .e2 !!xe2t 36.@dS eSt
37.@c4 lbc6 38 . .id 1 lbaSt 0-1
15 .. Jb8 16.tl:k3!N
This untested move looks like White's best.
In Krnjovsek - Lejlic, Ljubljana 1 993, Black
had no problems after 1 6.b3 !!b4 1 7.d3 .ib7
( 1 7 .. .f6! would have enabled Black to fight
for an advantage) 1 8.lbd2 !!d4 1 9.e2 lbe6.
Black went on to win with a crushing attack,
although the position at this stage is equal.
16 .. Jb4
1 6 . . . dS is a valid alternative which should
lead to approximate equality after accurate
play from both sides. Play continues 1 7.exd6
hd6 1 8.lbe4 !!b4 1 9.d3:
a b c d e f g h
Black now faces a choice:
a) 1 9 . . . .ifS does not enable him to equalise
after 20.xd6 xd6 (20 . . . !!xe4 2 1 .xf8t
xf8 22.he4 he4 23.!!xe4;!;) 2 1 .lbxd6 hc2
22.b3 !!d8 23.lbc4 lbe6 24 . .ib2;!;. Black faces
a difficult defence due to his inferior structure
and weak dark squares on the kingside.
b) 1 9 . . . .ie7 20 . .if4 lbe6! (20 . . . !!xb2 2 1 .xd8
!!xd8 22.hc7 !!d7 23 . .id6 hd6 [23 . . . !!xc2
24.lbf6t @g7 2S ..ixe7 !!xe7 26.!!xe7 @xf6
27.!!ee 1 ] 24 . .ia4! .ie7 2s .hd7 hd7;!;)
2 1 ..ieS lbd4o (2 1 . . .xd3 22.hd3;!;) The
position is dynamically balanced, although
plenty of complexity remains.
 Chapter 6 - Anti-Marshall 1 69

17.6'fl d5 18.exd6 .bd6

8
Black's active pieces compensate for his split
queenside pawns. 7
6
19.ttle4 ttle6 20.a3 gb6
From this slightly unusual square the rook is 5
ready to enter the action via d6. 4

21.ttlxd6 gxd6 22 ..ih6 ge8 23.6'c4 gd4 3

24.6'c3 .t.b7 2
1
8
a b c d e f g h
7
Black should not delay in challenging the
6 enemy centre.
5
13.6'8
4 1 3 .ttJd5 .ie6 1 4.'Wf3 c8 transposes to the
3 main line.
2 There is an important alternative in: 1 3 . .id5
1 b8 14 . .ie3 .ie6 1 5 ..ia7 hd5 1 6.hb8
a b c d e f g h
8
It seems to me that Black's active pieces 7
should be enough to compensate White's pair
6
of bishops.
5

B222) l 1 .c3 4
White fights for the initiative with the aid of 3
a pawn sacrifice. 2
1 1%({'="'w'q;t,;" /'{'="l
1 1. .. dxc3
a b c d e f g h
1 1 . . .d3 is sometimes played, but does not
quite equalise after 1 2.'Wxd3 .ib7 1 3.ttJd2 With a choice for Black:
d6 1 4.e6 fxe6 1 5 Jxe6. This position looks
somewhat risky for Black, although the game a) The high level game Dominguez - Aronian,
Dominguez - Aronian, Beersheba 2005, was Wijk aan Zee 2009, continued 1 6 . . . .ib7
soon agreed drawn after 1 5 . . . @h8 1 6 . .ic2 g6 1 7 ..ia7 'Was ( 1 7 . . . c5!? may have been better)
1 7.ttJb3 Yz-Yz. 1 8.exd6 hd6 1 9 ..id4 b4? ( 1 9 . . . c5 20 . .ie5
he5 2 1 .xe5 hg2 would have left White
12.ttlxc3 d6 with only a slight advantage) 20.ttJa4 hg2
2 1 .xe8 'Wxe8 22. @xg2 when Black did not
have sufficient compensation.
1 70 Attacking the Spanish

b) Stronger would have been:
1 6 . . . .iaB!N 1 7 . .ia7 c5
The bishop on a7 is in danger of being
trapped. Play might continue:
1 B.'IWe2
1 B .ttJd5 hd5 1 9.xd5 c7 20.a4 b4
2 1 .exd6 hd6 22.hc5 xc5 23Jad 1
c2+
1 B . . . d7 1 9.exd6 hd6 20Jad 1 b4!
20 . . . xa7? 2 1 .xeB lets White off the
hook.
20 . . . g6 is not clear after 2 1 ..ib6.
2 1 .hc5
After 2 1 .ttJe4 .ixe4 22.xe4 xa7 23.xeB
xeB 24.xeBt .ifB 25 .ddB g6 26.xf8t
@g7 Black has excellent winning chances.
2 1 . . .bxc3 22.hd6 ttJxd6 23.bxc3
23.d3?! c6 does not help White.
23 . . . c6+
Black escapes the pin, and his two pieces
should prove more effective than the opponent's
rook and pawn. White will have to struggle to
hold this position.
13 ....ie6 14JldS
1 4 . .if4 is another logical move, when Black
faces a choice:
White had regained his pawn while keeping a
slight initiative. The game proceeded with:
1 9 . . . ttJb5
Now White could have obtained the
advantage with Marin's suggestion of:
20.ttJxb5!
The game itself was soon agreed drawn after
20.ttJc6 ttJd4 2 1 .ttJxd4 xd4.
20 . . . axb5 2 1 .d 1
From here, Marin offers the following sample
continuation:
2 1 . . .bB 22.h4!? a2 23.h5 h6 24.e2;;l;
White's advantage is defined by his active
pieces and safer king.
b) In view of the above, I think that Black
should prefer:
1 4 . . . dxe5 1 5 .he5 .id6 1 6.ad 1
Marin proposed 1 6.he6 fxe6 1 7.c6 as an
improvement. He may be correct, although
Black can still obtain full equality with
1 7 . . . bB! 1 B . .ixd6 xd6 1 9.xe6 xc6
20.xc6 dB!.
16 . . . hb3 1 7.axb3 ttJf6 1 B.hd6 cxd6 1 9.c6
eB
The position is equal, Polzin - Stern,
Germany 2006.

a) Jakovenko - Karj akin , Foros 2007, 14 ... E!c8 lS ..if4 dxeS

continued:
1 4 . . . hb3 1 5 .axb3 dxe5 1 6.he5 .id6
8
1 7.ttJxb5 he5 1 B.xe5 ttJd6 1 9.ttJd4
7
6
5
4
3
2
1

a b c d e f g h
a b c d e f g h
16.tLlxe7t
 Chapter 6 Anti-Marshall
- 171

White also gains no advantage with: B . . . d6. We will now turn our attention to a
1 6.he5 .td6 very important anti-Marshall variation.
The game Henrichs - Gustafsson, Bad
Koenigshofen 2007, continued: C) 8.a4
1 7Jad 1
1 7 . .tc3 might lead to a draw after 1 7 . . . hd5
( 1 7 . . . '!Wg5! ? is a winning attempt) 1 B.'!Wxd5
b4 1 9.xeB '!WxeB 20.hg7 @xg7 2 1 .'!Wg5t
with a perpetual.
1 7 . . . he5 1 B.xe5 hd5
1 B . . . ltJd6 also deserves consideration.
1 9 .exd5 ltJd6
Black has no problems. The most that White
can reasonably hope for here is to regain his
pawn.

16 YHxe7
..

a b c d e f g h
8 b4
..

B . . . .tb7 is a respectable alternative, but I

a b c d e f g h
We have been following the game Adams -
Aronian, Wijk aan Zee 200B, which was soon
agreed drawn after:
17.,be5 lLld6 18.he6 fxe6 19.YHc6 YHe8
20Jac1 lLlc4 21 .YHe4 YHf7
1/2-1/2
White has enough play for a pawn and is
not in any danger, but at the same time he can
scarcely hope for any advantage.
As far as I can see, B . . . ltJxd4 looks like a
fully playable alternative to the more solid
choose to focus on the text. We now consider
Cl) 9.d4 and C2) 9.d3.
The following rarely played alternative should
not be too threatening:
9.c3 d6
In this position it would be a mistake for
Black to play 9 . . . d5 ?!, as compared to the
usual Marshall lines White has the extra
defensive option of .tc4-fl .
1 0.d3
1 O.d4 bxc3 1 1 .dxe5 ( 1 1 .bxc3 .tg4 can only
lead White to troubles, as the queenside is
too open) 1 1 . . .ltJxe5 is fine for Black.
In Karjakin - Svidler, Dresden 200B, White
preferred the unusual 1 O.a5 bB 1 1 ..tc4.
In this position Svidler came up with a very
interesting pawn sacrifice: 1 1 . . . .te6!? 1 2.ha6
aB 1 3 . .tb5 ltJxa5 1 4.cxb4 ltJb3 1 5 .xaB
'!WxaB, with excellent compensation.
1 O .. J%bB 1 1 .ltJbd2 ltJa5 12 . .ta2 c5
We have transposed to line C2) , page 1 72.
Cl) 9.d4 d6 10.dxe5
1 O.c3 reaches the above note.
1 72 Attacking the Spanish

10 ... tDxe5 Marin has pointed out that Black should have
1 O dxeS is possible, but it looks quite logical
preferred the preparatory l S . . . VNc8!, intending
to exchange the passive knight. . . . fS . In this case his position would have been
absolutely fine.
1 1 .tDxe5
In Anand - Naiditsch, Dortmund 2004, 12 ... Ae6 13.tDd2 Ac5 14.h3 tDd7 15.We2
White obtained no advantage with 1 1 .tDbd2 We7 16 ..ic4 a5
tDxf3t 1 2.tDxf3 i.b7 1 3 .eS tDd7 1 4.e6 We have been following the encounter
fxe6 l S .i.xe6t c;!;>h8 1 6.i.dS c6 1 7.i.a2 cS Dominguez - Navara, Turin 2006. Black
( 1 7 . . . dS also fine for Black) . The game soon eventually lost this game, but at this stage I see
ended in a draw after 1 8.i.dS i.xdS 1 9.'xdS no reason to evaluate his chances as worse.
tDb6 20.VNe6 i.f6 2 1 .aS tDd7 22.VNdS tDeS
23.tDd2 tDg4 24.tDf3 tDeS 2S.tDd2 tDg4 C2) 9.d3
Y2-Y2 . The problem with the previous line, from
White's point ofview, is that the early exchanges
l 1. .. dxeS removed a lot of tension from the position.
The text, on the other hand, leads to much
more subtle manoeuvring, with more chances
8
for both players to outplay the opponent.
7
6 9 ... d6
5
8
4
7
3
6
2
5
1
4
a b c d e f g h
3
12.W9
White has also failed to demonstrate an 2
advantage with other moves. 1

a b c d e f g h
In Jansa - Smagin, Copenhagen 1 992, he
achieved nothing with 1 2.i.gS i.cs 1 3.VNf3 10.tDbd2
i.g4! 1 4.,ixf6 VNxf6 1 S .VNxf6 gxf6. The doubled There has a major alternative in:
f-pawns were not really weak, and Black's 1 0.aS
powerful bishop pair gave him the advantage. With this move White fixes the pawn on a6
as a potential target. On the negative side,
After 1 2.VNe2 i.b7 1 3.f3 tDhS 1 4.g3 c;!;>h8 he spends a tempo and presents Black with a
l S .i.e3 fS?! 1 6.exfS xfS 1 7.tDd2 VNd6 1 8 .i.c4! square on bS, which might later be occupied
White stood better and eventually prevailed in by a rook.
Nisipeanu - Sokolov, Kerner 2007. However, 1 O . . . i.e6 1 1 .tDbd2 b8
 Chapter 6 - Anti-Marshall 1 73

From here both sides will patiently try 2S . . . W'b7!

to improve their positions. Although the The idea is to play . . . dS, with chances to take
chances are approximately balanced, it over the initiative in the centre.
is possible for either side to outplay the
opponent if they understand the position b) 1 2.i.c4 W'c8 1 3.'Dfl
better. We will now look at a few examples. Black has nothing to fear from the slower
In the present position White can occupy 1 3.b3 'Dd7 1 4.'Dfl i.f6 I S .'De3 hc4
the c4-square with either minor piece: 1 6.'Dxc4 'DcS 1 7.i.b2 'De6 with equal play,
Jakovenko - Zhang Zhong, Taiyuan 2006.
a) 1 2.'Dc4 h6 1 3 . . . h6
Black should safeguard the position of the 1 3 . . . hc4 1 4.dxc4 would give White a nice
bishop on e6. grip on the centre.
1 3.h3 1 4.'De3
After 1 3.c3 bxc3 1 4.bxc3 W'c8 I S .i.a3 E:d8 1 4.h3 E:e8 I S .'De3 i.f8 1 6.i.d2 hc4
1 6.i.a4 hc4 1 7.hc6 i.bS 1 8.hbS E:xbS 1 7.'Dxc4 W'e6 1 8.'Dh2 E:bS was equal in
Black has no problems, Sebag - Karj akin , Leko - Grischuk, Dubai 2002.
Mainz 2007. 1 4 . . . E:e8 I S .'DdS
13 . . .W'c8 1 4.i.e3 E:d8 I S .W'e2 i.f8
8
8 7
7 6
6 5
5 4
4 3
3 2
2 1
1 a b c d e f g h
a b c d e f g h
I S . . . i.d8!
Black has organised his position solidly, and It would be too early to exchange on dS
the chances are balanced. Let us follow the immediately, as the knight on c6 would have
high level game Adams - Anand, San Luis no good squares. We have been following the
200S. game Leko - Svidler, Moscow 2002, which
1 6.'Dfd2 'De7 1 7.d4 'Dg6 1 8.dS i.d7 1 9.i.a4 continued:
i.bS! 1 6.i.d2 hdS
Black is happy to exchange bishops, but it This is the correct timing, as the e7-square
will be on his own terms. is available.
20.b3 i.e7 2 1 .E:ec 1 c6 22.hbS E:xbS 23.'Db6 1 7.hdS 'DxdS 1 8.exdS 'De7 1 9.d4
W'b7 24.dxc6 W'xc6 2S .W'c4 Y2-Y2 After a period of accurate manoeuvring by
Evidently the players were content to share both players, there follows a sequence of
the point, although the game was far from exchanges resulting in a completely equal
dead. If anything, I would slightly prefer position.
Black's chances after: 19 ... 'DxdS 20.dxeS dxeS 2 1 . 'D xeS i.f6 22.'Dc6
1 74 Attacking the Spanish
a8 23.lDxb4
23.f3 d7 24.lDxb4 lDxb4 25 .hb4 hb2
is also equal.
23 . . . lDxb4 24.hb4 .txb2 25 .b 1 -
Both of the above games are fairly typical
for this line. The positions can sometimes be
deceptively difficult, and demand a high level
of understanding from both players. However,
if Black plays accurately then his chances are
by no means worse.
1 O ... tLl a5
1 O . . . .te6 is a valid alternative. The last trend
for White has been 1 1 .he6 fxe6 1 2.lDfl !?,
bringing the knight to g3 before advancing in
the centre. The position is playable for Black,
but I consider it to be strategically slightly
riskier than the main line.
1 1 .L2 i.e6
Compared with the previous note, I believe
that the extra move . . . lDc6-a5 improves Black's
chances by facilitating the space-gaining
. . . c5 advance. Black can also consider the
immediate:
1 1 . . . c5
In this case White should prefer a patient
build up. The following examples illustrate
the risks associated with a premature central
expansion.
a) 1 2.c3 b8 1 3.d4 .tg4!
This is a clever move. The pin on the knight
actually has a serious impact on the opposite
side of the board!
1 4.h3
No better is: 1 4.dxe5 dxe5 1 5 .h3 hf3
1 6.xf3 b3 ( 1 6 . . . c7 1 7 . .tc4 b6=)
1 7 . .tb 1 c4+
1 4 . . . hf3 1 5 .lDxf3 b3!
This is the idea: White's queenside pieces
will be trapped!
1 6 . .tb 1 c7
8
7
6
5
4
3
2
1
__=m
a b c d e f g h
White's queenside pieces are in serious
danger of being imprisoned.
1 7.d5
It looks logical for White to rescue his bishop
with 1 7 . .td3. However, in Corby - Mitchell,
England 1 998, Black secured an excellent
game with 1 7 . . . exd4! 1 8.cxd4 c4 1 9 ..tfl d5!
20.exd5 lDxd5 .
1 7.e2!? c4 1 8 . .te3 seems to be the best
choice for White. The position is rather
double edged. On the one hand, if Black
can maintain his blockade of the e4-pawn
while somehow managing to exchange
all of the kingside pieces, then he will be
winning automatically. Unfortunately this
is easier said than done. White has quite a
strong position on the kingside, and it will
hard for Black to achieve his ambitions
without opening the position. Overall I
would evaluate the position as unclear,
with perhaps a very slight preference to
Black.
1 7 . . . c4=F
Compared with the above note, the closing
of the centre is a big help to Black who no
longer has to worry about maintaining the
tension. Having said that, it is still not easy
for Black to achieve his goal of exchanging
the kingside pieces. In the end, White
managed to escape with a draw in Kulaots
Bartel, Gjovik 2009.
 Chapter 6 - Anti-Marshall 1 75

b) A more prudent option is: 1 5 . . . lLlc6 is also perfectly fine. Karjakin -

1 2. lLlc4 lLlc6 Alekseev, Khanty-Mansiysk 2007, was soon
From here the game Morozevich - Leko, agreed drawn after 1 6.lLlb3 YMeB 1 7.lLlg3
Monte Carlo 2006, continued: h5 1 B.h3 YMg6 1 9.YMe2 h4 20.lLlf1 lLlh5
1 3.c3 -Y2 .
This is not a bad move, although I would
prefer a more patient build-up with 1 3 .h3 16.tLlf3 tLlc6 17.tLlg3 tLlg4 18.e5
or a developing move with the dark-squared
bishop.
8
1 3 . . JbB
Black has good prospects on the queenside. 7
1 4.h3 h6 1 5 .d4 cxd4 1 6.cxd4 lLlxd4 1 7.lLlxd4 6
exd4 1 B.i.f4 i.e6
5
4
3
2
1

a b c d e f g h
So far we have been following the game
Svidler - Navara, Crete 2007. Black opted for
a b c d e f g h 1 B . . . adB, but failed to equalise after 1 9.exd6
i.xd6 20.i.g5 . Black' s position is still defensible
1 9Jk 1 ?! but clearly unpleasant. White eventually won.
White could have maintained the balance Far stronger would have been the natural
with: 1 9.e5 lLld5 ( 1 9 . . . dxe5 20.lLlxe5! is central strike:
slightly awkward) 20.i.g3 dxe5 2 1 .i.xe5 cB
22.YMxd4. 18 d5!

1 9 . . . lLld7 20.i.b 1 i.g5! 2 1 .i.xg5 YMxg5 22.i.d3
lLle5
Black was clearly better and went on to win.
12.,be6 fxe6
The knight on a5 discourages the usual c3,
d4 plan, as the knight would eye the potential
holes on b3 and c4.
13.d4
If Black were allowed to play . . . c5 then he
would be very comfortable.
13 exd4 14.tLlxd4 V;Yd7 15.tLlfl c5
One can only assume that Black was afraid
of:
19.h3
However, it seems that he has two satisfactory
responses here.
By far the simplest reaction would be
1 9 . . . xf3 20.YMxf3 lLlgxe5 2 1 .YMe2 i.d6 with
good compensation for a very small material
investment.
It also looks interesting to consider: 1 9 . . . lLlxf2!?

20.<tt> xf2 i.h4! 2 1 .e4! YMe7 22.f4 g5!

1 76 Attacking the Spanish

8 there is nothing to stop you from transposing

with 8 . . . d6 9.c3 etc.
7

6
9.d3
5 9.c3 can be met by 9 . . . d5! 1 0.exd5 xd5 .
4

3 a) 1 1 .xe5? is almost suicidal, as after 1 1 . . . xe5

1 2.xe5 we reach a normal Marshall except
2
that Black has been gifted the free developing
1
move . . . j,b7, while White has wasted time on
a b c d e f g h the weaking h2-h3. Black is almost winning
Two possible continuations: after 1 2 . . . f4!

a) After 23J:l:xh4 gxh4 24.f1 xe5 25. 1 h2 8

xf3 26. xf3 f5 Black is doing well. 7

6
b) 23.xh4 looks better, when there follows 5
23 . . . gxf4 24.VNg4t @h8 25 .e2 xe5 26.VNh5
4
VNf6 with a highly unclear position.
3

D) 8.h3 2

1 :L:=

8 a b c d e f g h

7 For instance, 1 3.d4 (after 1 3.f3 j,d6 the

black queen will shortly join in the fun)
6 1 3 . . . xg2 1 4.VNg4 h4 1 5 .d2 @h8!,
5 Laouini - Henni, Cairo 1 998. White faces the
extremely unpleasant threat of . . . f5 .
4
3 b ) 1 1 .d3 is not s o bad, but can hardly threaten
Black after 1 1 . . .VNd6 1 2.bd2 ad8 1 3.e4
2
VNg6, Fischer - Szabo, Portoroz 1 958.
1

a b c d e f g h c) 1 1 .d4 is well met by 1 1 .. .exd4 1 2.cxd4

a5!?:
This is the anti-Marshall of the new
millennium! The text became fashionable
during the past decade, and remains a common
occurrence at high level tournaments.

8 ...Ab7
This is the typical move for Marshall players.
Black retains the option of . . . d5 . Of course, if
your repertoire also includes the Gajewski then

a b c d e f g h
 Chapter 6 - Anti-Marshall 1 77

Black takes the opportunity to eliminate the
'Spanish bishop' from the board. Daulyte -
Azarov, Warsaw 200S, continued 1 3.i.c2 lDb4
1 4.i.e4 (If White is going to exchange this
bishop, he would sooner do it for a bishop than
a knight) 1 4 . . .i.xe4 I S .xe4 i.f6 1 6.lDc3 lDdS
1 7.lDeS cS!, when Black was even beginning to
take over the initiative.
9 ... d5!?
This was an important discovery. For several
years it was presumed that White was doing
enough to prevent this sacrifice, but it now
appears fully playable. The alternative was
9 . . . d6. This leads to positions similar to those
considered in variation C) above, with some
subtle differences. Black's position is perfectly
playable, but our main line is much more in
the spirit of the Marshall.
10.exd5 ttlxd5
8
I S .d4 exd4 1 6.cxd4 lDd8 Y2-Y2.
b) 1 2.c4!? was a creative idea, although
White also failed to obtain an advantage after
1 2 . . . bxc4 1 3.lDxc4 i.cs ( 1 3 . . . i.b4!?) 1 4.lDe3
lDaS I S .d4 exd4 1 6.lDxd4 lDxb3 1 7.xb3
i.xd4 1 8.xb7 i.xe3 1 9.i.xe3 d7. White's
bishop is not really any stronger than Black's
well centralised knight on dS , Adams - Bacrot,
Baku 2008.
Dl) 1 1.a4
This should not be too dangerous, although
if White is not feeling ambitious then he can
more or less force a draw.
1 1 ... ttld4! 12.ttlxd4
White can hardly take on eS , as after . . . lDxb3
the extra pawn would not make up for the
loss of the bishop pair and ruination of his
queenside structure. We now follow the game
Karjakin - Aronian Wijk aan Zee 2009.
12 exd4 13.axb5 axb5 14J:xa8 .bas
..

7 15.ttla3 AM
6
5 8

4 7

3 6

2 5

1 4

a b c d e f g h 3

White has three principal options: Dl) 2

1 1 .a4, D2) 1 1 .d, as well as, of course, D3) 1
1 1 .ttlxe5.
a b c d e f g h
l 1 .lD bd2 should be met by 1 1 . . .f6: The last few moves require little explanation.
Both sides are playing consistently, putting
a) In Lanka - Gustafsson, Graz 2008, White their pieces on the best available squares.
failed to put his opponent under any pressure
with 1 2.c3 <i>h8 1 3.lDfl d7 1 4.a3 ae8 16 ..id2
1 78 Attackin g the Spanish

1 6Je5 i.d6 leaves White nothing better We now follow the game Topalov - Leko,
than retreating, as 1 7 Jhd5 ? i.xd5 1 8.i.xd5 Morelia/Linares 2008.
i.xa3 works out badly for him.
15 .tf6 16.gel g;,h8

16 hd2 17.YlYxd2 YlYf6 18 .hd5 1/1-1/1

1 6 . . . fe8 1 7.i.d2 xe l t 1 8.i.xe l i.e5 1 9.f4
The players saw no reason to continue the i.d6 20.liJd2 @h8 2 1 .i.xd5 i.xd5 22.liJf3 e8
fight. Note that after 1 8.liJxb5 liJf4 1 9.f3 gives Black compensation according to Postny.
i.xf3 20.gxf3 h4 White has no more than a The evaluation is probably correct, although I
draw. After the move played, the game might do not believe this line to be an improvement
have continued: 18 ,ixd5 19.1Llxb5 .ixg2
over Leko's move.
20.g;,xg2 YlYc6t 21.g;,g3 YlYxb5 22.ge7=
White's exposed king should not change the 17.lLld2 b4!
evaluation, as the material is so limited.
8
D2) 1 l .c3 YlYd7!?
7
8 6
7 5
6 4
5 3
4 2
3 1

2 a b c d e f g h
1 An excellent move, which reminds us that
Black should always look for opportunities to
a b c d e f g h play across the whole board and not just on
1 1 . . .d6 is also possible, but I prefer to the kingside.
continue with the gambit theme.
18.lLlf3!
12.lLlxe5 It is pointless for White to try to keep his
1 2.d4 exd4 1 3.cxd4 i.b4 ( 1 3 . . . ae8 is also extra material. Instead he must continue to
logical) 1 4.i.d2 ad8 1 5 .liJc3 liJf6 leads to improve his pieces.
a complex fight where Black's chances are at
least not worse, Sokolov - Sargissian, Plovdiv 18 bxc3 19.d4 f4 20.YlYg4 YlYxg4 21 .hxg4

2008. We have reached a critical position. In the

12 lLlxe5 13Jxe5 gad8 14.YlYg4 f5 15.YlYg3
Less testing is 1 5 .h5 i.f6 1 6.e l @h8,
Zagrebelny - Giorgadze, Yerevan 1 996
( 1 6 . . . fe8 is also good) . White's queen is
achieving very little on h5.
game Black could find nothing better than the
slow 2 1 . . .g5 ?!, and after 22.liJe5 @g7 23.bxc3
liJxc3 24.i.b2 liJd5 25 .ac U White had
achieved a high level of piece coordination and
went on to win. Instead, it would have been
much more logical and principled to play:
 Chapter 6 - Anti-Marshall 1 79

21. .. c5! draw after 1 3 .tDdf3 tDxb3 1 4.axb3 f6 1 5 .tDg4

This was recommended by Marin. d7 1 6.i.d2 c5 1 7.e2 h5 1 8.tDgh2 i.d6
1 9.f1 lh-lh. White's position is too passive
22.bxc3 xc3 23.dxc5 for him to claim an advantage, but at the
White does not appear to have anything same time it is quite solid.
better: 1 3 . . . tDxb3 1 4.tDxb3 c5
As usual, Black's pair of bishops and generally
a) Marin gives the line: 23.i.a3 i.xf3 24.gxf3 active pieces ensure good compensation.
hd4+ Hou Yifan - Kosteniuk, Nalchik 2008,
continued:
b) 23.tDe5 f3 24.tDxf3 hf3 25.gxf3 tDb5 is 1 5 .h5 g6 1 6.f3 f6 1 7.tDxg6 hxg6 1 8.c4
equal. d7 1 9.cxd5 hd5 20.g3 g5

23 e2t
.

23 .. Jfe8!? is also quite playable.

24J:xe2 hal

8
7
6
5 a b c d e f g h

4 Black's kingside weaknesses cannot be

exploited, and the bishop pair remains a
3 potent force.
2 2 1 .i.e3 ac8 22.g4 xg4 23.hxg4 i.d6
Black's positional trumps remain, and if
1
anything it is White who has more reason to
a b c d e f g h worry about becoming worse. The game was
White probably has enough compensation eventually drawn.
for the exchange, but no more. Note that
25.tDg5 can be met safely by 25 . . . i.d5 . 1 2.i.d2 e8 1 3 .tDc3 tDxb3 1 4.axb3 i.d6

D3) 1 1 .xe5 d4!

After the moves 1 1 . . . tDxe5 1 2.xe5 Black's
compensation is questionable.

12.c3
Two other moves deserve our attention:

1 2.tDd2 e8 1 3.c3
Ivanchuk - Svidler, Foros 2008, resulted in a
a b c d e f g h
I BO Attacking the Spanish
1 5 .llJf3
1 5 .d4 llJb4 1 6.llJe4 enables Black to
liquidate to a drawn ending with 1 6 . . . he5
1 7.dxe5 xe5 I B.hb4 xe4 1 9.xdBt
xdB 20.xe4 he4 2 1 .xa6 hc2 22.c6
i.xb3 23.xc7 h6 24.c5 .ic4 25 ..ic3 d5
26.xd5 hd5 Y2-, Jakovenko - Grischuk
Elista 200B.
1 5 . . . xe l t 1 6.llJxe l
After 1 6.he l the simple 1 6 . . . llJf6 i s good
enough.
1 6 . . . llJf6 1 7 . .ig5
Now White's position becomes risky. Safer
would have been 1 7.d4, when Postny
analyses 1 7 . . . d7 I B.llJd3 eB 1 9 . .if4 c6
20.f3 b6 2 1 .hd6 xd4t 22.llJf2 xd6
23. xd6 cxd6 with equality.
1 7 . . . h6 I B . .ih4 g5 1 9 . .ig3 hg3 20.fxg3 d6
Black had more than enough for a pawn in
Zhigalko - Pashikian, Martuni 200B.
12 ... xb3
Black gains nothing by postponing this
exchange with 1 2 . . . llJb4 1 3.llJe4, as he can
hardly have anything better than 1 3 . . . llJxb3
1 4.axb3 transposing to the main line.
13.axb3 b4
8
7 sufficient.
6
5
4 b) 1 4 . .id2 f6 1 5 .llJg4
3 White, although he might be able to defend,
2 Schachinger - Freitag, Leoben 200B.
1
a b c d e f g h
14.e4
White has a wide choice of moves here,
although in most cases the general evaluation
of the position remains the same.
a) 1 4.d4
This is quite playable, although White
should always think carefully about where
to place this pawn. The text gains space in
the centre, but by committing this pawn to
a dark square White helps to increase the
potential of Black's unopposed light-squared
bishop. Quezada - Nielsen, Havana 2007,
continued:
14 . . . eB 1 5 . .ie3 .id6 1 6.llJd3 e6
8
7
6
5
4
3
2
1
a b c d e f g h
At this point White was content to return
the pawn to reach a drawn ending with:
1 7.d5
He could, of course, have played more
ambitiously with 1 7.f3, but after 1 7 . . . d7
intending . . . dB Black's activity should be
1 7 . . . llJxd5 I B.llJxd5 .ixd5 1 9.1lJf4 hf4
20.hf4 xe l t 2 1 .xe l c6 Y2-Y2
1 5 .llJf3 i.xf3 1 6.gxf3 is unpleasant for
15 ... f5 1 6.llJe5 .id6 1 7.d4
Too passive is 1 7.llJa2 he5 I B.xe5 llJc6
with good play for Black, Timofeev -
Sokolov, Sarajevo 2007.
1 7 . . . h4 I B.llJe2
 Chapter 6 - Anti-Marshall
a b c d e f g h
1 B . . . f4! ?
Simpler would have been 1 B . . . Wle4 1 9.ttJf4
Wlxc2 with equality.
1 9.f3 ttJd5 20.ttJc3 ttJe3 2 1 ..!xe3 fxe3 22Jhe3
adB
Black has compensation even for two pawns,
Ki. Georgiev - Ivanisevic, Valjevo 2007.
c) In Gormally - Beliavsky, Liverpool 200B,
White was successful with: 1 4.i.f4 Wld4 1 5 .i.g3
i.d6 1 6.ttJe2 Wlc5 1 7.c3 .!xe5 1 B . .!xe5 Wlxe5
1 9.cxb4 adB?! 20.Wld2 d7 2 1 .ac l The xe 1 t 23.xe 1 cB 24.e2 i.e4 25 .ttJxb4 a5
English Grandmaster went on to convert his
advantage.
Instead there were two decent alternatives on
move 1 9:
c l ) 1 9 . . . Wlxb2 was suggested by Postny, who
analysed as far as 20.d4 feB 2 1 .b 1 Wla3
22.Wld2 a5 23.bxa5 Wlxa5 . This looks like an
improvement over the game, although if we
continue a little deeper it seems that White has
some chances to be better after 24.Wlxa5 xa5
25.ttJc3 i.c6 26.ttJd5!.
c2) In view of the above, I believe that the
simplest solution would have been 1 9 . . . Wlg5!N
20.ttJg3 Wld5 2 1 .f3 Wld4t 22.@h 1 Wlxb4 when
Black is absolutely fine.
14 tvd5
.
The game Timofeev - Jakovenko, Moscow
1B1
200B, continued with the less precise 1 4 . . . f5?!
1 5 .ttJd2 i.f6 1 6.ttJdf3 when the outpost on
e5 gave White good chances of an advantage.
After 1 6 . . . eB 1 7.i.d2 c5 1 B.d4 he5 a critical
position was reached:
a b c d e f g h
White can recapture in three different ways:
a) In the game White squandered any real
chances ofplaying for a full point with 1 9. ttJxe5 ?
Wlxd4 20 . .!xb4 Wlxd 1 2 1 .axd 1 cxb4 22.ttJd3
26.ttJd3 .!xd3 27.cxd3 c l t 2B.@h2 d 1 .
Black easily defended this endgame. However,
both of the following alternatives would have
left Black struggling.
b) After 1 9.dxe5 hf3 20.gxf3 Wlh4 (20 . . . e6
2 1 ..!xb4 will leave Black facing a difficult,
though perhaps still tenable endgame) 2 1 . .!xb4
cxb4 22.Wld5t @hB 23.e6 the e-pawn is a very
serious threat.
c) Also promising would have been 1 9.xe5
i.xf3 20.gxf3. The doubled pawns, while
far from perfect, are not a serious liability.
Meanwhile the onus will be on Black to
demonstrate compensation for his lost pawn.
Perhaps he can draw the ending after 20 . . . Wlxd4
2 1 .xeBt xeB 22.c3 Wld3 23.cxb4 dB
24.bxc5 Wlxd2 25.Wlxd2 xd2 26.xa6 xb2
although White should definitely play on for a
while.
1 82 Attacking the Spanish

Returning to the main line, we will now Conclusion

follow the game Piscopo - Brunello, Cortina We have examined four different variations
2007. in which White can sidestep the main lines of
the Marshall Attack. Anti-Marshall systems
ls.tDa cS will always remain popular, especially amongst
I rather like this patient approach. Black has amateur players who may view them as a more
positional compensation and does not need to convenient solution than accepting the gambit
be in a hurry to prove anything. Nevertheless pawn.
there was a valid alternative in 1 5 .. f5, after
. Out of the four variations that we examined,
which Postny analyses 1 6.ttJeg5 i.d6 1 7.i.d2 variation A) , with 8.c3 d5 9.d4, should be
fe8 1 8 . .ixb4 i.xb4 1 9.c3 i.d6. the easiest for Black to meet. Line B) , 8.d4, is
somewhat more challenging, and I hope that
16.Af4 Ud8 my decision to cover two separate responses
1 6 . . . ae8 1 7.d2 f5 is a good alternative. will prove helpful to the reader. Variations C)
and D) are arguably the two most important
anti-Marshall variations that Black needs to
know. The former tends to lead to a subtle
manoeuvring game after 8.a4 b4. Generally
8
speaking, the high-level games seem to indicate
7 that accurate play from both sides will lead to
6 a balanced game. Line D) with 8.h3 has only
become popular relatively recently, and the
5 theory is still developing. The gambit approach
4 with 8 . . . i.b7 9.d3 d5!? seems to be holding up
at present, although I am sure that this line will
3
continue to be debated and tested at GM level
2 for some time. Generally speaking, I would
always recommend that the reader keeps an
1
eye on the theoretical developments, as well as
a b c d e f g h conducting his own original analysis at home.
Black has typical compensation for this line.
He has two strong bishops, active pieces and
no real weaknesses. Even if White manages to
exchange queens, it would be tough for him to
make anything of his extra pawn.
Chapter 7

Marshall: 12.d4
7
6
5
4
3

1
a b c d e f g h

l .e4 e5 2.tLla tLlc6 3.i.b5 a6 4.i.a4 tLlf6 5.0-0 i.e7

6Jel b5 7.i.b3 0-0 8.c3 d5 9.exd5 tLlxd5 1 0.tLlxe5
tLlxe5 I l Jxe5 c6 1 2.d4 i.d6 1 3Jel YNh4 14.g3 YNh3

Introduction page 1 84
Theoretical Highlights page 1 86
Theory page 1 87
A) 1 5 .i.e3 page 1 88
A I ) 1 7 . . . Wfh5 ! ? page 1 8 8
A2 ) 1 7 . . .E!e6 page 1 92
B) 1 5 .!%e4 page 207
B l ) 1 6.ltJd2?! page 208
B2) 1 6.Wff3 page 209
B3) 1 6.Wfe2 page 2 1 2
B4) 1 6.Wfe l page 2 1 3
B5) 1 6.Wffl page 2 1 5
1 84 Attacking the Spanish

1.e4 e5 2.lLla lLlc6 3.Ab5 a6 4.J.a4 lLlf6 preparing to attack the enemy rook on the
5.0-0 JJ.e7 6Jel b5 7.Ab3 0-0 8.c3 d5 next move.
9.exd5 lLlxd5
12.d4
This is the most natural move. It remains
8
a popular choice at all levels, although in
7 recent years 1 2.d3 seems to have become
6 the main line amongst the world's elite. I
think that it makes sense for us to begin by
5 analysing the most natural move. By doing so,
4 we will be able to appreciate certain nuances
of the position which have influenced the
3
subsequent developments in the theory. The
2 presently fashionable 1 2.d3 will form the
subject of Chapter 8, and White's other 1 2th
1
moves will be analysed in Chapter 9, along
a b c d e f g h with a few of the sidelines that can occur after
In the present chapter we will begin to 1 2.d4.
investigate the acceptance of the gambit pawn,
as occurs after: 12 ...Ad6

10.lLlxe5 lLlxe5 1 1 .gxe5 8

7
Introduction
The Marshall has always been a controversial 6
and divisive gambit - it seems that some Ruy
5
Lopez players love to try and refute it, while
others are so terrified that they avoid it at all 4
costs! There is an old adage which states that 3
"in order to refute a gambit, you must accept
it". So far the Marshall has stood the test of 2
time, and continues to be used at the highest 1
levels of chess. In the present and remaining
a b c d e f g h
two chapters I will endeavour to provide an
up to date summary of the most important The present position will serve as the
variations, as well as some challenging sidelines. starting point for the theoretical section of this
We will begin by discussing the next couple of chapter.
moves, which lead us to the starting point for
the present chapter. Strategic Themes

1 1 c6
..
This is firmly established as the most
promlsmg move. Black spends a tempo
to stabilise the knight in the centre, while
White consolidates his position
If Black were to follow his pawn sacrifice
with bog-standard developing moves, we might
reach a position resembling the following:
 Chapter 7 Marshall: 1 2.d4
- 1 85

something in between the two extremes shown

8
above.
7
6 Pawn power
Both sides will often look to open a file to
5 further their ambitions. The following is a
4 standard position, which could occur with
queens on a few different squares (d3 or f1 for
3
White, and h3 or h5 for Black) .
2
1 8
a b c d e f g h 7
Black has obtained no compensation 6
whatsoever for the sacrificed pawn. Needless
5
to say, he can do much better than this!
4
Black's kingside attack 3
Black should look to involve his queen in
the attack at an early stage. After provoking 2
the weakening pawn move, g2-g3, he will 1
often use his f-pawn to increase the pressure.
a b c d e f g h
The following would be a dream position for
him. White will look to open the a-file to activate
his rook and soften up the enemy queenside.
And we have already seen that Black's f-pawn
8
plays an important part in his plans. As a
7 general rule, if White cannot prevent the pawn
6 from advancing to f4 then he will be in big
trouble. Therefore he will either have to block
5 the position by playing f4 himself, or find
4 some other way to defend (e.g. force a queen
exchange) .
3
2 Exchanging the attacking pieces
White will sometimes try to exchange, or
1
at least drive away, some of the opponent's
a b c d e f g h attacking pieces.
Black is developing a huge attack, and
can easily involve the rest of his pieces with (see diagram next page)
. . . .ig4 and . . Jae8. Obviously White does not
have to end up in this nightmare scenario,
and the majority of games will resemble
1 86 Attackin g the Spanish

White is looking very solid on the kings ide,

8
but his pieces are not so well equipped to fight
7 for the centre. Therefore Black should change
6 plans with 22 . . . c5!, which gives him good
chances. This variation is analysed fully in line
5 A2 1 1 ) on page 1 99.
4
Theoretical Highlights
3
2 The main line
From the starting position mentioned at the
1
beginning of the chapter, the most popular
a b c d e f g h variation has been the following:
Having given up a pawn, Black must be sure
to play energetically, otherwise his initiative 1 1 . .. c6 12.d4 .ld6 13J:tel Yfh4 14.g3 Yfh3
may be extinguished by the exchange of his
active pieces. 8
7
Playing on both flanks
When the Marshall first became popular, 6
Black players usually had one thing on their 5
minds: to smash through the enemy defences
and checkmate the white king. Nowadays a 4
lot of defensive methods have been worked 3
out and, of course, checked with computers.
Thus it has become a lot more difficult for 2
Black to succeed with a crude attack against 1
a well-prepared opponent. For this reason the
a b c d e f g h
modern interpretation of the Marshall involves
playing across the whole board. The following This has been the main line for a long time
position is a good example: and it still enjoys great popularity at all levels.
White has tried several different ideas here, but
the most important move is:
8
7 15 .te3
.

6 This is the subject of line A) , beginning on

page 1 88. There are a number of sidelines,
5 which will be considered in Chapter 9.
4 Black should also be ready for the interesting
alternative: 1 5 J!e4!?, intending to drive the
3 queen away with h4.
2
The ge4 line
1
We will end the chapter by examining the
a b c d e f g h
 Chapter 7 Marshall: 1 2.d4
- 1 87

following important sideline (after the usual 13Jel

1 1 . . .c6 1 2.d4 j,d6 1 3Je 1 VBh4 1 4.g3 VBh3) : This is by far the most popular move,
although 1 3.e2 has also been used by
15Je4! some strong players. This alternative will be
analysed in Chapter 9. Let me reiterate that
the present chapter will focus on White's most
8
natural response to the gambit. By learning to
7 appreciate the nuances of these positions, the
6 reader will be in a better position to understand
the pros and cons of the various alternatives
5 and sidelines.
4
13 ...h4 14.g3
3 It is well known that White should avoid:
2 1 4.h3? i.xh3 1 5 .gxh3
Also unsatisfactory is: 1 5 .hd5 j,h2t!
I
L: 1 5 ... VBxh3
a b c d e f g h
This is an important alternative, which
demands accurate handling by Black. We
will see in line B) (page 207) that correct play
should enable him to reach a fully satisfactory
position.

Theory

1 .e4 e5 2.tLla tLlc6 3 ..ib5 a6 4..ia4 tLlf6

5.0-0 ILe7 6Jtel b5 7..lb3 0-0 8.c3 d5 a b c d e f g h
9.exd5 tLlxd5 10.tLlxe5 tLlxe5 1 1.xe5 c6 1 6.e5!N
12.d4 .ld6 This looks like the only chance, although it
is still not enough to make the line a viable
8 proposition for White.
Several games have ended with 1 6.hd5 ??
7
2t followed by mate in three.
6 After 1 6.f4, Nuschke - Quiring, Sackville
200 1 , the most efficient route to victory
5
would have been 1 6 . . . VBg3t!, when either
4 king move can be met by 1 7 . . . lLlxf4 with a
3 decisive attack.
1 6 . . . he5
2 1 6 . . . ae8 1 7.h5 enables White to defend.
I 1 7 .dxe5 ae8
1 7 . . . VBf5!?
a b c d e f g h
1 88 Attacking the Spanish
1 8.hd5 cxd5 1 9.i.f4 ge6 20.ttJd2
20.i.g3 h5 2 1 .f1 g4 wins.
20 . . . gg6t 2 1 .i.g3 f5!
White has no good defence.
I4...Wlh3
a b c d e f g h
In this position White has two main options:
A) I5 ..ie3 and B) I5.ge4. The former is the
most natural and by far the most popular
move, but the latter is also quite challenging
and demands attention.
A) I5.i.e3 i.g4 I6.Wld3
The queen must stay within touching
distance on f1 , in order to parry the potential
mating threats after . . . i.f3. Obviously 1 6.f3??
hg3 is no good for White.
I6 gae8 I7.tLld2
.
White needs to complete development.
Now we reach a very important moment. We
will consider AI) I7...Wlh5!? as well as the
main line: A2) I7 Jle6. The two moves can
.
often lead to the same position after White's
typical response: 1 8.a4. The main purpose of
line A I ) is to avoid line A2 1 ) below ( 1 8.f1 ) .
I do not believe that this move brings White
any advantage, but Black must nevertheless
be prepared for it. On the other hand,
line A I ) also offers White some additional
options, primarily 1 7 . . . h5 1 8.i.c2. Overall
I would rate the two move orders as being of
approximately equal strength, so I advise the
reader to study both options and decide which
he prefers.
Before moving on I will briefly mention the
other main option:
1 7 . . . f5
Historically this has been Black's main
alternative to 1 7 . . . ge6. While the position
remains complicated, if White defends
correctly then it is doubtful that Black can
break through using brute force alone. For
this reason, most Marshall experts now prefer
a more refined strategy. Instead of staking
everything on a kingside breakthrough, the
modern interpretation involves a more fluid
approach in which Black combines his attacking
potential with his positional trumps, including
his control over the e-file and the weakness
of White's light squares on the kingside. So
while the immediate advance of the f-pawn
may not necessarily be bad, I would prefer to
keep more options open at this stage of the
game.
AI) I7 ...Wlh5!? I8.Ac2
The usual response is 1 8.a4, when 1 8 . . . ge6
reaches line A22) . As mentioned previously,
the text is an attempt to take advantage of
Black's slightly unusual move order.
Other possibilities include:
a) 1 8.ttJf1 ge6 1 9.i.d 1
1 9.a4 will reach line A22) after a subsequent
exchange on b5.
19 . . f5 2o.hg4 xg4 2 1 .i.d2 gg6
.
Black has good compensation, and eventually
his threats became too strong in Ivanchuk
Adams, Terrassa 1 99 1 . It is worth playing
through the remainder of this game, as
Black's play was extremely instructive. I
especially like the way that Adams was able
 Chapter 7 - Marshall: 1 2.d4
to switch between attacking weaknesses on
both sides of the board.
22. i>g2 f4 23.f3 'MIhS 24.g4 'MIh4 2S Je2?!
a b c d e f g h
This looks like a mistake, although Black
had nice compensation in any case.
2S,..xg4t! 26.fxg4 f3t 27.i>h 1 fxe2 2B.'MIxe2
i>hB 29.i>gl h6 30.'MIg2 f4 3 1 .e 1 'MIgS
32.h3 'MIg6!
A very nice 'creeping' move, typical of the
British number one.
33.d 1 i.bB 34.d2 'MIb l 3S .f2 i>gB
a b c d e f g h
White is horribly restricted, and Ivanchuk is
unable to withstand the pressure.
36.b3 f4 37.e2 ltJxc3 3B.e6 'MIxa2 39.xc6
'MIxb3 40.xa6 ltJe2t 4 1 .i>h 1 bB 42.e 1
'MId I 0- 1
b) One of the clever ideas behind Black's
move order is to meet l B. 'MIfl with the careful
1 B, . .e7!:
1 B9
a b c d e f g h
Compared with line A2 1 ) the rook is less
exposed than on e6. Here we have:
b 1 ) The clever point behind of Black's last
move is revealed after:
1 9.f3? ltJxe3 20.'MIf2
Compared with the analogous position with
the rook on e6 instead of e7, here Black can
simply play:
20, . .feB!
The rook's position on e7 instead ofe6 means
that it is not under fire from the bishop on
b3.
2 1 .fxg4 ltJxg4 22.xe7 xe7
Only Black can be better here, for instance:
23.'MIg2
After 23.d 1 cS! Black has the initiative.
23, . .'MIxh2t
23, . .g6!? also looks promising.
24.'MIxh2 ltJxh2 2S.i>xh2 e2t 26.i>h3 xd2
27.fl xb2 2B.hf7t i>f8! 29.hSt i>e7
30.f7t i>e6 3 1 .xg7 c2
Black retains some winning chances,
although White can probably hold.
b2) Bacrot - Inarkiev, Elista 200B, continued
1 9.hdS 'MIxdS 20.f3 5 2 1 .b3 feB 22.f2
c2 23.xe7 xe7 24.e 1 e6 2S.xe6
'MIxe6 26.'MIe l 'MIxe 1 t 27.he 1 5. Black had
no real problems, as his light-squared bishop
was extremely strong. There is a simple plan
of , . .a3-c 1 xd2 leading to a drawn opposite
bishop ending.
1 90 Attacking the Spanish
b3) 1 9.a4 gfe8 20.hd5 xd5 2 1 .axb5 axb5
Peter Leko has achieved two comfortable
draws from this position against world-class
opposition.
b3 1 ) In Anand - Leko, Dortmund 2007,
White decided to exchange the rooks with:
22.f4 gxe l
But not 22 . . . hf4? 23.gxe7. This type of
possibility is the only real drawback of
having the rook on e7 instead of e6.
23.gxe 1 gxe 1 24.xe 1 f8
Black's strong pair of bishops mean that he
is is no real danger, even after a subsequent
queen exchange.
25 .e4 f6 26.xd5t cxd5 27.c7 @f7 28.b6 22.d3? ge2 23.h4 hf4! with a crushing
d6 29.c5 Y2-Y2
Anand obviously felt that it would be a waste
of time even to try to win this position.
b32) In Inarkiev - Leko, Elista 2008, White
instead elected to exchange queens with:
22.g2 xg2t 23.@xg2 f6
Black's pieces are all very active, so he sets
about improving his king and pawns.
24.b3
24.ga6 gets nowhere after 24 . . . c8!'
24 . . . h5 25.c4 b4 26.ga2 f5 27.gc l g5
28.h3 @f7 29.ltJfI gd7 30.@h2 d3 3 1 .ga6
ge6 32.ga8 ge8 33.ga6 ge6 34.ga8 ge8 Y2-Y2
Once again, Black had no problems
whatsoever in holding the draw.
18 ... f5
a b c d e f g
We already know that this is a typical
attacking move in the Marshall. In the present
position is could turn out to be even more
effective than usual, as the knight on d5 is no
longer pinned by the bishop on b3.
19.9
1 9.f4? is not really playable. Play proceeds
with 1 9 . . . ltJxe3 20.gxe3 and now:
a) At first it looks as though Black can win
with 20 . . . gxe3 2 1 .xe3 ge8, intending
attack. However, it turns out that White can
survive with 22.b3t! @h8 23.e6, although
even here it should be pointed out that after
23 . . . hf4! Black will regain his piece while
keeping the initiative.
b) The simplest move is probably 20 . . . e2,
when White loses the exchange for insufficient
compensation.
19 ....ih3
This looks more challenging than 1 9 . . . hf3
20.ltJxf3 xf3 2 1 .d2 xd3 22.hd3;!;; .
White's light-squared bishop became very
powerful in Ponomariov - Leko, Moscow
2008.
 Chapter 7 Marshall: 1 2.d4
- 191

20.W f4! 23 ... h8 24.fl hS!

This move highlights a real drawback of
White's 1 8th move - the lack of a pin on the
black knight. The next couple of moves are
forced.

21.gxf4 .bf4 22 ..ig3 g5 23.i.b3t

In an earlier game from the same event,
Efimenko - Sargissian, Dresden 2008, White
played:
23.tDfl h5 24.@f2
24.i.b3t makes little sense, as h7 is available
for Black's king now.
White should also not be attracted to:
24.@h l h4 25 J!e2 hxg3 26.hxg3 (26.tDxg3 a b c d e f g h
hg3 27Jgl xe2 28.'\&xe2 f4+) 26 . . . hfl Preparing to regain the piece while also
27.i.b3t @h7 28.h2t @g6+ creating an escape square for the king.
24 . . . h4 25.xe8 xe8 26.e l hxg3t 27.hxg3
2S.gxe8
25.@h l should lead to a draw after
25 . . . h4 26.e2 i.xfl 27.xe8 hd3 28.xf8t
@h7 29.i.g8t @h6 (29 . . . @g6 30.i.f7t @f6??
3 1 .e l +-) 30.i.f7 @h7 (30 ... g6?? 3 1 .hh4+-)
3 1 .i.g8 t etc.

2S ... gxe8 26.i.f'7 ge7 27..hhS i.xfl

28.xfl
28.@xfl is also level: 28 . . . hg3 29.hxg3
a b c d e f g h
xg3 30.xf5 h4! The only move, but good
enough. 3 1 .c8t @h7 32.i.g6t (32.f5t
Now Black erred with: @h8=) 32 . . . @xg6 33.g4t (33.xc6t??
27 . . . d8?! @h7-+) 33 ... xg4 34.fxg4 @g5=
Correct would have been 27 ... xe 1 28.@xe l
hg3t 29.tDxg3 xg3t with an easy draw. 28 ...i.e3t 29. hl xhS
28.@e2?! The dust has settled. White has retained his
White could have obtained a serious extra pawn, but Black's active pieces and safer
advantage with 28.e2!, although it is easy king ensure adequate compensation.
to miss such a move over the board.
28 . . . hfl t 30.g2
Black is okay again. Mikhalevski analyses the line: 30.i.e5 e6
29.xfl e8t 30.@dl xg3 3 1 .xf5 e l t! 3 1 .e l f4 32.e2 g6 33.g2 h3 34.e2
Forcing the draw. f5 35.d l xg2 36.@xg2 @g8 Black has
32.xe l xf3t 33.e2 fl t 34.e l 3t full compensation thanks to his active pieces,
Y2-Y2
especially the beautifully placed queen.
1 92 Attacking the Spanish

30 ... f4 31..ifl Yfd5i rank. We will now consider White's two main
moves: A2l) l8.Yffl and A22) l8.a4. Instead
1 5 .hd5 cxd5 reaches line A14) of Chapter 9,
8
page 264.
7
6
A2l) l8.Yffl Yfh5 19.3!
This is the clever point behind White's last
5 move. He offers a temporary piece sacrifice,
4 and hopes to obtain a positional advantage
after regaining the material. In this important
3 position we will investigate three alternatives:
2 A21 1) 19 ....if5, A2l2) 19 JU6 and A2l3)
.

19 tLlxe3. In my opinion the last option seems

..

1
to be best. The alternatives are unpromising:
a b c d e f g h
1 9 . . . .ih3? 20.f2 f5 2 1 .f4 leaves Black with
Black retained full compensation in Ni Hua
no real compensation, as the knight is coming
- Sargissian, Dresden 2008. In addition to the
to e5.
move played in the game, another good option
was 3 1 . . . c5 . 19 ... xe3 also fails to equalise after 20.fxg4
It seems that 1 7 . . . h5! ? is a perfectly valid xg4 2 1 .xe3 tDxe3 22.f2 tDd5 23.e 1 .
move order. Based on the evidence presented White has a slight but stable edge thanks to his
above, it seems unlikely that White has active pieces and extra central pawn.
anything better than 1 8.a4, transposing to line
A22) on page 1 98 after 1 7 .. Je6. A2l l) 19 ....if5 20.tLle4
20 . .if2 g6 100ks risky for White.
A2) l7 J: e6
.

20. f2 should be met by 20 . . . .ig6! (20 . . . fe8

8 2 1 .hd5 cxd5 22.tDb3;t) 2 1 .tDe4 (2 1 .a4
fe8 22.axb5 axb5=) 2 1 . . .he4 22.fxe4 xe4
7 23 . .id l h3 24 . .if3 xe3 25.xe3 tDxe3
6 26.xe3 d7 with equality.
5 20 ...,he4 21 .fxe4 gxe4
4
8
3
7
2
6
1
5
a b c d e f g h
4
This is still the main line. Black gives himself
the options ofdoubling on the e-file or swinging 3
the rook to an attacking position on the third 2
1
a b c d e f g h
 Chapter 7 Marshall: 1 2.d4
- 1 93

Material is level. The question is whether (27 . . . e3 2B.f1 xf1 29.xf1 ) 2B.g2;t
White will be able to make his bishop pair and 26.i.g5!
extra central pawn count for something. The
difficulty is that Black's pieces are rather active.
Still, if White can coordinate his pieces then he
may be able to torture his opponent for a long
time in the ending. At the present time, the
set-up employed for White by GM Etienne
Bacrot seems to be the last word in this line.

22 ..tdl
Another reasonable continuation is 22 . .if2! ?
feB 23.xe4 xe4 24.e l g6 25.xe4
a b c d e f g h
xe4 26.e 1 xe 1 t 27.he l . White must be
slightly better in this ending, although Black The only chance for an advantage, although
has good chances to hold. Black can still maintain the balance with
precise play.
22 ... YlYg6 23.i.c2 5 25 . . . h6!
23 . . . hg3? does not work in view of: 24.he4 26 . . . xg5? 27.e6t regains the piece, after
xe4 25.hxg3 ttJxe3 26.f2 eB 27.e2 e6 which White should win without too many
2B.ae 1 h6 problems.
26 . . . hg3 is also not quite sufficient after:
8 27.hxg3 xg5 2B.e6t @fB 29.xe4
7
xg3t 30.@h l g6 3 1 .gl c7 32.ge l
@g7 33.e5t xe5 34.xe5
6
27.cBt
5
After 27.i.h4 @h7!+ White's pieces are very
4 badly placed, while his material advantage is
3 not felt.
2 27 . . . @h7!?
27 ... .ifB also leads to a satisfactory position,
1
e.g. 2B.i.h4 ttJe3 29.i.e7 (29.ac1 f7
a b c d e f g h
30.xc6 f5=) 29 . . . ttJc2 30.f1 ttJxal
29.g2!+- 3 1 .xf3 exf3 32.xfBt @h7 33.xf3 b l t
34.@g2 xb2t 35.@h3 xa2 36.e4t
24 .td2!
@gB 37.xc6 f7! with equality.
This was a very important discovery by Bacrot. 2B.xc6 e6 29.i.e7
Instead, ifWhite takes the exchange then Black 29.i.d2?! e3 30.he3 xe3 3 1 .xe3 xe3t
seems to have enough compensation: 32.@f1 e4+
24.he4 fxe4 25 .h3 29 . . . h3 3o.hd6
25.g2 f3 26 . .if2 h5 gives Black enough After 30.xd6 ttJe3! (30 . . . f2 3 1 .@xf2
compensation. xh2t 32.@f1 e3 33.xe3 ttJxe3t 34.@e l
25 . . . f3 xb2=) 3 1 .xe3 xe3 the threat of . . . e2
Worse is: 25 . . . h6 26.@h l f3 27.i.gl @h7 forces White to find: 32.g4! xg4 t 33. @h l .
1 94 Attacking the Spanish

At this point Black has nothing better than Black's best chance was probably 25 . . . f7
33 . . . f3t with a perpetual. 26.f2 g6 27.e l , although even here White
30 . . J!f2! 3 1 .@xf2 xh2t 32.@f1 e3 33Jhe3 has a stable advantage thanks to the bishop
ttJxe3t 34.@e l ttJc2t pair as well as the slightly vulnerable position
The game will end in perpetual check - of the rook on g4.
unless, of course, White prefers to be mated
after: 26Jae1 Bf7?
35 .@dl ttJe3t 36.@c1 ?? Losing, although Black's position was already
36.@e l = difficult. For instance, after the superior
36 . . . c2#! 26 . . . h6 there follows: 27.f2 ttJe7 28.i.d l
e4 29.xe4 fxe4 30.xf6 gxf6 3 1 .h3 @h7
We now follow the game Bacrot - Jakovenko, 32.d7 f5 33.i.b3 h5 34.@f2!
Kallithea 2008.
27Je8t .if8 28JM8
24.. Jg4 25J3e2! There is no good defence against ee8.

8
28 ... e4 29 ..be4 fxe4 30.Bgl e3 31 ..be3
Be7 32Jxd5 ad5 33.Bxd5t
7 White was winning easily and soon converted
6 his advantage in Bacrot - Jakovenko, Kallithea
2008.
5
4 Al12) 19 .. JU6
3
8
2
7
1
6
a b c d e f g h
5
An excellent move, preparing to swing the
rook to g2 in the event of a sacrifice on g3. At 4
the same time White makes room for the other 3
rook to come to e 1 . Black is unable to break
through on the kingside, so White has time 2
to catch up on development, after which the 1
bishop pair will really come into its own.
a b c d e f g h
25 .. JU6 This is a tricky move, but ultimately it seems
25 . . . h5 26.i.d l ! is awkward. that White can maintain an advantage with
precise play.
The aggressive 25 . . . h5 is too slow, e.g. 26J!ae 1
h4? 27J!e6 h5 (27 .. Jf6 28Jhf6 followed 2o.Bgl
by hf5) 28Jhd6 hxg3 29.h3 h4 30.xc6 This looks more challenging than either of
xh3 3 1 .g2 ttJf6 32.i.f4 winning. the two alternatives:
 Chapter 7 - Marshall: 1 2.d4
a) 20.V;Ve2 leads nowhere for White after
20 . . . hf3 2 1 .lLlxf3 xf3 22.hd5 (22 . .if2?
lLlf4=t) 22 . . . V;Vxd5 . Black has nothing to worry
about, and his position may already be slightly
preferable, Anand - Ivanchuk, Bilbao 2008.
b) 20 . .id l e8 2 1 ..if4
In case of2 1 .V;Vf2 Black should play 2 1 . . . V;Vg6
(2 1 . . . .if5 22.lLle4 he4 23.fxe4 V;Vh3 24 . .if3
xe4 25.he4 xf2 26.hf2 h5;1;) 22 . .ib3
.id7 (22 . . . .if5 23.hd5 cxd5 24 . .if4;
22 ... h5 23 . .ig5! fe6 24.lLle4) 23.a4 h5,
with decent compensation.
2 1 . .. xe l 22.V;Vxe l e6 23 . .ie5 i.h3
23 . . . he5?! 24.dxe5 .if5 25.f4 White has
managed to consolidate his position and is
a clear pawn up, Becker - Henao, e-mail
1 995.
24.V;Vf2
24.f4 V;Vg6 25 .V;Vf2 .if8
24 . . . he5 25.f4 hd4 26 . .ixh5 .ixf2t Y2-Y2
Jakovenko - Svidler, Moscow 2007.
20 .ih3
. 30 ..if2 cxd5 3 1 .e3 V;Vh l t 32 . .igl intending
Black can regain his pawn with 20 . . . .ixf3,
but after 2 1 .lLlxf3 V;Vxf3 22.V;Vxf3 xf3 23 . .id2
White has a pleasant endgame advantage.
20 . . . xf3?! is certainly not an improvement:
2 1 .lLlxf3 hf3 22.V;Vf2 f5 23.f1 .ie2 24.V;Vg2
f4 25.xf4 hf4 26.hf4
21 .YlYfl .if5
Plenty of pieces remain on the board, and in
a practical game Black will keep some tactical
chances. However, the bottom line is that
White will be better if he continues to play
accurately.
22.a4!
A typical move, activating the rook while
softening up the enemy queenside.
22 .id3
to hold.
1 95
Perhaps Black should have been considered:
22 . . . g6 23.axb5 axb5 Two possible
continuations include:
a) 24.lLle4
This leads to a forcing sequence:
24 . . . he4 25.fxe4 hg3! 26.hxg3 V;Vh3 27.exd5
xg3t 28.V;Vxg3 V;Vxg3t 29.@f1
a b c d e f g h
29 . . . b4!
This disruptive move prevents White from
consolidating smoothly. Instead 29 . . . V;Vf3t
.id l -f3 gives White good chances to convert
his material advantage.
30.dxc6 V;Vf3t 3 1 . .if2
3 1 .@gl V;Vg3t repeats.
3 1 . . . bxc3 32.bxc3 V;Vd3t 33.@g2 V;Vg6t 34 . .ig3
V;Vxc6t 3 5 .d5 V;Vxc3 36.ab l h5!
White is better, but Black has counterplay.
b) 24.f4!?
This may enable White to obtain an edge
through simpler means. The idea is to return
the pawn in order to make way for the
manoeuvre lLlf3-e5. Play continues:
24 . . . V;Vh3 25.lLlf3 lLlxf4
And not: 25 . . . hf4? 26 . .ixf4 lLlxf4
27.lLle5+-
26.lLlg5! xg5 27.hf4 hf4 28.V;Vxf4 g6
White's better structure gives him the
advantage, although Black has decent chances
1 96 Attackin g the Spanish

We now follow the game Naiditsch - Sargissian, The German Grandmaster went on to
Kallithea 2008. convert his advantage convincingly:

23.axb5 axb5 24.%Vg2 .!.bS 25.g4! %Vg6 37 fxg5 3S.a6 .ib3 39.c7 .ic4 40.e6
.

26.,bd5 ad5 27.b3 The knight's dance forces a fatal weakening

of the enemy pawns.
8
40 ...g4 41.fxg4 hxg4 42.fl f7 43.c5
7 b4 44.ab4 f6 45.a4 e6 46.c3 .id3
6 47.g3 .iS 4S.f4 g6 49.b5 d6 50.b6
1-0
5
4 A213) 19 xe3..

As mentioned previously, I believe this to be

3
the most promising route to an equal game.
2
1
20.%Vfl

a b c d e f g h
8
By now it is clear that Black's strategy has
7
failed.
6
27 h5 2S.g5 e6 29.c5 e7 30.%Vd2 .ic2
..

5
31..if4 .ixf4 32.%Vxf4 gxel t 33Jxe1 f6
34.d7 4
Also tempting was 34.e7!? fxg5 3 5 .e5. 3
White will pick up the d5-pawn, with excellent
winning chances. 2
1
34 .. J;e8 35Jlxest %VxeS 36.%Ybs %VxbS
a b c d e f g h
37.xbS
Black now has two minor pieces plus the
rook on e6 en prise, so he can hardly hope to
8
maintain his material advantage.
7
6 20 d5 21.fxg4 %Vxg4 22.%Vf3 %Vg6
.

I think that this is Black's most promising

5 move.
4 In Kramnik - Aronian, Yerevan 2007, Black
fell into trouble after:
3 22 . . . g5 ?! 23.xe6 fxe6 24.tL\e4! g6 25 .e2
2 tL\f4?!
The lesser evil would have been 25 ... i.f4,
1
although even here after 26.<;t>h l i.e3
a b c d e f g h
 Chapter 7 - Marshall: 1 2.d4 1 97

27 . .!xd5 exd5 28.lLlc5 e8 29.e U ; Black's 27.cxd5 xb3

troubles are far from over. 27 . . . xf3 28 . .tc4 cxd5 29.@g2 e3 30 . .!xa6;1;;
26.c2 lLlh3t?! (Marin) gives White decent winning chances,
26 ... .te7 also leaves Black clearly worse after as the a-pawn is a real threat.
27.e l or 27.lLlf2!? 28.lLld2 d3
27.@g2 Black does not solve his problems with
Black has two pieces hanging. 28 . . . xb2 29.lLlc4 c2 30.lLlxd6 b3
3 1 .e l !. Thanks to the mate threat, White
Black can also exchange queens with: wins a crucial tempo which enables him to
22 . . . xf3 23.lLlxf3 position his rook behind the enemy pawn
while also driving the enemy king further
away. Following the likely continuation of
3 1 . . . h5 32.e8t @h7 33.b8 b2 34.xb2
xb2 35.dxc6 c2 36.d5 @g6 37.@f1 !,
White is winning.
29.lLlc4
We have been following the analysis of
Marin. White has excellent winning chances.

23Jxe6 Wxe6 24.We4 ge8

a b c d e f g h
8
This endgame is deceptively dangerous for
Black. White has the better bishop and an 7
extra central pawn. The placement of the 6
b5-pawn also presents the first player with
an obvious target with which to prise open 5
the queenside. Play may continue: 4
23 . . . fe8
Perhaps it is worth considering 23 . . . h6 3
24.xe6 fxe6 25.lLld2;1;; . The position is 2
somewhat unpleasant for Black, but he has
1
chances to hold, Naiditsch - Ivanisevic,
Subotica 2008. a b c d e f g h
24.xe6 xe6 25.a4! b4 25J: e l
25 . . . h6? 26.axb5 axb5 27.a8t @h7 25.xe6 xe6 is identical to the previous
28.c8 (Marin) . note (22 . . . xf3 23.lLlxf3 fe8 24.xe6 xe6)
25 . . . e2? 26.axb5 axb5 27.a8t .tfB except that the white knigh t is positioned
28.lLle5+- (Marin) . on the less active d2-square. This certainly
26.c4 helps Black. e.g. 26.a4 is well met by 26 . . . b4!
26.cxb4 f6 27.lLle5 lLlxb4 is okay for (26 . . . e2 27.axb5) 27.c4 e2! 28.cxd5 xd2
Black. 29.dxc6 xb2 30 . .td5 @fB with equality.
26 . . . e3
26 . . . lLlf6 27.c5 e3 28.cxd6 xb3 29.lLle5 25 c!>f8
..
1 98 Attacking the Spanish

I believe this to be the most precise move,
although Black also has good drawing chances
after 2S . . . xe4 26.lLlxe4 <j;f8 as suggested by
Mihail Marin.
26. Yxe6 xe6 27.xe6 fxe6
Black is marginally worse, but he should
be able to hold this ending without much
difficulty.
To summarise, 1 8 .f1 is certainly an important
option which deserves serious attention. We
saw in line A I ) that Black can avoid it with
1 7 . . . hS!? if he so desires. Therefore my
advice to Black players would be to compare
lines A I ) and A2 1 ) and choose whichever one
you prefer.
A22) 18.a4 Yh5
Black anticipates the white queen's arrival
on f1 , which will no longer come with gain of
tempo. A secondary point is that an exchange
on dS will no longer force a compromising of
Black's queens ide pawns.
19.axb5 axb5
possibilities for both sides, after first dealing
with a few rare alternatives:
a) In Catapano - Warrick, corr. 1 998, White
made the mistake of trying to utilise the newly
opened a-file too early with 20J%a6? Here the
strongest reaction would have been: 20 . . . lLlf4!
with a dangerous attack. 2 1 .gxf4 can be met
by 2 1 . . J%g6!, when White must give up his
queen.
b) 20.i.d 1 ?? was played in Hellers - 1. Sokolov,
Haninge 1 989, as well as quite a few other
games. Amazingly, so far no-one seems to
have noticed that Black can win by force after
20 . . . lLlxe3!N 2 1 .fxe3 (2 1 .xe3 i.xd 1 wins a
piece for nothing) 2 1 . . .i.xg3! 22.hxg3 h6
with a mating attack.
c) 20.i.xdS xdS is not a serious option for
White, who has needlessly ceded the bishop
pair without good reason. Parma - Geller,
Sukhumi 1 966, concluded 20.c4 bxc4
2 1 .lLlxc4 i.b4 22.ec 1 i.e2 23.i.d 1 i.xd3
24.i.xhS i.xc4 2S.xc4 lLlxe3 26.fxe3 i.d2
Y2-Y2.

d) 20.lLle4
8
This is quite a reasonable move and has been
7 tested at elite level.
6 20 . . . i.f5 2 1 .i.d2 xe4!
Naturally Black captures in this way so as
5 to catch the enemy rook in an awkward
4 pin.
22.xe4 lLlf6
3
Also possible is 22 . . . g6, when Di Berardino
2 - Vescovi, Sao Paulo 2006, concluded
23.ae l lLlf6 24.f3 i.xg3 2S.hxg3 xg3t
1
26.<j;h 1 h3t 27.<j;gl g3t 28.<j;h 1
a b c d e f g h Y2-Y2 .

Now White normally chooses between 23.f3 g6 24.f1

A221) 20.Yfl and A222) 20.tLlfl . Both 24.i.c2 or 24.ae 1 both allow Black to force
moves are quite sensible, and can be met in a draw immediately with 24 . . . i.xg3 etc.
different ways. We will explore all of the major 24 . . . lLlxe4 2S.fxe4 i.xe4 26.i.f4
 Chapter 7 Marshall: 1 2.d4
- 1 99

2 1 .Vf1e2 i.g4 gives White nothing better

than 22. Vf1f1 repeating.

21. VN5 22.VNe2 c5!

8
7
6
a b c d e f g h 5
White's extra central pawn gives him a 4
symbolic advantage, but no real chances to
3
play for a win. Svidler - Kamsky, Groningen
1 995, was soon agreed drawn after: 2
26 . . . .id3 27.Vf1f2 hf4 28.Vf1xf4 h6 29.Vf1f2 e8 1
30.e l xe l t 3 1 .Vf1xe l .ie4 32.Vf1f2 - .
a b c d e f g h
A221) 2o.VNfl Black should always keep in mind the
possibility of combining play in the centre
8 and both flanks, rather than just the kingside.
Note that 22 . . . tDf4?! does not help Black after:
7 23.Vf1f3! tDd3 24.Vf1xfS hfS 25 . .ic2!
6
23.lLla
5
Alternatives give White no chance of an
4 advantage.
3
Sax - I. Sokolov, Haninge 1 989, soon resulted
2 in a draw after 23.tDf1 cxd4 24.cxd4 (24 . .ic2?
1 tDxc3 25.bxc3 Vf1d5+) 24 . . . tDb4 25 .a3 tDc6
(stronger would have been 25 . . . c8!+) 26.d3
a b c d e f g h .ib4 27.d5 d6 28 ..id2 xd5 29.hb4 tDxb4
This looks slightly less purposeful without a 30.f3 Vf1d7 3 1 ..ib3 .ig4 32.hd5 Vf1xd5
black queen on h3, but nonetheless it is quite 33.tDe3 Vf1xf3 -Y2 .
playable and has been used repeatedly at high
levels. We will analyse both of Black's principal Mauro - Coscia, corr. 1 996, proceeded with
responses: A221 1) 20 .ih3 and A2212) 23.tDb3 tDf4! Now this idea works much
20 JUeS. Both are sound, although if one
. better. 24.Vf1f3 .ig2 25.Vf1g4 Vf1d5 26.gxf4 g6
wishes to play ambitiously then it should be 27.Vf1xg6 fxg6 28.tDxc5 hc5 29.dxc5 . At this
noted that the former invites White to repeat point Black should have played: 29 . . . .ih l
moves immediately should he wish it. 30 . .ib3 Vf1xb3 3 1 .@xh l Vf1d5t 32.@gl Vf1f3,
with more or less a guaranteed draw. Instead he
A221 1) 20 .ih3 21 .i.dl
quickly self-destructed with 29 . . . .ih3? 30.f3;;1;
200 Attacking the Spanish

g5? 3 1 ..id4 !%xf4 32Je5+- c4 33 ..ie2 b3 26 JUe8 27.i.c2

34Je7 @f8 35 Jb7 1-0.

The alternative is 27.ltJe5 xf2t 28.@xf2 f6
23 J.f4!
. with a choice for White:
Black takes the opportunity to transfer the
bishop to a more purposeful diagonal, from a) Leko - Kramnik, Monte Carlo (blind) 2007,
which it can menace the weakness on e3. continued: 29 . .ig4?! gxe5! 30 . .ixh3 he3t
and Black went on to convert his advantage.
24.YlYd2
24 . .ic2 g4 25 .d3 g6 leads to a forcing b) Stronger would have been:
sequence: 26.ltJe5 he5 27.dxe5 f3 28.e4 29.ltJg4
xe4 29.he4 ltJxe3 30J%xe3 gxe5 The I think that Black can maintain the balance,
position is equal, and Svidler - Kramnik, although he must be precise.
Monte Carlo 2007, was soon agreed drawn 29 . . . .ig5!
after 3 1 . .ic6 gxe3 32.fxe3 gd8 33.hb5 gd2 29 . . . hg4?! makes Black's task unnecessarily
34.b4 cxb4 3 5 .cxb4 gg2t 36.@h l gb2 37 . .ifl complicated, although he can probably still
Y2-Y2. hang on with accurate play. 30.hg4 f5
3 1 . .ih5!? g6 32 . .if3 @f8 (Black's task is not
24 ti)xe3 25.fxe3 i.h6
made any easier after 32 . . . he3t 33. @g2 @f8
[no better is 33 . . . @g7 34.d5 gd6 35.ga7t
@h6 36.gb7, or 33 . . . gd6 34.dxc5 gd2t
8
35.@h l ] 34.d5 gd6 35.ga7 h5; White is
7 clearly better, but Black may be able to hold
6 with precise defence.) 33 . .id5 (33.d5 gxe3=)
33 . . . he3t 34.@f3 gd6 35.gxe3 gxd5.
5 Black ought to hold this endgame, although
4 White can certainly continue 'playing for
two results' for a while longer.
3
30.dxc5 ge4
2 30 . . . h5 3 1 .i.b3 hg4 may be playable, but is
1
hardly necessary.
3 1 ..if3
a b c d e f g h
It is not easy for White to defend the e3- 8

pawn. 7

6
26.YlYfl 5
26.ltJe5 leads to equality after 26 . . .f6
4
27 . .ic2 (27.ltJd7?? he3t! wins) 27 . . . h5
3
28.ltJd3 (28.ltJd7? gfe8+) 28 . . . gfe8 29.ltJf4
hf4 30.exf4 ge2 3 1 .gxe2 gxe2 32 . .id l gxd2 2
33 . .ixh5 g6. Black guards against mate, and 1
will regain his pawn to reach a completely level a b c d e f g h
endgame.
 Chapter 7 Marshall: 1 2.d4
- 20 1
3 1 . . . f5!
3 1 . . Jhg4 32.hg4 hg4 would be less
accurate, as the f-file remains closed which
means that White's king will be safer.
32.he4 fxe4
Black will capture the knight on g4, thus
obtaining two bishops for a rook. However,
White has some extra pawns on the queenside.
I think that the correct result should be a draw,
although the position is rather unbalanced and
I suppose that the stronger player might try to
win with either colour.
27...fth5 28.lLle5
Also possible is 28.e4 f6 29 . .id l :
a b c d e f g h
a) In Wang Hao - Grischuk, Dagomys 2008,
Black fell into trouble after 29 ... g6? 30.eS f5
3 1 .e2 cxd4 (3 1 . . . .ig4 32.xbS+-) 32.cxd4
d8 33.e4 g4 34.xg4 hg4 3S.lLlh4 gS
36.hg4 xg4 37.lLlf3 1 -0.
b) As our repertoire should not lose by force, I
want to recommend an improvement.
29 . . . .ig4 30.fl g6!
30 . . . .ih3 does not equalise after 3 1 .eS!
hfl 32.exf6 (32.xfl fe6 33.xbS cxd4
34.cxd4;l;;) 32 . . . h3 33.lLlh4 .id3 34 . .if3
d7 3S.lLlg2 (3 S.g2 e6=) 3S . . . cxd4
36.xd4 xd4t 37.cxd4;l;;
3 1 .eS fS
Black has fair counterplay. Play might
continue:
32.h4!? cxd4 33.cxd4 c8
The game will more than likely be drawn, as
it will be very difficult for White to unpin the
knigh t in a favourable way.
28 6 29 .idl ftg5 3o.lLld3 gxe3 31..ib3t
..
c4 32.lLlf4 gxel t 33Jxel .id7
8
7
6
5
4
3
2
1
a b c d e f g h
Material is equal. Black has a pair of
bishops, although the one on h6 is not well
placed at present. White can probably claim
a faint edge, although if Black can relocate
this problem piece to a better square then he
may have chances to take over the initiative
himself
34 .ic2
34.xe8t he8 3 S .e l .if7 36.e7 g6!
completely solves Black's problems, as the
bishop will come to a better square with gain
of tempo.
34... 5 35.gxe8t Axe8 36.ftf3
White should try to create some problems
before Black can coordinate his pieces.
36 ... g6 37.ftd5t jp 38.ftxb5
Winning a pawn, but allowing Black to
create counterplay.
38 ... fte7 39.ftb8t q;; g7
202 Attacking the Spanish

21..ixd5
Usually White feels obliged to make this
exchange in order to relieve the pressure against
e3. Of course, the downside is that his light
squares will become more sensitive.

2 1 .g2!? has seldom been played, but looks

like a reasonable alternative. Kobe - Roskar,
Slovenia 1 992, continued 2 1 . . .xe3 22Jhe3
E:xe3 23.fxe3 E:xe3 24.xc6 E:e2 25 .h4 .if8
26.E:f1 .ie6? 27.he6 E:xe6 28.f3 g6 29.h5
f6 30.xf6 gxf6 3 1 .@f2 and White went on
a b c d e f g h to win the ending with ease.
40.We5t Instead Black should have preferred
White has nothing better than exchanging 26 . . . E:xd2 27 . .txf7t xf7 28.E:xf7 @xf7
queens, but now he will be left with a weak 29.d5t .ie6 30.xb5 with rather a double
e-pawn. For instance, after 40.@f2 hf4! edged ending. White has a slight material
4 1 .xf4 (4 1 .gxf4?! h4t can only be advantage but his pawns are not far advanced,
dangerous for White) 4 1 . . .b7! White has no so I suspect that Black should at least be able to
advantage. draw without too many problems.

40 Wxe5 41.dxe5 <tflf8 42.,ibl <tfle7 43.i.a2

..
Alternatively, he could even consider the much
Ag7 44.e6 he6 45.li:)xe6 <tflxe6 46.hc4t earlier deviation: 2 1 . . .h6! ?N This safeguards
<tfle7 against any back rank mates, thus turning
Unfortunately for White this endgame is . . . xe3 into quite a serious threat as Black will
rather easily drawn, Relic - Equidistance, no longer have to worry about his king after
Internet 2006. the enemy queen arrives on c6. In this case
Black also seems to be fine.
A2212) 20 J U e8
.

21. Wxd5 22.h3

.

White can also play:

8 22.g2 f5
7 Black might consider avoiding the queen
exchange with 22 . . . h5! ? 23.f3 .ih3 24.f2
6 f5 .
5 23.E:ec 1
23.h3 xg2t 24.@xg2 .ih5 transposes to
4
the main line.
3 23 . . . E:g6
2 23 . . . xg2t 24.@xg2 f4 25 . .txf4 hf4
26.gxf4 E:e2 should also be alright for
1 Black.
a b c d e f g h 24.xd5t cxd5 25.@h l b4
 Chapter 7 Marshall: 1 2.d4
- 203

suspect that Black can hold this position as

well.

24. c.t>xg2 f5!

By playing actively, Black easily solves his
problems. We will follow the model game:
]. Polgar - Adams, San Luis 200S.

25.tna
2S.a6 f4 26.gxf4 hf4 27.hf4 xe l
a b c d e f g h
2B.xc6 dl 29.cS .te2 Black should be
Black's powerful light-squared bishop fine, as long as he is not careless enough to
remains an important asset in the endgame. allow the enemy pawns to advance too easily.
2S . . . hS!? also looks playable. For instance, if 30.dS then 30 . . . f8 3 1 .i.e3
eB 32.tDf3 (32 . .if4 repeats) 32 . . . dB with a
22 i.h5 23.ftg2 ftxg2t!
.. safe position.
In Gurevich - Benjamin, New York 1 992,
Black allowed his structure to become damaged 25 4 26 .ld2
..

after 23 . . . fS 24.xdS cxdS . This looks slightly 26.gxf4 hf4 is equal.

riskier, although even here he may still be
able to hold the balance. In the game he was 26 &g3 27 .xe6 l1xe6 28Ja8t .if8
.

able to survive after 2S.tDfl f4 26 . .id2 xe l

27.xe l xe l 2B.he 1 .ie2, when White had 8
no chance to make his extra pawn count.
However, if White had played 2S .aS! then 7
things would not have been quite so easy, e.g. 6
2S . . .hg3 26.fxg3 xe3 27.xe3 xe3 2B.tDfl
e2: 5
4
8
3
7
2
6
5 1
4 a b c d e f g h
3 29.tne5
2 After 29.i>xg3? g6t 30.tDgS h6 3 1 .i>h4
1 .id l 32.tDh7 i>xh7 33.xf8 g2 White's king
a b c d e f g h
would be in serious danger.
29.g4! (29.xb S ? .if3) 29 . . . .ieB (29 . . . fxg4? 29 ...grll 30.c.t>xf2 ge8 31.a6 i.d6
30.tDg3) 30.tDg3 xb2 3 1 .tDxfS .ig6 32.tDe7t Adams decides to secure the draw by
i>f7 33.tDxdS .ie4 34.tDf4 White retains an swapping down to an opposite bishop ending
extra pawn, although his king is cut off so I with equal material. If Black wished to prolong
204 Attacking the Spanish

the game he could have tried 3 1 .. .c5!? although 25.h4

the position is still just equal. This may not be completely forced, but it
is certainly understandable that most White
32.i.f4 Axe5 33 ..he5 e6 34.M 'it>f7 players have preferred to halt the advance of
35.a7t e7 36.xe7t 'it>xe7 37..bg7 the enemy pawn.
:z-:z
25 ... c5!
A222) 20.t[}fl Once again we see Black turn his attention
towards the centre.
8
In Gashimov - Shirov, Poikovsky 2008, he
7 failed to equalise with 25 . . . g6 26.b3 b4?!
6 (26 ... llJf6 is better, but I would still prefer the
main line continuation) 27.c4 llJc3 28 . .ixc3
5 bxc3
4
3
2
1

a b c d e f g h
This has been a higher scorer for White than
20.f1 , but Black should still be fine.
Once again there are two decent moves
to consider: A2221) 20 .. JUe8 and A2222) a b c d e f g h
20 ...i.5.
29.c5!? (playing against the bishop, although
A2221) 20.. JUe8 21..idl Axdl 22.Wfxdl 29Jk 1 also looks good) 29 . . . c2 30.d2 .!c7
WffS 23 ..id2 3 1 Jk 1 f3 32J!xc2 and White eventually
White continues his policy of simplification. converted his advantage.
The alternative of23J!a6, A. Sokolov - Hellers,
Reykjavik 1 990, should probably be met by 26.b3
23 . . . .!f8N, with the possible continuation In Elburg - Simmelink, Netherlands 1 999,
24 . .!d2 !!xe 1 25 . .ixe l . At this point Black can Black became very active after 26. @g2 cxd4
choose between the straightforward 25 . . . g6 27.xd4 .!c5 28.d 1 e5 .
and the more subtle 25 . . . h6!?, intending
26J!xc6 e4 27 . .!d2 llJe3 with approximate The alternative 26.a6 .!f8 is also satisfactory
equality. for the second player.

23 .. Jxel 24..b:el h5! 26 ... cxd4

This creates a square for the king while also I will keep this as the main line as it has
threatening an awkward advance to h4 and been used in two games. However, there is a
perhaps h3. tempting alternative in 26 . . . b4!?:
 Chapter 7 Marshall: 1 2.d4
- 205

obvious 27 .. Je4?! 28.VBd3 when Black's pieces

are slightly unstable. Play continues 28 . . . j,c5
and now:

a) In Chandler - Nunn, Hastings 1 987, White

attempted to take immediate advantage of
the pinned rook with 29Ja8t @h7 30.:ge8.
However, it seems that he overlooked the
tactical resource: 30 . . . xc3! when the position
was just equal.
a b c d e f g h

White has a choice: b) In R. Ward - Garcia Rojas, e-mail 2000,

he improved with 29.@g2!, after which I do
a) 27.c4?! lDc3 28.,ixc3 bxc3 29.d5 c2 30.VBd2 not see any route to a satisfactory position for
VBf3! Black combines play on both flanks. In Black. The game continued 29 . . . :ge5 30.:ga8t
view of the threatened rook invasion on e2, @h7 3 1 .VBxf5t :gxf5 32.f4 :gf6 33.:gc8 j,d6
White has only one choice: 3 1 .lDh2 VBxb3 34.j,f2 and White eventually won.
32Jkl VBxc4, when Black stands better.
The point of my improvement is to improve
b) 27.cxb4 looks better, when 27 . . . cxb4 is the rook while safeguarding the bishop on
satisfactory for Black, whose active pieces make d6. In general I think that Black should have
up for the missing pawn. enough activity to compensate the missing
pawn. A possible continuation is:
27.YlYxd4
28.c4
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
1
a b c d e f g h
a b c d e f g h
We have reached an important position, in
which I believe Black can improve on earlier Obviously White is not obliged to rush
games. with this, but it is perhaps his most forcing
continuation so we should consider it.
27 Je6!?N
.

Hitherto Black has preferred the more 28 bxc4 29.bxc4 lll f6

206 Attacking the Spanish

The game is likely to end in a draw. Black's 25 . . . g5

pieces are still very active. It will be difficult for This leads to a draw, which is fine from a
"White to make his pawn advantage count for theoretical perspective. Black can also
anything, especially considering the reduced maintain the tension with 25 . . Jfe8!?, when
material. I think he has enough for a pawn.
26.fxg5
A2222) 20 .if5 26.g2 xg2t 27.i>xg2 fe8 28.i.d2 xe 1
29.xe 1 xe 1 30.,ixe 1 gxf4 was equal in
Karpov - Short, Tilburg 1 99 1 .
8
26 . . . f4 27.,ixf4 ,ixf4 28.gxf4 tLlxf4 29.tLlg3
7 29.xe6?? xe6 leaves "White defenceless.
6 After the text, the game Ljubojevic - Nikolic,
Belgrad 1 99 1 , was soon agreed drawn after:
5 29 . . . g4 30.xe6 tLlh3t 3 1 .i>g2 tLlf4t 32.i>gl
4 tLlh3t 33.i>g2 -
3 21 J1fe8 22.,lc2

2 "White achieved nothing in Ivanchuk -

1
a b c d
21 .%Yd2
The alternative is:
2 1 .d 1
2 1 .e2 i.g4 22.d2 is the same.
2 1 . . .i.g4 22.d2
22.d3 i.f5 repeats.
22 . . . h3 23.i.d 1
Black was threatening . . . i.f3, and 23.f4
would be very weakening.
I. Sokolov, Biel 1 989, after 22.,ixd5 cxd5
23.i.f4 xe 1 24.xe 1 xe 1 25 .xe 1 i.e4
e f g h 26.tLld2 hf4 27.tLlxe4 dxe4 28.gxf4 g4t
29.i>f1 Y2-Y2
22.i.d 1 g6 23.i.f3 keeps some more tension
in the position, although 23 . . . h5, Fleetwood
- Olofsson, corr. 1 999, maintains Black's
compensation. The game was eventually
drawn.
22 Ah3

23 . . . ,ixd 1 24Jaxd 1 Threatening . . . f3 mating.

No-one seems to have tried 24.xd 1 , when
24 . . . f5 looks like a good reply. 23.%Ydl .t.g4 24.%Yd3 %Yh3
24 . . . f5! 25.f4
8
8 7
7
6
6
5
5

4 4
3 3
2 2
1
1
a b c d e f g h
a b c d e f g h
 Chapter 7 Marshall: 1 2.d4
- 207

Once again Black threatens to construct a
mating net. We will follow the game: Svidler
- Jakovenko, Foros 2008.
25.Ad2
Another possibility is 25 .i.d l i.fS 26.'@Fd2
h6! ? Black does not need to rush to prove
anything, so he takes time out to safeguard
his king's position. White is defending his
kingside, but it is not easy for him to utilise
his extra pawn.
to both 20.'@Ffl and 20.itJfl will find favour
with the reader. Should a problem arise in one
of these variations, you will have a dependable
backup on which you can rely.
Having devoted a considerable amount of
space to the traditional 1 5 .i.e3 variation,
we will now consider quite a respectable
sideline.
B) 15Jfe4

25 Je2 26Jxe2 he2 27Jel

.

8
After 27.'1WfS Black can repeat the position
with 27 . . .i.g4 28.'@Fd3 i.e2. 7
6
27...hd3 28Jxe8t Am 29.hd3 'iNd7
5

8 4

7 3

6 2

5 1

4 a b c d e f g h

3 This important variation has been tested in

approximately a third of the number of games
2 as 1 5 .i.e3. White intends to drive the enemy
1 queen away with h4.
a b c d e f g h 15 ... g5
30Jel This is the most common response. Black
1/2-1/2 utilises a tactical feature ( 1 6.,ixg5 ?? '@Ff5
At the time of writing the above game seems wins material) to cover the h4-square. The
to be the last word in the 20 . . . i.fS variation. drawback is the permanent compromising
of his kingside. On the other hand, he may
I have no doubt that the 1 5 .i.e3 variation will be able to exploit the exposed position of
continue to be debated at a high level for many the white rook in order to gain time for the
years. At this stage it seems to me that Black forthcoming attack. We now consider five
is holding his own in all the major lines we moves: Bl) 16.tLld2?!, B2) 16.'iNa, B3)
have covered, although I have no doubt that 16.'iNe2, B4) 16.'iNel, and B5) 16.'iNfl .
proponents of the white side will continue to The first is weak, but the other four are all
unearth new ways to fight for an advantage. I playable with line B5) being perhaps the most
hope that my decision to cover two responses challenging.
208 Attackin g the Spanish

B1) 16.d2?!
This turns out to be too serious a mistake.

16 ... 5
White faces a potent attack.

17Je3
The rook can also move to a different
square:

In AI Sayed - Kaplan, Bad Wiessee 2006,

White lost in just three more moves: 1 7Je l
f4 1 8.e2 J.g4 (already White is without a b c d e f g h
a satisfactory defence) 1 9.d3 (White also 20Je2?
loses after: 1 9.f1 h5 20.d3 [or 20.g2 It was necessary to play 20.e l although
f3 2 1 .f1 J.h3 22.d3 h:g3-+1 20 . . . fxg3 Black still has an advantage after the simple
2 1 .fxg3 f2 22.lDf1 af8-+) 1 9 . . . fxg3 20.fxg3 20 . . . J.d7.
f2! 0- 1
20 f4
..

Carlsson - Deltin, Sweden 2000, continued From here the game: Stone - Knol, e-mail
1 7 .e2 f4 1 8. f1 h5 when White was under 200 1 , ended beautifully after some fine
pressure. He then blundered with 1 9.1De4?? attacking play from Black.
J.h3 and was forced to give up his queen in
view of the ... f3 mating net. 21.Yfe1 i.g4 22.f3 fxg3! 23.fxg4 Yfh3! 24.h4
ae8!
Perhaps it was already time to sacrifice the
exchange with: 1 7.e5! ? h:e5 1 8.dxe5
when White can perhaps claim some
compensation, and should at least avoid being
mated.

17 ... h8!
There is no need to rush with 1 7 . . . f4?!
1 8.gxf4 h6, Giaccio - Vescovi, Sao Paulo
2006, 1 9.1De4! h:f4 20.g3! ,hg3 2 1 .hxg3
when White may even be better.

1 8.,bd5
Perhaps White could consider moving a b c d e f g h
his rook along the third rank, although this
appears rather artificial and Black would retain 0-1
excellent compensation regardless. The finish might have been 25 .J.b2 xe2
26.xe2 f2 27.e8t i>g7 28.d7t f7
18 ... ad5 19.Yfn Yfh6 when the checks run out.
 Chapter 7 - Marshall: 1 2.d4
B2) 16.Ba
White prepares to drop his queen back to
g2. The move is not a bad one, but nor is it a
serious try for an advantage.
16 ...Af5
a b c d e f g h
17.Ac2
Sacrificing the exchange. A few other moves
have been tried.
1 7 .e I ? This is far too slow. After the automatic
1 7 . . . ae8 Black's initiative is already nearing
decisive proportions. White has to develop his
pieces.
Lima - Grischuk, Khanty-Mansiysk 2007,
continued:
1 7.ttJd2 ae8!
Not 17 . . . ,ixe4? 1 8.ttJxe4 and White is clearly
better.
1 8.,ixdS cxdS 1 9.e3 xe3 20.'lWxe3
White is not helped by 20.fxe3?! e8. Here
the most important factor is the weakness
of White's kingside. Black's light-squared
bishop is the best-performing piece on
the board. It protects his weaknesses and
may become useful also in the attack, e.g.
2 1 .a4 bxa4 22.xa4 i.d3 with a strong
initiative.
20 . . . f6 2 1 .ttJfl .tg6 22.f3 e8
209
Black has achieved a comfortable position.
23.V9f2 .td3 24.i.e3
24.ttJe3 f5 also gives Black enough
counterplay.
24 . . . hfl
It is not necessary to force the draw, and
Black could also have considered 24 . . . hS!?
2S.xfl ,ixg3 26.hxg3
White decides to allow a perpetual. Instead
26.V9xg3 V9xg3t 27.hxg3 xe3 would have
led to an equal rook ending.
26 . . . xe3 27.V9xe3 Y2-Y2 .
Finally, another reasonable option is:
1 7.,ixdS cxdS 1 8.e3 i.e4
1 8 . . . ad8! ? is also possible, with a balanced
position.
1 9.xe4 dxe4 20.V9f6 V9g4!
a b c d e f g h
Indirectly defending the bishop, as White
can hardly allow the black queen to penetrate
to d l and c l .
2 1 .ttJd2
Harmless is 2 1 .V9xgst V9xgS 22.,ixgS f5.
Material is approximately equal, but Black's
pieces are more actively placed. I think it is
White who should be fighting for a draw.
2 1 . . .ae8
Now a couple of moves deserve attention:
a) Timman - Ivanchuk, Linares 1 99 1 , was
soon agreed drawn after:
22.ttJfl .te7 23.V9xa6 fS 24.V9xbS f4:
210 Attacking the Spanish

25.lLlf6t @h8

a b c d e f g h

Ih-Ih
a b c d e f g h
The position is extremely double-edged, and
the players were evidently not in a mood A fascinating position has arisen. Black can
to take risks. Personally I would tend to always force a draw without difficulty, but I
favour Black as the side with the initiative. A think he can achieve more if he can combine
plausible continuation might be: the most accurate moves with some strong
25 . .id2 nerves!
After 25 .Wld5t @g7 26.Wlxe4? .id8 the 26 . .if4
opening of the e-file is too dangerous for Hopeless is 26.lLlxe8? Wlfl t (26 . . . e l =lLlt?
White. 27.@g l =) 27.@f3 e l =lLlt! when Black wins,
25 . . . e3 26.he3 fxe3 27.lLlxe3 Wlf3 28.Wld5t e.g. 28.@g4 Wle2t 29.@f5 Wld3t 30.@xg5
Wlxd5 29.lLlxd5 g4 g8t and White can resign.
White has five pawns for a rook, but they 26 . .id2 is a better attempt, but still not quite
are not far advanced. Furthermore, White will sufficient after 26 . . . Wlxal 27.lLlxe8 Wlfl t
have to worry about the safety of his king in 28.@f3 e l =lLlt! 29.he l Wlh l t 30.@g4
the event of a rook invasion on e2. h5t! etc.
26 . . . Wlxal 27 . .ie5 Wlfl t 28.@f3 e l =lLlt!
b) A critical alternative was: This underpromotion is an important theme
22.Wlxd6 e3 23.lLle4! in several of the above variations.
This looks best, although my analysis 29.@g4
indicates that Black can stay on top if he
demonstrates study-like precision. Instead,
after the feeble 23.fxe3 e3 24.lLlfl e l-+
White's pieces are unable to develop, and he
is helpless against the development of Black's
remaining rook.
23 . . . Wld l t
23 . . . exf2t only leads to a draw after: 24.lLlxf2
e l t 25.@g2 Wle2 26.Wlf6 Wlfl t 27.@f3
Wle2t 28.@g2=
24. @g2 e2
a b c d e f g h
24 . . . xe4?! 25.Wlf6 g4 26.fxe3 is not at all
dangerous for White. 29 . . . h5t!!
 Chapter 7 Marshall: 1 2.d4
- 21 1

It seems incredible that this can work, yet
nonetheless it seems to.
30.lLlxh5t
30.rxh5 '!Wh3t 3 1 .@xg5 lLlf3t 32.@f4
lLlxe5 33.dxe5 '!Wxh2 is winning for Black.
30 . . . @h7 3 1 .'!Wf6
After 3 1 .lLlf6t @g6 White is amazingly
helpless, e.g. 32.h4 (or 32.lLlxe8t f6
threatening . . . '!We2t mating) 32 . . . '!We2t
33.@h3 e6 34.h5t '!Wxh5t!? 35.lLlxh5
xd6 36.hd6 d8. Black will emerge with
an extra rook after winning back one of
White's minor pieces.
3 1 . . .'!We2t 32.f3
32.@f5 xe5t wins easily.
32 . . . '!Wxf3t 33.'!Wxf3 lLlxf3 34.@xf3 f6
Black should win the ending without too
many problems.
17....he4 18 .he4 Yfe6 19.i.xgS
exchange, but while the quantity was fine, the
quality was poor. Play continued 24.@h l e7
25 .gl t @h8 26.lLlfl '!We4! 27.@g2 '!Wxf3t
28.@xf3 e l ! 29.h l e4, after which it
did not take Black long to wrap up the full
point.
20 ... 4
We will now follow the game: Hallengren -
Lakatos, e-mail 200 1 .
21 .Yfe4 Yfd7!
Black's queen should avoid being exchanged,
as her opposite number will soon come under
fire after . . . ae8.
22.ti)d2 gae8 23.Yfbl
White has no choice but to make this
awkward retreat. 23.'!Wg2? f3 24.'!Wfl would be
even worse after: 24 . . . '!Wg4! 25 ..th4 lLlf4-+

8
23 Yfg7! 24.ti)e4 J.c7 2S.i.b4 f3
..

7
8
6
7
5
6
4
5
3
4
2
3
1
2
a b c d e f g h
1
White has two pawns for the exchange, plus
a b c d e f g h
a pair of powerful bishops. If he is allowed to
coordinate his pieces then he will stand clearly Now the queen on h I is hopelessly
better, so Black must react energetically. misplaced. White faces severe difficulties, and
soon succumbed after:
19 ... f5! 20.J.d3
In Topalov - Adams, Sarajevo 2000, White 26. c.t>fl b4!
got nowhere with 20.hd5 cxd5 2 1 .lLld2 Opening a second front is the easiest way to
f4 22.hf4 h4 23.gxf4 a7. White had take advantage of White's wayward pieces on
temporarily gained three pawns for the the h-file.
212 Attacking the Spanish

27J d l bxc3 28.tLhc3 tLlxc3 29.bxc3 .la5 22 . . . g4! 0-1

0--1
White's position is hopeless, e.g. 30Jkl
,ixc3! (there are other ways to win, but the
text is simplest) 3 1 .xc3 "lWxd4 32 . .ic4t @h8
33.cl e2! (33 . . ."lWd2 is also good enough) .
B3) 16.YlYe2 f5
1 7.e5 ,ixe5 1 8 ."lWxe5 "lWh5 1 9.,ixd5t cxd5
20."lWxd5t @h8 gave Black good chances in
Lindberg - Nyback, Stockholm 2005. For
instance, 2 1 ."lWxa8?? "lWd l t 22.@g2 f4! wins for
Black, and 2 1 . ttJd2 a7 also gives the second
player an active position.

1 7.e8, M. Hansen - Virag, Mureck 2007,

8
should probably be met by 1 7 .. .f4!N 1 8.xfBt
7 @xfB 1 9.a4 .if5 20 . .id2 (20.axb5?? e8-+)
6 20 . . . e8 2 1 ."lWfl "lWh5 with excellent play for
a pawn.
5
4 17 cxd5 18J e6 f4!
..

As the saying goes: sometimes the best form

3 of defence is to attack. Black must certainly
2 avoid: 1 8 . . . ,ixe6?? 1 9."lWxe6t @h8 20.,ixg5
when White is simply winning. After the main
1
move, the game is headed for a forced draw.
a b c d e f g h We will follow the game Ponomariov - Anand
There is no reason for Black to refrain from Linares 2002 (as well as several others) .
this natural attacking move.
19.E:xd6 i.g4 20.YlYfl YlYxfl t 21 .mxfl gae8
17..bd5t 22.Ad2 i.h3t 23. mgl fxg3 24.hxg3 ge2
White has also tried moving the rook without
the preliminary exchange: 8
7
In Smirin - Grischuk, Panormo 200 1 , he was
quickly crushed after 1 7 .e6?? ,ixe6 1 8."lWxe6t 6
@h8 1 9."lWxd6 ( 1 9.,ixg5 ae8-+) 1 9 . . . ae8
5
20 ..id2 f4 2 1 .,ixd5 cxd5 22.f3:
4
3
2
1
a b c d e f g h
White can only defend his kingside at the
cost of a perpetual check.

25.Ae3
 Chapter 7 Marshall: 1 2.d4
- 213

2S.f4?? allows 2 S . . J3g2t 26.i>h l e8 2 1 . i>g2 is also unpleasant for Black) 1 9 ..bd 1
intending . . . ee2 with a mating net. lLlxe4 20.lLlxe4 .ie7 2 1 .,LgS ,LgS 22.lLlxgS;\;,
Motylev - Tkachiev, Kazan 200S. White had
2S .h6 does not appear to have been tested. a pleasant endgame advantage, which he went
Play continues 2S . . . fxf2 26.xh3 g2t on to convert successfully.
27.i>h l gf2 28.lLla3 xd2. White can hardly
hope to win this position, as his rooks are 17.d2
horribly passive and his extra knight is trapped 1 7.,LdS cxdS 1 8.e3, Gutierrez - Moline,
on the edge of the board. corr. 1 999, is playable for White, but Black's
domination over the light squares will provide
25 Jxe3 26.fxe3 lUI t 27.<iI?h2 g4!
. long-lasting compensation.
Securing the draw.
17 ... h6!
28Jxd5 Black decides to strengthen his kingside
V2-1/2 before doing anything else. This approach
It seems that 1 6. e2 is only likely to represent was favoured by Etienne Bacrot, so it deserves
a problem to players who are not content to serious attention although I must add that it is
achieve a draw with the black pieces. not the only playable continuation. Here are a
few alternatives.
B4) 16.%Yel
17 . . . ,Le4?! 1 8.lLlxe4 only helps White to
develop.
8
7 1 7 . . . lLlf6
6 This is more sensible.
1 8.f3 cS 1 9.e3 h6
5 Now a couple of moves deserve our
4 attention:
3 a) 20.dxcS i.xg3 2 1 .hxg3 xg3t 22.i>fl
2 h3t 23.i>e2
23.i>f2 ad8 100ks dangerous for White.
1
23 . . . h l !
a b c d e f g h
The point of the text is that, compared with 8

the 1 6.e2 variation, 1 6 . . . f5 ? no longer works, 7

as after 1 7.,LdSt cxdS 1 8.e6! f4 1 9.xd6, 6
the move . . . .ig4 will not attack the queen and 5
so White is just winning.
4

3
16 -*.5

The experimental 1 6 . . . lLlf6 1 7.lLld2 hS 2

turned out unfavourably after 1 8.d l xd l t 1
( 1 8 . . . .ig4 1 9.xg4 lLlxg4 20.lLlfl ae8 a b c d e f g h
214 Attacking the Spanish

White is a full bishop up, but his king is f-pawn, so it makes a lot of sense to remove
in danger and he will have a hard time the king from the watchful eye of the bishop
completing development. It would take a on b3.
great deal of space to analyse the position
exhaustively, so I will instead offer the 19.a4 Axe4!
following sample variation which I believe Now is the right time to take the rook.
represents a good standard of play from
both sides: 20.fxe4 gaeS 21 .axb5 axb5
24Je5 aeB! 25.xfS xe3t 26. c;t>xe3 dB
27.rtfe2 c;t>g7 2B.tLle4 tLlxe4 29.xf7t c;t>hB
8
30.fxe4 YMxe4t 3 1 .i.e3 eB 32.8 g4 33.g3
h5 34.c;t>f2 h4 3 5 .i.d4t c;t>h7 36.e3 YMf4t 7
37.c;t>e2 YM8t 3B.c;t>d3 xe3t 6
3B . . . YMfSt 39.c;t>d2 YMf2t 40.c;t>d3 repeats
the position. 5
39.he3 h3 40.c;t>d4 g3 4 1 .i.d5 YMg4t 4
42.i.e4 t c;t>g7
3
Black has some advantage, although it will
not be easy to promote the kingside pawns. 2
1
b) In Shchekachev - Gustafsson, Austria
2005, White preferred the safer continuation: a b c d e f g h
20.YMf2 cxd4 2 1 .cxd4 adB White still White's idea has backfired, and he must even
has to develop his queens ide, while Black struggle to equalise.
is nearing the stage where he can consider
capturing the rook on e4. Therefore the 22.f1
game continued: 22.e2 i.d3 23.e3 YMfS 22.YMf2 is an alternative, when Black may
24.tLle4 White returns the pawn in order to try:
free his position. 24 . . . tLlxe4 25.fxe4 he4
26.YMxfS hfS 27.i.d2 c;t>g7 -Y2 Neither a) 22 . . . tLlf6 might lead to a nice perpetual
side can claim any real advantage in this after 23.YMf3 tLlg4 24.tLlf1 fS 25 .e5 f4 26.YMg2
ending. YMxg2 t 27. c;t>xg2

After the main line of 1 7 . . . h6!?, we will follow

the game Anand - Bacrot, Sofia 2006.

IS.8
This is a consistent move. Anand is confident
that he will obtain fine compensation in the
event that Black takes on e4.

IS 'tt> g7!

A very useful move. We will see that in certain

positions Black may profit by advancing his a b c d e f g h
 Chapter 7 Marshall: 1 2.d4