Starting Out The Trompowsky Attack - Palliser PDF

starting out:
the Trompowsky
attack RICHARD PALLISER
EVERYMAN CHESS
Gloucester Publishers pic www.everymanchess.com
First published in 2009 by Gloucester Publishers pic (formerly Everyman
Publishers pic), Northburgh House, 10 Northbu rgh Street, London ECl V OAT
Copyright 2009 Richard Palliser
The right of Richard Palliser to be identified as the author of this work has been as
serted in accordance with the Copyrights, Designs and Patents Act 1988.
All rights reserved. No part of this publication may be reproduced, stored in a re

trieval system or transmitted in any form or by a ny means, electronic, electrostatic,
magnetic tape, photocopying, recording or otherwise, without prior permission of
the publisher.
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.
ISBN: 9781 85744 562 6
Distributed in North America by The Globe Pequot Press, P.O Box 480,
246 Goose Lane, Guilford, CT 06437-0480.
All other sales enquiries should be directed to Everyman Chess, Northburgh House,
10 Northburgh Street, London EC1 V OAT
tel: 020 7253 7887; fax: 020 7490 3708
email: info@everymanchess.com: website: www.everymanchess.com
Everyman is the registered trade mark of Random House Inc. and is used in this
work under licence from Random House Inc.
For the chessplayers at the Minster Inn
EVERYMAN CHESS SERIES

Chief Advisor: Byron Jacobs
Commissioning editor: John Emms
Assistant editor: Richard Palliser
Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteverde.
Printed and bound in the US by Versa Press.
Contents
Bibliography 4
Introduction 5
1 The Classical 2 ds
...
9
2 2 g6 and Minor Lines

... 39
3 The Positional Choice: 2 e6 ... 54
4 The Uncompromising 2 cs ... 100
5 The Popular 2 ... Ne4 142
6 The Modern Preference: 2 ... Ne4 3 Bf4 163
7 The Main Line: 2 ... Ne4 3 Bf4 cs 201
Index of Variations 260
Index of Complete Games 268

Bibliography
Chess Openings for Black, Explained, Lev Alburt, Roman Dzindzichashvili & Eugene
Perelshteyn (CIRC 2005)
Dealing with d4 Deviations, John Cox (Everyman 2005)
El Ataque Trompowsky, Jesus De Ia Villa (Evajedrez 2001)
Fighting the Anti-King's Indians, Yelena Dembo (Everyman 2008)
Secrets of the Trompovsky, Julian Hodgson (Hodgson Enterprises 1997)
The Soviet Chess Conveyor, Mikhail Shereshevsky (Semko 1994)
The Trompowsky, Joe Gallagher (The Chess Press 1998)
The Trompowsky (second edition), Nigel Davies (Everyman Chess 2005)
The Trompowsky: The Easy Way, Andrew Martin (ChessBase DVD 2006)
The Veresov, Nigel Davies (Everyman 2003)
Trends in the Torre and Trompowsky, Julian Hodgson (Trends Publications 1995)
Winning with the Trompowsky, Peter Wells (Batsford 2003)
Other Sources
Good use was made of Chess Informant, Mega Database 2008 (ChessBase), New in
Chess Yearbook andTim Harding's UltraCorr 2 CD. I can also heartily recommend
Eric Prie's Trompowsky coverage in his 'd-Pawn Specials' section of the ChessPub
lishing website.
Last, but by no means least, I am most grateful to John Emms and Peter Wells for
their help with this project.
4
I nt ro d u ct i o n
Introduction
1 d4 Nf6 2 Bg5 (Diagram 1)
Diagram 1 (B)
The Trompowsky
Few openings allow White t o stamp his authority on the game a s early a s the
Trompowsky. At a stroke Black discovers that his favourite King's Indian, Nimzo
lndian, Modem Benoni, Griinfeld or even Benko Gambit has been side-stepped
and without his obtaining easy equality in the process. The Trompowsky is no
longer quite the shock weapon it once was, but even nowadays it carries a fair
amount of surprise and practical sting; opponents might have decided on a re
sponse to 2 BgS, but generally they have studied that line much less than the more
critical parts of their repertoire after 2 c4.
Whether one is looking for an occasional weapon for surprise use or a full-time
opening to be employed against 1 d4 N f6, the Trompowsky fits the bill. It is not
that hard to learn and White can expect to be somewhat more familiar with the
resulting complex lines and often unusual middlegames than his opponent. I hesi
tate to describe the attribu tes of a typical Trompowsky player, though, as I believe
5
Sta rt i n g Out: The Trom powsky Attack
that the opening appeals to a wide range of tastes. That said, in connection with
the Trompowsky one tends to think first and foremost of that most creative of
English Grandmasters, Julian Hodgson. Even in 2009 a number of unbalanced and
fascinating lines remain far from worked out, and so the creative and original
player will still find 2 BgS an enjoyable and successful opening to adopt, but so
too will those looking for something a little more solid and positional.
Throughout this work we will see that White has at least a couple of fully viable
options against Black's main defences to 2 BgS. Which path to follow is often a
matter of taste, and nowadays one can broadly see two main Trompowsky
schools. One contains the aforementioned creative types; the other the more posi
tional, 'systems-orientated' player. These are players who often have some experi
ence of openings like the Torre and London, and like to employ simple but by no
means innocuous set-ups. Many with a 1 e4 background have taken up the Trom
powsky too, and even they will find some structures in which they have experi
ence. After all we should not forget that an opening which arises after just two
moves is likely to give rise to a wide range of middlegames and structures.
I don't want to over-theorize on these two schools, though, as the lines between
them can become quite blurred at times and quite a few Trompowsky players have
a foot in both camps. Indeed, many players will choose to include a mixture of fairly
positional and much sharper lines in their Trompowsky repertoire. Moreover, even
those who revel in sharp positions sometimes have to play more quietly and posi
tionally, especially in the case of 2 ...d5 and 2 ...g6. Indeed, we should never lose sight
of the fact that by playing 2 BgS White prepares to double Black's pawns, and after
both 2 ... d5 and 2 ...g6 he is probably best advised to carry out that threat.
The Structure of this Book

Opening books often consider the more critical and theoretical lines first, moving
on to the less forcing and theoretical lines towards the end. That will not, how
ever, be our approach here. Rather I have begun with a discussion of two fairly
simple defences, 2 .. d5 and 2 .. g6. These are especially popular at club level and
. .
there is no point taking up the Trompowsky if one isn't happy handling the re
sulting middlegames. The doubled f-pawn structure (Diagram 2) might give rise
to fewer complications than some would like, but even here White has a choice of
set-ups, including some quite pleasant ones in my view.
Our journey then continues with 2 .. e6, which is especially popular with Nimzo
.
players. Black is happy to put up with a slightly cramped position in return for
avoiding doubled f-pawns, although he usually gains the bishop-pair in any case
after 3 e4 h6 4 Bxf6 Qxf6. Of course, White obtains a strong centre that way and
there's always the systems-based alternative, 3 e3, for those who prefer to keep
their bishop.
6
I ntrod u ct i o n
The most critical sections o f this book deal with 2 ... c5 and Black's most popular
response, 2 ... Ne4. Then the play can become quite forcing and a certain body of
theory has built up. However, in both cases White has a choice between some fas
cinating approaches, can easily gain an early initiative and still needs much less
theoretical knowledge than to play 2 c4! Moreover, in keeping with the Starting
Out format, here too I have included a number of tips, notes and warnings to help
the reader understand their chosen lines as well as possible.
Diagram 2 Diagram 3 (B)

An important structure A fascinating gambit
In the case of 2 ... Ne4 3 Bf4 c5, which one can consider to be the main line of the
Trompowsky, I haven't shirked on supplying theoretical detail, largely because
White has a wide range of options. These include a fascinating gambit which was
unknown until introduced by Konstantin Chemyshov in 2005, namely 4 f3 Qa5+ 5
c3 Nf6 6 d5 Qb6 7 e4!? Qxb2 8 Nd2 Qxc3 9 Bc7! (Diagram 3), trapping the black
queen mid-board. In other sharp variations too I have aimed to supply up-to-date
theoretical coverage, which should help both those new to the opening and even
long-term Trompowsky practitioners. Even after 2 Bg5 theory continues to evolve;
there are now more than 40,000 Trompowsky games in my various databases!
Balancing plenty of explanation with supplying enough theoretical coverage for
even the stronger club player to adopt 2 Bg5 doesn't come without its drawbacks.
Chiefly I have had to include less new analysis than I would have liked, and much
less than one finds in two of my all-time favourite opening works, De Ia Villa and
Wells's respective works on the Trompowsky. Still, this work is hardly devoid of
new ideas and suggestions, at least in the Trompowsky proper. Please be aware
that there's no coverage whatsoever of 1 d4 d5 2 Bg5, the so-called Pseudo Trom
powsky, largely because I consider it to be a much less challenging and fun rela
tive: Black is fine in the complications after 2 .. .f6, and the solid and popular 2 ...h6
7
Sta rti n g Out: T h e Trom powsky Attack
3 Bh4 c 6 i s a pretty tough nut to crack.

Garry Kasparov has enjoyed success with the Trompowsky in his many simulta
neous exhibitions, while the creative young Azeri, Shakhriyar Mamedyarov, has
made good use of the opening on occasion throughout his career, as has the lead
ing English Grandmaster, Michael Adams. More regularly, while Hodgson is
sadly no longer playing actively, the 2600-rated Serb Grandmaster, Igor Miladino
vic, continues to play his favourite Trompowsky against all-comers, as does the
former Women's World Champion, Antoaneta Stefanova. We will see plenty of
inspiring games from these players and many more leading grandmasters
throughout this work. I hope that their efforts will inspire both the new and exist
ing Trompowsky player alike, just as they have helped to remind me why the
Trompowsky has long been one of my favourite openings.
Happy Tromping!
Richard Palliser,
York,
June 2009
8
C h a pter One
Th e Cla s s i c al 2 ds
m I nt rod uction
m The Sol i d 3 exf6
m The Dyn a m ic 3 . gxf6
m Wh ite Dec l i n e s to Exc h a n ge

Sta rti ng Out: The Trom powsky Attack
I ntrod uction
1 d4 Nf6 2 Bgs dS (Diagram 1)
Diagram 1 (W) Diagram 2 {W)

A simple and solid defence White has the better structure
A solid approach which has been played in 20% of all the Trompowsky games in
my database. 2 ... d5 is regularly employed by several solid Russian Grandmasters
and is an especially popular choice at club level.
NOTE: Even in 2009 there are some without a prepared line against
the Trompowsky, and such careless players often fall back on 2 ...ds
followed by seemingly natural development.
It must be said that if White can't reach a position he's happy with after 2 ... d5,
there's not much point playing the Trompowsky! However, the resulting different
types of position offer something for everyone, whether White begins by carrying
out his positional threat on f6 or aims for a kind of improved Torre with 3 e3. The
latter is fairly popular with the more positional school of thought, but there is ab
solutely no reason for the more solid Trompowsky player to avoid exchanging on
f6. Likewise those of ambitious bent might be drawn towards the less-theoretical 3
e3 - there really is something for everyone here! That also applies to Black: 2 ... d5
is only a solid choice if he recaptures on f6 with his e-pawn; one could never ac
cuse 3 . . . gxf6 of being a little dull.
The Solid 3 ...exf6

1 d4 Nf6 2 Bgs ds 3 Bxf6 exf6 (Diagram 2)
10
The C l a ss i c a l 2 ... d s
The obvious recapture, opening a standard path for the bishop on f8. Already we
have a classic Trompowsky debate before us: will Black's bishop-pair prove useful
or ineffective? As we will see, it is hard for Black to make his bishops felt for some
time and patience really must be his watchword in this line. Moreover, White's
knights often find good roles, putting pressure on Black's centre from such
squares as c3, c4 (the c2-c4 advance is White's main pawn break) and f4. It is far
from clear that White has any objective advantage here, but his game has always
struck me as being the easier to play.
4 e3 Be6
Black takes steps to prevent White rushing through with c4, generating some early
pressure. However, 4 ... Be6 is by no means universally played:
a) 4 ... Bf5 5 Bd3 (White is happy to trade bishops and speed up his development; 5
Ne2 c6 6 Nd2 Qb6 7 Ng3 followed by b3 and c4 is a decent alternative, but Black
should allow such an expansion, whereas 7 ... Bg6 8 Rb1 Bb4 9 a3 Bxd2+?! 10 Qxd2
0-0 1 1 f4! Be4 12 Nxe4 dxe4 13 c4 Nd7 14 Be2 gave White a pleasant advantage
with the better prospects on both flanks in R.Palliser-G.Bak, Leeds 2007) 5 ... Bg6
(less compliant than 5 ... Bxd3 6 Qxd3 c6 when 7 Nf3 Bd6 8 Nbd2 Na6 9 0-0 0-0 1 0
c4 dxc4 1 1 Nxc4 Bc7 1 2 Rfd1 Re8 1 3 a 3 Qd5 14 b 4 Rad8 15 Rab1 f5 16 Ncd2! Bb8 1 7
Rdcl was rather depressing for Black despite his apparent solidity i n J.Hodgson
E.Teodoro, Winnipeg 1 994; White's control and queenside prospects give him
clearly the upper hand) 6 Ne2 Bd6 7 0-0 Nc6?! (Black has the ambitious ...Nb4 in
mind, but he must have underestimated White's next) 8 Bb5! 0-0 9 Bxc6 bxc6 1 0
Nbc3 Rb8 1 1 b 3 (Diagram 3) gave the pure knight-pair the upper hand in
M.Shereshevsky-E.Barkovsky, Minsk 1981; White will expand with Na4 and c4,
and can always hinder the bishops with Nf4 if necessary.
Diagram 3 (B) Diagram 4 (B)

White is in control White attacks the isolated pawn
11
TIP: Even when White begins by coordinating his minor pieces, the
c4-break should remain at the front of his mind. Pawn breaks are the
key to most middlegames and this variation is most certainly no ex-
ception.
b) The pseudo-active 4 ... c5 appears a little at odds with Black's last, but might be
tried by those with little Trompowsky experience. Here White has a pleasant
choice between 5 Nc3 Nc6 6 Nge2 Be6 7 g3, as played by Smyslov and investi
gated further via the move order 2 ... c5 3 Bxf6 exf6 4 Nc3 dS in Chapter Four, and
the even simpler 5 dxcS Bxc5 6 Nc3! Be6 7 Bb5+ Nc6 B Nge2 0-0 9 0-0 Ne5 10 Nf4!
(Diagram 4), which retained a pleasant edge in M.Gurevich-P.Wolff, Palma de
Mallorca 19B9.
c) 4 ... Bd6 has actually been a more popular choice in practice than our main line
and introduces an important dilemma: how best to deploy White's pieces. Essen
tially White has three options: a quick c4 followed by rapid and often aggressive
development; an outwardly-aggressive set-up with Bd3, Nd2, Qf3 and Ne2, which
may pack some punch if Black is careless; and a sensible set-up with g3, Bg2, Ne2
and often Nd2, preparing c4 while keeping the position under control. Thus we
have:
cl) 5 c4 dxc4 6 Bxc4 0-0 7 Nc3 is an easy-to-play approach. Game 1 will reveal why
I feel Black is probably best advised to avoid this line with 4 ... Be6.
c2) 5 Bd3 0-0 6 Qf3 c6 (the aggressive 6 ... c5! ? becomes a better bet with White's
queen already committed; Black fears an isolated d-pawn less as he will obtain
counterplay on the queenside and in the centre) 7 Nd2 ReB (theory has generally
approved of 7... Na6! B a3 Nc7 when long castling would be risky to put it mildly
and 9 Ne2 Ne6 1 0 c4 Ng5 1 1 Qh5 g6 1 2 Qh4 ReB 13 cxdS cxd5 14 Nc3 Be7 gave
Black sufficient counterchances in J.Hodgson-M.Taimanov, Yerevan 19B6) B Ne2
Qb6 9 0-0-0!? (Diagram 5) 9 ... Na6 10 c3 Be6 1 1 g4! Nc7?! (there is no time to waste;
Black has to get on with 1 l . ..c5!) 12 h4 RecB 13 gS NeB 14 Rdg1 QdB 15 Nf4 gives
White strong pressure, A.Zubarev-A.Bets, Obninsk 2007.
c3) 5 g3 c6 (or 5 ... 0-0 6 Bg2 c6 7 Ne2 and now Black should hinder White's break
with 7...Be6, transposing to the notes to his 7th move in our main line, whereas
7... Bf5?! B 0-0 Nd7 9 b3 ReB 10 c4 dxc4 1 1 bxc4 Be4 1 2 Bxe4 Rxe4 13 Nd2 ReB 14 cS
Bc7 15 Qc2 left him under unpleasant queenside pressure in R.Palliser-P.Swallow,
Leeds 2004) 6 Bg2 (if White is worried about ... Qb6 ideas, he might consider 6
Nd2) 6 ... Nd7 (6 ... Qb6 has been proposed by some commentators, but after 7 b3 I
have been unable to find a good way for Black to disrupt White's development,
and 7...0-0 B Ne2 BfS 9 0-0 aS!? 10 c4 dxc4 1 1 bxc4 Qa6 12 cS Be7 13 Nbc3 again saw
White seizing the advantage on the queenside in K.Rusev-V.Kukov, Blagoevgrad
2009) 7 Ne2 0-0 (Howell later preferred 7 .. .f5 when B b3 Nf6 9 Nd2 Be6 transposes
to our main line) 8 0-0 fS 9 b3 Nf6 10 c4 (Diagram 6) sees White's aims become
clear. Sometimes he can exchange on dS and attack a resulting isolani (if there's no
12
The C l a s s i c a l 2 . . . d s
knight on f6 to recapture on d 5 or i f White has a knight on c3), but more often he

is after a direct queenside assault with c5, b4-b5, etc.

Highly-aggressive play White's key pawn advance
NOTE: Due to the omission of ...Be6, White has managed to break

with c4 without resorting to the preparatory Nd2. This is definitely a
gain, as White would like a knight on c3 and can now keep his king's
knight flexibly placed on e2.
Note too that White shouldn't fear an exchange on c4, opening up the b-file for his
use. Just take a look at the game P.Wells-D.Howell, Halifax (rapid) 2004: 10 ... dxc4
1 1 bxc4 Qe7 12 Nbc3 ReB 13 Rb1 Ne4 14 c5! Bc7 1 5 Nxe4! (a well-judged trade to
leave Black low on counterplay and the remaining knight with good prospects in
the resulting structure) 15 .. .fxe4 16 Nc3 f5 17 Qa4 Rd8 18 Rb2 h5 (perhaps the grim
18 ... a6!? 19 Rfb1 Ra7 had to be tried) 19 Rfb1 h4 20 Rxb7! (not so hard to find, but
still beautifully logical) 20... Bxb7 21 Rxb7 Qd7 22 Bh3 (Diagram 7) and White was
in control with rich pickings in prospect on both flanks.
d) Finally, before returning to 4 ...Be6, we should note that 4 ... Be7 is rather passive
and does little to help the dark-squared bishop: for example, 5 c4 (White plays as
per Game 1; again a set-up with 5 g3 0-0 6 Bg2 is very possible too, as, indeed,
Hodgson later employed: 6 ... c6 7 Nd2 Be6 8 Ne2 Nd7 9 0-0 f5 10 c4 Nf6 1 1 Nf4
Qd7 1 2 Rcl gave White an edge in J.Hodgson-T.Upton, Moscow Olympiad 1994)
5 ... dxc4 6 Bxc4 0-0 7 Nc3 c6?! (ultra passive) 8 Nge2 Nd7 9 Qc2 Bd6 10 Bd3 g6 1 1
h3! (White prepares to meet Black's plan of ... f5 and ... Nf6 with an undermining g4
thrust) 1 1 ...Qe7 12 0-0-0 (Diagram 8) 1 2... a5 13 Kb1 Nb6 14 h4! saw White whip up
a strong attack in J.Hodgson-J.Gokhale, British Championship, Dundee 1993.
13
Sta rti n g O u t: The Trom powsky Attack

A strong exchange sacrifice Black's kingside is vulnerable
5 g3
...
There's no need for 5 Nd2 just yet if White is after a g3 set-up, although 5 c6 6 g3
has also been seen.
'
WARNING: White must not hurry with c4 once Black's bishop is on
e6: 6 c4?1 dxc4 7 Bxc4? Bxc4 8 Nxc4 QdS (Diagram 9), forking c4 and
X g2, is most certainly one to avoid.
Diagram 9 (W) Diagram 10 (B)

One to avoid A harmonious set-up
White might, th ough, begin with 5 B d3 fS an d n ow 6 N d2 is sensible, but a lea ding
14
The C l a ssica l 2 . . . d s
Trompowsky authority preferred to plough a creative furrow i n I.Miladinovic

M.Krivokapic, Pancevo 2006: 6 h3 Nd7 7 g4!? (who'd have thought that h7 would
be an early target?) 7 ...Qf6 8 Qe2 hS (the immediate 8 ...0-0-0! ? might be stronger) 9
gxfS BxfS 10 Nd2 Bd6 1 1 0-0-0 0-0-0! 12 BxfS QxfS 13 Qf3 Qe6 14 h4 and a complex,
m anoeuvring middlegame lay ahead. I would, however, be in less of a hurry to
undouble Black's pawns.
s c6 6 Bg2 Bd6 7 Ne2 (Diagram 10) 7 ... Nd7
...
The most flexible, although there can't be too much wrong with 7 ... 0-0 - the sub
ject of Game 2.
8 Nd2
White too refuses to commit his king just yet. Moreover, the knight must go to d2
either here or after 8 0-0 fS 9 b3 Nf6 to support the c4-break.
s ..fs
.
Common. Inexperienced Trompowsky players have been known to worry about

8 .. h5, but after 9 h4 Black's advance is likely to cause him at least as many prob
.
lems as White: for example, 9 ... 0-0 1 0 0-0 Bg4 1 1 c4 dxc4 12 Nxc4 Bc7 13 Qc2 Re8
and now White might well elect to get his queenside play under way with 14 b4.
9 b31?
Peter Wells has done much to support this approach, which has an ambitious fol
low-up in mind. More routine is 9 0-0 Nf6 10 b3 when Tiviakov's 10 ... Ne4! dem
onstrates that Black is alert to the needs of the position. Now 1 1 c4 QaS!? gives
Black decent counterplay and 1 1 f3 Nf6 12 c4 0-0 13 cS Bc7 14 b4 Re8 saw Black
playing most aggressively to target the weakness on e3 with 15 Qb3 Bc8 16 Rfe1
aS!? 1 7 a3 Qe7 18 Kf2 gS, which was rather unclear in M.Cebalo-D.Solak, Portoroz
2004.
White has also tried 9 Rcl Nf6 10 0-0 0-0 1 1 c4, but then 1 l ...dxc4! 12 Nxc4 BdS 13
Nxd6 Qxd6 14 Nc3 Bxg2 1 5 Kxg2 Rfe8 was extremely solid for Black in
R.Wojtaszek-K.Landa, German League 2006. Indeed, the fS-pawn can hardly be
considered a weakness here, binding down most effectively on the e4-square.
9 Nf6 10 c4 (Diagram 11)
...
White fights for the initiative with a pawn sacrifice. Despite good coverage in
Winning with the Trompowsky, this position remains quite unexplored and I must
confess that the premature end at this point to a game of mine with Jon Speelman
(British League 2005) did not help to flesh out the theory! White really should con
tinue and here we have:
a) 10 ... Bb4 11 0-0!? (okay, only now do we actually have a pawn sacrifice, but 1 1
cxdS BxdS would be very solid for Black a Ia Landa) l l ...Bxd2 12 Qxd2! (12 cxdS
Bxd5 1 3 Qxd2 Bxg2 14 Kxg2 QdS+ 15 Kg1 Ne4! is again fine for Black) 12 ... dxc4 13
Nf4 cxb3 14 Nxe6 fxe6 15 Rfb1 ! 0-0 16 Rxb3 gave White good pressure for his
pawn in J.Hodgson-S.Tiviakov, Groningen 1994, but Black should not be worse.
15
Sta rt i n g Out: The Tro m powsky Attack
b) 10 ... Ne4 11 c5! ? Bc7 1 2 b4 g6 13 Qb3 Kf8 14 a4 led to a rather complex manoeu
vring battle in A.Walton-M.Schaefer, Bad Worishofen 2003.
c) 10 ... h5 1 1 h4 Ne4 12 c5 Bc7 13 b4 g6 14 a4 (Diagram 12) is similar and after
14 ... Kf8 1 5 Qc2 Kg7 1 6 Qb2 Bd7 1 7 Nf4! Rb8 18 Ra3 Qe7 19 Nf3 a6 20 Ne5 Be8 21
Ned3 White had manoeuvred well, but the position remained extremely rich in
potential in P.Wells-J.Parker, British League 2003.

White will sacrifice a pawn! Heavy manoeuvring beckons
Conclusion
White's score of 55% from 2,800 games with 4 e3 suggests that his position is the
more pleasant to play. Indeed, he often has a decent choice of set-up. I quite like
an approach with g3, but must concede that Black's precise play in our main line
should enable him to equalize. However, at sub-grandmaster level Black is some
times on his own even as early as move 4, and one will often encounter the alter
natives to 4... Be6.
Illustrative Games
Gamel
D V.Kramnik V.Tkachiev
Ta l Memor i a l B l itz, Moscow 2008
It's often hard to include a blitz game in a book. There are bound to be mistakes,
but here White's approach is highly thematic and his overall play still of high
16
The C l a ssica l 2 . d s
..
q uality, as one would expect from such a strong player.

1 d4 d5 2 Bg5
This is often referred to as the Pseudo-Trompowsky. I must admit that I'm not a
fan (White should play 2 c4 or 2 Nf3!), but Tkachiev generously returns play to
standard Trompowsky waters.
2. .Nf6 3 Bxf6 exf6 4 e3 Bd6 5 c4
.
And why not with Black having made no attempt to dissuade this ideal break?
s ..dxc4 6 Bxc4 0-0 7 Nc3 (Diagram 13)
.
Diagram 13 (B) Diagram 14 (W)

White enjoys easy development How to net a pawn ?
7 . ..c6
Solid and fairly popular, but hardly essential a t this stage. Thus practice has also
seen:
a) 7 .. .f5 8 Nf3 Nd7 9 Qc2 (the more solid 9 0-0 c6 10 Rei Nf6 1 1 Qc2 Qe7 12 Bd3 g6
13 a3 Be6 appears fine for Black, although both Hodgson and Miladinovic on oc
casion have been happy to play this way and then manoeuvre) 9 . Nf6 10 0-0-0!?
. .
Qe7 11 h4 (11 Bd3 would be consistent with Kramnik' s approach in our main
game; then Black might block things up with l l ...g6 12 h4 h5, but 13 Ng5 gives
White something to play with: Rhel and e4 may follow) l l ...h5 12 Ng5 c6 13 Kbl
g6 14 Bd3 (thus we reach the same sort of position after all) 14 ...a5 15 Rdel ! ?
(White wants to keep his king's rook on the h-file) 1 5 ... Nd5?! (this doesn't fit i n too
well with Black's 14th and further weakens his structure) 16 Nxd5 cxd5 1 7 Qa4!
Bb4 18 Re2 saw White abandon his kingside lust for a positional edge in V.lotov
}.Borisek, Dresden Olympiad 2008.
17
Sta rti ng O u t: The Trom powsky Attack
TIP: Never forget that White's superior pawn structure often allows
him a decent degree of flexibility, as lonov fully exploited here.
b) 7... Nd7 8 Bd3 cS! ? 9 Nge2 cxd4 10 Nxd4 NeS 1 1 0-0 Bd7 12 Be4 restricted White
to just a pull in I.Miladinovic-D.Ivanovic, Mataruska Banja 2007.
c) However, the immediate 7... c5? is misguided for the same reason that 7...a6 8
Rei cS? (Diagram 14) 9 dxcS BxcS 10 Bxf7+! Kxf7 1 1 QhS+ Kg8 12 QxcS cost Black a
pawn in J.Hodgson-K.Arkell, London 1991 .
8 Nf3
This can't be faulted, especially as White is happy to chase the bishop in the event
of ...Bg4, but a decent case can also be made for 8 Qf3 followed by Nge2.
s...f5
More recently 8... Nd7 9 Qc2 ReB 10 Rg1 ! ? (I'm not too sure why White rejected 1 0
0-0-0, although after 10. . . f5 h e must, o f course, avoid 1 1 QxfS? o n account of
1 l ...Ne5) 10 ... Nf8 11 g4 aS 12 h4 a4 13 a3 QaS 14 gS fS 15 hS led to a rather unbal
anced middlegame in M.Bosiocic-F.Berkes, European Championship, Budva 2009.
9 Qc2 Nd7 10 Bd31 (Diagram 15)
Diagram 15 (B) Diagram 16 (W}

Only now is fS en prise Prising open the kingside
White avoids the aforementioned tactic and forces Black to weaken his kingside.
10 ... g6 11 h41
Kramnik wastes no time exploiting the hook on g6 to launch a strong attack.
11...Qe7
Risky. More solid would have been 1 1 ...h5, although after 12 NgS Nf6 13 0-0-0 Qe7
14 Kb1 play has actually transposed to the fairly pleasant waters (from White's
point of view) of Iotov-Borisek .
18
The C l a ssica l 2 .. d s
.
12 h 5 Nf6 13 hxg6 fxg6 (Diagram 16)

13 ... hxg6 isn't as risky as it looks, since Black can bring a rook to the h-file in time,
although with 14 0-0-0 Be6 1 5 Qd2!? Kg7 16 Ne5 White retains the upper hand and
has aggressive ideas of e4 as well as f4.
14 Bc4+ Be6?1
Bla ck trades his supposedly bad bishop, but this leaves his light squares vulner
able and enables White to maintain the initiative with some vigorous play. Thus I
suspect that Black should have taken his chances with the more active 14 ... Kg7 1 5
0-0-0 b5!.
15 Bxe6+ Qxe6 16 Ng5 Qc4 17 g41?
Olt! Kramnik presses on with his attack, although a decent case might be made
for 17 b3 Qa6 18 0-0-0 when White remains slightly for choice due to his central
(Kb1 and e4 is a plan) and kingside options.
17 ... Bb41?
Critical. Black had to avoid 17 .. .fxg4?? 18 Rxh7!, but might have considered the
solid 1 7... Nxg4 18 Nxh7 (18 b3 Qa6 is possibly worth flicking in, but then White
must avoid 19 Rxh7? Rae8! when it's suddenly Black who has all the nasty threats)
18 ... Rfe8 19 0-0-0 Kg7 20 Ng5 Rh8, restricting White to just a pull.
18 gxf5 Nd51
Tkachiev has pinned everything on the pin, but Kramnik is up to the challenge.
19 fxg61 Nxc3 20 bxc3 (Diagram 17)

A critical position Black's king is overwhelmed
20...Qxc3+?
Too greedy. Black had to prefer 20... Bxc3+! 21 Kdl QdS when matters wouldn't
19
Sta rti ng O u t: The Trom powsky Attack
have been a t all clear: 22 f3 Bxal (22 ...Qxg5!? 23 gxh7+ Kh8 24 Qxc3 Rxf3 i s a
calmer defence) 23 Qh2 Qd7 24 Qe5 (this queen manoeuvre appears po werful, but
White has invested a whole rook) 24 ... Rae8 25 gxh7+ Qxh7 26 Nxh7 RxeS 27 dxe5
Rd8+ 28 Ke2 Bxe5 29 f4 and White retains the initiative, but the resulting ending is
far from clear.
21 Qxc3 Bxc3+ 22 Ke2 Bxa1 23 gxh7+ Kh8 24 Rxa1
Taking stock we can see that White has two good pawns for the exchange. Even
more importantly, his knight is extremely well placed, guarding the fishbone on
h7 and assisting the central pawns' advance.
24... Rf5?
White's prospects were already quite bright, but there was no need to force him to
improve his f-pawn and king. Much better would have been 24... Rae8 when 25 e4
RfS! 26 Rgl Ra5 would have given Black some counterplay with his resulting
passed a-pawn.
25 f4 ReS 26 Kf3 c5?
This loses, but Black was already in huge trouble. A more solid defence was
26 ...Ref8, but after 27 Kg4 Rb5 28 e4 Rb2 29 Kf3! Rd2 30 e5 White's pawns are roll
ing.
27 e4 Rxgs
Alternatively, 27... Rf6 28 dxc5 Ref8 29 f5 and there's no stopping those pawns.
28 fxgs cxd4 29 g6 Kg7 30 Rh11 (Diagram 18}
Kramnik has everything worked out and has realized that the pa wn ending will
be an easy win.
TIP: When a pawn or two ahead in an endgame, remember to con

sider ways to return some or all of the extra material to simplify to a
straightforward winning task.
30... Rh8 31 e5 Kxg6 32 Ke4 Rxh7 33 Rxh7 Kxh7 34 Kxd4 Kg7 35 Kd5 Kf7 36 Kd6
White wins by a tempo.
36 ... Ke8 37 Kc7 as 38 a4 Ke7 39 Kxb7 Ke6 40 Kb6 Kxes 41 Kxas Kd6 42 Kb6 Kd7 43
Kb7 1-0
Gamel
D K.Georgiev A.Horvath
European Club Cu p, Fuegen 2006
1 d4 Nf6 2 Bg5 d5 3 Bxf6 exf6 4 e3 c6 5 g3 Bd6 6 Bg2 0-0 7 NeZ Be6

Related play to our main game occurred with 7... Re8 (there's no need to commit
this rook so early) 8 b3 f5 9 0-0 Be6 10 Qd3!? in R.Palliser-S.Mannion, British
20
Championship, Scarborough 2004, which continued 1 0... Nd7 1 1 c4 Nf6 1 2 Nbc3 a6

13 c5! Bc7 14 b4 (White presses ahead with his typical queenside expansion)
14 ... h5! (Diagram 19) 15 a4 (I wasn't so keen on 15 h4 Ne4, but this is a reasonable
defensive concept so long as White continues 16 Nf4, rather than 16 Bf3? g5! when
suddenly Black whips up a strong attack) 15 ... h4 16 b5 axb5 1 7 axb5 Rxa1 18 Rxa1
hxg3 19 hxg3 Ng4 and now the calm 20 Bf3 followed by Kg2 would have retained
a pleasant edge.

Black must seize some cou nterplay The plan is b3, c4 and Nc3
8 0-0
Georgiev castles having a specific follow-up in mind. More usually White has pre
ferred 8 Nd2 Nd7 9 b3 along the lines of the main line of the theory section.
8 ... Nd7 9 Qd31? (Diagram 20)
A rare but notable idea from the strong Bulgarian Grandmaster. White wants to
force through c4 without needing to develop his queen's knight to d2; he has the
more aggressive c3-square in mind for this piece.
9. .fs
.
The only other example of 9 Qd3 which I could find continued 9 ... Re8 and now I
dare say Georgiev would have played 1 0 b3, rather than follow the 10 e4 dxe4 1 1
Bxe4 Nf8 of V.Zhikharev-S.Bystrov, correspondence 2002; White does have the
superior structure here, but Black is extremely solid and exploiting that extra cen
tral pawn will be far from easy.
10 b3 Nf6 11 C4 Qd7
The alternative was l l ...Ne4, but I dare say that the Hungarian Grandmaster
wasn't so sure how to follow up after 12 Nbc3. Then White might well roll for
wards with c5 an d b4-b5, but a plan of exchanging on d5 followed by playing
21
S ta rti ng O u t: The Trom powsky Atta c k
down the c-file i s also possible.

12 Nbc3 Rae&
Black readies himself for an exchange on d5, but White prefers to keep the queen
side closed.
13 cSI Bc7 14 b4 hSI
NOTE: Black's position might appear quite solid, but it is also rather
passive -just look at the prospects of that bishop on e61 He thus
needs counterplay and as the ... b6-break would only help White,
Horvath must turn to the kingside.
15 h4
Deciding when to block Black's kingside play in this manner is never easy, but
Georgiev has seen deeply into the position and discovered a strong follow-up.
15 ... Rfe8 16 Nf41 Bxf4 17 exf4 (Diagram 21)

White retains a pleasant pull White has options on both flanks
It might seem strange at first to have allowed Black to damage White's kingside
structure, but just compare the remaining bishops and pawn breaks! White's
bishop has clearly the better prospects and just where is Black's counterplay
against White's long-term advance on the queenside?
17 ... Ne4 18 Ne2 bsl?
Radical but by no means bad defence. White will now gain the better prospects
down the a-file, but at least Black remains fairly solid. Note that Black couldn't
prepare a counterattack with ... g5 because 18 ... f6? 19 f3 would trap and win the e4-
knight, and 18 ... Rb8 19 a4 Rec8 20 Rfcl looks like a pleasant pull for White,
whether or not Black tries to counter on the queenside with ... b6.
22
T h e C l a ssical 2 . . d5
.
19 a4 a6 20 f3 Nf6 21 Ra3 Qb7 22 Qd2 RaB 23 Rfa1 Reba

Bo th sides have deployed their major pieces along logical lines, but what now for
Whi te?
TIP: It's rarely a bad idea to improve your worst-placed piece (or
even pieces). Note too how Georgiev doesn't hurry here - Black
hasn't any counterplay.
24 Bf1 NeB 25 Nc11 Nc7 26 Nd3 f6 27 Be2 Qc8 28 Nc1
Having provoked a small weakness ( .. .f6), the knight heads back as White remains
content to slowly probe. Moreover, it's not only the queenside on which he has
options, whereas Black, for his part, has the rather unenviable task of having to sit
tight and suffer.
28 ... Bd7 29 Bd3 Qf8 30 Ne2 Qe8 3 1 Kf21
Continuing to improve his pieces and this will help should White later decide to
open the kingside with g4.
31 ...g6
Horvath's position was rather unpleasant, but I wonder whether he would have
been better off trying 31...bxa4!? 32 Rxa4 Qc8 at some point. This does leave a6
ra ther weak, but with b4 at least a little sensitive, White might not have found it as
easy as in the game to retain control while opening the kingside.
32 Nc3 Kg7 33 Nd11 Be6 34 Ne3 Qd7 35 Qc2 (Diagram 22)
Black is being outmanoeuvred. Too late it appears he realized White's kingside
designs.
35 ... Rh8
White's play has been impressively patient thus far and only now does he invade
on the queenside.
36 axb5 axb5 37 Ra7 Rxa7 38 Rxa7 RaB?
Black continues to underestimate White's plans. Trading the rooks is natural, but
he had to keep things solid with 38 ... Bf7! when White would have had to gradu
ally probe with his queen; a2-a5-b6 being one possibility.
3 9 Rxa8 Nxa8 40 g41
Bla ck's bad bishop and fixed queenside pawns arguably constitute a weakness,
but to win White would like to classically attack a second weakness. The text pre
pares to do just that: Black's king now becomes exposed and his knight can but
spectate from a8.
40...hxg4 41 fxg4 Qc7 42 Ng2 (Diagram 23) 42 ...fxg4?1
Not the stiffest of defences. Instead 42 ... Qd7 43 hS gxhS 44 gxhS Nc7 45 Nh4
would have left Black under heavy pressure, but at least this way he would have
given White some issues to ponder: to cash in on fS or try to mate down the g-file.
23
43 Bxg6 Qa7 44 f51

The final blow. White's queen and knight will enter on the kingside with decisive
effect.
44...Bf7 45 Qe2 Bxg6 46 Qxg4 Qa2+ 4 7 Kg3 1-0

Black cannot hold fS Much more active than 3 ... exf6
The Dynamic 3 ... gxf6

1 d4 Nf6 2 Bg5 d5 3 Bxf6 gxf61? (Diagram 24)
The much more dynamic recapture and quite a challenging approach. Indeed,
those who like dynamic, unbalanced positions could do far worse than make
2 ... d5 and 3 ...gxf6 their defence to the Trompowsky.
4C41
A challenging response; White wants to open the position to take advantage of
Black's potentially vulnerable kingside.
WARNING: Despite having been used by a number of strong players,

I feel quite strongly that White should avoid 4 e3?1 as 4...c51
(Diagram 25) gives Black easy and fairly effective counterplay.
White has tried a number of ideas here, but without enjoying all that much suc
cess:
a) 5 c3 risks leaving White a little passive and even the active attempt S ... Nc6
(S ... Qb6 is a good alternative, after which 6 Qc2 Nc6 7 Nf3 cxd4 8 exd4 Bg4 9 Be2
Rc8 10 Nbd2 e6 1 1 Qb3 Bh6 was about equal in S.Conquest-G.Jones, Hastings
2008/09, but might Black not have gone 8 ...e5?) 6 dxcS!? e6 7 b4 aS 8 Qb3 f5!? 9 Nd2
24
The Cla s s i c a l 2 .. d s
.
Bg7 10 Rcl f4! failed to dent the notion that Black was having the greater share of
the fun in A.Walton-R.Palliser, Crewe 2001.
b) 5 dxc5 e6 6 c4 treats the position like a Queen's Gambit, but 6 ... dxc4! is an easy
equ alizer; Black's bishop-pair fully compensates for his fractured kingside.

Black fights for the initiative It's a Veresov!
c) 5 c4 cxd4 6 exd4 is White's most active try, but probably his least convincing
one too: 6... Nc6! (stronger than both 6 ...Qb6 7 Nc3 Qxb2 8 Nxd5 Bf5 9 Qcl Qxcl+
10 Rxcl Na6 1 1 c5 0-0-0 1 2 Ne3 Be4 13 Bxa6 bxa6, which reached a pretty unbal
anced middlegame in E.Torre-S.Tiviakov, Turin Olympiad 2006, and 6 ... dxc4 7
Bxc4, which enables White to escape into our main line) 7 Nc3 dxc4 and White is
yet to demonstrate anything better than 8 d5 Ne5 9 Bxc4 Nxc4 10 Qa4+ Bd7 1 1
Qxc4 ReS, but practice has shown Black to be at least equal here with his bishops.
d) Taking play out of the Trompowsky and into the Veresov with 5 Nc3 (Diagram
26) is perhaps not such a bad idea. Nevertheless, I consider 1 d4 Nf6 2 Nc3 d5 3
Bg5 c5! ? to be a decent, dynamic line for Black. Those who disagree and like
White's better structure and ideas of a quick Qh5 should consult Nigel Davies's
The Veresov (Everyman, 2003).
e) 5 Bb5+!? Bd7 (I wonder what Morozevich intended after 5 ... Nc6!?; perhaps 6
Ne2, since 6 dxc5 e6 7 c4 dxc4 8 Nc3 Qxd1+ 9 Rxd1 Bxc5 1 0 Nf3 Ke7 is comfortable
for Black, Z.Rahman-H.Mas, Kuala Lumpur 2008) 6 Bxd7+ Nxd7 7 Nf3 e6 8 0-0
Bg7 9 Nc3 f5 10 Ne2! Rc8 1 1 c3 0-0 12 Nf4 wasn't entirely clear, but gave White
good chances of emerging with an edge in A.Morozevich-P.Leko, Monaco (blind
fo ld) 2007.
Having seen why 4 e3 is probably best avoided, we tum our attention to 4 c4:
4...dxc4
25
Sta rt i n g Out: The Trom pows ky Attack
Black opens the centre i n a bid to assist his bishops and saddle White with a cen
tral weakness. Practice suggests that this is his best approach, but he has been
known to prefer something more solid on occasion:
a) 4 ... c6 5 e3 e6 (the bishop might come out first: 5 ...Bf5 6 Nc3 e6 7 Nge2 Qb6 8 Qd2
dxc4 9 Ng3 Bd3 10 Bxd3 cxd3 and now immediately going long is sensible, al
though 1 1 Nh5!? Nd7 1 2 g4 0-0-0 13 0-0-0 Qa5 14 Kb1 Rg8 15 h3 f5 16 f3 left Black
facing certain structural difficulties in A.Stefanova-E.Kovalevskaya, Russian
Women's Team Championship 2006) 6 Nc3 f5 7 Nf3 Bg7 8 Qc2 Nd7 9 h3!? (9 cxd5
cxd5 10 Bd3 followed by playing on the queenside would be a simple route to a
pleasant edge) 9 ... dxc4 10 Bxc4 Qc7?! 1 1 0-0-0 b5?! 12 Bb3 Ba6 13 g4! left Hodgson
well on his way to a classic Trompowsky crush in J.Hodgson-A.Martin, British
Championship, Plymouth 1992.
b) 4 ... c5 5 cxd5 Qxd5 6 Nf3 cxd4 7 Nc3 Qa5! (more active than 7 ... Qd8 when Hodg
son has shown the way: 8 Nxd4 e6 9 e3 a6 10 Qh5! and Black cannot be happy
with the outcome) 8 Nxd4 Bg7 9 e3 0-0 10 Bd3 (10 Bc4!? with the idea of IO Rd8?!
...
1 1 Qb3 is another possibility) IO... Rd8 11 0-0 f5 ( 1 l ...Nc6!? 12 Nxc6 bxc6 would at
least give Black the b-file for counterplay, albeit at the cost of a further structural
weakness) 12 Qh5! e6 (12... Bxd4 13 exd4 Rxd4 14 Qg5+ Kf8 15 Rad1 gives White a
strong initiative for the pawn) 13 Rad1 Qe5 (Diagram 27) gave White various
pleasant options in P.Wells-S.Sulskis, Port Erin 2003, including the simple 14 Rd2
followed by doubling rooks and the somewhat more ambitious 14 g4!?.
Diagram 27 (W) Diagram 28 ( B)

A little shaky for Black Unexplored aggression
0
NOTE: Black's third move isolated and weakened his h-pawn. Thus
hs is often an excellent outpost for the white queen; targeting that
pawn while posing serious difficulties for the black king.
26
The C l a s s i c a l 2...d s
5 e3
Despite some attention from Wells, 5 e4!? (Diagram 28) hasn't really taken off. In
d eed, much remains to be discovered here and those who revel in sharp, unbal
anced positions could do worse than give 5 e4 a closer look. One critical line runs
s .. Nc6 (5 ... f5!? is an ambitious alternative; the game S.Fowler-C.Crouch, London
.
2007, continued 6 Bxc4 fxe4 7 Nc3 Bg7 8 Qh5 0-0 9 0-0-0 Qd6! with a double-edged
fight ahead and forcing ...e6 with Wells's 7 Qh5! ? is very possible too) 6 d5 Ne5 7
f4 and now practice has chiefly debated 7... Nd3+ 8 Bxd3 cxd3 9 Qxd3, which looks
fairly attractive for White, but the untried if risky 7 ... Ng4!? might be more of a
concern. Then 8 Bxc4 c6! leaves White a little vulnerable on the dark squares and
even the superior 8 Qf3! can hardly be described as anything but rather murky
after a continuation like 8...Qd6 9 h3 h5!?.
s . ..cs
The greedy 5 ... Rg8 has some logic, but is rather ambitious and has been rather rare
in practice. With good reason I would say: 6 Nc3 c6 7 Bxc4 Rxg2 8 Nge2 Rg7 9 Qc2
gave White decent value for his pawn in J.Fries Nielsen-N.Huschenbeth, Ham
burg 2005, and Wells's 7 g3! ? most certainly has its logic too- just where is Black's
king going to reside?
6 Bxc4 cxd4 7 exd4 Bg7 (Diagram 29)
Diagram 29 (W) Diagram 30 (W)

Both sides have wea knesses Shou ld the d-pawn be advanced?
Thus the contours of the forthcoming struggle begin to take shape: Black has in
flicted an isolani and will pressure it with ... f5, but White has free piece play and
would like to build up a kingside attack.
8 Ne2
Considering what we said about h5 being an ideal square for the white queen, it's
27
Sta rti n g Out: The Trom powsky Attack
n o surprise that Wells's idea o f 8 QhS!? has received some attention since the pub
lication of his fine work. However, here the queen has perhaps been committed a
touch early and Black appears to have the answer in 8 ... 0-0 9 Ne2 Nc6 10 Nbc3
QaS!, as first played in Cao Sang-Z.Gyimesi, Hungarian Championship, Budapest
2004.
8 0-o 9 Nbc3 fs (Diagram 30)
...
Both sides have developed logically: White's knight went to e2 rather than f3, for
instance, to avoid being pinned and because it dreams of landing on hS with
aplomb. At this point White has two main approaches:
a) 10 Qd2 Nc6 1 1 Rd1 keeps the pawn on d4 for the time being and will be dis
cussed in Game 3.
b) 10 0-0!? Nc6 1 1 dS is a more ambitious try, after which ll ... NeS 12 Bb3 Qd6! (a
more critical defence than 12 ... a6 13 Qd2 Qd6 14 Qf4! Kh8?! 15 Rfe1 Qf6 1 6 Rad1
Rg8 17 Nd4 which enabled White to take control in R.Palliser-K.McPhillips, Street
2004) 13 h3 Bd7 (Diagram 31) 14 Ba4! (White prevents ...bS and sensibly ends any
hopes Black may have harboured of one day seeing his bishops operating in tan
dem) 14 ... Nc4 15 Bxd7 Qxd7 16 b3 Nb6 17 Qd3 Rac8 18 Rad1 ReS 19 Qf3 Rfc8 20
Rd3 Kh8 was finely balanced and agreed drawn in a leading Trompowsky battle,
I.Miladinovic-K.Georgiev, Bar 2008.

Dynamically ba lanced Black attacks d 4
Conclusion
The dynamic recapture on f6 i s quite a challenging idea. White should avoid 4 e3
cS when he has only scored 44%. He has done better (56%) with 4 c4, although that
falls to just 51% after 4 ... dxc4. Indeed, both sides should be fairly happy to contest
28
T h e C l a s s i c a l 2 ... ds
our main line here, which gives rise to a rather unbalanced middlegame and a
ty pical Trompowsky battle between White's knights and Black's bishops.
Illustrative Games
Game3
0 A.Bigg K.McPhillips
Street 2004
1 d4 Nf6 2 Bg5 d5 3 Bxf6 gxf6 4 c4 dxc4 5 e3 c5 6 Bxc4 cxd4 7 exd4 Nc6 8 Ne2 Bg7 9
Nbc3 o-o 10 Qd2 f5 11 Rd1 a6
Black usually flicks this in, although it's far from clear why he has been so keen to
avoid 1 1 ...Qd6 12 NbS Qb4.
12 0-0
This might be delayed further and 12 Qf4!? Bd7 (12 ...Qd6 13 Qxd6 exd6 14 0-0 Bd7
15 NdS is perhaps a little better for White) 13 0-0 Qb8 14 Qh4 Qd6 15 Rfe1 Rad8 16
a3 Bc8 was seen in G.Van Perlo-E.D' Adamo, correspondence 1988. This unbalanced
position is unclear and White might like to look into both 1 7 Rd3 and 1 7 BdS!?.
NOTE: High-level correspondence games are an excellent source of

opening ideas and inspiration. Readers who like to study their open
ings in some depth could do far worse than purchase a decent corre
spondence database, such as Tim Harding's UltraCorr 2.
12 ...Qd6 (Diagram 32) 13 Bd51?
White finds a creative way to keep the pawn on d4 and deal with Black's pressure
down the d-file. Again he might also advance and 13 dS NeS 14 Bb3 Ng4 (proba
bly too ambitious and 14 ... Bd7!? might well be preferred) 15 Ng3 Bd7 16 h3! Bh6!?
17 Qe2 Ne3 18 fxe3 Qxg3 19 Rf3 QeS would have favoured White in M.Galyas
A.Naiditsch, Budapest 2000, had he now taken aim at the kingside with Wells's 20
Bc2 and 21 Qf2.
13 .. Rda 14 Qg5 Qg6
.
I suspect that Black feared the white knight joining the attack with 14 ... Bd7 15 Nf4
Kh8 16 NhS, but this might well be manageable for him after 16 ... Bh6.
15 Qe3 Bf6?1
B la ck's bishops probably don't supply quite enough compensation after 15 ... Bd7
16 Bxc6 Qxc6 (16 ... Bxc6 is well met by 17 d5) 17 Qxe7 Re8 18 Qb4, but he might
have returned with 15 ... Qd6!? when 16 Nf4 Nxd4! is an important tactical point.
Indeed, the tactics appear to hold up for Black and leave him with rough equality
29
Sta rt i n g Out: The Trom powsky Atta c k
after 1 7 NhS f4 18 Nxf4 e5.

16 f41 (Diagram 33)

The fl-rook will be lifted Black ca n activate his q ueen
Bigg wants to attack, although a decent case might also be made for 16 Bxc6!? bxc6
1 7 Nf4, leaving White with a pleasant edge; Na4-c5 may follow.
16 ... Qg7 17 Rf3 Kf81l
Creative defence. Few people would have allowed their queen to be buried in the
comer, but McPhillips wants to maintain pressure against d4 and is confident that
his queen will later re-emerge. The alternative was 1 7... Kh8 18 Rg3 Qf8 when
White might trade on c6 and leave Black a little tied up or prefer the more aggres
sive 19 Bf3!? followed by dS.
18 Rg3 Qh8 19 Bxc6 bxc6 20 d5ll
White opens the position, but this fails to trouble the black monarch and actually
increases the black queen's prospects. Better was the simple approach: 20 Na4! Be6
21 b3, retaining a grip and a fairly pleasant edge.
20...cxd5 21 Rxd5 Be6 22 Rxd8+ Rxd8 23 Qb6 (Diagram 34)
White may have been relying on this queen raid, but now Black is able to solve all
his problems down the a1-h8 diagonal.
23 ...Bd4+ 24 Nxd4 Qxd4+ 25 Qxd4 Rxd4 26 Rf3 Bd7 27 Rf2
White's queenside majority and slightly better structure still give him a few hopes,
but McPhillips moves quickly to activate his central majority.
27 ...f6 28 Kf1 e5 29 Ke1 Ke7 30 Rd2
Trading rooks is the obvious way to attempt progress, but Black doesn't need fear
their exchange.
30
The C l a ssica l 2 . . . d s
3o... Rxd2l 31 Kxd2 Kd6 32 Ke3 Kc51?

A sli ghtly ambitious try. Black might have drawn more simply by keeping
Whi te's majority in check with either 32 ... a5 or 32 ... Be6!? 33 b4 Bc4 34 a4 Bfl 35 g3
Bc4.
3 3 h4
Whi te must have looked long and hard at 33 g4!?, and presumably concluded that
33 . exf4+ 34 Kxf4 fxg4 35 Ne4+ Kb4 is fine for Black because he is most certainly
. .
not worse in the race situation after 36 a3+ Kc4 37 Nxf6 Bb5 38 Nxh7 Kb3 39 Kxg4
Kxb2 40 h4 Kxa3.
33 ... Be6 (Diagram 3 5)

Tense if d rawish Torre-li ke play
34g3
Once again 34 g4!? deserved attention, although 34 .. .fxg4! (34 ... Kd6 35 gxf5 Bxf5 36
Ne4+ Ke7 37 Nc5 gives White a pull) 35 Ne4+ Kb4 36 Nxf6 exf4+ 37 Kxf4 Bxa2 re
veals the bishop's advantages over that of the knight. Both sides have a couple of
small tricks here, but 38 Nxh7 Bf7 39 Kxg4 Kb3 40 Ng5 Bg6 should really lead to a
draw.
3 4.. .Kb43 5 Kd3 Yz-Yz
Wh ite Declines to Exchange

1 d4 Nf6 2 Bg5 d5 3 e3 (Diagram 36)
This is a reasonable choice for those with some experience of the Torre Attack or
31
those who like to g o their own way a s soon a s possible, not that the exchange on
f6 can really be described as being especially theoretical. The text is White's main
alternative to that exchange. Instead 3 Nf3 reaches a poor form of the Torre,
against which 3 ... Ne4 followed by 4 ... c5 gives Black active, early and good coun
terplay. A more important alternative is 3 Nd2 when 3 ...e6 is likely to lead into our
main line with 3 e3. Black might well, though, prefer one of:
a) 3 ... c5!? is quite a challenging riposte. Then 4 Bxf6 (not forced, but 4 dxc5 e6 5
Nb3 Nbd7 is fine for Black and I'm not overly convinced by the gambit which 4 e3
cxd4 5 exd4 Qb6 entails; moreover, Black doesn't even have to grab the pawn and
6 Ngf3 Ne4!? 7 Nxe4 dxe4 8 Ne5 Nc6 9 Nxc6 bxc6 10 Be2 g6 gave him promising
counterplay in S.Fowler-M.Hebden, Halifax (rapid) 2006) 4 ... gxf6 (Diagram 37)
should be compared to our last section. White has tried to avoid an early e3 here,
but both 5 dxc5 Na6 6 e4 Nxc5 7 Bb5+ (7 Qh5!?) 7... Bd7 8 Qe2 dxe4 9 Nxe4 Nxe4 1 0
Bxd7+ Qxd7 1 1 Qxe4 f5 (A.Sokolov-A.Veingold, Upplands Vasby 2000) and 6 g3
Nxc5 7 Bg2 Bg7 8 Ngf3 0-0 9 0-0 f5, as in S.Dishman-A.Law, British League 2003,
are fine for Black.

Dynamic play from Black It's a Torre Attack proper!
b) However, 3 ... Nbd7 4 Ngf3 g6 doesn't seem as good as against 3 e3 in view of

Hodgson's vigorous reaction 5 c4!? dxc4 6 e4 unless Black can make the ambitious
5 ... Bg7!? 6 cxd5 Nxd5 7 e4 h6 8 Bh4 Nf4 work.
3 e6
...
This solid choice has been Black's most popular in practice, but it is the alterna
tives which worry me more:
a) 3 ... c5! is a serious challenge because 4 Bxf6 gxf6 is quite pleasant for Black as we
saw in the last section. Instead White might contest a Veresov with 4 Nc3 or try 4
c3, but Black has a few good responses to that, including 4 ... Nc6 and 4 ... Ne4 5 Bf4
32
Nc6, after which 6 Nd2 Nxd2 7 Qxd2 Bf5 8 Nf3 e6 saw Black equalize without any
d i fficulty whatsoever in N.Sedlak-Y.Dembo, European Championship, Budva
2009 .
b) 3 . . Nbd7 4 Nf3 (Hodgson's more ambitious 4 c4!? Ne4 5 Bh4 is well met by the
.
acti ve 5 ... g5! ? 6 Bg3 e6 according to Davies) 4 ...g6 (far from forced, but a tempting
move because White is likely to lose a tempo with his e-pawn if he now goes in for
5 c4 ) 5 Nbd2 Bg7 6 Bd3 (6 c4 does, indeed, appear less challenging here; after
6 . . 0-0 7 Be2 c5 8 cxd5 Nxd5 9 0-0 h6 10 Bh4 b6 1 1 Nb3 Bb7 Black was fairly com
.
fortably placed in Zhang Pengxiang-Chao Li, Jiangsu Wuxi 2008) 6... 0-0 7 0-0
tra nsposes to a solid line of the Torre Attack (1 d4 Nf6 2 Nf3 g6 3 Bg5) in which
Black usually prepares ...e5 with 7...Re8 (Diagram 38).
TIP: Never forget about transpositions when planning your reper

toire. Not only do they occur between various Trompowsky varia
tions, but sometimes it's even possible to take play into entirely dif-
ferent openings!
c ) Indeed, one should not assume that just because he didn't move the knight last
go Black won't here: 3 ... Ne4 4 Bf4 transposes to a position more usually reached
via a 2 ... Ne4 3 Bf4 d5 4 e3 move order - see Chapter Six.
4Nd2
White keeps his options open and refuses to commit his king's knight just yet. In
stead 4 Nf3 would take play into a normal Torre Attack and it's also possible to
reach the Queen's Gambit! Indeed, 4 c4 Be7 5 Nc3 h6 6 Bh4 0-0 7 Nf3 takes play
into a key tabiya of the Queen's Gambit Declined. There really are a lot of possible
transpositions after 3 e3!
4 cs 5 c3 (Diagram 39)
...

Black still has options An improved Stonewa ll
33
s . Be7
..
Black will develop his knight to d7 in our main line, shoring up the defence of f6
and refusing to block the path of a fianchettoed bishop on b7. However, the text is
far from essential and S ... Nc6 has also been seen: 6 f4! ? Be7 (6 ... Bd6!? looks like an
idea to me; the point is that 7 Ngf3 h6 8 Bh4 cxd4 forces White to recapture with
the c-pawn, rather than the more ideal recapture with the e-pawn, and so White
might consider giving up the bishop-pair on f6) 7 Bd3 0-0 (Black castles into it; a
more creative defence is 7...Qc7 8 Ngf3 Bd7 9 Ne5 h6 10 Bxf6 gxf6!?, but with the
thematic manoeuvre 1 1 Nxd7 Qxd7 12 Qh5! White obtained the upper hand in Bui
Vinh-Nguyen Van Hai, Hanoi 2009) 8 Ngf3 b6 and now I would attack in standard
Torre fashion with 9 Ne5 followed if possible by Qf3 which appears fairly danger
ous. Moreover, 9 h4! ? NeB 10 Ne5 Nxe5 1 1 fxe5 f5 12 QhS Rf7 13 Nf3 also left
White for choice in J.Ehlvest-F.De Ia Paz Perdomo, Santo Domingo 2006.
6 Bd3 Nbd7 7 f41 (Diagram 40)
Thus we get to see why White held back the development of his knight from gl .
By first placing the pawn on f4 White essentially obtains an improved Torre At
tack (in which Ne5 followed by f4 is a common theme), or a turbocharged Stone
wall if one prefers - that bishop is certainly much better on g5 than cl.
7 ... b6
Related play occurred after 7...h6 8 Bh4 b6 9 Ngf3 (9 Nh3!? and Qf3 might be tried
a Ia our next note) 9 ... Bb7 10 Ne5 in I.Miladinovic-O.Cvitan, Zadar 2007, in which
Black demonstrated a typical exchanging ploy: 10 ... 0-0 1 1 0-0 Ne4! (Diagram 41)

An equalizing attem pt White has some pressure
12 Bxe7 Qxe7 1 3 Ndf3 Nxe5 14 Nxe5 Qe8 and the exchanges left him close to
equality.
34
The C l a ss i c a l 2 . . d s
.
TIP: When slightly cramped it's often a good idea to exchange pieces
to free one's position.
Be fo re returning to our main line, let's consider an interesting twist on this im

p roved Torre theme which Hodgson liked to employ: 7...0-0 8 Nh3!? b6 (better the
mo re active 8...b5!) 9 Nf2 Bb7 10 Qf3 cxd4 1 1 exd4 Qc7 12 0-0 aS 13 Rfe1 Ba6 14 Bc2
bS 15 Qh3 gave White a very dangerous attack in J.Hodgson-T.Hebbes, Southend
1 998; Ng4 will follow.
s Ngf3 Bb7 9 Nes (Diagram 42)
Whi te sinks his knight into its favourite outpost and now Black must be a little
care ful:
a) 9. . .0-0 10 0-0 Ne4! can be compared with the notes to Black's 7th move, but per
haps here White can consider 10 Bxf6!? Nxf6 1 1 Qf3, still hoping to attack.
b) However, ... Ne4 per se does not equalize: 9 ... a6?! 10 0-0 Ne4?! 1 1 Bxe4! dxe4
( 1 l . . .Bxg5? runs into 12 Nxf7!) 1 2 Ndc4 Nf6 13 dxcS BxcS 14 Qxd8+ Rxd8 15 b4 Be7
16 Nxb6 netted a pawn in I.Miladinovic-A.Dragojlovic, Zlatibor 2007.
c) 9. . . Nxe5 10 fxeS Nd7 is another equalizing try, as we'll consider in Game 4.
Conclusion
3 e3 will appeal to those with some experience of related system-based openings,
especially the Torre. Overall, White has scored well with 57% from over 1,600
games, but that's largely because of 3 ... e6 which sees his score rise to a whopping
64%; these improved Torre lines pose definite problems. However, 3 ... c5! is much
of a theoretical hurdle, albeit one which won't appeal to everyone; Black really has
to be happy to then meet 4 Bxf6 with 4 ... gxf6.
Illustrative Games
Game 4
0 Tu Hoang Thong A.Vul
Ottawa 2007
1 d4 Nf6 2 BgS dS 3 e3 e6 4 Nd2 Nbd7

B lack delays ... cS and prepares to meet 5 Bd3 with 5 ... e5. Thus White must ad
va nce the f-pawn without delay (or acquiesce to a normal Torre with 5 Ngf3).
5 f4 Be7 6 Bd3 cs 7 c3 b6 8 Ngf3 Bb7 9 Nes Nxes?I 10 fxes Nd7 (Diagram 43)
Th is type of exchanging method is quite common in the Torre, but here it fails to
35
Sta rti n g Out: The Trom pows ky Attack
convince. However, it's easy t o see i t being popular with a panicking black player
concerned about White's kingside intentions.

Black's kingside can be pressured Keeping up the pressure
11 Bf41?
White keeps the bishops on, thereby leaving Black a little cramped. However, ex
changing them also leads to a pleasant advantage: 1 1 Bxe7 Qxe7 1 2 QhS 0-0-0 13
0-0 f6 1 4 BbS (14 Qh4!? looks quite reasonable too, causing problems with the pin
as well as against h7) 14 ... Rdf8 15 a4! (Diagram 44) favoured White in V.Georgiev
S.Barrientos Chavarriaga, Turin Olympiad 2006. It's worth noting how Georgiev
intended to meet 15 .. .fxe5: 16 Bxd7+ Kxd7 17 QxeS Rxfl + (or 17 ... Ba6 18 Rxf8 Rxf8
19 aS with strong pressure) 18 Rxfl Rf8 19 Rxf8 Qxf8 20 Nf3 and White's classic
queen-and-knight tandem are favourite against Black's queen and bishop.
11 . 0-0?1
. .
Highly risky, but Black's king position was a serious concern for him in any case.
Previously 1 1 ... c4 12 Bc2 Qc7 had been tried, but after 13 QhS g6 14 Qh6 0-0-0
(14 ... Bf8?! 15 Qh3! only assists White's cause, as shown by, for example, 15 ...Be7 16
Bh6 0-0-0 1 7 0-0 and a pawn will fall) 1 5 0-0 Rdf8 16 Nf3 Qd8 1 7 Rf2 Bc6 18 Rafl
Kb7 19 BgS White was in complete control and it wasn't long before he won a
pawn in A.Miles-P.Van der Sterren, Linares Zonal 1995. Observe too how 1 1 ...Qc7
12 0-0 0-0-0 13 Bg3! Rdf8 14 QhS leaves White clearly better.
0
NOTE: Once again h5 proves to be a great square for the white
queen! Black's king might be safe on the queenside, but his kingside
pawns can be attacked regardless.
I suspect that Black really ought to try 1 l . ..Bh4+!? 12 Bg3 (12 g3 gS! isn't so clear)
36
The C l a s s i c a l 2 . . . d s
t z Qg5, although after 1 3 Qf3 White still enjoys a n edge.

. . .
12 Qg4 Kh8 13 0-0

The Vietnamese Grandmaster wastes no time attacking. The text not only enables
White to pressure f7, but also introduces ideas of a timely rook lift. White most
certa inly has a dream Torre/London attacking situation!
13 ... Bc8?l
Bla ck wants to break with .. .f5. Perhaps he already should have tried the desperate
13 f5!?, although after 14 exf6 Nxf6 1 5 Qxe6 Bc8 16 Qe5 White's queen will es
. . .
ca pe, leaving him a clear pawn to the good.

14 Qh31 (Diagram 45)

A simple and strong attack There's no defence
Less clear would have been 14 Rf3 f5!, but the text forces a serious kingside weak
ness.
14...g6
Black 's problem is that 14 ... f5? fails to the tactic 15 exf6 Nxf6 16 Bc7! Qxc7 17 Rxf6,
destroying the defence of h7.
15 Bh6 f5
The only move; he couldn't let the floodgate on f7 open.
16 exf61
White is so confident in his attacking chances that he spurns winning a clear ex
change and who can blame him?
16 ... Rxf6 17 Qg3
Taking control of the eS-square and now White is all set to include his remaining
37
pieces in the attack.

11 ...as11
Black wants to trade the light-squared bishops. This is far too slow a plan to save
him, but even after 17 ... Kg8 18 e4! his chances of survival would have been slim
indeed.
18 Nf3 Ba6
Neither would 18 ... Ra7 19 Ng5 Qe8 20 Bb5! have enabled Black to defend, and it's
also almost impossible to find a reasonable move for him after 18 ... Kg8 19 Ng5.
19 Bxa6 Rxa6 20 Ne5 (Diagram 46) 20 ... Nxe5
The check on f7 had to be dealt with and Black's vulnerable back rank would have
cost him after 20 ...Qe8 21 Nxd7 Qxd7 22 Rxf6 Bxf6 23 Rfl .
21 Qxes b5 22 Rf3
The pin is decisive.
22 ... Kg8 23 Raf1 1-0
Crushing stuff.
38
C h a pter Two
2 g6 a n d M i n o r Li n e s
I nt roduction a n d Ra re Defe nces
Black Pl ays 2 g6
I ntrod uction a n d Ra re Defences

1 d4 Nf6 2 Bg5 (Diagram 1)

Black has severa l unusual tries Oops!
In this chapter we continue our study of those lines in which Black doesn't fear the
exchange on f6 with the notable exception of the dynamic 2 ... c5, which can be
found in Chapter Four. By far Black's most common alternative to both that and
2 ... d5 is 2 ... g6, but before we look at the fianchetto we should examine some un
usual but not implausible tries:
a) 2 ... d6 3 Bxf6 exf6 is very similar to 2 ...g6, since Black's dark-squared bishop
hardly has a better square than g7 unless he wants to lose a tempo with ... d5. In
deed, here 4 c4 g6 5 Nc3 fS 6 e3 Bg7 7 Nge2 0-0 8 g3 transposes to the main line of
our next section.
b) 2 . . . h6?! is a clear error, albeit one not unknown at lower club level. White
should be quite happy to be provoked into 3 Bxf6 exf6 (3 ... gxf6 4 e3 is also ex
tremely pleasant for White; already that hS-square is calling the queen) 4 e3 when
the inclusion of ... h6 is undoubtedly questionable.
c) 2 ... c6 forces White to display some care if he isn't a fan of exchanging on f6 in
general.
WARNING: 3 e3?? Qa5+ (Diagram 2), exploiting the loose bishop, is

most certainly a trap to avoid.
Thus those after a Torre-like position should prefer to begin with 3 Nd2. How
ever, there's no reason whatsoever to avoid 3 Bxf6 exf6 (we saw that 3 ... gxf6 4 c4
40
2 . ..g6 a n d M i n o r Lines
d 5 i s pleasant for White i n note ' a ' to Black's 4th move in the third section of the
p reviou s chapter) 4 c4, since 4 ... d5 5 cxd5 cxd5 6 Nc3 looks like an inferior version
of the Slav for Black with his d-pawn a little weak. Instead 4 ...Bb4+ 5 Nd2 d5 6 e3
Be6 7 cxd5 Bxd5!? saw Black avoiding such a scenario in J.Hodgson-R.Slobodjan,
Ge rm an League 1999, but 8 a3 Bd6 9 e4 Be6 10 Bc4! Bxc4 1 1 Nxc4 0-0 12 Ne2 still
le ft him worse as it wasn't so easy to undouble those pawns with ... f5.
d) 2 ...b6 has its logic as we can see by exploring 3 Bxf6 exf6 when White would
like to go 4 g3?, but 4 ... Bb7 beats him to the long diagonal and, rubbing salt in the
wound, 5 Nf3 Qe7! actually wins a pawn. Unlike previous commentators, though,
1 am not so unhappy with 3 Nd2 when Black has two main tries:
d 1 ) 3 .. e6 4 e4 h6 5 Bxf6 Qxf6 6 Ngf3 reaches a position which can come about via a
.
number of move orders; the rare 2 ... e6 3 e4 h6 4 Bxf6 Qxf6 5 Nf3 b6 6 Nbd2, as per
our next chapter, being one. After 6 ... Bb7 7 Bd3 (Diagram 3) I quite like the control
White's centre affords him: for example, 7...Qd8 (the queen scurries out of the
way; this isn't forced, but 7... a6 8 Qe2 d6 9 0-0-0 Nd7 10 Kb1 e5 1 1 c3 Be7 12 Nc4
gave White an edge too in no lesser clash than V.Korchnoi-A.Karpov, Hastings
1971/72) 8 Qe2 Be7 9 0-0-0 d6 10 h4! (discouraging Black from going short while
seizing some useful space) 10 ... a6 11 Nc4 Nd7 12 Ne3 was a little more pleasant
for White in K.Lerner-L.Yudasin, USSR 1 983.

A handy centre White is in control
d2) 3... Bb7 4 Bxf6 exf6 (or 4 ...gxf6 5 Ngf3 f5 when White is able to get his fianchetto
in with 6 g3 Bg7 so long as he then spots the threat and continues 7 e3 c5 8 c3 d6 9
Bg2 with good chances for an edge; a timely Nh4 may cause problems and White
can also press with a4 ideas) 5 Ngf3 g6 (the logical try as Black doesn't want to
obstruct his other bishop with ... d5) 6 e4!? (more ambitious than the far from im
plausible 6 g3) 6 ... Bh6 7 Bd3 0-0 8 0-0 Re8 9 Rel d6 10 a4! (Diagram 4) left White
41
slightly for choice with his centre secure in J.Timman-E.Rozentalis, Malmo 1997.
NOTE: Timman-Rozentalis actually began with the move order 1 d4

Nf6 2 Nf3 b6 3 Bg5 to which the Trompowsky player might trans
pose by meeting 2 ... b6 with 3 Nf3.
This probably allows Black more options than one would like, but White should
emerge with a pull in all lines, as I demonstrated in Starting Out: d-pawn Attacks.
Finally after 2 ... b6, we should note Wells's preference for 3 Nc3!? Bb7 4 Nf3, again
being prepared to trade the bishop-pair for the centre along the lines of our next
chapter with 4...e6 5 e4 h6 6 Bxf6 Qxf6. Then 7 Bd3 Bb4 8 0-0!? Bxc3 9 bxc3 Nc6 had
certain similarities to the Nimzo-Indian in R.Palliser-J.Radanovic, London (rapid)
2004, and now I should probably have freed the f-pawn without delay with 10 Nd2 .
e) 2 ... Nc6?! echoes the Tango, but I'm not convinced with White not committed to
c4, and 3 Nc3 e5 (or 3...e6 4 e4 Bb4 5 e5! h6 6 Bd2 Bxc3 7 bxc3 Ne4 8 Qg4 with a
terrible version of the McCutcheon French for Black) 4 d5 Ne7 5 d6! (not the only
good approach, but this positional gambit will leave Black badly cramped for a
long time to come) 5 ... Nf5 (5 ... cxd6 6 e4 Nc6 7 Bxf6 Qxf6 8 Bc4 had previously
given White excellent compensation in L.Trent-S.Gregory, British Championship,
Torquay 2002) 6 dxc7 Qxc7 7 Bxf6 gxf6 8 Nd5 (8 Qd3!? Bb4 9 0-0-0 is also strong)
8 ... Qc6 9 e4 Ne7 10 Bb5! (Diagram 5) saw White making good use of some tactics to
retain control in A.Walton-S.Gregory, British Championship, Scarborough 2004.

Punishing overly-creative play Sol id but a touch passive for Black
Conclusion
The lines covered here amazingly account for 5% of all Trompowsky games in my
database, so they are not as rare as one might imagine. Nevertheless, White
42
2 . . . g6 a n d M i nor Lines
should not b e unhappy to see any o f these offbeat tries and h e has a good score
aga inst them.
Bla ck Plays 2 ... g6

1 d4 Nf6 2 Bg5 g6
The move which a King's Indian or Griinfeld player would like to make work, but
the resulting positions tend to be a little more solid for Black than such dynamic
p layers would like. 2 ...g6 was employed by a number of grandmasters when the
Trompowsky still packed a certain surprise punch in the Eighties and early Nine
ties, but in general such players, including even Kasparov, subsequently moved
on to other defences. However, this variation remains fairly common at lower lev
els and has occurred in 8% of all Trompowsky games in my database.
3 Bxf6
The consistent follow-up. White might take play into the Torre Attack with 3 Nf3
or go 3 Nc3 when 3 ... d5 is a Veresov and 3 ... Bg7 4 e4 d6 a Pirc, but to play along
such lines rather begs the question: why did White select 2 BgS in the first place?
3 ...exf6 (Diagram 6) 4 e3
White's play in this variation tends to be quite strategic and largely driven by
general ideas, but there are some move order issues.
NOTE: Almost everyone agrees that White's best set-up involves a

combination of the moves e3, c4, Nbc3 and Nge2 allied to a fi
anchetto of the king's bishop, but not over what order to make
these moves.
These move order options boil down to the optimum way to prevent an early ... cS
causing too many difficulties. The text has been by far White's most common
move order, but Wells has made a good case for the alternative 4 c4 Bg7 (it's too
early for 4 ... c5 to do much damage; White continued 5 e3 Bg7 6 Nc3 in M.Navarro
Cia-J.Escofet Llongueras, Pamplona 2007, and might well have recaptured with
the queen had Black now exchanged on d4, rather than allow White an easy pull
with 6 ... a6?! 7 dxcS!) 5 Nc3 0-0 6 e3 d6 (6 .. .f5 7 Nge2 cS!? might be Black's best at
tempt, as Adam Hunt played against me at Blackpool in 2003; that game contin
ued 8 dxcS Na6 9 g3 NxcS 10 Bg2 d6 1 1 0-0 Be6 12 b3 a6 13 Rb1 Rb8 14 Nd4 ReB 15
Qc2 B d 7 16 Rfd1 and a draw was agreed, but plenty of play remains in the posi
tion and I slightly prefer White, who might combine NdS with an advance of the
a-pawn to aS) and now not the common 7 g3, but rather 7 Nge2 (Diagram 7) when
7 . . . f5 (White might now be happy to meet 7 ... c5 with 8 dxcS!? dxcS 9 Qxd8 Rxd8 10
NdS Nc6 1 1 0-0-0 Be6 1 2 Nef4, as he did i n I .Schneider-D.Schulze, Bad Lauterberg
2005) 8 g3 Nd7 9 Bg2 N f6 takes play back into our main line.
43
Start i n g Out: The Trom pows ky Attack

Not hurrying with g3 and Bg2 Black remains structurally worse
4 Bg7
...
Black's usual continuation. He can also delay developing the bishop here until
after 4 ... d6 5 g3 f5 6 Bg2. Ideas involving ... Bh6 have been tried too on occasion,
but White should maintain a slight edge, such as with 6 ...h5!? 7 h4 (it makes good
sense to keep Black's kingside ambitions under lock and key) 7... Nd7 8 c4 Nf6 9
Nc3 Bh6 10 Nge2 0-0 1 1 Qd3 ReB 12 b4 which was followed by the standard
queenside assault in Hoang Thanh Trang-Huang Qian, Beijing (rapid) 2008.
5 g3 0-0
Play will often transpose after 5 .. .f5 6 Bg2 unless Black prefers 6 ... d5, which is
similar to the ...exf6 variations considered in our previous chapter. I doubt that
Black's bishop is better placed on g7 than d6 and certainly 7 Ne2 0-0 (7... c5?! 8
Nbc3! takes good aim at d5) 8 0-0 c6 once again sees White playing to advance
with c4. Following 9 Nd2 (not essential, although after 9 b3 Be6 10 Nf4 Nd7 I
would move the knight to support c4, since the 1 1 c4?! of M.Kanep-S.Kukk, Tal
linn 2006, might well have been met by 1 1 . ..dxc4! 12 Nxe6 fxe6 13 bxc4 e5 14 Nc3
exd4 15 exd4 Nc5, thereby obtaining decent counterplay) 9 ... Nd7 10 Nf4!? (a slight
change of approach to exploit the bishop's absence from the h2-b8 diagonal)
10 ... Nf6 11 c4 Be6 12 Qb3! Qd7 13 cxd5 Bxd5 14 Nxd5 Nxd5 15 Racl Rfd8 Black
was solidly placed in M.Bezold-H.Reitz, Wuerzburg 1995, but the instructive ma
noeuvre 16 Nf3 Qe7 1 7 Ne1 ! Rac8 18 Nd3 (Diagram 8) left White with a pleasant
edge.
6 Bg2 d6
Consistent with Black's dark-square approach, although he might also take play
back into the waters of our last note with 6 ... d5.
44
2 ... g6 a nd M i nor Lines
7 N e2 f5 8 c4 (Diagram 9)

Black should try ... cs An harmonious white set-up
This move order looks very sensible and has been used by some strong players,
but I remain uncertain whether is it is more precise than beginning with 4 c4 fol
lowed by prioritizing central development.
8.. Nd7
.
Black manoeuvres his knight to f6, but 8 ... c5!? might well be more critical. Play
continues 9 Nbc3 cxd4 (9 ... Nc6 10 0-0! is another key line when I agree with De la
Villa that even 10 ... cxd4 1 1 exd4 f4!? fails to equalize after 12 Nxf4 Nxd4 13 Re1 ) 10
exd4 (Wells's 10 Nxd4!? Nc6 11 Nde2 deserves attention when the issue is
whether 1 l ...Be6 12 b3 aS! gives Black quite enough counterplay) 1 0... Nc6 11 Qd2
f4! 12 gxf4 Qh4 13 dS Ne7 and at this stage Black had decent compensation in
I.Sokolov-R.Har-Zvi, Wijk aan Zee 1993.
WARNING: Don't ignore this ... cs issue just because many black
players are unaware of itl As we have seen, White has a number of
ways to both dissuade and meet ...cs, and it's good to find a method
which you are happy with.
9 Nbc3 Nf6
Transpositions are rife at this point; Black is just as likely to begin with either 9 ...c6
or 9... Re8 before transferring his knight to f6. However, White should be a little
careful after 9 ... Re8 10 0-0 c6 when 1 1 Qd3 appears best, since 1 1 b4 permits
1 1 . . . Nb6!? 12 cS Nc4 when even the positional treatment 13 cxd6 Nxd6 14 a4 Be6
15 bS Rc8 16 bxc6 bxc6 didn't seem too bad for Black in V.Akopian-L.Yurtaev,
Calcutta 2000.
45
10 o-o (Diagram 10)

White has also delayed this in favour of accelerating his queenside approach.
That's an interesting idea, but White is likely to have to castle at some point and I
can't see too much wrong with the text.
10 Re8
...
Black usually flicks this in, but as ... Ne4 isn't obligatory, he might prefer 10 ... c6 1 1
b4 and then one of:
a) 1 1 ...h5!? wastes no time developing counterplay, although 12 Nf4 (White rightly
doesn't overly fear Black's counterplay; 12 h4 would hold up the advancing h
pawn, although perhaps an ambitious black player might try his luck here with
12 ... g5!? 13 hxg5 Ng4) 12 ... h4 13 b5 Bd7 14 Rb1 Rc8 15 bxc6 bxc6 (Diagram 11) 16
Rb7 (16 Qd3 also looks logical, but quite possibly White was hoping to provoke
Black into a committal ...g5 when he still had Nh5 available in response) 16 ... Rc7
1 7 Rb3 Re8 18 Qd3 Bc8 19 Ra3 Ng4 20 Rb1 left Black struggling to break through
on the kingside and White slightly for choice with Rb8 next up in T.Nalbandian
A.Pashikian, Yerevan 2006.

White seems to be ahead in the race Holding up ... Ne4 for a move
b) 1 l ...Be6 tries to tempt d5 out of White. I feel he does better to keep the a1-h8
diagonal closed and 12 Qd3 d5 13 c5 sees him racing ahead on the queenside.
11 Qd3 (Diagram 12)
White keeps the knight out of e4 for the time being, although it doesn't really
make too much difference whether he starts with 1 1 b4 and only then goes 12
Qd3, such a s after 1 l . ..Ne4.
After 1 1 Qd3 we will examine the typ i cal plans for both sides by studying 1 1 ...Qe7
1 2 b4 c6 1 3 b5 Ne4 in Game 5 and the more ambitious l l ... c6 1 2 b4 h5!? in Game 6.
46
2 . . . g6 a n d M i nor L i n es
co n clusion
z g6 is a fairly solid line, but it's probably a little too simplistic unless Black
...
closely investigates early ... c5 ideas. Our main line appears more pleasant for
White, although I was still surprised to discover that he has scored an impressive
65'Yo from the 58 games which I could find after 1 1 Qd3.
Illustrative Games
Game S
o M.Galyas G.Voiteanu
Buda pest 2006
1 d4 Nf6 2 BgS g6 3 Bxf6 exf6 4 e3 fS 5 g3 Bg7 6 Bg2 0-0 7 Ne2 d6 8 0-0 ReS 9 c4 c6
10 Nbc3 Nd7 11 b4 Nf6 12 Qd3 Qe7 13 bs (Diagram 13)
Diagram 1 3 (B) Diagram 14 (B)

Making good use of the g2-bishop Excellent play for the excha nge
1 3 ... Ne4
Bla ck plugs the long diagonal in a bid to reduce White's queenside pressure. He
has also been known to defend with 13 ... Bd7, but in both cases he lacks counter
play while White can play down the b-file.
14 a41?
W hite decides that he even has time to include his a-pawn in the attack. A more
popular choice has been 14 Rabl, which also looks good:
a) 14 ... Be6 relies on the rook's position on bl, but White can actually fal l headlong
47
into the trap: 1S bxc6 bxc6 16 Nxe4!? (16 Nf4 is also promising, but the transforma
tion generated by the text is a very tempting one) 16 ... fxe4 17 Bxe4 dS 18 cxdS cxdS
19 Bf3 BfS (this skewer is the point of Black's play) 20 Qd2 Bxb1 21 Rxb1 Rab8 22
Rxb8 Rxb8 23 BxdS (Diagram 14) gave White excellent compensation for the ex
change in A.Aleksandrov-E.Janev, Sas van Ghent 1992. Perhaps Black might now
have tried 23 ...Qb4! ?, but after 24 Qxb4 Rxb4 2S Bb3 aS 26 Nc3 I would still take
White: he has control of the position, two good pawns for the exchange and Black
faces a long defence, although he does have chances to make it a successful one.
TIP: When your opponent is reliant upon a long, forcing line to keep
his position together, always try to look a little further. There's often
a sting in the tail or the opponent may just, as here, have misjudged
the resulting position.
b) 14 ... hS is a more active try, albeit less dangerous than when played in our next
game: 1S h4 (this is definitely a case where White doesn't have to block; the 1S a4!?
h4 16 aS Bh6 1 7 a6 cxbS 18 NdS Qd8 19 cxbS of E.Torre-S.Safin, Aden 2002, being a
tempting alternative) 1S ...Bd7 (the extremely ambitious 1S ... Bh6 16 Nf4 Kh8 17 Bf3
gS? 18 hxgS QxgS failed to convince in K.Berg-J.Ibarra Padron, Skanderborg 200S,
and now White might even have gone in for 19 Nxe4!? fxe4 20 Bxe4 h4 21 bxc6 bxc6
22 Bxc6!) 16 bxc6 bxc6 1 7 Rb7 Nxc3 18 Nxc3 Rec8 19 Rfb1 left Black on the back foot
in K.Rusev-D.Dochev, Pleven 200S; Ne2 and Qa3 will increase the pressure.
14... a5
Black has no desire to allow aS-a6, but this will hardly keep the queenside closed.
I also like White after 14 ... Bd7 1S aS!? cxbS 16 cxbS, since his knight can come to
dS, and 14 ... gS!? 1S aS hS 16 a6 h4 1 7 axb7 Bxb7 1 8 RaS cxbS 19 RxbS also leaves
him in control of proceedings, M.Navarro Cia-J.Ibarra Jerez, La Algaba 2001 .
15 Rab1 h5 16 bxc6 bxc6 17 Rb6 Qc7?1
The solid 17 ... Ra6!? 18 Rb8 Qc7 19 Rfb1 Nf6 would have been a better defence,
keeping the entry squares down the b-file covered for the time being.
18 Rfb1 (Diagram 15) 18 ... Ba6
By now it was too late for 18 ... Ra6?! on account of 19 cS!; that queen really can be
pretty effectively placed on d3.
19 Nxe4
White might also have improved his pieces with 19 Nf4!?, but Galyas is happy to
transform the structure.
19 ...fxe4 20 Qc21
Remarkably this had all been seen before in S.Martinsen-C.Hoi, Kiel 2004. There
White won material with 20 Bxe4?! dS 21 Rxa6 dxe4 22 Rxa8 exd3 23 Rxe8+ Kh7 24
Nf4, but matters were actually far from clear after 24 ... c5!, activating the queen
with some effect.
48
2 .. g6 a n d M i n o r Lines
.

White has good pressure Spot the breakthrough !
2o...f5?
Too static. Black would have done better with 20 ... d5, even if 21 Nc3! would have
facilitated Bfl and enabled White to meet 2l...Bxc4 with 22 Nxe4, thereby retain
ing a pleasant pull.
21 Nf4 Kh7 22 c51 d5 23 Bf1
Black's dark-squared bishop might nominally be his 'good' bishop, but his struc
ture is horribly static and White is in complete control.
23 ... Bc8? (Diagram 16)
It's natural to want to keep b7 covered, but this allows a decisive blow. Voiteanu
had to exchange even if 23 ... Bxfl 24 Kxfl Ra7 25 Qb3 Re7 26 Rb8 would have left
White calling all the shots and with excellent winning chances.
24 Rxc61 Qxc6 2 5 Bb5
Now White regains his material with a crushing advantage.
NOTE: Even quiet manoeuvring positions are not completely devoid

of tactics. Always try to remain alert to tactical opportunities, espe
cially those which exploit loose pieces.
2S . ..Qe6
A desperate try as 25 ...Qc7 26 Bxe8 Ra6 27 Qb3 would have been curtains for
B lack.
2 6 Nxe6 Rxe6 27 Qb3 Rb8 28 Qxds h4 29 Qd8 1-0
49
Sta rt i n g O ut: The Trom powsky Attack
Game 6
D E.Mensch N.Giffard
French League 2004
1 d4 Nf6 2 Bg5 d6 3 Bxf6 exf6 4 e3 f5 5 g3 Nd7 6 Bg2 Nf6 7 Ne2 g6 8 c4 Bg7 9 Nbc3
o-o 10 o-o Re8 11 b4 c6 12 Qd3 hSI (Diagram 17)

Black begins his counterplay The f4-knight is a good defender
A decent attempt to acquire some counterplay and a better try than the 12 ... Qe7 of
our previous game.
13 h4
Once again I wonder if this is definitely best; it all depends on the strength of the
positionally risky but potentially dangerous ... gS-break. In M.Svensson-C.Holst,
Swedish Team Championship 2007, White preferred 13 a4 h4 14 bS hxg3 15 hxg3
Bd7 16 aS a6 1 7 bxc6 bxc6 (17... Bxc6!? would at least enable Black to weaken
White's kingside with a later trade of bishops, although 18 Bxc6 bxc6 19 Kg2 Ng4
20 Nf4 should keep matters under control there) 18 Rfb1 Ng4 19 Rb7 Bh6 20 Nf4!
(Diagram 18) when his alert knight helped to keep Black at bay on the kingside
and after 20 ... Rb8 perhaps 21 Rxb8 Qxb8 22 Rbl Qd8 23 Rb6 was possible, since
23 ... Bxf4 24 gxf4 Qh4 25 Qd2 Kg7 26 Rxa6 Rh8 doesn't actually threaten very
much at all.
NOTE: In this variation we really get to see the power of the fi

anchettoed bishop at work. Not only does White's key piece pres
sure the black queenside, but it is also an excellent defender and can
keep some rather powerful opposition pieces at bay.
50
2 .. g6 a n d M i n o r Lines
.
13 .. N g4 14 Nf4
.
A nn otating in his excellent 'd-pawn Attacks' column for the ChessPublishing web
site, E ric Prie points out that this wasn't yet necessary. Instead 14 bS! ? Bd7 15 Rfbl
g5 16 hxgS QxgS 17 bxc6 bxc6 18 Rb7 Bc8 19 Rc7 h4 20 Bxc6 hxg3 21 Nxg3 Qh4 is a
highly-instructive line given by the French Grandmaster. Black's attack looks
strong, but White still controls the key kingside light squares and 22 Ndl Rxe3!?
23 fxe3 Qxg3+ 24 Kfl Rb8 25 Rbl, for instance, appears to defend. It's never easy
not to panic when facing such an attack, but in general it does seem that Black of
ten doesn't have quite enough, scary though his sacrificial possibilities may ap
pear .
14 ... g51 15 hxgs Qxgs (Diagram 19)

Black has saved a tempo White's defences hold tight
And so the queen is developed to g5 in one go, without having to come via e7.
1 6 Rae1
White abandons his queenside ambitions for the time being, but perhaps he might
have got away with the more straightforward 16 Rfe1 ! ?, after which 16 ... h4 1 7 Nh3
Qh6 18 gxh4 Qxh4 19 Ne2 echoes the game and leaves f2 sufficiently well de
fended.
l6... h4 17 Nh3 Qh6 18 gxh4 Qxh4 19 Ne21
Mensch rushes pieces over to help the defence.
19 . Re6 20 Nef4 Rh6 21 bS Bd7 22 bxc6 bxc6
. .
Trading the bishops would only have helped White because of one downside to
... gS: 22 .. . Bxc6? 23 QxfS Bxg2 24 Kxg2 and White has won a pawn while keeping
his defence together.
51
23 Rb11 Kh8 (Diagram 20) 24 Rb7

White might have lost a tempo with his rook, but I wouldn't be surprised if he
was still better here.
NOTE: Knights are often excellent defenders and here White's

knight-pair are proving something of an obstacle to Black.
24...Nf6 25 Qa3?1
Active, but counter-attacking with the queen will prove rather risky. Instead a
decent case might have been made for trading a pair of rooks with 25 Rfb1 !, after
which 25 ... Rg8 26 Kfl Bf8 27 Rxa7 Qg4 28 Ke1 still leaves Black chasing that elu
sive tactic. Moreover, we should not forget that Black's own king might eventually
become weak should he never get anywhere himself on the kingside.
2S ... Rg8 26 Rxd7?
White panics and I'm not sure why. Like Prie, I haven't found anything especially
devastating after 26 Qxd6 Bf8 27 QeS when 27...Be6 28 Rcl Qg4 29 Kfl Bxc4+ 30
Kel BbS 31 Rb1 is extremely unclear; admittedly with his king on el White must
remain pretty vigilant, but Black too has his problems.
26 ... Nxd7 27 Qxa7? Besll (Diagram 21)

A stunn ing blow The attack is too strong
One can only assume that Mensch missed this amazing if logical move; Black
simply wants to remove the knight on f4.
28 dxes Rxg2+?
Black suffers a rush of blood to the head. He shouldn't have rejected the straight
forward 28 ... dxe5 29 Qxd7 exf4 when there's no defence, as 30 Qxf5 Qxh3 31 Qxh3
Rxh3 32 exf4 Rh4 merely leads to a hopeless ending for White.
52
2 . . . g6 a n d M i nor L i n es
2 9 N xg2?
Re turning the favour. A much better defence was 29 Kxg2 dxeS!? (29 ...Qg4+ 30
J( h 2 NxeS looks very strong_ but Prie points out that those knights keep every
th ing together after 31 f3 N xf3+ 32 Rxf3 Qxf3 33 Qd4+ Kh7 34 Qd2) 30 Qa8+ Kh7
( 30 . . . Kg7 31 Rgl is fine too for White) 31 Qe8! when Black has nothing better than
3 1 . . .Qg4+ 32 Kh2 exf4 33 Qxf7+ Kh8 34 Qe8+ Kh7, repeating moves.
29 Qxh3 30 f4 Ncsl (Diagram 22)
...
N ow Black is back on track and there's no hope of salvation for White's ex

tremely-vulnerable king.
31 QaB+
Alternatively, 31 Qxf7 Qh1+ 32 Kf2 Nd3+ 33 Ke2 Qxg2+ 34 Kxd3 Qe4+! and the
king will eventually be hunted down.
31 Kh7 32 QfB Qh2+ 33 Kf2 Nd3+ 34 Ke2 Qxg2+ 35 Kxd3 Qxf1+
...
35 . . .Qe4+ would have been more precise, but the text is plenty good enough.
3 6 Kc3 Qc1+ 37 Kd3 Qb1+ 38 Kc3 Qb7 39 exd6 Rh2 40 Kd3 cS 41 Qe7 Qb1+ 0-1
53
C h a pter Th ree
T h e Pos it i o n a l C h o i ce : 2 e6
m I nt rod uction
m W h ite Pl ays 5 Nc3
m Wh ite Pl ays 5 c3
m The Torre- l i ke 3 e 3
The Position a l Choice: 2 . e 6
. .
1 n trod uction
1 d4 Nf6 2 Bg5 e6 (Diagram 1)
Diagram 1 (W} Diagram 2 (B)

A respectable defence Spot Black's little tactic!
A sensible approach and the only alternative to moving the knight if Black wishes
to avoid the doubling of his pawns. This positional approach is Black's second
most popular response to the Trompowsky, occurring in 23% of all the games in
my database, and was recommended in John Cox's popular 2005 work Dealing
with d4 Deviations.
3 e4
White takes up the challenge. This is critical and by far his most popular response,
although it's also possible to again aim for a kind of improved Torre with either 3
e3 or 3 Nd2, as we'll see in the final section of this chapter.
3 ... h61
Bes t. Black obtains the long-term advantage of the bishop-pair. There are other
possibilities, though, of which 'a' is pretty popular at lower levels and 'd' quite a
creative defence:
a) The solid 3 .. Be7 does little to combat White's spatial advantage, but does set a
.
l ittl e trap.
WARNING: One might be forgiven for thinking that 4 Nc3 (Diagram

2) will lead to the French Defence after 4...d5, but Black also has
4... Nxe41 5 Bxe7 Nxc3 6 Bxd8 Nxd1 7 Bxc7 Nxb2 8 Bd6 Na4 when
White has decent compensation for the missing pawn, but no more
than that.
55
Thus White might prefer 4 Bd3 when 4 ... c5 (4...d5 5 e5 Nfd7 6 Bxe7 Qxe7 7 c3 c5 8
Ne2 gives White a good version of the French Tarrasch, whereas here 6 Be3 c5 7 c3
Nc6 8 Nf3 Qb6 doesn't seem so clear - compare with variation 'b') 5 dxc5! is an
aggressive idea to make use of White's active set-up. Play might continue 5 ...Qa5+
6 Qd2 (by no means forced: 6 Nc3 Nxe4!? 7 Bxe7 Nxc3 8 Qd2 Kxe7 9 Ne2 Qxc5 10
Nxc3 offers White decent value for his pawn, and 6 Nd2!? Qxc5 7 Be3! might be
considered by those with some Open Sicilian experience) 6 ... Qxc5 7 Nf3 d6 (Dia
gram 3) 8 c4!? Nc6 9 Nc3 a6 10 Rcl Bd7 11 0-0, which gave White a small edge
thanks to his cramping bind in M.Nikolov-P.Masouros, Aghios Kirykos 2008.

Sicilian-like play White should play a Ia Steinitz
NOTE: In this variation we've reached positions quite similar to both

the French and the Open Sicilian. Some experience of those openings
can't do any harm, but is far from essential; White has quite a pleas
ant version of both openings and hasn't directly transposed to any
thing horribly theoretical.
Finally, we should observe that those who wish to force play into French lines
should give serious consideration to the closely-related 4 Nd2!?; the point being
that 4 . . . c5?! 5 e5! Nd5 6 Bxe7 Qxe7 7 Ne4 is pretty dangerous, G.Kasparov
D.Doukas, Patras (simul) 2001 .
b) 3... d5 4 e5 h6 5 Be3 Nfd7 6 f4 c5 7 c3 is another position which rather resembles
the French, although it cannot arise from a 1 e4 move order. Here Black has had to
spend a tempo on ...h6, which may make it harder for him to break with ... f6. Play
might continue 7... Nc6 (7...cxd4 8 cxd4 Nc6 is similar, but enables White to develop
his knight to c3 and 9 Nf3 Qb6 10 Qd2 Bb4?! 1 1 Nc3 Qa5 12 Rcl Nb6 13 a3 Bxc3 14
Rxc3 Na4 15 b4! Qd8 16 Rcl left him in control in P.Wells-R.Auschkalnis, Bad
56
T h e Posit io n a l C h o i c e : 2 ... e6
zwes ten 2005) 8 Nf3 Qb6!? (more positional is 8...Be7 9 Bd3 cxd4 10 cxd4 Nb4 1 1 Be2
N b6! ?, which restricted White to an edge after 1 2 a3 Nc6 13 Nc3 Nc4 14 Bel ! ? Bd7 15
b3 N b6 1 6 0-0 in M.Adams-V.Ivanchuk, Moscow (blitz) 2007) 9 b3!? (this weakens
the dark
squares and provokes a sharpening of the struggle; 9 Qd2 might be more
r u den t, leaving the issue of the queen's knight u ntil later) 9 ... cxd4 10 cxd4 Bb4+
p
(Diagra m 4) 1 1 Kf2 (essential by this point, but White's king often goes to f2 in the
relate d variation 1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 5 f4) 1 1...0-0 12 a3 Be7 13 Nc3
(the king mustn't show too much ambition: 13 Bd3 f6 14 Kg3? fxe5 15 fxe5 Ncxe5! is
on e to avoid) 13 ...Qd8 (sharper and more principled would have been 13 ... f6!? when
1 4 N a4!? Qd8 15 Bd3 fxe5 16 fxe5 Ndxe5!? 17 dxe5 Nxe5 18 Kg1 Nxf3+ 19 gxf3 gives
B lack some typical compensation, but not quite enough for his piece) 14 Bd3 f6 and
now in R.Palliser-l.Lewyk, York 2005, 15 Ke2! would have left White's Steinitzian
king quite safe in the centre and Black fairly cramped.
c) 3 ... d6?! 4 Nc3 Be7 fails to contest the centre, but is another line far from un
know n at club level. In P.Cech-W.Uhlmann, German League 2007, White showed
littl e respect for his legendary opponent: 5 Nf3 (5 f4! ? h6 6 Bxf6 Bxf6 7 Nf3 is
rather tempting too) 5 ... h6 (White also attacked after 5 ...0-0 6 Bd3 c5 7 dxc5 dxc5 8
h 4 ! Nc6 9 e5! in V.Baklan-O.Dolzhikova, Schwaebisch Gmuend 2000; note Black's
lack of counterplay and the Greek Gift possibilities) 6 Be3 b6 7 Bb5+! (disrupting
Black's development as the c8-bishop would prefer not to go to d7) 7 ... c6 8 Bd3
Ba 6 9 Bxa6 Nxa6 10 Qe2 Nc7 1 1 0-0-0 Nd7 12 h4! Qc8 13 h5 Qb7 and having seized
some useful kingside space, White decided it was time to press ahead in the centre
with 14 d5!?.
d) 3 ... c5!? (Diagram 5)

Sharpening the struggle Uncompromising stuff!
is a provocative idea which I both played and explored in an article for CHESS a
few years back. The critical response runs 4 e5 (4 d5 d6 5 Nc3 Be7 is possible but
57
Sta rt i ng Out: The Trom powsky Attack
pretty solid for Black, who was certainly fine after 6 Bb5+ Bd7 7 Bxd7+ Nbxd7 8
dxe6 fxe6 9 Nf3 0-0 in R.Pert-A.Collinson, British League 2002, and the unchal
lenging 4 c3 cxd4 5 cxd4 Bb4+! is even more pleasant for him) 4 ... h6 5 Bel! .
NOTE: It may seem strange to retreat the bishop back to base, but
this does keep it away from the clutches of the black knight and also
keeps b2 usefully protected in the event of any ...Qb6 ideas.
Black now has two options:
a) 5 ... Nd5 6 c4 (Diagram 6) refuses to let Black settle and gives White good chances
of emerging with the initiative:
a1) 6 ... Nc7?! 7 dxc5! (White opens the position as he enjoys smooth development
and wants to target g7) 7... Bxc5 (7... Nc6 8 Nc3 Nxe5 9 Bf4 Ng6 10 Bd6 Na6 1 1 Bd3!
was promising too in R.Eames-T.Abergel, British League 2003) 8 Qg4 (Diagram 7)
8 ... Kf8 9 Nf3 d5 10 Be2 Nc6 1 1 0-0 Qe7 12 Rd1 dxc4 13 Bxc4 gave White a pleasant
edge in B.Macieja-R.Wojtaszek, Polish Championship, Lublin 2008.

It's not so easy to cover g7 Black's king is a concern
a2) 6 ... Ne7 7 dxc5 Nbc6 (G.Sargissian-M.Sorokin, Ubeda 2000) prevents such Qg4
ideas, but Wells's suggestion of 8 Nf3!? Ng6 9 Be3 Ngxe5 10 Nxe5 Nxe5 1 1 Nc3
appears to give White the edge.
a3) 6... Nb4!? is the most ambitious try and quickly leads to quite sharp play, as
we'll see in Game 7.
a4) 6. . . Nb6 7 dxc5!? (again the most challenging approach; 7 d5 d6! breaks up
White's centre with easy play for Black) 7 ... Bxc5 8 Qg4 g6!? (8 ... Bf8 avoids creating
a weakness, but fails to equalize after 9 Nf3 Nc6 10 Nc3 d6 1 1 Bf4) 9 Nc3 Nc6 1 0
Nf3 d 6 (Diagram 8 ) might be Black's best try, since White's queen isn't ideally
placed, although 1 1 exd6 (1 1 Be2!? dxe5 12 0-0 is another gambit which might be
58
The Positio n a l Choice: 2 . e6
. .
investigated) l l ... Bxd6 1 2 cS! BxcS 13 BbS gave White decent play for his pawn
d u e to Black's kingside weaknesses in P.Wells-Z.Efimenko, Hastings 2003/04.
b) s . Nh7!? is very modem, countering one strange-looking retreat with another.
..
We now have a final divide:

b 1) 6 Bd3?! is a radical way to deal with the h7-knight, but doesn't fully convince:
6. . . cx d4 7 Nf3 (7 Bxh7 QaS+ 8 Nd2 Rxh7 9 Ngf3 Nc6 is also good for Black) 7... Ng5!
reveals Black's main strategic idea, after which 8 0-0 Nc6 9 Re1 d6 10 Nbd2 Bd7 1 1
a 3 Qc7 12 exd6 Bxd6 left him somewhat for choice i n T.Clarke-V.Bologan, Euro
pean Club Cup, Rethymnon 2003.
b2) 6 dS has been White's most popular response and 6 ...exd5 (or 6 ... d6 7 Nc3 exdS
8 QxdS Nc6 9 BbS dxeS 10 QxeS+ Qe7 1 1 Bxc6+ bxc6 1 2 Bf4 with a clear edge,
M.Cebalo-G.Franchini, Reggio Emilia 2006) 7 QxdS Nc6 8 Nf3 (preferable to 8
Bc4?! Qe7 9 Nf3 Nf6!) 8 ...Qe7 9 Be3! (Diagram 9) certainly looks quite dangerous,
even if 9 . .f6!? (9 ... d6 is more solid, but fails to equalize after 10 BbS Bd7 1 1 exd6
.
Qxd6 12 Nc3 Nf6 13 Qxd6 Bxd6 14 0-0-0 Be7 1 5 Rhe1 0-0-0 16 Na4!) 10 Nc3 (far
from essential and both 10 Bd3 and 10 Nh4!? NxeS 1 1 Nc3 deserve closer scrutiny)
lO . . . fxeS 1 1 Bd3 N f6 12 Bg6+ Kd8 13 Qc4 Nd4 might be okay for Black thanks to
his strong centre.
Diagram 9 {B) Diagram 10 {B)

Overly-a rtificial play? More French-like play
b3) 6 dxcS!? is a less-explored idea, although then 6 ... Nc6 7 Nf3 (7 Be3!? NxeS 8
N c3 is an attempt to speed up White's development) 7... Ng5 8 NxgS hxgS 9 Be3 b6
10 cxb6 axb6 1 1 Nc3 Bb4 gave Black sufficient counterplay in A.Smirnov
V.Iordachescu, Russian Team Championship 2008.
After that long but important digression, we return to Black's main move, 3 ...h6,
ex p loiting the undefended e-pawn:
4 Bxf6 Qxf6 5 Nf3
59
Sta rti n g Out: T h e Trom pows ky Attack
White's old choice, reaching a position which often comes about via the Torre: 1 d4
Nf6 2 Nf3 e6 3 Bg5 h6! 4 Bxf6 (4 Bh4 d6 followed by ...g5 is fine for Black) 4 ... Qxf6 5
e4. Nowadays White usually prefers to keep his f-pawn free and so selects either 5
Nc3 or 5 c3, the respective subjects of our next two sections. The only other move is
5 Qd2, angling to avoid the variation 5 Nc3 Bb4, but I suspect that such avoidance
only makes things worse if Black responds actively with 5 ...c5! 6 e5 Qd8.
s d6
...
By far Black's most popular choice, although with White's f-pawn blocked 5 ... d5 is
far from illogical. Following 6 e5 (the less committal 6 Nbd2 is probably a better
try) 6... Qd8 7 Bd3 (I have seen the position after 7 c4!? dxc4 8 Bxc4 Nd7 9 Nc3 Nb6
10 Bb3 Bd7! 1 1 0-0 Bc6, as in J.Hodgson-M.Lodhi, London 1987, assessed as fa
vouring White, but Black shouldn't be more than a touch worse with his light
squared bishop well placed on c6) 7 ... c5 8 c3 Nc6 9 a3 cxd4 10 cxd4 (Diagram 10)
the position resembled an Advance French in K.Chernyshov-T.Abergel, Cappelle
Ia Grande 2006. Exactly who this favours is not clear: Black's kingside has been
slightly weakened and it's tempting for White to aim for a Bc2 and Qd3 battery,
but then again White may miss his dark-squared bishop in any kingside attack.
6 Nc3
Via the Torre move order White often hasn't been so keen on this aggressive
move, but quieter choices also fail to trouble Black: for example, 6 Bd3 Nd7 (hold
ing up ...e5 before fianchettoing the dark-squared bishop) 7 Nbd2 g5! 8 c3 g4 9
Ng1 h5 1 0 h3 Bh6! (Diagram 11) 1 1 hxg4?! (not liking Black's initiative, White pre
pares an overly-aggressive counterstrike, but he would have done better to house
his king with 1 1 Qe2 a6 12 0-0-0) 1 1 ... hxg4 12 e5?! dxe5 1 3 Ne4 Qg7 was superb for
Black, who enjoyed his usual dark-square pressure as well as an extra pawn in
H.Temmerman-S.Vanderwaeren, Ghent 2003.

Wh ite has lost the i nitiative Black is fine
60
The Pos i t i o n a l Cho ice: 2 ... e6
NOTE: White might have a small lead in development after 5 Nf3,

but it's extremely hard to do anything with it, which is why grand
master interest has rather waned in this variation.
It w o uld be ideal to open the centre with d5, but that is usually well met by ...e5,
or w ith e5, which in tum can often be countered by ... d5. Meanwhile it's not all
th a t difficult for Black to advance and develop on one wing or the other, just as
w e'll see in our main line.
Bearing these problems in mind, one can see why it would be nice to make 6 e5!?
w ork, but unfortunately 6 ... dxe5 7 dxe5 Qe7 8 Nc3 Nc6 appears fine for Black too:
9 Bb5, Burgess's suggestion from The Gambit Guide to the Torre Attack, can be met
by just 9... Bd7 (Diagram 12) followed by ...0-0-0 and then either ... Qc5 or ... g5, and
9 Bd3 g5! 10 Qe2 Bg7 1 1 Bb5 (saving the e-pawn, but Black cannot complain about
acq uiring a second bishop) 1 l ...Bd7 1 2 Bxc6 Bxc6 13 0-0-0 Qb4! gave Black the
edge due to his pressure against e5 in K.Shirazi-L.Christiansen, US Champion
ship, Berkeley 1984.
6 Nd7 (Diagram 13)
...

Black may play for ... cs Both sides press forwards
Bl a ck's most flexible and probably best approach, keeping the e5-square closely
monitored. Karpov once preferred 6... g6 against Korchnoi, but this may allow
White to better time 7 e5!?, as well as to consider the more usual 7 Qd2 Bg7 8
0-0-0. Surprisingly Karpov has also dabbled in 6 ... g5!? at this stage, which may
Wel l be taking the whole provocation angle a little too far. In K.Chernyshov
B. Thorfinnsson, Budapest 2005, White handled things calmly and the instructive
m a noeuvre 7 eS Qe7 8 h3!? (rather than Timman's 8 BbS+) 8 . Bg7 9 Qe2 a6 1 0 0-0-0
..
dS 11 Qe3 bS?! 12 Ne2! Nd7 13 Ng3 cS 14 NhS saw him seizing the upper hand.
61
7 Qd2
Here too White has tried 7 Bd3, but 7...Qd8 (moving the queen away from a poten
tial e5-advance while facilitating the development of the bishop to e7 is often a
good plan in this variation) 8 Qd2 a6 9 0-0-0 Be7 1 0 h4 b5 1 1 g4 c5! gave Black
good counterplay in A.Stefanova-M.Gurevich, Antwerp 1997.
7 ...a6
WARNING! Black should not consider this variation totally harmless:

7 ...c6 8 o-o-o es?l 9 dxesl dxes? 10 Nbsll, as in J.Kiinger-D.King,
Lucerne 1989, is a trap which continues to claim vidims.
8 0-0-0 Qd8
Any reader unconvinced that White can't put his better development to good use
should pause at this point and try to find an acceptable plan. It isn't so easy and in
practice White has usually resorted to pushing his kingside pawns.
9 h4 bs (Diagram 14)
Black too is allowed to mobilize his own potential attackers.
10 Bd3
Hodgson's choice. White later attempted to improve with 10 Rh3!? in C.Crouch
M.Adams, European Club Cup, Slough 1 996, although after 10 ... Bb7 1 1 Qf4 b4 12
Ne2 c5! 13 dxc5! (White must open the position, since 13 d5?! e5! 14 Qe3 Qa5 15
Kb1 c4 would have already left Black with a pretty useful attack) 13 ... Nxc5 14 e5
Black had a number of reasonable options, including 14 ... Bd5!?.
Once again 10 d5 e5 doesn't especially help White either, but 10 e5! ? might be his
best chance to pose at least a few problems, as we'll see in Game 8.
10... Bb7 11 Kb1 cs 12 dxcs Nxcs
We've followed the game J.Hodgson-A.Miles, Kuala Lumpur 1992; two highly
creative Trompowsky users in opposition here. Unfortunately for Hodgson Black
had managed to obtain a pretty comfortable version of the Sicilian, as was borne
out by the further 13 Qe3 Qc7 14 Nd4 0-0-0! (Diagram 15) 15 f4 Kb8. It's notable
that it didn't take Hodgson long after this encounter to switch to 5 Nc3 (and later
to 5 c3).
Conclusion
I've covered Black's 3rd-move alternatives to 3... h6 in fair detail because they crop
up quite often at sub-international levels. Only 3 ... c5! ? should really concern
White, though, when it's important to be prepared and so avoid joining the fairly
large ranks of those who have been caught by surprise.
Bagging the bishop-pair with 3 . .. h6 is a classical and fairly reasonable response.
62
The Posit i o n a l Cho ice: 2 . . . e 6
f-{e re too White must have some ideas and these days 5 Nf3 just looks pretty
ss. Indeed, I'm not surprised that White has scored a little under 50% here;
too th le
he rnu st undertake more active measures to counter Black's long-term advantage
of th e tw o bishops.
Diagram 15 {W) Diagram 16 (B)

An acceptable Sicilian for Black Can White maintain control?
Illustrative Games
Ga me 7
D R.Palliser P.Taylor
Golders G reen Rapid play 2004
1 d4 Nf6 2 Bg5 e6 3 e4 c5 4 e5 h6 5 Bel Nd5 6 c4 Nb41? 7 a3

The best move order if White plans to exchange on c5.
WARNING: I later preferred the careless 7 dxc5, only realizing during

my opponent's subsequent think that 7 .. Bxc5 8 a3? is not the way
.
to go on account of 8 .Qh41 9 Qd2 Be31? 10 Qe2 Bxf2+ when White

..
is struggling, albeit perhaps not quite as much as he deserves to be

after 11 Kd11 Bxg1 12 axb4 o-o 13 Nd2.
Even more fortuitously 7 ... N8c6? appeared on the board in R.Palliser
M.Carpenter, York (rapid) 2008, and after 8 a3! Na6 9 b4 Nxe5 10 Bb2 Qc7 1 1 Nc3
N g6 12 NbS Black found himself horribly bound.
Pl ay ing to hit g7 is not the only plan, though: 7 dS!? being a worthy alternative.
Then 7... d6 (7 ... exd5 8 cxd5 already looks a little precarious for Black and 8 .. Qe7? .
63
9 f4 d6 10 a3 N4a6 11 Nf3 Bg4 1 2 Bb5+ saw his Pseudo-Benoni brutally annihilated

in B.Savage-J.Bick, Las Vegas 2006) 8 a3 N4a6 9 Bf4! (Diagram 16) may well retain
an edge: 9... dxe5 10 Bxe5 Nd7 1 1 Qa4!? ( 1 1 Bg3 is a simpler, respectable alterna
tive) 1 1 . .. Qg5?! (Black faces some difficulties in any case, but this tempting idea is
probably too ambitious) 12 f4 Qg6 13 Nf3 Be7 was quite complex in M.Karttunen
O.Sepp, Finnish League 2004, but I suspect that 14 d6!? Bf6 15 Nbd2 favours
White.
7 ... N4c6 8 dxcs Bxcs
The critical move, but not the only approach. Black might also try his chances in
the extremely murky structure which 8.. . Nxe5!? 9 Be3 Na6 10 b4 b6 gives rise to.
9 Qg41 (Diagram 17)

Again g7 is pinpointed It's getting messy...
Ambitious, but also quite critical and far from easy to deal with at a fast time limit.
I no longer agree with all my earlier (2005) analysis for CHESS, but do still feel
that 9 Nf3 Qc7 should be comfortable for Black and that 10 b4!? Nxe5 1 1 Bb2 Bd6!
12 Nbd2 0-0 13 c5 Nxf3+ 14 Nxf3 Be7 15 Be5 Qd8 doesn't offer White quite enough
for his pawn.
9 . Qb6?
. .
I'm not a huge fan of analyzing with an engine, but must admit that they can be
pretty helpful at times! Previously I had quite liked Taylor's choice, but now I
suspect that Black does much better with 9 .. Qa5+!, which is quite a disruptive
.
check. Nevertheless, this is far from being all doom and gloom for White, not least
because 10 Nc3! Bd4 1 1 Bd2 Qxe5+ 12 Be2 offers a decent lead in development in
return for the pawn.
64
The Posit iona l C h oice: 2 . . . e6
NOTE: It's far from uncommon for White to decide on an ambitious

course of action in the Trompowsky, and when he does so he must
often be prepared to sacrifice a pawn or two to mainta in the initia
tive.
Here an early Qg4 raid is risky with Black ahead in development, but White still
obtain s decent play for his pawn thanks to his safer king and Black's slightly
ja m med-up queenside.
10 Qxg7 Bxf2+ 11 Kd1 Qd4+ (Diagram 18)
The alternative was 1 l ...Rf8 12 Bxh6 Qb3+, but after 13 Kcl Be3+ 14 Nd2 Bxd2+
(14. . Bxh6? 15 Qxh6 Qb6 16 Ne4! pinpoints Black's serious dark-square weak
.
nesses) 15 Kxd2! Qxb2+ 16 Ke3 Black must lose the exchange for clearly insuffi
cien t compensation.
12 Ke21
This required some calculation and good nerves, but I was determined to avoid
block ing my key bishop with 12 Nd2 unless I absolutely had to.
12 ... Bxg1?
Taylor plays for the attack, but this is far too ambitious. He had to save the rook,
even if the accurate sequence 12 ... Rf8 13 Nc3! (13 Bxh6?! Qxb2+ 14 Nd2 Bc5 looks
more dangerous for White than Black) 13 ... Bxg1 14 Rxg1! Qxe5+! (14 ...Qxg1? 15
Bxh6 Qc5 16 Ne4 Qe7 1 7 Rd 1 i s crushing) 1 5 Qxe5 Nxe5 16 Bxh6 Rg8 1 7 Ne4
would have left White with a strong initiative.
13 Qxh8+ Ke7 14 Qf6+ Ke8 15 Rxg11
The simplest. Black had no doubt considered this, but presumably missed the
sting in the tail down the h-file.
1 5 . Qxg1 16 Qh8+ Ke7 17 Qf6+ Ke8 18 Be3 Qxh2
. .
Now the backwards pin costs Black his queen, but 18 ...Qh1 19 Qh8+ Ke7 20 Bxh6
w ould also have been completely crushing.
19 Qh8+ Ke7 20 Bg5+1 (Diagram 19) 1-0
Game B
0 M.Popovic M.Andersen
Rey kjavi k 2009
1 d 4 Nf6 2 Bgs e6 3 e4 h6 4 Bxf6 Qxf6 5 Nf3 d6 6 Nc3 Nd7 7 Qd2 Qd8 8 h4 a6 9

O-o-o bs 10 es d5 (Diagram 20) 11 Rh31?
The sa me idea which we saw Crouch try against Adams, but here the centre is
cl osed. In this French-like position transferring the rook to generate potential
P ress ure against g7 make a fair amount of sense. White had previously employed
another typical French manoeuvre with 1 1 Ne2 and after 1 1 . . .Nb6!? (Black hurries
65
Sta rti n g O u t : The Trom pows ky Attack
his knight towards c4, but 1 1...c5 i s also quite possible, a s pointed out b y Cox, af
ter which White might perhaps go 12 Rh3 anyway or occupy the d4-square with
12 dxcS BxcS 13 Ned4) 12 Nf4 Nc4 13 Bxc4 bxc4! (Black is alert to the danger; with
White so far ahead in development he would like to open the position for his
knight-pair and 13 ... dxc4?! 14 dS! would have given him just such an opportunity)
14 NhS Qd7 15 Qc3 Rb8 1 6 a3! (good defence against ... Qa4 and ... Bb4 ideas)
16 ...Qb5 1 7 Rh3 a complex, manoeuvring struggle was well underway in
O.Annageldyev-I.Rogers, Istanbul Olympiad 2000.

A backwards pin! Another French centre
11 C5...
Black plays in true French-style, but I suspect that the Rogersesque 1 1 ...Nb6!?
would have been stronger. Then 12 Rg3 Nc4 13 Bxc4 bxc4 should probably be fol
lowed up by 14 Ne2!, again aiming to apply some pressure against g7.
12 dxcs Nxcs
Black might also have gone in for 1 2 ... Bxc5!? 13 Rg3 Kf8, since it's not then so easy
for White to increase the kingside pressure.
13 Qf4 b4 14 Ne2 Qc7
White is happy to redeploy his queen's knight, but at least this way Black is able
to develop his queenside in comfort.
15 Kb1 Bb7 16 Nfd4 Be7 17 Qe31 (Diagram 21)
Thus White has made good use of the d4-square and now he introduces another
typical French motif, the advance of the f-pawn.
17 0-0-0
...
Andersen's play strikes me as being just a little too compliant in this game. One
66
T h e Posit i o n a l C h o i ce: 2 . . . e 6
can u nderstand his desire to avoid 17. . .0-0 18 g4!, but I'm b y n o means convinced
th at the active 1 7... a5!? was such a bad approach: for example, 18 NbS Qb6 19 Nf4
(W hite wants to generate some kingside pressure) 19 . . a4 20 Rg3 gS! 21 NhS b3
.
gives Black decent-enough counterplay.

1s Rc1! Kb8 19 C3
Now White is control and enjoys an edge thanks to his grip on d4 and slightly
safer king, although Black remains pretty solid with his own knight also control
li ng some key squares.
19 bxc3 20 Nxc3 Qb67
.
..
N a tural enough, but this leads to tactical problems unless Black wants to part with
hi s f-pawn. Somewhat better was 20...Rc8, intending 21 Na4 (21 Qf4? is now rather
well met by 2l...f6) 21...Qa5, which seems to keep everything together.
21 Qf4 Rhf8 22 Na41
It may seem strange to activate Black's bishop thus, but Popovic has realized that
Bl a ck's queenside situation is more than a little shaky.
22 Qa7 23 Nxcs Bxcs (Diagram 22) 24 Nc6+7
...

Thematic control of d4 How best to attack?
Flashy and it appears that White overestimated his attack. However, he wouldn't
have been at all guilty of doing that had he found 24 Bxa6!, leaving Black pretty
rnu ch defenceless: 24 ... Bxd4 (or 24 ... Qxa6 25 RxcS and both rooks will swing into
the attack) 25 Bxb7 Kxb7 26 Ra3 and the black queen is a goner.
24 Bxc6 25 Rhc3 Bb67
...
B l a ck preserves the wrong bishop. Correct was 25 ... Bxf2 26 Rxc6 (or 26 Rlc2!?
Qd4! 27 Rxf2 Qxf4 28 Rxf4 Bb7 29 Ra3 f6 and Black should hold the endgame)
67
26... Qd4 27 Qf3 Qe4+, forcing the queens off and leaving everything still to play
for.
26 Rxc6 Qb7?
Completely collapsing. Black had to try 26 ... Ka8 27 Rlc3 Rb8, which Andersen
may have thought failed to 28 Bxa6? Qxa6 29 Ra3?, missing the rather large
counter-tactic 29 ... Qxa3! 30 bxa3 Be3+.
WARNING: It's amazing how often serious mistakes go together in

pairs. Thus if you susped that you have just erred, do be extremely
vigilant over the next few moves!
27 Qa4 Ka7 (Diagram 23)

It's a l l over The centre vs. the bishop-pair
28 R1c3
Rather powerful too would have been 28 Rxb6 Qxb6 29 Rc6.
28 ... Rb8 29 Rc71 Bxc7 30 Rxc7 Kb6 31 Rxb7+ Kxb7 32 Qxa6+ Kc7 33 BbS 1-0
Wh ite Plays 5 Nc3

1 d 4 Nf6 2 BgS e6 3 e4 h 6 4 Bxf6 Qxf6 5 Nc3 (Diagram 24)
White refuses to obstruct his f-pawn and this aggressive approach has become his
most popular choice.
s Bb4
...
There's something very consistent to my mind about this move which has certain
68
The Pos i t i o n a l Cho ice: 2 . . e6
.
Ni rnzo-Indian echoes. Black can also play as in our last section with 5 ... d6, which
portant alternative, and after 6 Qd2 we have:
is an im
a) 6 .. . a6 is less convincing than when played in the early Nf3 lines and 7 f4! Nd7
(fa r rn ore radical is 7 ... g5!?, but 8 fS! exfS?! 9 NdS Qd8 10 exfS BxfS 1 1 Qf2 gave
W hite great play for his pawn in V.Akopian-E.Iljushkin, Novgorod 1999, and even
the su perior 8... Bg7 9 0-0-0 exfS leaves White with good compensation after
Wells's 10 exfS BxfS 1 1 Bd3!) 8 Nf3 Qd8 (the queen scurries out of danger; 8... g6 9
o-0-0 Bg7 is now fairly risky, since White can consider such direct ideas as 1 0 eS!
Qe7 11 h4 hS 12 dS!?) 9 0-0-0 cS? (Black fails to grasp just how helpful having ad
va nced the f-pawn is to White; he would have done much better to have devel
oped with 9...Be7) 10 dS eS (desperately trying to keep lines closed, whereas
10 .. . Nf6? 1 1 dxe6 Bxe6 1 2 fS! Bd7 13 eS dxeS 14 NxeS Be7 15 Bc4 would have led to
an early bath) 1 1 fxeS NxeS 12 NxeS dxeS 13 d6! (Diagram 25) 13 ... Be6 14 Bc4! saw
the cramping d-pawn leave Black in serious trouble in R.Palliser-R.Britton, York
2004.

The d-pawn is rather awkward Black prepares ...es
NOTE: White's three abreast pawn centre offers him a number of

aggressive ideas, as well as keeping Black cramped. Do keep an eye
out for breakthroughs with all of the e5-, f5- and d5-advances.
b) Thus I suspect that 6 ... c6 (Diagram 26) is a better try and this has, indeed, been
advocated by Cox. Now 7 f4 eS (7.. . g5 8 fS! exfS 9 0-0-0 again gives White a strong
initiative; note how this sacrifice opens lines to the detriment of the black king) 8
dxeS dxeS 9 fS has been White's main try in practice, but while he is cramped
Bl ack' s position is quite solid and 9 ...Bb4! (Diagram 27) has become established as
the best way for him to regroup his forces: ... Nd7, ...Qe7, ... Nf6, ... Bd7 and ... 0-0-0
69
i s the plan. The game L.McShane-S.Gordon, E U Championship, Liverpool 2006,

continued 10 Nf3 Nd7 1 1 a3 BaS 12 Bc4 (the pin is annoying and, indeed, no lesser
player than Mamedyarov has tried 1 2 Bd3 Qe7 13 b4!? Bb6 14 Na4 in a rapidplay
game; an approach which deserves further attention) 12 ...Qe7 13 0-0-0 Nf6 14 Rhet
Bd7 and already it was clear that if anyone was better it was Black as he had a
clear plan: queenside expansion.

It's roughly balanced It's all a bout the dark squares!
Instead White might make a further positional concession to develop with 9 fxeS!?
QxeS 10 Nf3, but once again 10 ... Qa5! 1 1 Bc4 Bb4 is quite an annoying pin. I've
spent some time here over the years looking for a way to exploit White's superior
development, but am yet to find it and 12 0-0 0-0 13 Kh1 Nd7 14 Rad1 led to noth
ing more than a fairly complex middlegame in M.Bezold-R.Dive, Wrexham 1 994.
c) Bearing in mind the problems posed by White's f-pawn, perhaps it should not
come as a surprise that 6 ... g5! ? (Diagram 28) has been a popular choice in practice,
at least at higher levels. White must then decide how aggressive a mood he is in:
cl ) 7 Bc4 Nc6 8 Nge2 Bg7 9 Rd1 sees White developing sensibly. This does keep
Black a little cramped and tends to give rise to a full-blooded battle, as it did in no
lesser game than V.Anand-A.Karpov, FIDE World Championship, Lausanne 1 998:
9 ... Bd7 10 0-0 0-0-0! (Black goes this way because White is happy to prise the king
side open with f4) 1 1 NbS! a6 12 Na3 g4 (more recently 12 ...Qg6!? posed some
tricky central questions and after 13 f3 hS 14 Qe3 g4 15 f4 dS!? 16 fS Bh6 Black ob
tained decent counterplay in N.Vitiugov-B.Savchenko, Bornholm 2008) 13 f4 gxf3
14 Rxf3 Qe7 15 c3 hS 16 Rdfl Rdf8 1 7 b4 Na7 18 Nc2 Bh6 and both sides had their
chances in this unbalanced middlegame.
c2) A more restrained approach is 7 g3!? Bg7 (it's notable that 7 g3 doesn't commit
White to quiet play; he might just use the move as preparation for f4 and 7... Bd7 8
70
The Positiona l C h oi ce : 2 .. e6
.
o-0-0 g4 9 f4! gxf3 10 Qf2 Bg7 11 Nxf3 Nc6 12 Be2 a6 13 Rhfl gave him an edge in
E . Co rdova-L.Bruzon Bautista, Merida 2007, since even long castling wouldn't
ha ve been safe for Black with dS-ideas on the agenda) 8 NbS!? Qd8 9 c3 a6 10 Na3
b6 1 1 Bg2 Bb7 12 Nc2 Nd7 13 Ne3 Qe7 14 Ne2 (Diagram 29), which saw White re
tain decent-enough central control in M.Adams-P.Leko, Cap d' Agde 1996.
c3 ) White has often begun too with 7 0-0-0, but probably the best move for those
w ho want to drum up early complications is 7 h4!?, as we'll see in Game 9.

White manoeuvres Avoiding being pinned
We now return to S ... Bb4:

6 Qd2
White's main approach, but in view of the problems he faces in the main line, he
has also explored:
a) 6 Nge2 prepares to recapture with the knight in the event of an exchange on c3,
but is quite slow and 6... d6 7 Qd3 (inserting 7 a3 BaS is no panacea for White, not
least because the calm 8 g3?! Nc6 9 Bg2?! actually blunders the d-pawn to 9 ... Bb6!)
7. . . cS 8 dxcS dxcS (I can't see too much wrong either with 8 ... BxcS!?) 9 a3 Bxc3+ 10
Qxc3 Qxc3+ 1 1 Nxc3 Nc6 12 0-0-0 Ke7 13 g3 b6 14 Bg2 Bb7 saw Black equalize in
Z.Rahman-Li Ruofan, Kuala Lumpur 2007.
b) 6 eS Qe7 (the more flexible 6 ...Qd8 is also a decent choice) 7 a3 Bxc3+ 8 bxc3 is
p retty committal from White and now Black has a choice between the calm,
Nimzo-like 8... b6 and 8 ... d6, which quickly led to quite an unbalanced situation
a fter 9 f4!? dxeS 10 fxeS Qh4+! 1 1 g3 Qe4+ 12 Kf2 Qxh1 13 Bg2 Qxh2 14 Qg4 g6 1S
Nh3 Bd7 1 6 Rh1 Qxg2+ 1 7 Kxg2 Bc6+ 18 Kgl Bxh1 19 Kxh1 i n T.Nalbandian
A.Anastasian, Armenian Championship 2007; White's queen has some targets
here, but I would be surprised if Black was worse with his extra kingside pawns.
71
Sta rti n g Out: The Trom pows ky Atta c k
c ) I've long been o f the view that 6 Qd3!? (Diagram 30) i s a reasonable try, because
0-0-0 will completely break the pin. Black has quite a wide choice in this unforcing
position, but he usually opts to challenge in the centre:
cl) 6 ... c5 7 e5 Qe7 (best; Wells points out that 7... Qd8 8 dxc5! Qa5 9 Nge2 Nc6 10 4
Bxc5 1 1 0-0-0 gives White useful ideas of Qg3 and Ne4) 8 a3 leads to a further di
vide:
el l) 8 ... Ba5 9 0-0-0 cxd4 10 Ne4 0-0 1 1 f4!? (an attempt to improve over 1 1 Qg3 Nc6
12 Nf3 Bc7!, which was quite awkward for the attack in A.Bigg-C.Pritchett, British
League 2004) l l ...Nc6 12 Nf3 Bb6 1 3 g4 (Diagram 31) gives White decent attacking
chances for his pawn and could really do with a test.

Decent compensation Cleverly delaying the excha nge
c12) The solid 8 ... Bxc3+ 9 Qxc3 b6 might well be Black's best. At any rate 10 dxc5
Qxc5 11 Qxc5 bxc5 12 Nf3 (in New in Chess Yearbook 78 I proposed the more ambi
tious 12 f4!? Bb7 13 b4) 12 ... Nc6 13 0-0-0 f6 14 exf6 gxf6 15 Bb5 Bb7 was pretty level
in M.Taylor-J.Curtis, British League 2005.
c13) However, 8... cxd4?! is less convincing, because 9 axb4 dxc3 10 bxc3! gives
White a grip on the position, which soon turned into a strong attack after 10 ...0-0
1 1 Nf3 Rd8 12 Qe4 f5 13 exf6 Qxf6 14 Bd3!? in R.Palliser-P.Van Voorthuijsen, Cork
2005.
c2) 6 ... d5 is a solid, French-like approach, which we'll see in Game 10.
c3) 6... Nc6 7 Nf3 0-0 8 0-0-0 d6 9 Kb1 Qg6 10 h4 f5! 11 h5 Qf7 was about even in
L.McShane-V.Kramnik, European Club Cup, Saint Vincent 2005, but I would pre
fer the more aggressive 7 0-0-0!?, after which 7... Bxc3 (7...Qxf2? 8 Nf3 traps the
queen) 8 Qxc3 d6 9 Kb1 Bd7 10 Qe3! e5 1 1 d5 Ne7 12 g3 c5 13 h4 h5 14 Bh3 slightly
favoured White in N.Legky-N.Spiridonov, Sautron 2002.
72
The Positio n a l C hoice: 2 . . . e6
6 c5
...
The solid 6... d6 7 a3 BaS! (Diagram 32) also has quite a respectable reputation.
Bl ack s main point is that 8 f4 (the more restrained 8 Nf3 is probably called for, al
'
thou gh 8... Nd7 9 Be2 c6 10 0-0 e5 is fine for Black) 8 ... Bxc3! 9 bxc3 (9 Qxc3? Qxf4
w ins a pawn) 9...e5 gives good counterplay. White is loathe to close the centre with
his q ueenside damaged, but 10 g3 0-0 1 1 Nf3 Bg4 12 Bg2 Nc6 saw Black keeping up
the pressure with at least equality in T.Tolnai-I.Stohl, Hungarian League 1999.
1 a3
White tries to break the pin. Instead 7 e5 Qf5! leaves his centre vulnerable and 8 a3
(8 Bd3? Qg4 is Black's idea) 8...cxd4 9 axb4 dxc3 10 Qxc3 Nc6 1 1 Nf3 Qe4+ 12 Be2
Qx b4 13 Qxb4 Nxb4 14 0-0-0 offers no more than sufficient compensation,
o.Gutsche-N.Davies, correspondence 2003.
Another try has been 7 dxc5, but 7... Bxc5 (Davies points out that 7... 0-0 8 f4 Bxc3 9
b x c3 e5!? is also possible) 8 Nf3 Nc6 9 Bb5 Nd4! 10 Nxd4 Qxd4 1 1 Bd3 Bb4 equal
ized without difficulty in A.Schenk-A.Schmitt, French League 2004.
7 . Bxc3
. .
This has good pedigree, but once again I can't see too much wrong with 7... Ba5!?,
maintaining the pin.
8 bxc3 d6 (Diagram 33 )

Ni mzo-like play It's a bout eq ual
NOTE: The position most certainly has a fair amount in common

with the Nimzo-lndian: White must try to exploit his extra central
space, while Black would like to keep the centre closed with ... es
and to gradually exploit White's weakened queenside.
73
9 Nf3
Hodgson switched to this more sedate approach than White might like after no
doubt discovering 9 f4 e5! (9 ... 0-0 10 Nf3 Nc6 1 1 Bb5 Na5 12 Bd3 b6 13 0-0 Bb7 1 4
Rae1 kept control of the central situation and left White well mobilized in
J.Hodgson-L.Yudasin, New York (rapid) 1994). Then 10 Bb5+ Bd7 1 1 Rb1 (prefer
able to 1 1 Bxd7+?! Nxd7 1 2 Ne2 Qh4+ 13 g3 Qe7 14 fxe5 dxe5 15 d5 Nb6 16 a4 0-0,
which is something of a model for Black, A.Schirbel-J.Rowson, German League
2001) 1 1 ...exd4 12 cxd4 cxd4 1 3 Nf3 Nc6 14 0-0 0-0 (Wells) doesn't give White m ore
than equality.
g ...0-0 10 Bel Nc6 11 0-0 e5 (Diagram 34)
Black keeps up the central pressure and this position appears quite comfortable
for him. Two opposing experts clashed in J.Hodgson-J.Rowson, York 2000, and
after 12 dxc5 dxc5 13 Qe3 Bg4!? 14 Qxc5 Rfc8 15 Qe3 Qf4 16 Qxf4 exf4 Black en
joyed promising play for his pawn.
Conclusion
White has scored fairly well after 5 Nc3 d6 6 Qd2 (53% from almost 550 games).
It's easy for him to whip up a strong, early initiative, but it must be said that 6 ... c6
7 f4 e5 appears to equalize.
One reason for the growth in popularity of 5 c3 (see our next section) is that 5 Nc3
Bb4 gives Black decent counterplay. White is down to 50% here and try as he
might, he is yet to find a way to establish a strong centre without making signifi
cant concessions. I suspect that 6 Qd3!? is the best try, although objectively this
leads to nothing more than an unclear middlegame.
Illustrative Games
Game 9
D N.Nuesken F.Sawatzki
Ger m a n League 2007
1 d4 Nf6 2 Bg5 e6 3 e4 h6 4 Bxf6 Qxf6 5 Nc3 d6 6 Qdl g5 7 h411 (Diagram 3 5)

7 ... Bg7
Black's most natural move. He has also tried to keep lines closed with 7... g4, but
after the 8 f4 gxf3 (8 ... Be7 9 0-0-0 Qg7 was preferred in S.Grayland-J.Suto, corre
spondence 2004, and now I would be tempted to transfer the king's knight to g3
before advancing with e5 and/or fS) 9 Nxf3 Nc6 10 BbS Bd7 1 1 Rfl Qe7 12 0-0-0
Bg7 of N.Eliet-V.Chuchelov, Belgian League 2002, White could have seized the
74
The Positiona l C hoice: 2 . . . e6
u p per hand with 13 Nd5! Qd8 (13 ...exd5 14 exd5 creates problems on the e-file) 14
N f4 .
s h xg 5
once again 8 0-0-0 is possible, but weakening first the defence of g5 is more con
si stent with the approach begun by 7 h4.
s h xg5 9 Rxh8+ Bxh8 10 Nf31
. ..
on ce again White makes the g-pawn his focus. Here 10 0-0-0 is quite sensible too,
albei t a less-testing approach than the text: 10 ... Nc6 (10... Bd7 1 1 Nf3 Nc6 12 d5!
levered open the position to White's advantage in S.Lputian-V.Tukmakov, Tilburg
(b litz ) 1994) 1 1 Bb5 Bd7 1 2 Nge2 0-0-0 13 Qe3 Kb8 1 4 g4 was perhaps a touch bet
ter for White thanks to his extra space in F .Hoffmeyer-M.Schuster, German
League 2006.
10 g4 (Diagram 36)
...

Challenging Black head-on How to open the centre?
Cf
TIP: This strike may still need more testing, but I've long been a fan
and feel that it packs a certain punch.
The most natural reaction, but not the only possibility:
a) 10 ...Qf4?! 1 1 NbS! Qxe4+ (1 1 ...Qxd2+ 12 Kxd2 either wins a pawn or leaves
White clearly better in the endgame after 12 ... g4 13 Nxc7+ Kd8 14 Nxa8 gxf3 15 c3
b6 16 gxf3 Bb7 17 Nxb6 axb6 18 Bb5) 1 2 Be2 Kd7 13 Nxg5 Qg6 14 Bd3 Qg7 15 0-0-0
left White enjoying a strong initiative in J.Gather-G.Marcotulli, correspondence
2003.
b) 10 ... Nc6 1 1 NbS Qe7 12 QxgS QxgS 13 NxgS Nxd4 14 Nxc7+ Kd7 15 c3! (Wells)
a lso leaves White with the initiative and slightly for preference.
75
11 est Qg7?
Black does better to keep his queen in contact with the c7-point with 1 l ...Qe7, but
it seems that White can go 12 NbS!? regardless! The point being that 12 ... gxf3
(12 ... Kf8!? 13 Qh6+! Bg7 14 Qg5 Nc6 doesn't look so clear and could do with a test)
13 Qh6 regains the piece and after 13 ... Bxe5 14 dxe5 fxg2?! 15 Bxg2 d5 16 0-0-0
White's superior development far outweighs the sacrificed pawn, H.Graef
S.Roemling, German League 2005.
NOTE: Black is the only side with a dark-squared bishop in this varia
tion, but he can still find himself struggling on the dark squares! In
deed, White should usually look to combine kingside pressure with
an early attack on c7.
12 Ng5
White's knight is safe here with the enemy bishop stuck on h8 for the time being.
Moreover, Black must now be careful with his queenside yet to be developed.
12 ...dxes?l 13 dSI Bd7
White's play in this game is pretty brutal, but from his point of view 13 .. .f6? 14
Bb5+?! c6 15 Nxe6 Bxe6 16 dxe6 cxb5 1 7 Nxb5 was even more aesthetic in the ear
lier J.Schlenker-C.Pfrommer, Waldshut 2002. Black might have had an extra piece,
but he was defenceless with White's well coordinated pieces pouring in as they
did with 1 7... Nc6 18 Qd6 Rc8 19 Rd1 Nd4 20 Nxd4 exd4 21 Rxd4 in the game.
Observe too how 13 ...exd5? 14 Nxd5 f5 1 5 0-0-0 gives White far too strong an ini
tiative, and he is also the heavy favourite after 13 ...Qh6 14 dxe6 (the sharper 14
Nce4!? might be even more promising) 14... Bxe6 15 Nd5 (Wells).
14 o-o-o (Diagram 37)

Black is in trouble The floodgates open
76
The Positio n a l C h o i c e : 2 . . e6
.
14 e4
.
..
a
B l c k des
perately tries to return some material to distract White. It's unlikely to
o r k, but his position would also have been rather disgusting after 14 ... Na6 1S
w
d:xe6 Bxe6 1 6 Nxe6 fxe6 1 7 Ne4.
1s Bc4
pretty effective too is 1S Ngxe4 QeS 16 dxe6 fxe6 1 7 Qh6 when Black's queenside
con tinues to sleep. Indeed, it hasn't taken White long to whip up a crushing initia
ti ve - 7 h4 and 10 Nf3 is an approach well worth remembering!
1s e3 16 Qxe3 Qe5 17 Qd2 Bf6 18 Nge4 Be7
...
One can understand why Black wanted to avoid 18 ...Bg7 19 NcS, but perhaps this
follo wed by 19 ... Kf8 was relatively the best defence.
1 9 f4!?
I suspect that 19 Qh6 might have been even more clinical, but by now it's clear
tha t most roads lead to Rome.
19 . gxf3 20 gxf3 Na6 21 f41 Qf5 22 Bxa6 bxa6 23 d61 (Diagram 38)
. .
The final nail in Black's coffin. His king isn't going to survive the coming on
slaught.
23 ... cxd6 24 Nxd6+ Bxd6 2 5 Qxd6 Rd8 26 Rh1 Qf6 27 Nd5 1 1-0
Game 10
D A.Bigg C.Hanley
Street 2004
1 d4 Nf6 2 Bg5 e6 3 e4 h6 4 Bxf6 Qxf6 5 Nc3 Bb4 6 Qd3

It's less established after 6 Qd2 than both 6 ... cS and 6 ... d6, but here too the French
approach, 6 ... dS, is by no means a bad one: for example, 7 eS Qd8 8 a3 Be7 9 f4 cS
1 0 Nf3 a6 (Black prepares swift counterplay with ... c4 and ...bS should White go
long) 1 1 dxcS BxcS 12 fS! (White too would like to open lines against the opposi
tion king) 12 ... Nc6 13 0-0-0 bS 14 fxe6 fxe6 1S Qf4 gS! saw both sides fighting for
the initiative and was rather murky in Zhang Pengxiang-E.Aiekseev, Nizhnij
Novgorod 2007.
6 d5 (Diagram 39) 7 Nge21?
...
White borrows an idea from the Winawer to keep his structure intact. He has
rnore often preferred 7 eS, but 7.. .Qd8 (7 ... Qg6!? 8 Qxg6 fxg6 9 h4 cS 10 0-0-0 0-0 1 1
Nbs a6 12 Nd6 cxd4 also gives Black decent counterplay, l.Schneider
D . Bunzmann, Boblingen 2008) 8 a3 (or 8 0-0-0 cS 9 dxcS 0-0 10 f4 Nd7 1 1 Nce2 aS! ?
12 Ng3 a4! 13 a3 BxcS which left Black's attack looking the more potent in
R.Eames-J. Emms, British League 2004) 8 ... Be7 9 Qg3 0-0 10 Nh3 cS 1 1 dxcS BxcS 12
Bd3 Nc6 13 0-0 Bd7 (Diagram 40) gives Black a comfortable version of the French
77
Sta rt i ng Out: T h e Trom powsky Attack
with White missing his dark-squared bishop, S.Fowler-I.Gourlay, Coventry 2005.
NOTE: Black's queen does best to retreat to d8 rather than e7 in this

line; both to enable his bishop to retreat if kicked and so that he can
later increase the dark-square pressure with ...Qb6 or ...Qas.

The French Defence to 6 Qd3 There's no bishop to ta rget h6
As well as the text, 7 a3!? Bxc3+ 8 bxc3 deserves attention, after which 8...0-0
(8 ... Nd7 9 eS Qe7 10 f4 Nb6 1 1 Qe3 Bd7 12 Nf3 cS 13 Bd3 was similar in F.Nijboer
M.De Jong, Dieren 200S) 9 eS Qe7 10 f4 cS 1 1 Nf3 b6 is unclear according to the
Hungarian Grandmaster, Tibor Fogarasi, but those with some Winawer experi
ence may be fairly happy on the white side.
7 ...dxe4
This whole variation remains quite unexplored and 7 ...0-0!? 8 a3 BaS can't be too
bad an alternative when 9 exdS (I would probably prefer to facilitate queenside
castling with 9 Qe3) 9 ... Rd8 10 dxe6 Bxe6 offers Black the freer development for
his pawn.
8 Qxe4 o-o 9 a3 Bas 10 f4 Bd7
Far from bad, but 10 ... cS! must be more critical. Then 1 1 dxcS (White probably
does better to take up the challenge than to go in for the fairly tame 11 0-0-0 cxd4
12 Qxd4 Qxd4 13 Nxd4) 1 1 ...Rd8 1 2 b4 Bc7 13 Rd1 Rxd1+ 14 Nxd1 aS is a pawn
sacrifice which definitely deserves a test.
11 o-o-o Bc6 12 Qe3 Nd7 13 Rg11 (Diagram 41) 13 ... b51
Hanley correctly counters in kind.
14 g4 b4 15 axb4 Bxb4 16 g5 hxgs 17 Rxgsl Qh6
Strangely just a few weeks later Black preferred 17 Rfe8?! in A.Bigg-D.Tan, Brit-
...
78
T h e Pos i t i o n a l C h o ic e : 2 ... e6
i sh C hampionship, Scarborough 2004, bu t after 18 Ng3 g6 White could have ob

ta ined the advantage with 19 Nge4! Qe7 (19 ...Qh8 is met in the same manner and
19 . . . Bxe4?! 20 Nxe4 Qg7 21 BbS would only make matters worse for Black) 20 BbS
BxbS 21 NxbS Nb6 22 Rdg1 when his attack looks pretty dangerous. Tan's desire
to improve was, though, quite understandable: the queen is not entirely happy
having to defend from an isolated spot on the rim. However, it's almost certainly
earlier that Black should be looking for an improvement.
18 Ng3 (Diagram 42)
Diagram 41 ( B) Diagram 42 (B)

White attacks Black can hold up the attack
TIP: When both sides castle on opposite sides, he who attacks first
often seizes the upper hand! That certainly seems to be the case
here and White could hardly have hung around.
However, now Black gets to cover some key squares, if not to equalize. Instead 18
Bh3! followed by Rdg1 would have been even more dangerous.
18... Nf61
Keeping White's knight out of gS and now the attack is slowed down.
19 Bc4 Bxc3 20 bxc3
Bigg must have been sorely tempted by 20 Bxe6!?, which would also have forced
Black to tread carefully: 20...Bb4 (20 .. . fxe6 21 Qxe6+ Kh8 22 Qxc6 regains the piece)
21 N fS Qxh2 22 Rxg7+ Kh8 23 dS fxe6 24 Rg3 sees Black pick up enough material
for the queen, but without helping his king to feel any more secure in the process.
2 o ..Rfe8
.
Running a second member of the royal family on to the h-file with 20 ... Kh8? looks
suspect and even though it's not so easy for White to get his rook to h5, 21 d5!
79
Sta rti ng Out: The Trom pows ky Attack
exdS 22 NfS Qh7 23 Bd3 should do the trick.

21 Rg1 Kf8 (Diagram 43)

Is Black escaping? Strengthening the centre
Prudent defence, although White will continue to press.

22 Qesl Bf31 23 Nf1 Nds Yz-Yz
Hanley must have been relieved to see his draw offer accepted. Black has de
fended pretty well, but even so after the long, forcing line 23 ... Nd5 24 BxdS exdS
25 Qxg7+ Qxg7 26 Rxg7 Rel + 27 Kd2 Rxfl 28 Rxfl Kxg7 29 Rxf3 aS, 30 c4! would
still have given White some chances in the ensuing rook endgame.
1 d4 Nf6 2 Bgs e6 3 e4 h6 4 Bxf6 Qxf6 5 c3 (Diagram 44)

Determined neither to impede his f-pawn nor to allow a pin on his queen's knight,
White began to experiment with this set-up towards the end of the Nineties. In his
fine 1998 survey of the Trompowsky, Joe Gallagher gave just one game reference
with 5 c3, but did comment: 'perhaps we will be seeing more of this line in the
future.' How right he was! I now have more than 500 examples of 5 c3 on my da
tabase and it can at the least be considered a serious alternative to 5 Nc3.
s ... d6
By some margin Black's most popular response, beginning a dark-square strategy
just as we saw him often employ after 5 Nf3. However, 5 c3 hardly forces the pace,
so there are unsurprisingly alternatives:
80
The Posit i o n a l Cho ice: 2 ... e6
a) The French-like 5 ... d5 has quite a good pedigree and was used in some impor
glish Trompowsky clashes when 5 c3 was still finding its feet. White must
tan t En
whether or not to close the centre:
then decide
a l ) 6 e5 Qd8 7 Nf3 (similar is 7 Nd2 c5 8 dxc5 Bxc5 9 Nb3 Bb6 10 Nf3 Nc6 1 1 Bd3
B d7 12 Qe2, but with 12 ... 0-0 13 0-0 f6! Black obtained sufficient counterplay in
J . De la Villa Garcia-D.Garcia Ilundian, Pamplona 1992; breaking with ... f6 makes a
lo t of sense with White missing his dark-squared bishop) 7... c5 8 dxcS (White has
no desire to find himself tied down defending b2 and d4 after ... Nc6 and ... Qb6)
8 . . . Bxc5 9 Bd3 Nc6 (9...Qb6 has been quite a common reaction, but makes it easier
fo r White to stir up trouble and 10 Qe2 Nc6 1 1 0-0 Bd7 12 b4! Be7 13 Nbd2 Rc8 14
a 3 g5!? led to a pretty complex manoeuvring struggle in J.Hodgson-J.Rowson,
Southend 2001 ) 10 0-0 {Diagram 45) 10 ... Bd7 (despite John Cox's best efforts in
Dealing with d4 Deviations, many black players don't seem entirely at home in this
line and continue to reject the straightforward and good 10 ... 0-0 1 1 Nbd2 f6! break;
here 11 Bc2!? might be White's best, but even 1 1 . ..f6 12 Qd3 f5 is hardly uncom
fortable for Black) 1 1 b4! ? Bb6 12 Nbd2 Rc8 13 Qe2 a6 was about equal in
M.Kanep-Y.Rantanen, Finnish League 2006; White would like to manoeuvre his
queen's knight to d4 or to break with c4, but ... f6 will again give Black enough
counterplay.

A comfortable French for Black? Only White has a knight
\
la2) 'arious Trompowsky authorities have been happy enough to close the centre,
but I suspect that 6 Nd2!? might be a better try; it certainly gives White a position
With more options, as we'll see in Game 1 1 .
b) S . . c!5. enables White to demonstrate one strength of 5 c3: his centre i s secure.
.
Thus 6 Nf3 is sensible and with 6 ... Nc6 (or 6 ... cxd4 7 cxd4 Bb4+ 8 Nc3 when Black
is fair ly pretty solid, but White should be a little better after 8 ... Nc6 9 e5!? or 8 ... b6
81
Sta rti n g Out: Th e Trom pows ky Attack
9 Bd3 Bb7 1 0 0-0 Nc6 11 eSt after which 1 l ...Qf4?? 1 2 Ne2 Qg4 13 h3 won the er
rant queen in M.Bosiocic-M.Cebalo, Sibenik 2006) _?_g;; (Davies's gambit sugges
tion 7 BbS!? Qg6 8 0-0 Qxe4 9 Re1 is a dangerous alternative; Tzermiadianos con
tinues 9 ... QdS 10 c4 QhS 1 1 dxcS! QxcS 12 Nc3 when the initiative most certainly
belongs to White) 7 ... NeS 8 Be2 Nxf3+ 9 Bxf3 eS ( Diagram 46) Black keeps White's
queen out of dS by closing the centre rather than exchanging there. However,
White has the superior minor piece and 10 a4 g6 1 1 Na3 d6 12 Be2 hS!? 13 BbS+
Ke7 14 aS Qh4 1S Qe2 fS 16 Bd3 saw him retain the upper hand in V.Milov
B.Macieja, European Championship, Batumi 2002.
c) S . . . Nc6 is a little mysterious, but it has been used by some strong grandmasters.
Black's point appears to be 6 Bd3 eS 7 dS BcS!, but White does much better with 6
eS! Qe7 7 Nf3 d6 8 BbS Bd7 9 exd6 cxd6 10 0-0, which gave him a pull in J.Rowson
A.Chemiaev, London 1999.
We now return to S ... d6:
6 Bd3 (Diagram 47)
Diagram 47 (B) Diagram 48 ( B)

Ne2, o-o and f4 will follow White is in control
NOTE: We can now appreciate the main reason White chose 5 c3: the
d4-point is secure and so he can develop with Bd3, Ne2, 0-0 and f4; a
straightforward but also quite dangerous set-up.
6 ... es
Black wastes no time countering in the centre, but not everyone has been keen to
play ...eS so quickly:
a) 6. . . Nd7 7 Ne2 g6 8 0-0 Bg7 9 f4 reveals White's set-up in action and he will usu-
ally aim for a rapid, cramping eS-advance: 9 . .Qe7 10 Nd2 (10 fS!? is an ambitious
.
82
The Posit i o n a l C h oice: 2 ... e6
pressure,
try to punish Black's move order: 1 0 ...0-0 11 Nf4 generates some early
bu t 1 0 ... e5 1 1 fxg6 fxg6 1 2 Nd2 Rf8 13 Qb3! Nb6 14 a4 was hardly the answer for
B la ck in M.Kanep-A.Chuiko, Moscow 2005) 10 ... 0-0 (Black may do better with
JO . . e5, transposing to our main line) 1 1 e5! aS 12 Qe1 b6 13 Qg3 Ba6 14 Bxa6 Rxa6
.
1 5 h4! h5 16 Ne4 (Diagram 48) is a good illustration of how White can build up on
the kingside while maintaining control, I.Rogers-K.Arakham ia Grant, Wijk aan
zee 2002.
b) 6 ... g6 will transpose to either our last variation or to our main line depending
on when Black breaks with ... eS.
c) 6 . Qg5!? is a radical attempt to disrupt White's smooth development. The criti
. .
cal reaction is 7 Nf3!? (there's no compulsion to sacrifice; 7 g3 being a safe alterna

tive a nd after 7... Nd7 8 f4 Qd8 9 Ne2 b6 10 Nd2 Bb7 1 1 a4 a6 12 0-0 White's weak
ened kingside was hardly a problem in A.Stefanova-S.Vijayalakshmi, Kavala 2006,
but 7... g6 8 Ne2 Bg7 9 Nd2 eS is more of a challenge: there are issues with ... Bh3
and the absence of the Qe1-g3 manoeuvre after f4) 7... Qxg2 8 Rg1 Qh3 9 Nbd2, but
Black remains very solid here and it's not so easy to exploit White's superior de
velopment despite his possible central advances: 9 ... Nc6 (9 ... Qh5 10 Nf1 ! QaS 1 1
Ne3 makes i t difficult for Black to develop smoothly, a s pointed out b y Wells) 1 0
Rg3!? QhS (Diagram 49) 1 1 Qb3 e5?! 1 2 0-0-0 g5?! 13 BbS! saw White seizing the
initiative in D.Doric-H.Stevic, Djakovo 2006, but Black should be happier to react
rather than challenge in the centre.

Sufficient compensation Activating the king's rook
Returning to 6 ... e5:

7 Ne2 g6 8 o-o Bg7 9 f4 (Diagram 50)
White puts his main pawn break into operation.
83
TIP: Always keep a close eye on all the possible pawn breaks in a po
sition, especially should the structure change. Appreciating the
strengths of your main pawn break(s) will often help you decide
where the pieces should best go.
9 Qe7
...
Black keeps things solid. He has also been known to target d4, which is a decent
alternative: 9 ...exd4 10 cxd4 Bg4! (10 ... Nc6 1 1 eS Qd8 12 Nbc3 0-0 13 Rcl saw
White maintain his central wedge without too much difficulty in J.Hodgson
C.Ward, British Championship, Millfield 2000) 1 1 Qa4+ Nd7 12 Nbc3 0-0 13 eS
dxeS 14 fxeS and now in P.Wells-G.Fish, German League 2002, 14 ... Qd8! would
have prepared ... cS and left the position rather murky.
10 Nd2 Nd7
Black shores up eS before castling and we'll return to this position in Game 12.
1 0. . .0-0 is also fine, since White hasn't anything better than 11 Nf3 (11 fS!? Nd7
fails to really trouble Black) 1 1 ...Nd7 (Diagram 51), reaching a position which can
also be found in the notes to that illustrative game.

Black strongpoints es for now Tarrasch-like development
Practice has also seen the more provocative 1 0... Nc6, after which 1 1 Nf3 0-0 12
Qd2 Bg4 13 Bc2! exd4!? 14 cxd4 dS 1 5 eS Qb4 restricted White to just a small edge
in N.Sulava-S.Fedorchuk, French League 2006; once again we can see that Black' s
dark-squared bishop can find itself a little redundant on g7 should White manage
to keep his central pawn-wedge well supported.
84
T h e Pos ition a l Cho ice: 2 . . e6
.
c o n clusion
White's easy-to-employ set-up is certainly a decent alternative to 5 Nc3. That he is
then probably best advised to remain flexible in the event of 5 c3 dS is borne out
bv the statistics: 6 eS having made just 43% from 56 games, as opposed to 65% for
6 -N d2, albeit from only 36 games.
The main line with 5 ... d6 and ...eS leads to an unbalanced middlegame with
chan ces for both sides. Indeed, White has scored 50% from the position after 10
Nd2 in our main line. He can hope to suffocate Black, but Black's position is pretty
solid and the second player can aim to obtain counterplay on the dark squares.
Illustrative Games
Game 1 1
o Z.Rahman M.Carlsen
Dresden Olym piad 2008
1 d4 Nf6 2 Bg5 e6 3 e4 h6 4 Bxf6 Qxf6 5 c3 d5 6 Nd217 (Diagram 52) 6 ... cs

Black's most natural and popular choice. Instead 6 ... dxe4 7 Nxe4 Qd8 8 Nf3 re
sembles the Rubinstein French. Black is fairly solid and does have the bishop-pair,
but is a little too passive for my liking and after 8 ... Nd7 9 Bd3 Be7 1 0 Qe2 0-0 1 1
0-0-0 c5 12 dxc5! Qa5 13 Kb1 Nxc5 1 4 Nxc5 Qxc5 1 5 Bc2! White retained a pull in
J.Salimaki-P.Span, Kerner 2007; Ne5 as well as Qe4 may follow.
The more creative 6 ... Nc6 7 Bd3 Bd7 8 Ne2 0-0-0 9 eS Qe7 10 b4 (10 f4!? g5 11 0-0 is
also tempting) 10 .. .f6 was preferred in E.Rosenberg-G.Kacheishvili, Ledyard 2009,
but with 1 1 b5 NaS 12 Nf3 White would have retained the advantage.
7 Ngf3
Aga in the most flexible approach is the most challenging. Otherwise 7 e5 Qd8 8
dxc5 Bxc5 takes play back towards lines we've already considered and 7 exdS
exdS 8 Bb5+ Bd7 9 Bxd7+ Nxd7 10 Ne2 0-0-0! failed to trouble Black in P.Wells
C. Ward, Southend 2002.
7 Nc6
...
Bla ck too is happy to maintain the central tension, but again there are alternatives:
a ) 7... cxd4! ? might actually be Black's best bet and it's not so easy to prove any
thi ng for White here: 8 cxd4!? (an active try; otherwise 8 Nxd4 BcS 9 Qa4+ Bd7 10
Bbs shoul dn't really lead anywhere and 8 exdS exdS 9 Nxd4 Nc6 10 N2b3 Bd6 1 1
Qf3 Qxf3 12 Nxf3 Bg4 1 3 Be2 0-0-0 gave Black enough counterplay to offset his
IQP in A.Bigg-D.King, British League 2004) 8 ... dxe4 9 NeS (Diagram 53) is an ag
g ressi ve idea stemming from Nigel Davies, although Cox is probably right that
85
Sta rti n g Out: T h e Trom powsky Att a c k
9 .. Bb4!? (9 ... Bd6 10 BbS+ Bd7 1 1 Nxd7 Nxd7 1 2 Qa4 was preferred in
.
M.Karttunen-M.Agopov, Finnish League 2006, and now the simple 12 ... Rd8
would have been about equal) 10 BbS+ Bd7 1 1 Nxd7 Bxd2+ 12 Qxd2 Nxd7 should
be okay for Black.
Diagram 53 {B) Diagram 54 {W)

Fighti ng for the i nitiative The d-pawn is poisoned
b) 7 ... a6 rules out the check on bS, bu _;> (8 B with the idea of 8 ... cxd4 9
cxd4 dxe4 10 Bxe4! is a good alternative, as pointed out by Tzermiadianos) 8...Bxc5
9 Bd3 Nc6 1 0 0-0 0-0 1 1 Qe2 Ba7 12 Rfe1 Bd7 13 Rad1 Rad8 14 Nfl is perhaps a
little better for White, M.Karttunen-J.Hellsten, European Team Championship,
Plovdiv 2003.
NOTE: These variations suggest that even 6 Nd2 may not promise
White any advantage, but at least Black is forced to make some
early decisions rather than just being able to develop along standard
French lines.
B dxc5
Those happy to accept an isolated queen's pawn (IQP) in return for chances to
seize the initiative should again consider 8 BbS!? cxd4 9 cxd4.
B ... Bxc5 9 Bd3
IQPs are often a highly divisive issue, especially at club level: some players love to
attack with them; others hate such a scenario and only want to play against them.
A player who falls into that latter category is unlikely to contest this variation as
Black, because White might continue 9 exdS exdS 10 Nb3 Bb6. However, this again
looks like a fairly comfortable version of the French Tarrasch for Black after 1 1 Be2
0-0 1 2 0-0 ReB 1 3 Re1 BfS! (Diagra m 54).
86
The Positi o n a l C h oice: 2 . . . e 6
0.o 10 o-o Bb6 (Diagra m 55)

9
Diagram 5 5 (W) Diagram 56 (W)

Roughly level Black's king is safe enough
The position continues to have certain similarities to the French Tarrasch, albeit
this time the modem 3 ... Be7 variation, as well as to the Colle. White has a pleasant
central set-up and while Black has the bishop-pair, he is still to solve the problem
of how to activate his light-squared bishop.
11 Qe2 Qf41?
Carlsen wants to meet eS with .. .f6. There was nothing wrong with the simpler
l l ...Bd7, though, when White should probably keep the tension with 12 Bc2!?,
rather than go in for 12 eS Qf4 13 Bc2 f6! 14 exf6 Rxf6 15 Qd3 g6.
12 Rad1 Bd7 13 Bc21
White prepares to build a queen and bishop battery and this accurate retreat high
lights Black's difficulties.
13 ... Rfd8
Carlsen will have to defend a slightly worse middlegame after this, but it was al
ready too late to find a way to equalize and 13 ... dxe4 14 Nxe4 Rfd8 15 Nd6 would
have been pleasant for White.
14 exd5 1
The Bangladeshi Grandmaster spots the chance to inflict an isolated pawn while
d oub ling on the bl-h7 diagonal, and so seizes the upper hand.
14...exd5 15 Qd3 Be6
Fa irly essential, as 15 ... g6 16 QxdS Bf5 1 7 Qb3 NaS 18 Qa4 wouldn't have given
B lack enough for his pawn.
16 Qh7+ Kf8 17 Nb3
87
0
NOTE: Rahman is beginning to exploit one of the main drawbacks to
an IQP: the way in which the holder of that pawn often doesn't con
trol the square in front of it.
17 ...Qf6 18 Nbd4 Nxd4 19 Nxd4 Rac8 20 Rd3
The aggressive 20 f4! ? was also an option. Then Black should probably defend as
in the game with 20 . .. Bxd4+ 21 Rxd4 Rc4 22 Rfdl Rxd4 23 Rxd4, although White
can continue to probe here, such as with 23 ... Bd7 24 h3 followed by tidying the
king and Bb3.
20... Bxd41 21 Rxd4 Rc4 (Diagram 56)
Good, classical defence from Carlsen. Being slightly worse Black is happy to trade
pieces and knows that any endgame should be tenable as he only has the one
weakness.
22 Rfd1 Rxd4 23 Rxd4 Qe5 24 Rd1?1
This allows Black to solve all his problems at a stroke, although even 24 Bd3 Qel +
25 Bfl f6 would by now have been quite tenable for the second player with ... ReS
on its way.
24...d41
WARNING: Never forget that the weakness of an isolated pawn

becomes more felt the more pieces that are exchanged. Thus one
should often try to liquidate any such pawn unless passed when an
endgame approaches, just as Carlsen does here.
25 Qd3 Bxa2 26 cxd4 Qa5? (Diagram 57)

A Fischer-like bishop on a2 How to build up on the kingside?
88
The Pos i t i o n a l Choice: 2 . ..e6
u pri sing slip, but perhaps Carlsen was trying too hard to win. Correct was
As r
26 . .Qc7 when White's d-pawn cannot advance and 27 b3?! can be countered by
.
z7 . R c8.
. .
27 h3?
Rah man returns the favour and perhaps both players were already low on time.
Be mu st surely have considered 27 b3 and I'm not sure why this was rejected: for
ex a mple, 27... Rd6 (27...Qa3 is another attempt to avoid a fatal pin on the a-file, but
then 28 Qc3!? renews such a possibility) 28 Ra1 Ra6 29 d5 and White's passed
pa w n is extremely powerful.
27 . Be6
. .
Now the bishop is out of danger.

28 Bb3!
Before he becomes tied down defending an isolated pawn of his own, Rahman
sensibly liquidates.
28 ... Bxb3 29 Qxb3 Rxd4 30 Rxd4 Qe1+ 31 Kh2 Qe5+ 32 Kg1 Qe1+ 33 Kh2 Qe5+ 34
Kg1 Qxd4 3 5 Qxb7
Of course the queen ending is a total draw.
35 ...gs 36 g3 Kg7 37 Kg2 as 38 b3 Yz-Yz
Game 12
D J.Coleman P.Skacelik
B rno 2004
1 d4 Nf6 2 Bg5 e6 3 e4 h6 4 Bxf6 Qxf6 5 c3 g6

A slightly unusual move order, but we soon transpose to familiar territory.
6 Nd2
Th ose of a highly aggressive bent might wish to consider the original positions
which quickly arise after 6 e5! ? Qe7 7 f4 g5!?. Moving the knight to d2 is a more
restrained approach, as is 6 Bd3 Bg7 7 Ne2 d6 8 0-0 0-0 9 f4 Qe7 and now 10 Nd2 is
usually seen in any case. Following 10...e5 11 Nf3 Nd7 12 Qd2 (12 fxe5 dxe5 13
Qb3 !? is a decent, slightly more unbalanced alternative, which certainly worked
out well after 13 ... exd4 14 cxd4 c5 15 Nf4 cxd4? 16 Nxg6 in V.Hiinka
R.Briestensky, Slovakian League 2000) 12 ... c5 13 Rae1 a6 14 b4!? cxd4 15 cxd4 Qd8
1 6 Kh1 b5 the position was about equal in L.Vasilescu-R.Milu, Bucharest 1999, and
now it was time for 17 Rcl followed by the creation of a passed d-pawn and/or
arra nging the a4-break.
6 ...d6 7 Bd3 Nd7 8 Ne2 Qe7 9 0-0 Bg7 10 f4 eS (Diagram 58)
B lack counters consistently in the centre, but Coleman is ready with a small sur
p rise.
89
Sta rti n g Out: The Trom powsky Atta c k
11 Qe117
A very aggressive approach, but also one which I have some time for. More usual
has been 1 1 Nf3, transposing after 1 1 ...0-0 to the notes to White's 6th move, above.
11 ...0-0
In view of what follows perhaps Black should consider accelerating his counter
play with 1 l ...exd4!? 1 2 cxd4 c5.
12 Qg3 exd41
Black often aims to open up his extra dark-squared bishop with this sequence. A
solid alternative is 12 ... Nf6, but that gives White a free hand and both 13 f5 and 13
fxe5 dxe5 14 Bc4!? are quite tempting.
13 cxd4 c5 14 f51 (Diagram 59)

White is determined to attack Black must be pretty ca reful
White too remains true to his plans and now Black must take some care as his
king position begins to open.
WARNING: The king is rarely happy when chunks begin to appear in

its pawn shield. Indeed, White has won many a Trompowsky game
by attacking such a weakened king.
14...gxf5
It looks a little ugly, but a decent case might well be made for keeping the king
side closed with 14 . .. g5!?. Play might continue 15 Nf3 (trying to prevent ...cxd4
and ... Ne5 ideas with 15 Qf2!? is the alternative) 15 ... cxd4 16 Nexd4 Ne5 when
Black is pretty strong on the dark squares, but White has the h4-break and is most
certainly not without his chances too in this rather unclear situation.
90
The Position a l C h o i ce: 2 ... e 6
,_ 5 exf5 Kh8?1
This enables White to seize the initiative. Less clear would have been 15 ... Qg5 16
xgS hxg5 17 Ne4 dS 18 Nxg5 cxd4 19 Rad 1, although here too I would prefer to
Q
take White not least because he is likely to emerge a pawn to the good.
}-low ever, 16 f6 wasn't yet a threat, so Black should have claimed some central
space with 15 ... d5!. Following 16 Nf4 Qg5 17 Nf3 Qxg3 1 8 hxg3 c4 19 Bc2 Nf6 the
dS-p awn might be weak, but is far from easy to win and so Black is probably
okay .
16 Ne4 (Diagram 60)
In view of our next note, I suspect that beginning with 16 Rae1!? was a touch more
accura te.
16. .cxd4?
.
A further slip. Black might have thought that his dark-square control was suffi
cient, but it isn't. Again Skacelik had to advance in the centre: 16 ...c4! 17 Bc2 d5
was the way to go, even though by now White would have been slightly for
choice after 18 N4c3 Nf6 19 Nf4.
17 Rae11
Coleman wants to attack, not regain his material.
11 ..ds?
.
A natural-enough move, but White has a strong and thematic riposte. Black might
like to play 17 ...Be5?!, but then 18 Qh3 pushes the bishop back and 18 ... Bg7 19 Nf4
looks rather dangerous. Thus Black had to block the f-pawn with 1 7... Nf6, al
though after 18 Nxd6 Nh5 (even 18 ... Qe3+!? 19 Qxe3 dxe3 20 Nf4 leaves Black
rather cramped and likely to lose a pawn on e3 or f7) 19 Qf3 Qxd6 20 Qxh5 Qf6 21
Nf4 White would have been well in charge of the proceedings.
18 f61
TIP: White's f-pawn might have initially gone to f4 to support an es

advance, but it can become a dangerous battering ram in its own
right should Black be careless.
18... Nxf6
There's no defence: 18 . .. Bxf6 19 Nxf6 Nxf6 20 Qh4 Qe3+ 21 Kh1 Ne4 22 Nf4 would
ha ve been completely crushing too.
1 9 Nxf6 Bxf6 (Diagram 61) 20 Nf4 Be6
In stead 20 .. .
Qd8 21 Ng6+! fxg6 22 Qxg6 would have been a beautiful illustration of
the superiority of White's light-square play over Black's dark-squared pieces, but
the text allows White to finish
with another neat sacrifice.
2 1 Rxe61 1-0
91

The knight is bound for g6 A simple, untheoretical choice
The Torre-like 3 e 3
1 d4 Nf6 2 Bgs e6 3 e 3 (Diagram 62)
A more flexible choice than the immediate 3 Nf3. Those looking to take play into
Torre waters often begin too with 3 Nd2 when 3 ... c5 4 e3 usually transposes to
lines considered below. White also has 4 c3, which can transpose too, although
4 ...cxd4 5 cxd4 Qb6 6 Qc2 Nc6 is a possibility.
3 ...cs
Black wastes no time challenging in the centre. Instead ... d5 either here or over the
next couple of moves is a little inflexible and will transpose to our coverage in the
final section of Chapter One. Black has also been known to begin with 3... Be7, but
a slightly more thought-provoking option is 3 ... h6. Following 4 Bh4 (I'm not so
keen on 4 Bxf6 Qxf6 5 f4 with the pawn back on e3 and Black pretty flexible about
how he counters on the dark squares - even 5 ... d6 6 Nf3 g5!? is possible) 4 ...c5
(4 . . . d6!? echoes a line of the pure Torre Attack; here White isn't committed to Nf3,
but even so 5 c4 g5 6 Bg3 Ne4 7 Nc3 Nxg3 8 hxg3 Bg7 might well give Black a suf
ficient share of the play in an unbalanced manoeuvring struggle) 5 c3 the position
is clearly pretty similar to our main line. Black might find the inclusion of ... h6
helpful, although it might also tempt White to align his bishop and queen on the
bl -h7 diagonal.
4 c3 cxd4
Black resolves the tension, but there is certainly no compulsion to:
a) 4 . . b6 5 Nd2 Bb7 (Diagram 63) is a very sensible approach and one popular with
.
92
The Posit i o n a l C h oice: 2 . . . e6
B l a c k in the overall world of the d-pawn systems.

Queen's Indian-like development A double-edged decision
NOTE: The term 'd-pawn systems' has become fairly common in re

cent years and refers to a number of openings popular at club level:
the Torre, the London, the Colle and nowadays even the Trom
powsky variations 2 ... d5 3 e3 and 2 . ..e6 3 e3.
Here White has nothing better than to take play into the Torre proper with 6 Ngf3,
which is why a 3 e3 or 3 Nd2 approach usually appeals most to those with some
existing d-pawn system experience. After 6... Be7 White has tried a number of
ideas, including 7 Bd3, 7 a4 and 7 h3 (to safeguard a retreat for the bishop). On his
DVO The Trompowsky: The Easy Way, Andrew Martin draws attention to 7 Bxf6!?
Bxf6 8 Bd3 Nc6 9 Qe2, which is not without a bit of sting and has been employed
by both Miladinovic and the creative Spanish Grandmaster, Bellon Lopez. This
id ea might appeal to those keen to get off the beaten Torre track, although 9 ... d5
(9. . . Rc8 10 h4!? h6 1 1 Rd1 Qc7 12 a3 d5 13 Kfl g6 14 e4 is a good example of the
creativity as well as the desire to attack which is required from White in this sub
vari ation, I.Miladinovic-S.Martinovic, Subotica 2008) 10 dxc5 (not ideal, but un
doubtedly White would like to at least semi-stabilize the central situation before
push ing anything on the kingside) 10 ... bxc5 1 1 e4 0-0 12 e5 Be7 13 h4 (Diagram 64)
1 3 . . .Qc7 14 0-0-0 Rfd8 gave Black his fair share of the play in I.Miladinovic
A. Kovacevic, Sibenik 2008.
b) 4 ...Qb6 5 Qc2 Ne4 fails to disturb White too much (this is chiefly why he went
c3 so early), and 6 Bf4 d5 7 Bd3 cxd4 8 exd4 Nc6 9 Nf3 Bd6 10 Bxd6 Nxd6 1 1 0-0
gives him an edge.
5 exd4 b6
93
Black usually fianchettoes thus, although once again he might first flick in 5 ...h6 6
Bh4.
6 Nd2 Bb7 (Diagram 65)

Play will transpose to the Torre White may play aggressively
NOTE: With Black's bishop well placed on b7, once again White has
nothing better than to return to standard Torre waters.
7 Ngf3 Be7 8 Bxf61?
This variation should really fall a little outside our scope, but I'll give it some cov
erage as Miladinovic has often played this way and then exchanged on f6. Instead
8 Bd3 0-0 9 0-0 (here too Miladinovic has oft-preferred 9 Bxf6 Bxf6 and 10 h4! g6 1 1
Qe2 Bg7 12 h 5 has given him some useful pressure i n a couple o f games) 9 . .d 6 10
.
Qe2 Nbd7 is a more traditional approach, and one which usually gives rise to a
long manoeuvring struggle. Black is very solid and might eventually arrange to
advance with ... e5 or ... b5-b4, but White can hinder such plans while aiming to
attack on one flank or the other.
B .Bxf6 9 Bd3 d6 10 Qe2 Nd7 (Diagram 66)
..
Both sides have developed sensibly. White has slightly the freer position, but
Black the long-term asset of the bishop-pair. That helps to explain why White of
ten nowadays pushes his kingside pawns, as we'll see him doing in Game 13.
Conclusion
Meeting the solid 2 ... e6 with the Torre-style 3 e3 was a fairly neglected approach
until recently. If White wishes to play in such vein, exchanging on f6 with attack-
94
The Posit i o n a l C hoice: 2 . e6
. .
jog d esi gns in mind is a welcome addition to the Torre stable. It's still very much
ea rl Y days, bu t thus far White has scored quite well in practice and so perhaps
B l a ck shouldn't be quite so keen to hurry to exchange on d4, common though that
ha s been.
Illustrative Games
ca me 13
o I.Miladinovic A.Martorelli
Rome 2005
1 d4 Nf6 2 Bg5 e6 3 e3 c5 4 c3 b6 5 Nd2 Bb7 6 Ngf3 Be7 7 Bd3 cxd4 8 exd4 d6 9

Bxf6 Bxf6 10 Qe2 Nd7 (Diagram 67)

Should ... bs be prevented? Black is a touch passive
11 a4!?
Thi s advance is far from unknown in the Torre: White wants to restrain any ex
pansion with ... bS and sometimes may even be able to attack himself on the
queenside with aS. However, it is a surprise to see Miladinovic employing it, as he
usually likes to push his kingside pawns in this line. Indeed, the earlier game
l. Mila dinovic-E.Dervishi, Bratto 2004, had gone 1 1 h4 h6 12 g4 Qc7 13 Rg1 g6!,
Which saw Black responding well on the kingside and preparing to go long. White
a ttem pted to dissuade such schemes with 14 a4! ?, but after 14 ... a5 15 gS hxgS 16
hx gS Bg7 17 0-0-0 0-0-0 1 8 Ba6 Bxa6 1 9 Qxa6+ Kb8 2 0 QbS d S the position re
rn a ined about equal until White arranged c4 and eventually won.
95
Miladinovic has also played without a 4 and 1 1 0-0-0!? Qc7 12 Kb1 0-0-0 13 Ba 6!
Bxa6 14 Qxa6+ Qb7 15 Qd3 dS 16 h4 h6 1 7 g4 (Diagram 68) 17 ... Be7 (17 ... g6 18 Rei
Kb8 19 Qe3 is also a touch better for White) 18 gS hxg5 19 hxgS Rxh1 20 Rxh1 gave
him a pull in I.Miladinovic-R.Fontaine, Serbian Team Championship 2005.
TIP: White should usually seize the chance to exchange the light
squared bishops on a6 in this line. Not only is Black's bishop quite
active on b7, but its exchange prevents Black from later operating
his two bishops in tandem (the dreaded strength of the bishop-pair
in action).
11...a6
Not bad, but Black might call White's bluff with 1 l ...Qc7!? 12 Ba6 Bxa6 13 Qxa6
0-0. Perhaps Miladinovic is still happy to go long here, although there's nothing
whatsoever wrong with the safer 14 0-0.
12 h41
Now that Black can no longer go long, the new Mr Trompowsky is happy to re
veal his true colours.
12 ... h6
Black's position isn't so easy to handle as even a grandmaster was later to dis
cover: 12 ... Be7 13 Ne4 Qc7 14 Kfl ! (White's king will be safest on the kingside, but
he would also like to keep a rook there; note too that 14 0-0-0?! bS! is a sacrifice
well worth avoiding) 14 ... h6 1 5 Re1 Kf8!? 16 Kg1 Bc6 1 7 Rh3! Qb7 (White retains
an edge too after 1 7... Bxa4 18 Bxa6 Bc6 19 Bd3 dS 20 Ng3 Bd6 21 NeS, although
this solid approach might have been a better defence) 18 c4 Nf6 19 Nc3 Bd7 20
Nd2 saw both sides engage in some heavy manoeuvring, but only White had
gained some handy central options in I .Miladinovic-V.lordachescu, Serbian Team
Championship 2008.
13 g4 (Diagram 69) 13 ... Be7
A decent way to sidestep the onrushing pawns, although one can also make a case
for the more provocative 13 ... g6!?, intending quick counterplay after 14 0-0-0 Bc6.
14 o-o-ol?
Very brave, but with the g-pawn having advanced, the king would no longer have
been entirely happy on f1 either.
14...g6
Once again we see Black doing his utmost to keep the kingside closed.
15 Rde1 Nf6!
The veteran Italian IM continues to play well, improving his knight while provok
ing White into closing the kingside.
16 gS Nds 17 Be4 hS 18 BxdSI
96
The Positi o n a l C h oice: 2 . . . e6
d caL but White had to prevent the position from stabilizin thereby allowing
Ra i
Bl ack to exploit that potential weakness on a4.
:s. s. .. BxdS 19 C4 Qc7 20 Kb1 Bxf3
fhis i s a
little compliant. Presumably Martorelli didn't like the look of 20... Bb7 21
d5!?, but the pawn can be declined and 2l.. .e5! 22 Ne4 0-0 doesn't seem at all clear
to me with a ...bS sacrifice high on the agenda.
2 1 Qxf3 o-o (Diagram 70)

Here come those pawns! Brave but correct from Black
Risky, but a decent choice so long as Black is brave and confident in his defensive
resources.
22 Qd 3
White might have tried to exploit his central pressure with 22 dS!? eS 23 Qe4, but
a fter 23 ... b5! 24 f4 Rab8 Black has enough counterplay as the position begins to
open up.
22... bs 23 Rxe61
is only leads to a draw with best play, but was undoubtedly White's best prac
hcal approach: he will no longer potentially suffer as the queenside opens and
Black will have to tread something of a defensive tightrope.
23 ...fxe6 24 Qxg6+ Kh8 25 Qh6+ Kg8 26 Qxe6+ Kg7 27 Qh6+ Kg8 28 Qg6+ Kh8 29
Qxhs+
White collects as many pawns as possible before deciding how best to include fur
ther u nits in
the attack.
29... Kg7 30 Qh6+ Kg8 31 Re1
Mi ladinovic must have looked long and hard too at 31 Qe6+ Kh8 (31 ...Kg7?! 32 h5!
97
only accelerates White's attack) 3 2 Ne4, but presumably wasn't a t a l l sure what
was going after 32 ... d5! 33 g6! Qf4! (33 ... dxe4? 34 g7+ Kh7 35 Rgl wins) 34 QxdS
Qf5 35 Rgl Kg7. Indeed, while he might have five pawns for the piece, I would be
surprised if White wasn't worse here: too many of his pawns are vulnerable.
31... Rxf21
Active and good defence, albeit only if followed up correctly.
32 Ne4 (Diagram 71)

Where shou ld the rook go? White's initiative is too strong
32 ... Rf4?
Making contact with the knight looks tempting, but it is actually defended hy the
check on g6 and so the white rook can move away. Correct was 32 ... Rf7! when I've
failed to find anything better for White than 33 Qg6+ (33 Rgl ? Rg7 34 Qe6+ Kh8
defends) 33 ... Rg7 34 Nf6+ Bxf6 35 ReS+ Rxe8 36 Qxe8+ and perpetual or some
variant on that theme.
33 Qe6+ KhB 34 Qh6+ KgB 35 Rg11
Now Black must prevent the massive check on f6 by returning an exchange.
3S ... Rxe4 36 Qg6+ KhB
The game score indicates 36 ... Kf8?, but I can't believe that the two players would
have overlooked the strength of 37 Rfl + even in a manic time scramble.
37 Qxe4 ReB
Alternatively, 37... Rf8 38 cxb5 axb5 39 aS and White is the clear favourite with
three pawns and the initiative for the piece.
38 g6?
There must have been something of a scramble in progress, otherwise White
98
The Pos ition a l C h oice: 2 . e6
..
\v o uld first have captured twice on bS before pushing the g-pawn.

3 s Bf6l
...
Black too overlooks the defensive resource 38 ... Qxc4 39 g7+ Kg8 40 hS BgS! when
any thing might still have happened.
39 hSI {Diagram 72)
No w White is back on track.
3 9 Qxc4 40 h6 1-0
...
The onrushing pawns are far too strong.
99
C h a pte r Fou r
T h e U n co m p ro m i s i n g 2 cs
m I ntrod uction
m The Da ri ng 3 Nc3
m 3 d 5 a nd the Va ga n i a n G a m bit
m Wh ite Exch a n ges on f6
m The 2 c 5 Ma i n Li ne: 5 Qc1
m The 2 c 5 Ma i n Li ne: 5 Qc 1 f5 6 g3
The U ncom p ro m i s i ng 2 . . c s
.
I n trod uction
1 d4 Nf6 2 Bgs cs (Diagram 1)

Striking out on the dark squares As ever b2 is a natural ta rget
With one or two exceptions we've thus far generally considered quite solid ap
p roa ches by Black; something which 2 ... c5 most certainly cannot be described as
being. Indeed, by advancing his c-pawn and freeing his queen at such an early
stage, Black shows that he is keen to fight for the initiative. The resulting play
tends to be quite dynamic and often revolves around whether Black's dark-square
play fully compensates for whatever concessions he has made (especially in the
case of 3 Bxf6 gxf6).
2 c5 might have only occurred in 1 1 % of all the Trompowsky games on my data
. . .
base, but it is well worth being ready for. Not only are there some sharp and rela
tively theoretical lines, but both sides can quickly drift into trouble should they be
unfamiliar with the resulting complex early middlegames.
While I have some sympathy for meeting 2 ... d5 and 2 ...e6 with 3 e3, here such an
a pproach is frankly unchallenging at best. Indeed, 3 e3 Qb6 already forces White
into either a grave dark-square weakness (4 b3) or an unpromising gambit, such
as 4 dxc5 Qxb2 5 Nd2 Qc3. Neither do I recommend allowing Black plenty of early
a ctiv ity with 3 c3 Ne4 4 Bf4 Qb6 (Diagram 2).
WARNING: Already White must be careful and 5 Qb3?1 cxd4 6 Qxb6

axb6 7 BxbS?I dxc3 8 Bes? Rxa211, as in D.Terentiev-J.Gallagher,
Liechtenstein 1990, is most certainly a trap to avoid.
Occasi onally 3 dxcS has been touted as a way to avoid theory, but this accelerates
101
Black's development and 3 ...e6 4 Nd2 (4 e4? Bxc5 5 e5 Qb6! sees the tactics backfire
on White) 4 ... Bxc5 5 e3 (and not, of course, 5 Ne4?? Nxe4!) 5 ... Nc6 6 Ngf3 0-0 is
hardly a particularly inspiring position to head for. Thus White usually plumps
for either 3 Bxf6 or 3 d5, but first we must consider something even more radical.
The Da ring 3 Nc3

1 d4 Nf6 2 Bg5 cs 3 Nc3 (Diagram 3)

Active and a mbitious Sicilian-like play beckons
White wastes no time developing and shows himself happy to bring his queen
into action. This approach is quite sharp and may appeal especially to players
with some Veresov and/or Open Sicilian experience. However, it's never really
caught on at grandmaster level and has been played in under 10% of all the games
I could find with 2 ... c5.
3 ...cxd4
A natural move, but by no means the only choice:
a) One drawback to 3 Nc3 is that 3 ... d5!? takes play into a line of the Veresov
which is generally considered to supply decent counterplay. I'm afraid that I must
really draw the line here - those interested in further details of this reversed Chi
gorin position are referred to Nigel Davies's 2003 work on the Veresov.
b) 3 . . .Qb6 reveals another move order issue: White has nothing better here than 4
d5, taking play into the Vaganian Gambit and our next section.
c) However, 3... Qa5 isn' t so challenging and allows White a fairly pleasant choice
after 4 Bxf6 gxf6 between 5 d5 and 5 e3.
102
The U n co m p ro m i s i n g 2 ... c s
4 Qxd4 Nc6 5 Qh4 (Diagram 4)

The most active square for the queen and White's main choice by some margin.
s e6...
Bl ack prepares to neutralize the pressure down the h4-d8 diagonal with ... Be7. A
sensible approach, but not the only one:
a ) 5 . . d6 6 e4 e6 7 0-0-0 Be7 resembles the Richter-Rauzer Attack in the Sicilian, and
.
is likely to transpose to our main line after 8 f4 Qa5.

b) 5 . Qa5 6 0-0-0 d6 7 e4 Be6!? sees Black prioritizing counterplay over develop
. .
ment, as advocated in Chess Openings for Black, Explained. I agree with Alburt,
Dzindzichashvili and Perelshteyn that 8 Bxf6?! gxf6 9 Kb1 f5 1 0 exf5 Qxf5! (Dia
gram 5) is good for Black (although 10 ... Bxf5? 1 1 RdS Bxc2+ 12 Kcl ! Qb4 13 Bc4
leaves Black in serious trouble), but White doesn't have to help develop Black's
kingside so. Indeed, 8 a3 looks like a better bet, after which 8 ... h6 (8...Rc8 9 f4
Nd8?! saw Black desperate to land a blow on c3 in I.Nataf-A.Houriez, Cannes
1996, but De la Villa's 10 Bxf6!? gxf6 1 1 BbS+ keeps control and leaves White com
fortably better) 9 Bxf6 gxf6 10 Nd5 Bg7 1 1 Nh3 f5 12 Nhf4 was quite complex, but
probably a little better for White in ].Ibarra Jerez-J.Valmana Canto, La Roda 2005.

Black's bishops look good White's king isn't all that safe
c) S . . bS?! is another active idea from Black, but one rather well met by 6 e4 a6 7
.
0-0-0 when he faces a strong central initiative: for example, 7... h6 (7...Qa5 8 Kb1 d6
9 Nge2 Be6 1 0 Ncl! ? Rc8 1 1 Nb3 calmly leaves Black suffering in a Rauzer-type
position, A.Jackson-O.Gladyszev, Port Erin 2000) 8 NdS Bb7 9 Bxf6 gxf6 10 Ne2
QaS 1 1 Kb1 Rg8 12 f3 gave White an edge in I.Nei-B.Diesen, Helsinki 1990, and 1 2
a3!? might b e even stronger with the direct idea o f 1 2 ... b4 13 Qh3!.
6 e4
103
White can also castle first, but after 6 0-0-0 Be7 7 e4 we're back in our main line.
6 ... Be7
Prie draws attention to 6... Bb4!?, which is logical enough, but has only been
played twice from what I can see. Indeed, it might even be worth sidestepping
with 6 0-0-0.
TIP: When planning your repertoire never forget about move-order

and transpositional tricks. Indeed, swapping two moves around to
avoid a certain nuance is key to a number of opening lines.
7 0-0-0
White wants to bring his f-pawn into play before developing his king's knight, but
is probably best off castling first. Admittedly 7 f4 Qb6!? (7 ...b5!? is another idea) 8
0-0-0 Qe3+ 9 Kb1 Nxe4 10 Nxe4 Qxe4 is far from clear, but I'm not entirely sure
that White has enough compensation after De Ia Villa's 1 1 Bd3 Qxg2!? (1 1 ...Bxg5
12 QxgS Qd4 13 Nf3 Qf6 14 QhS h6 1 5 fS! dS 16 g4 gave White enough pressure in
J.Gil Gonzales-F.Arnold Perier, Spanish Team Championship 1996) 1 2 Bxe7 Nxe7
13 Ne2 Ng6 14 QhS.
7 ... Qa51 (Diagram 6)
Black wastes no time activating his queen while also sidestepping potential trou
ble on the d-file. The alternatives are perhaps not so bad, but do allow White to
showcase his assets:
a) 7...0-0 8 f4 h6?! looks rather risky, but does require some precision from White
with 9 Nf3! being the way to go: 9 ... hxg5?! 10 NxgS Qc7 (or 10 ... g6 1 1 Qh6! Re8
when Prie analyses the crushing continuation 1 2 eS Bf8 13 Qh3 Nh5 14 g4 f6 15
gxhS! fxgS 16 hxg6) 1 1 eS! NxeS 1 2 NdS! exdS 13 fxeS QxeS 14 Re1 QfS! ( 14...Qd6?
15 Bd3 g6 16 Qh6 wins) 15 Rxe7 Qg4! 16 Qxg4 Nxg4 17 BbS saw Black defend well
but still come out behind in M.Krasenkow-J.Degraeve, Cappelle Ia Grande 1990.
b) 7... a6! ? 8 f4 bS might be met by the calm 9 a3, but in practice White has usually
gone forwards with 9 eS! b4 10 exf6 gxf6 1 1 Ne4 fxgS 12 fxgS, which is hard to as
sess, but probably slightly favours him, R.Vaganian-R.Knaak, Tallinn 1979.
8 f4 d6 (Diagram 7)
Black contests the eS-square and has tricky ideas of ... h6 in mind.
9 Nf3
White has also retreated his queen, but Dembo is surely right that 9 Qe1 0-0
(9 . . . h6!? 10 Bh4 gS is a tempting, very Rauzer-like alternative) 10 Nf3 Rd8! cannot
be bad for Black. Perhaps White's best try is thus 9 BbS!?, cutting off the black
queen from eS, as we'll consider in Game 14.
9... h6
Black puts the question to the bishop without delay. Instead 9 ... Bd7 1 0 BbS trans-
104
The U ncom pro m i s i n g 2 ... c s
poses t o Game 14, but also allows White the option o f 10 Nd2!?.

It's Richter-Rauzer-like! Sharp but okay for Black
10 Bb5
White continues to develop as quickly as possible. He has also tried to open the
position with 10 e5?!, but 10 ... dxe5 1 1 fxe5 Nxe5 appears to offer insufficient com
pensation: 12 Bb5+ (or 12 Nxe5 Qxe5 1 3 Bb5+ Kf8 14 Bf4 Qc5 and while Black can't
castle, there doesn't appear to be a particularly promising way for White to take
advantage) 12 ... Bd7! 13 Nxe5 Bxb5 14 Nxf7!? Kxf7 1 5 Bxf6 Bxf6 16 Qh5+ Kg8 1 7
Qxb5 Bxc3 18 Qxb7 Bxb2+! 19 Qxb2 Kh7 leaves White i n trouble, S.luldachev
G.Serper, Tashkent 1993.
10 Bd7
...
Black threatens to go long and so White must act, but 11 e5 dxe5 12 Bxc6 Bxc6 13
Nxe5 (Diagram 8) 13 ... Rc8 (Serper) leaves the onus on White to prove it's equal.
Perhaps 14 Rhe1 !? Bxg2 (14 ... 0-0? 15 Bxh6 gxh6 16 Qxh6 prepares a fatal rook lift,
but 14 ... Qc5!? is possible) 1 5 Bxf6 is best, although here Black has a pleasant choice
between 15 .. . gxf6 16 Ng4 Rxc3 1 7 Nxf6+ Kf8 18 Nd7+ Ke8 19 Nf6+ (Diagram 9)
when White must force perpetual and 15 ... Bxf6!? 16 Qg3 Bc6 17 Nc4 Qc5 18 Nd6+
Ke7 19 Nxc8+ Rxc8 with sufficient compensation for the exchange.
Conclusion
The rare but dangerous 3 Nc3 might well catch out even quite an experienced
2 . . c5 practitioner. Indeed, White has scored 56% with it from just under 400
.
games. However, if Black knows his stuff, he obtains decent counterchances and
W hite has made just 35% from 20 games after 9 Nf3 in our main line. I must admit
10 5
Sta rt i n g O u t: T he Trom pows ky Attack
that 3...d5 also puts me off 3 Nc3, although this has actually been quite rare in
practice, possibly because most players answer the Veresov with 3 ... Nbd7 and not
3...c5.

White forces a d raw White wants to enforce es
Illustrative Games
Game 14
D J.Brandics F.ott
Kecskemet 1990
1 d4 Nf6 2 Bg5 cs 3 Nc3 cxd4 4 Qxd4 Nc6 5 Qh4 e6 6 o-o-o Be7 7 e4 d6 8 f4 QaS 9
BbSI?
White not only pins the knight, but may even exchange it in some lines as he aims
to force through the e5-advance.
9 o-o?l
...
Either an extremely brave or a rather naive choice. Black has more usually pre
ferred 9 ... Bd7 10 Nf3 (Diagram 10) and then:
a) 10 ... 0-0-0?! 1 1 e5! dxe5 12 fxe5 Nd5? (De Ia Villa points out that even the varia
tions 12 ... Nxe5 13 Nxe5 Bxb5 14 Nxf7 Rxd1+ 15 Rxd1 Rf8 16 Nxb5 Qxb5 1 7 Bxf6
gxf6 18 Qxh7 and 12 ...h6 13 Rxd7! Kxd7 14 exf6 hxg5 15 Qxg5! Bb4 16 Bxc6+ bxc6
1 7 Qxa5 Bxa5 18 Ne5+ leave White firmly in charge of proceedings) 13 Nxd5 exd5
14 Bxc6 Bxg5+ 15 Nxg5 Bxc6 16 Nxf7 won the exchange and the game in M.Illescas
Cordoba-Rivera, Barcelona 1984.
b) 10 ... a6 also fails to hinder White: 1 1 Bxc6 Bxc6 12 e5 dxe5 13 fxe5 Ng8?! 14 Qg3
106
T h e Uncom p ro m i s i n g 2 ... c s
was horrible for Black i n J.Johansson-I.Kanko, Helsinki 1991, and while the brave
13 ...Bxf3 improves, it does allow White the choice of two tempting sacrifices in 14
gxf3 Qxe5 15 Rhg1 and 14 exf6!? Bxd1 15 fxe7, with dangerous play in both cases.
c) 10 ...h6! is again critical: 1 1 e5 dxeS 12 Bxc6! (best; note that the tempting 12
Rxd7? falls short after 12 ... Kxd7! 13 Nxe5+ Kc7) 12 ... Bxc6 13 Nxe5 Rc8 (Black
wisely sidesteps 13 ... 0-0? 14 Bxh6! gxh6 15 Qxh6), and we've transposed to Ser
per's analysis in the main line of our last theory section. There I suggested that
Black could grab and that 14 Rhe1 Bxg2 is roughly equal.
10 Nf3 h6? (Diagram 11)

Can White sacrifice? The attack is crushing
This fails to convince, but with the black queen cut off, so too does 1 0... Rd8?,
which fails to the simple 1 1 e5 dxe5 12 fxe5 Rxd1+ 13 Rxd1, winning material. It
looks a little ugly, but quite possibly Black had to take refuge in 10 ... e5, thereby at
least holding White up in the centre.
11 Bxc6?
So was White correct to reject the standard sacrifice 1 1 Bxh6! gxh6 12 Qxh6? I
rather think not! Just look at the lines 1 2 ... Nxe4 13 Bd3 Nf6 14 Ng5 (Diagram 12)
followed by a lethal bishop check, 12 ...Qb6 13 NgS Qe3+ 14 Kb1 and there's no
d efence to the threat of 15 e5 and 16 Nce4, and 12 ... Rd8 13 NgS when the f6-knight
will be shifted, either by e5 or 13 ...e5 14 Nd5.
TIP: In this variation ... h6 might be a critical way for Black to play,
but it usually only works when he hasn't castled. If he has and then
goes ... h6, do immediately get to work seeing whether the standard
bishop sacrifice is onl
107
Sta rting Out: The Trom powsky Attack
11... bxc6
Ott correctly avoids 1 l . .. hxg5? 12 Nxg5 when 12 ...bxc6 1 3 e5 transposes to the
game.
12 esl?
Critical as White clearly has no desire to trade on f6 and retreat his queen to el .
Again we must also consider the alternative sacrifice, but here 12 Bxh6 gxh6 1 3 e5!
(13 Qxh6? Qh5 defends) 13 ... dxe5 14 fxe5 Nd5 is much less clear than before and
15 Rxd5 exd5 1 6 Qxe7 Rb8 1 7 Qh4 Qb4 gives Black enough counterplay.
12 ... hxgs?
Black might have been rated 2265, but evidently all the early pressure had got to
him, as the text badly underestimates White's attacking chances. Instead Ott had
to find 12 ... dxe5! 13 Bxh6 gxh6 when 14 fxe5 transposes to our last note, but White
also has the surprising 14 Nxe5!?. Then 14 ... Kh7 (14 ...Qb6!? might well be possible
too, as it's not so easy to lift a rook after 15 Qxh6 Rd8 16 Rxd8+ Bxd8) 15 Nxc6 QcS
16 Nxe7 Qxe7 1 7 g4 Bb7 18 g5 Ng8 19 Rhg1 gives White ongoing pressure for the
piece, although I would be surprised if it was objectively quite enough after
19 ... Rad8.
13 Nxgs dxes?
This loses, although Black's long-term chances for survival after even 13 ... Nh5 14
Qxh5 Bxg5 15 Qxg5 dxe5 16 fxe5 f6 would hardly have been all that high.
14 Rd3 (Diagram 13)
Diagram 13 (B) Diagram 1 4 (B)

A killer rook l i ft Not such a quiet choice
The arrival of a further piece into the attack heralds the end.
14...Qb4
108
The Unco m p rom i s i n g 2 ... c s
Desperation, but there was n o defence a s shown by the line 14. . .g6 (to play ... Nh5)
15 Qh6 Ba3 16 Rh3 Bxb2+ 17 Kb1 and mate follows.
15 Nce4 Qxe4 16 Nxe4 Nxe4 17 Qxe7 exf4
Black might have two pieces and a pawn for the queen, but his position is far too
loose and Brandies was never going to fail to convert from here:
18 Rh3 Ba6 19 Qh4 f5 20 Rd1 Nf6 21 Qxf4 Rac8 22 Ra3 Be2 23 Re1 Bh5 24 Rxa7 Bf7
2 5 a4 Ne4 26 as g5 27 Qe5 Rfd8 28 Rxe41 fxe4 29 Qxg5+ 1-0
3 dS a nd the Vaga nian Ga mbit

1 d4 Nf6 2 Bg5 cs 3 ds (Diagram 14)
White pushes on and shows that he is happy to contest a Benoni structure. It
might not appear so to the uninitiated, but the text is actually quite a sharp ap
proach - with his next move Black can pretty much force White to gambit a pawn.
3 .Qb61?
..
And this is the critical response, hunting down White's b-pawn. Alternatively:
a) 3 ... Ne4 is by some margin Black's second main choice.
NOTE: White should be happy enough to meet 3 ... Ne4 by transpos

ing to the first section of Chapter Seven with 4 Bf4 .
Hodgson did once wonder about 4 Bel, but this is probably best avoided as 4 ... e6!
begins slightly annoying, early counterplay.
b) 3 ... d6 4 Nc3 Qb6 (investing a tempo with 4 ...h6 5 Bxf6 exf6 is the independent
try; Black's kingside is much safer than in the 3 Bxf6 gxf6 variation, but White has
free development and 6 e4 g6 7 Bd3 Bg7 8 Nf3 0-0 9 0-0 Nd7 10 Nd2! gave him a
pleasant edge in J.Gallagher-S.Knott, British Championship, Scarborough 2001) 5
e4 Qxb2 6 Bd2 transposes to our main line, but also allows White the extra option
of 5 Qd2!? Qxb2 6 Rb1 Qa3 7 e4.
c) 3 .. g6 4 Nc3 (best; Black's position is just too easy to play after 4 Bxf6 exf6 5 c3
.
Bg7 6 e3 0-0 when ... f5 will follow and White will find it hard to restrain an even
tua l ...b5) 4 ... Bg7 5 e4 (Diagram 15) takes play into an unusual type of Schmid Be
noni. This isn't such a bad approach from Black, but neither do I think that
White's bishop has been committed too early: for example, 5 ... d6 (Hodgson has
al so shown the way against 5 ... h6: 6 Bf4! d6 7 h3 Qb6 8 Bb5+ Nfd7 9 Rb1 Qa5 10
Bd2 0-0 1 1 Nf3 a6 12 Be2 Nf6 13 0-0 Qd8 14 a4 left Black solid but typically a little
too cramped in J.Hodgson-P.Cramling, Bern 1996) 6 f4 0-0 7 Nf3 a6 8 a4 Nh5?! 9 f5!
Nd7 10 Be2 Rb8 1 1 0-0 Nhf6 1 2 Qel and Black was mown down by the direct at
ta ck in J.Hodgson-J.Hjartarson, Bermuda 1997.
109
Sta rti n g Out: T h e T ro m pows ky Attack

It's a bit Pirc-like Wisely halting Black's pawns
d) 3 ...e5 4 Nc3 d6 5 e4 might also come about via a l . ..cS move order. Here Black's
knight has been developed a little early ( ... Be7-g5 will take time to arrange), and
White has access to the c4-square, factors which he exploited with 5 ... Be7 6 Nf3 0-0
7 Be2 (White must be alert to tactics on dS, but as Gallagher points out, they sim
ply don't seem to work) 7... Nbd7 8 0-0 Kh8 9 a4 Ng8 10 Be3 g6 11 Nd2 fS 12 exfS
gxfS 13 f4 (Diagram 16) to obtain control of the position in A.Miles-C.Matamoros
Franco, Ubeda 1997.
4 Nc31 (Diagram 17)
Nothing else makes any sense. Indeed, White must sacrifice and with the text he
introduces the Vaganian Gambit, named after the famous Armenian Grandmaster
who crushed all-comers with it in the Seventies.
WARNING: Anyone who even considered 4 b3?1 must be rather

inexperienced in the Trompowskyl White must always beware
weakening his dark squares thus and here 4 ... Ne4 5 Bf4? Qf6 is most
certainly a position to avoid.
4...Qxb2
Critical. Again an unprepared opponent might back out of the challenge, though,
and 4 ... d6 5 e4 g6 has been seen. With b2 still hanging, 6 f4 a Ia Hodgson is now
too ambitious, but 6 BbS+ (or just 6 Rb1) 6 ... Bd7 7 a4 Bg7 8 Nf3 a6 9 aS Qc7 10 Bd3
0-0 1 1 0-0 Bg4 1 2 h3 Bxf3 13 Qxf3 gave White a pleasant Schmid Benoni-style edge
in I.Cheparinov-E.Iturrizaga Bonelli, online blitz 2009.
5 Bd2 Qb6
Fairly essentia l. The queen scurries back towards safety on c7 or d8.
110
The Unco m p r o m i s i n g 2 . . . c s
6 e4
Already White's compensation is quite evident: a strong centre and the freer de
velopment. Indeed, it's not unknown for him to blow Black away with a quick f4
and eS advance; an attack which has helped to attract many to the white cause.
Black, on the other hand, remains quite solid and can hope to gradually unravel
and exploit his extra pawn should he manage to keep the central situation under
control.
6... d6

White must sacrifice b2 Trying to slow White u p
Black's most popular choice, but a very important alternative i s 6. . .e5 7 f4 d 6 (Dia
gram 18), logically closing the centre. This approach has been recommended of
late in Yelena Dembo's Fighting the Anti-King's Indians and is a fairly critical chal
lenge to the Vaganian Gambit. White must now choose between:
a) 8 Rb1 appears tempting, but driving the queen back seems to help Black: 8 ... Qc7
9 Nf3 (compared with our main line, 9 BbS+?! Bd7 1 0 fxeS dxeS 1 1 Nf3 a6 only
helps Black to develop) 9 ... a6!? is Prie's still untested idea. His analysis continues
10 a4 exf4! 1 1 Bxf4 Nbd7 12 Nd2 Be7 13 Nc4 0-0 when Black will regroup with
... Ne8 and ... Bf6. That won't completely unravel his forces, but does leave the onus
on White to demonstrate that he has enough for the pawn.
b) 8 Nf3 Nbd7 (8 ... exf4 appears risky, but certainly has its merits as shown by the
solid line 9 Bxf4 Nbd7 10 Rb1 QaS 1 1 Qd2 Be7; thus I wonder if White wouldn't do
better to investigate 9 eS!? dxeS 10 NxeS Be7 11 BbS+, fighting hard for the initia
tive) 9 fxeS (Wells's 9 Bb5!? a6 10 Bxd7+ Nxd7 1 1 0-0 Be7 12 Rb1 has in mind the
breakthrough 12 ... Qd8?! 13 fxe5 dxe5 14 d6! Bxd6 15 Bh6, but as he observes,
12 ... Qc7 is a much tougher nut to crack) gives Black a tough decision.
111
sta rting Out: The Trom pows ky Att a c k
NOTE: The Dutch grandmaster Friso Nijboer, a leading proponent of

Black's set-up, has pointed out that the seemingly freeing exchange
of knights actually only helps White: 9... Nxe5?1 10 Nxes dxes 11
Rb1 Qd8 12 Bb5+ Bd7 13 Bg5 with strong pressure.
Thus Black should prefer 9 ... dxe5 10 Bc4 Bd6 (we'll see in Game 14 why 10 ... Be7 1 1
Rb1 ! QdB 12 BgS is dangerous for Black, but 10 ... a6!? 1 1 0-0 Be7 may well improve;
this gives the queen the c7-square and 12 Rb1 Qc7 13 Nh4 Nb6 14 Be2 0-0 15 Qe1
NeB appears acceptable enough for the second player) 1 1 Nh4! (White must not
delay: 1 1 0-0 0-0 12 Nh4 QdB! 13 NfS Nb6! 14 Be2 BxfS 15 RxfS NeB saw the then
prodigy supply a model for Black in B.Soos-P.Leko, Budapest 1993) 1 1 ...0-0 1 2 NfS
(Diagram 19) 12 ... NeB 13 0-0, which reaches quite a critical and amazingly still un
tested position. White is better off than in Soos-Leko, but after 13 ... QdB! still needs
to find a counter to Black's idea of ... Nb6 and ... BxfS. Perhaps 14 Qf3 Nb6 15 BbS
might be tried, while I also wonder about the earlier idea of 13 a4!? QdB 14 aS.

White has decent pressure Enough for the pawn
c) The jury is also still out on the cramping B fS!?. One critical line, as recom
mended by Dembo, runs B ... hS! (an essential prophylactic inclusion: B ... Nbd7?! is
just too easy for White who can expand with 9 g4 h6 10 h4 before completing his
development) 9 Nf3 Nbd7 10 Bd3 a6. Black remains very solid here, but White is
most certainly not without his compensation. If push comes to shove, I'd guess
that the chances are roughly balanced with 1 1 a4!? (Diagram 20) probably the way
to start, keeping a careful eye on the queenside situation ( ... c4 can be an issue and
White would like to be in a position to win a6 should Black break with ...b6 after
aS).
Before we return to 6 ... d6, keeping Black's options open, it's worth noting that
6... e6 7 f4 d6 is another route into our main line. Here 7... exd5 has also been seen,
112
The U ncom p ro m i s i n g 2 . . . c s
bu t i s risky and 8 e5 d4?! (8. . .Ne4!? 9 Nxd5 Qd8 might be a better try, although I
thi nk that Wells's 1 0 Bel !? with the ideas of 10 ... d6 1 1 Bd3 and 10 ...Qa5+ 1 1 c3 of-
fers White challenging compensation) 9 exf6 dxc3 10 Bxc3 gxf6 (10...Qe6+?! 1 1 Kf2!
only makes matters worse for Black) 1 1 Bc4 left Black with a rather ragged struc
ture in D.Boros-D.Howell, Copenhagen 2006.
7 f4 (Diagram 21)

White th reatens 8 eS Model play from Vaga nian
7 .e6
..
Again Black counters in the centre, but he has also tried a number of alternatives:
a) 7...g6? 8 e5 just gives White everything he could want. It's oft-quoted but to dem
onstrate the power of White's initiative I can do no better than to give the classic
demolition R.Vaganian-V.Kupreichik, USSR Championship, Leningrad 1974:
8 .. dxe5 (8... Nfd7 9 Nf3 Bg7 10 Rb1 Qd8 1 1 e6! fxe6 12 Ng5 gave White too strong an
.
initiative too in R.Vaganian-V.Jansa, Kragujevac 1974) 9 fxe5 Nfd7 10 Nf3 Bg7 1 1

Rb1 Qd8 1 2 e6! (Diagram 22) 1 2... fxe6 1 3 Ng5 Nf6 1 4 Bb5+ K8 1 5 dxe6 a6 1 6 Be3!
Qa5 1 7 0-0 h6 (17... axb5 18 Rxb5 Qc7 19 Nd5 Qe5 20 Bxc5 is crushing, as pointed out
by Wells) 18 Qd3 Kg8 19 Qxg6 Bxe6 20 Nxe6 Rh7 21 Rxf6 Nd7 22 Bxd7 1-0.
TIP: This inspirational game is well worth studying again, not least
because the key blows occurred on the light squares; a far from un
known scenario when Black doesn't close the centre with ...es.
b) One can understand the logic behind 7 ... Bg4?, but Black doesn't have time to
develop his queenside before touching his kingside pawns: 8 Be2! Bxe2 9 Qxe2
Na6 (9 ... a6 10 Rb1 Qc7 11 e5 is excellent too for White) 10 Rb1 Qc7 1 1 e5! Nd7? 12
NbS Qd8 13 exd6 Nb6 14 Nf3 NxdS 15 Ne5 was already pretty much all over in
R.Palliser-M.White, York 2004.
113
Sta rti n g Out: The Tro mpows ky Attack
c) 7...e5 is a much better idea, transposing to the notes to Black's 6th move, above.
d) 7... Nbd7?! doesn't, though, transpose there too because the straightforward 8
Nf3 e5?! 9 dxe6 fxe6 10 Ng5! is rather strong.
e) 7...Qd8 8 Bc4!? cuts across any plans of ...e5 Black might have. That can now be
met en passant and 8 ... Nbd7 9 a4!? Nb6 10 Bb5+ Bd7 1 1 Qe2 (Wells) gives White
good compensation.
f) If Black wishes to retreat his queen out of harm's way, 7 ...Qc7 is probably a bet
ter method. Black's point is revealed after 8 Nf3 (the calm 8 Bb5+ Nbd7 9 Nf3 a6 1 0
Bxd7+ Nxd7 1 1 0-0 g6 1 2 e5 Bg7 appears fine for Black, a s i n J.Vilela d e Acuna
Y.Gonzalez Vidal, Barcelona 2007, but De Ia Villa's 8 Bc4!? a6 9 a4 g6 1 0 Qe2
would probably be my choice when Wells suggests that Black must preface ... g6
with 10 ... Nfd7!?) 8 ... Bg4 when even the vigorous 9 e5! (White cannot afford rou
tine play: 9 Bb5+ Nbd7 10 0-0 a6 1 1 Bxd7+ Nxd7 12 Qe1 Bxf3 13 Rxf3 g6 left the
Trompowsky expert on the way to victory against our favourite opening in
N.Povah-P.Wells, Portsmouth 2003) 9... dxe5 10 fxe5 Bxf3 11 Qxf3 Qxe5+ 12 Be2
(Diagram 23) was no more than unclear after the highly-creative defence 1 2...g5!?
13 Rb1 g4 14 Qd3 Bg7 15 Rxb7 0-0 in J.Ehlvest-S.Kudrin, Reno 2005.

It's stil l unclear A well-judged decision
After that important theoretical interlude we return to 7 ... e6:

8 Rb11 (Diagram 24)
NOTE: This move is well worth flicking in. It wasn't after 6...es 7 f4
d6, but here White would like to meet Bb5+ Bd7 with an exchange
on e6, thereby opening the centre for his more actively-placed
pieces.
114
The U ncom pro m i s i n g 2 .. c s
.
White has also tried the immediate 8 dxe6, but 8 ... Bxe6! is a good response: for
example, 9 fS (or 9 Rbl Qc7 1 0 Nf3 Be7 and Black is already slightly for choice
with White's central options failing to impress) 9... Bd7 10 Nf3 Be7 1 1 Bc4 Nc6 1 2
0-0 Ne5 13 NxeS dxeS and White's control o f d S gave him no more than some
compensation in A.Fier-E.Tsuboi, Sao Paulo 2007.
s . .Qc7
.
Black keeps his queen in touch with both the b7- and eS-points. He has also re
treated with 8 ...Qd8, but 9 dxe6 (9 BbS+!? Bd7 1 0 dxe6 fxe6 1 1 eS is also pretty
dangerous, as once played by Hodgson) 9 .. .fxe6 (quite possibly Black should re
turn the pawn for some development with 9 ... Bxe6!?) 10 eS! is very dangerous:
10 ... dxe5 (or 10 ... Nd5 1 1 Bd3 Nxc3 12 Bxc3 dS 13 QhS+ Kd7 14 fS! with a strong
initiative in V.Golubenko-O.Sepp, Eesti 1996) 1 1 fxeS Nd5 12 Bd3 (Girchurkin's 12
Qg4!? is yet another possibility) 1 2... Nxc3 1 3 Bxc3 QgS 14 Qf3! Nc6 15 Nh3 Qh4+
16 Nf2 Be7? 1 7 g3 Qa4 18 Bb5 soon forced Black to resign in A.Lahiri-V.Kamble,
Hyderabad 2006.
9 Bb5+
White's main choice. He has also tried 9 dxe6 when 9 .. .fxe6 10 Nf3 a6 11 a4 Nc6 12
Bc4 maintains decent compensation, but again 9 ... Bxe6! is much more of a prob
lem.
9... Nbd7
Black has also blocked with 9 ... Bd7 when 10 Qe2! a6 (Prie observes that 10 ... Be7 1 1
dxe6 fxe6 1 2 e5 dxeS 1 3 fxeS Nd5 14 Bd3 once again appears dangerous for Black,
and even 1 0...e5!? 1 1 fxeS dxe5 fails to wholly convince after 12 Bc4 Bd6 13 Nf3 0-0
14 0-0 when White will be quick to increase the kingside pressure) 1 1 Bxd7+ Qxd7
(11 ...Nbxd7 1 2 dxe6 fxe6 13 Nf3 leaves Black's light squares rather tender) 12 dxe6
fxe6 (Diagram 25) isn't totally clear, but White has at least decent compensation
here, such as with Palos's 13 e5 (13 Nf3 Nc6 14 0-0 is a simpler alternative with the
idea of 14 ... Be7 15 Na4! Qc7 16 Nb6 Rd8 17 NgS Nd4 18 Qd3) 13 ... dxe5 14 fxe5
Nd5 15 Nh3!?.
10 dxe6
White hurries to open the position and will now aim to increase the pressure
against the vulnerable light-square defender on e6.
10 ...fxe6 11 Nf3 (Diagram 26)
White's superior development and safer king position offer him good compensa
tion, but matters remain far from clear as we'll see in Game 1 6.
Conclusion
The Vaganian Gambit remains a dangerous weapon in the hands of an attacking
115
player . That said, a well-prepared opponent who is happy to undertake some de

fending might accept the gambit. White has scored 55% from over 300 games a fter
6 e4; a figure which was once higher until the evolution of some decent defences
for Black in the shape of 6 ...e5 and 6 ... d6 7 f4 e6/e5. They have necessitated a fair
amount of coverage, but even against them White retains decent compensation.

White m ight target e6 A ha ndy lead in development
Il lustrative Games
Game 15
D L.Bruzon Bautista F.Nijboer
Wijk a an Zee2004
1 d4 Nf6 2 Bg5 c5 3 d5 Qb6 4 Nc3 Qxb2 5 Bd2 Qb6 6 e4 e5 7 f4 d6 8 Nf3 Nbd7 9 fxe5
dxe5 10 Bc4 Be7 11 Rb11
White might play 1 1 0-0 0-0 12 Qe1 with ideas of Nh4 and Qg3, but the text is even
more powerful with the black position temporarily uncoordinated.
11 ...Qd8
The most natural retreat. Instead 1 1 ...Qd6?! 12 Nh4! is pretty strong, especially as
12 ... Nxe4?! 13 Nf5 Qg6 14 0-0 Nxd2 15 Qxd2 Bg5 16 Qe2 0-0 17 h4 leaves Black in
all sorts of trouble, but even with the b5-square unguarded 1 1 ...Qc7!? may not be
so bad.
12 Bg5 (Diagram 27)
White wastes no time unobstructing his queen and now the d6-advance is most
certainly on the agenda.
116
T h e U ncom p ro m i s i n g 2 . . . c s
12 ... h6?
B lac k surprisingly underestimates the danger. He had to prefer either 12 ...Qa5 1 3
Qd2 a 6 14 0-0 0-0 (and not 14 ...b5? 15 d6!) 15 a4 when Bruzon i s surely right that
White has good compensation (the d-pawn is mobile and Nh4 on the cards in any
case), or Prie's sensible 12 ... Bd6!?, acquiescing to a tempo loss in order to blockade
the key pawn.

Watch that d-pawn ! How to undermine the defences?
13 d61 hxgs 14 Nxgs

This is presumably what Nijboer had missed. White will now regain the piece un
der highly favourable circumstances.
14...0-0
Black's cause is far from ideal too after 14 ... Rf8 15 dxe7 Qxe7 16 NbS when he will
surely lose material in short order.
15 0-01
Once again White can do better than recapture the piece.
Cf
TIP: When calculating an exchanging sequence, always keep an eye
out for intermezzos or in-between moves.
15 ... Nb6
Easy to criticize, but I don't think there was a defence: 15 ... Bxd6 16 Qxd6 illus
tra tes White's point and after 16 ...Qe8 (or 16 ... Nb6 17 Bxf7+ Kh8 18 QxeS) 17 NdS!
Nxd5 18 BxdS Black is completely helpless against the threat of 19 Qg6, which
18 ... Nf6 19 Rxf6 merely renews.
Finkel meanwhile points out that 15 ... Qe8 16 dxe7 Qxe7 1 7 Qe1 followed by Qh4 is
117
Sta rt i n g O ut: The Trom pows ky Attack
completely crushing too: there's no defence to Nd5 followed by removing the de

fenders of h7.
16 dxe7 Qxe7 (Diagram 28} 17 Rxb61
Very aesthetic, albeit not so hard to find when one considers that White would
like to remove the defensive knight from f6.
17 ... axb6 18 Nds Nxds
It would have been all over too after 18 ... Qd8 19 Nxf6+ gxf6 20 Bxf7+ Kg7 21 Qh5
(Bruzon).
19 Qh5
The point of White's combination. Black can now give up his queen to avert mate,
but even this isn't going to keep him in the game for long.
19 ... Qxgs 20 Qxgs Nf4 21 Qxes Be6 22 Rxf41?
The Cuban wants to carry on attacking. A fair enough policy, although a pretty
good case can be made for 22 Bd5.
22 ... Bxc4 23 Qhs Rxa2 24 Rh4 Ra1+ 25 Kf2 (Diagram 29)

The attack rages on The a2-g8 diagonal will be key
2s ...fs?
A weak defence. Much more resolute would have been 25 ... f6 26 Qh7+ Kf7 27 Rg4
Rg8, although the attack should be decisive after 28 e5!: for example, 28 ... Rfl + 29
Kg3 Be6 30 Qg6+ Ke7 31 Qd3!! Bxg4 32 Kxg4 and Black cannot save both his loose
rooks, as pointed out by Bruzon.
26 Qh7+ Kf7 27 exfs 1-o
118
The U ncom p rom i s i ng 2 . . . c s
Game 16
0 W.Siewert M.Noble
Correspondence2007
1 d4 Nf6 2 Bg5 cs 3 d5 Qb6 4 Nc3 Qxb2 5 Bd2 Qb6 6 e4 d6 7 f4 e6 8 Rb1 Qc7 9 Bb5+
Nbd7 10 dxe6 fxe6 11 Nf3 a6 12 Bc4 (Diagram 30)
WARNING: White must retreat: 12 Ng57 axb5 13 Nxe6 Qc6 14 Nd5

leaves his knights looking pretty, but Black a piece up for
insufficient compensation after 14... Ra41, V.Bondarenko-D.Scherbin,
Podolsk 1993.
12... Nb6
Aroshidze suggests 12 ... d5 1 3 exd5 Nb6, but after his further 14 Rxb6! Qxb6 15
Qe2 Be7 16 dxe6 0-0 I would be pretty happy to take White with such a powerful
fishbone in Black's throat on e6.
13 Rxb617
A critical sacrifice, which sees White fighting for the initiative and determined not
to retreat his bishop. Hodgson did once reject this possibility, but his 13 Bd3 c4 14
e5?! (White can maintain sufficient compensation with 14 Be2 Qc5 15 Qcl)
14 ... cxd3 15 exf6 dxc2 16 Qxc2 didn't tum out well after 16 ... gxf6 1 7 f5 Bg7! 18 fxe6
0-0! with advantage to Black in J.Hodgson-H.Alber, Neu Isenburg 1992.
13 ...Qxb6 14 es
The only move tried so far, but 14 Ng5!? (Diagram 31) deserves serious considera
tion too: for example, 14 ... d5 (Prie's 14 ...Qb4!? is another critical response, after
which he feels that 15 Qe2 b5 16 Nxb5 Qb1+ 17 Kf2 Qxh1 18 Nc7+ Ke7 is unclear
and this fascinating position could really do with a test) 15 exd5 Be7! (15 ... exd5 16
Nxd5 Nxd5 1 7 Bxd5 traps Black's king in the centre) 16 0-0!? (the simple 16 dxe6
0-0 17 f5 cannot be a bad alternative) 16 ...Qd8 17 Re1 ! 0-0 (Aroshidze also notes
that 17 ...exd5? 18 Nxd5 Nxd5 19 Bc3 leaves Black in some trouble, since 19 ... Nxc3?
fails to 20 Bf7+ Kf8 21 Qxd8+ Bxd8 22 Re8 mate) 18 Nxe6 Bxe6 19 Rxe6 Bd6 20 a4
and White's control of the position supplies decent compensation for the ex
change, as analysed by Aroshidze in Informant 87.
Furthermore, 14 0-0!? is possible too, albeit likely to transpose to our last variation
after 14 ... Be7 15 Ng5 d5 16 exd5.
NOTE: Even with all the analytical labours of Aroshidze, Prh! and
Wells, much remains to be discovered in this critical variation. In
deed, the stronger reader who delves deeply into this line will surely
be rewarded over the board.
14. .dxes
.
119
The amazing game S.Mamedyarov-L.Aroshidze, Baku 2002, saw Black prefer to

return material with 14 ... d5!? 15 exf6 gxf6 (15 ... dxc4? 16 Ne5 is rather nasty for the
black king) and then 16 BxdS? exd5 1 7 Nxd5 Qe6+ left White struggling. However,
White might have tried Wells's 16 Ne5! (Diagram 32), throwing more wood on to
the fire, even if 16 ... fxe5 17 Qh5+ Kd7 18 Qxe5 Rg8 19 Nxd5! exd5 20 QxdS+ Kc7!
(Wells only gave 20 ...Bd6? 21 Qxg8 Qb1+ 22 Kf2 Qxh1 23 Qe6+ Kc6 24 BdS+ Kc7 25
BaS+ with a winning attack) 21 Qxg8 Qb1+ 22 Kf2 Qxc2! (Aroshidze) is no more
than roughly balanced with both kings exposed.

Pressuring e6 White is determined to attack!
15 fxes Ng41
A recent development, although even nowadays a less brave or underprepared
opponent might well prefer one of:
a) 15 ... Nd7 16 0-0 Be7 (or 16 ...Qc6 1 7 Nh4 g6 18 Bg5! and one has to like White's
compensation) 1 7 Bg5! Bxg5 18 Nxg5 Nxe5 19 Qh5+ g6 20 Qh6 Qd6 21 Qg7 Qd4+
22 Kh1 Nxc4 23 Qf7+ Kd8 24 Rd1 and at the end of this impressive piece of analy
sis by Peter Wells, White's active queen is superior to the inactive black rooks.
b) 15 ... Nd5 16 Bxd5 exd5 1 7 Nxd5 Qd8 18 Bg5 QaS+ 19 Nd2! Be6 20 Nf4 Qb6 21 0-0
again leaves White with a strong initiative for the exchange, W.Siewert
J.Simmelink, correspondence 2007.
16 0-01?
A novelty. Previously White had erroneously allowed Black to demonstrate his
main point: 16 h3?! Nh6 17 Bxh6 gxh6 18 0-0 Bd7 and White found himself low on
compensation with the black king set to find sufficient safety on one flank or the
other, E.Cordova-J.Granda Zuniga, Santiago 2006.
16 Be7 (Diagram 33)
...
120
The U ncom p rom i s i n g 2 . . . c s

Enough for the exchange Anything m ight still ha ppen
17 h3
Once again this is far from forced. Indeed, Prie has analysed 17 Nh4!? Qd8! (an
other point behind Black's 15th) 18 Qxg4 Qd4+ 19 Qxd4 cxd4 20 Ne4 Bxh4 21
Nd6+ Kd7 22 Rf7+ Kc6 23 Rxg7 Bd7 when White retains compensation for the ex
change, although assessing exactly how much is not so easy.
17 ... Nh6 18 NgSI
Siewert is determined to attack and prevent Black's king from ever castling.
18 ...Qc61
Noble too is up to the challenge. It looks like it will take Black too long to develop
his queenside, but he actually has just enough time.
19 QhS+ KdS 20 Nf7+ Nxf7 21 Rxf7 bS 22 BgS
White might have tried 22 Rxg7!?, hoping for the line 22... bxc4 23 Rxe7! Kxe7 24
Bg5+ Kf8 25 Bh6+ Ke7 26 Qg5+ Ke8 27 Qg7 when he regains a rook with check, but
I dare say that Black would have preferred 22 ... Ra7 when 23 Bg5 transposes to our
next note.
22 . Ra7 (Diagram 34) 23 Bd31?
. .
I'm not totally sure about this bold choice, but it is clear that Siewert was deter
mined to avoid the repetition which would have occurred after 23 Rxg7 bxc4 24
Qdl + Ke8 (24 ... Rd7?! 25 Bxe7+ Kc7 26 Bd6+ Kb7 27 Qg4 gives White at least
enough for the exchange) 25 Qh5+.
23 . Rd7 24 Be4 Qb6 25 Bxe7+ Rxe7 26 Qgs Rhe8 27 Bxh7 Bb7 28 Bg6
..
Good defence; White must hassle Black's rooks is he is to prevent him from mov
ing his king and at least one rook a file to the west.
121
Sta rt i ng Out: The T ro m pows ky Attack
28 ...Qc7 29 Rf2 RgB 30 Rf7 Bc6 31 Ne2 Rge8 32 Rf1 Rh8 33 Rf7 (Diagram 35)

B lack remains tied down B lack plans . . fs and ... Bg7
.
33 ... Rhe8
Black is happy to draw, but another player might have been happy to continue
with 33 ... Be8!? 34 Rxg7 Bxg6 35 Rxg6 Kc8, although this would have been risky:
White's knight will come to e4 and his h-pawn is a pretty useful long-term asset.
34 Rf1 Yz-Yz
White Excha nges on f6

1 d4 Nf6 2 Bg5 c5 3 Bxf6
Just as against 2 ... d5 and 2 ...g6, one can argue that this is the consistent Trom
powsky move. Unsurprisingly it's again also been the most popular one, having
been played in 52% of all games with 2 ... c5.
3 ...gxf6 (Diagram 36)
Black's invariable choice, but about 10% of the time he prefers 3 ... exf6 and that per
centage might grow after this unusual line was recommended by Alburt,
Dzindzichashvili and Perelshteyn. In response White should avoid 4 dS since
Black's kingside is much more solid than in our main line, preferring 4 Nc3 (4 c3 dS
5 e3 Nc6 6 Nd2 is a slower alternative) 4...d5 (4 ... cxd4 5 Qxd4 Nc6 6 Qe4+ Be7 7 e3
fails to trouble White, so Black takes play into a rare line of the Veresov) 5 e3 Nc6 6
Nge2! (the Americans only consider 6 Nf3, but Stefanova's set-up, first played by
Smyslov, makes much more sense in terms of targeting d5) 6...c4 (perhaps Black
should investigate the more energetic 6 ... Be6 7 g3 cxd4 8 exd4!? Qb6!?, although
here 9 Bg2! Qxb2 10 Rb1 Qa3 1 1 Rxb7 Bb4 12 0-0 Bxc3 13 Rb3 Qxa2 14 Rxc3 looks
122
The U n com pro m i s i n g 2 . . . c s
like pretty good compensation, a s pointed out b y Wells) 7 a3 Qa5 8 g3 Bd6 9 Bg2
(Diagram 37) 9 .. Ne7 10 b4! cxb3 1 1 cxb3, which leaves the isolated d-pawn in some
.
trouble, A.Stefanova-I.Chelushkina, Turin Women's Olympiad 2006.

That backwards pawn is pretty wea k Once again b2 is targeted
4 d5
White's invariable choice, helping himself to a spatial advantage. This isn't his
only approach, but the fairly-rare alternatives appear less challenging:
a) 4 dxc5 Qa5+ (4 ... Na6 also has a good reputation) 5 Nd2 Qxc5 6 c4 f5 7 e3 Bg7 8
Rb1 aS 9 Ngf3 Bf6 10 Bd3 d6 1 1 0-0 Nc6 12 Nb3 Qb6 was fine for Black in
R.Slobodjan-P.Bauer, German League 2006; his unopposed bishop and useful cen
tral control balance out his structural defects.
b) 4 c3 looks sensible enough, aiming to restrict Black's dark-squared bishop. It's a
fav ourite of the Russian Grandmaster, Konstantin Chemyshov, but is a little pas
sive and 4 ... Qb6! 5 Qd2 d5 6 e3 Nc6 (Black might also develop with 6 ... Bf5 7 Nf3
e6, but playing for the central advance is highly tempting) 7 Nf3 eS 8 Be2 Be6 9 0-0
0-0-0 10 Qc2 Kb8 was at the very least okay for Black in K.Chemyshov
A.Chernobai, Moscow 2008.
4 Qb6 (Diagram 38)
...
In turn this is Black's invariable choice and a very logical one at that. 2 ... c5 pre
pared to strike out on the dark squares and against b2, and so why not do so at
once? There are alternatives, though, but as the best way to respond often de
pends on how White handles the main line, the reader may wish to begin by
s tu dying 4 ... Qb6 before returning to this juncture:
a ) 4 ... d6 5 e3 (those who prefer a c4 set-up should play 5 c4 without delay) 5... Bf5
(unusual; instead 5 ... Qb6 6 Qcl takes play back into our main line, while another
123
instructive game ran 5 ... Na6 6 BbS+! Bd7 7 Bxd7+ Qxd7 8 Ne2 Qg4 9 0-0 Rg8 1 0
Ng3 Qxd1 1 1 Rxd1 e6 12 Nc3 with a pleasant edge i n P.Wells-J.Sugden, Ports
mouth 2004) 6 Ne2 (in keeping with our last example, a decent case might be
made too for 6 Bd3!? Bxd3 7 Qxd3 Qb6 8 Nc3) 6 ... Bg7?! (rather planless; Black
should prefer to fight on the kingside and 6 ... h5! 7 Nf4 Qb6 8 Qcl Nd7 9 a4 h4 1 0
aS Qc7 was no more than a touch worse for him in M.Adams-A.Hon, London
1992) 7 Nf4 Na6 8 BbS+ Bd7 9 a4 BxbS?! 10 axbS Nc7 1 1 Qg4! 0-0 12 NhS NeB 1 3
Nc3 gave White a dream opening i n R.Palliser-R.Westra, Middlesbrough 2004.
NOTE: Black often struggles to activate his light-squared bishop

early on after 3 ...gxf6, but this variation is a classic example of the
usefulness of a passive bishop. If the bishop finds itself exchanged
too early, Black's kingside light-square weaknesses are easy for
White to target and the ... e6-break may well be much less effective.
b) 4 ... f5 5 c3 (5 c4 Bg7 6 Qc2 followed by Nc3 would be the c4 approach) 5 ...Qb6 6
Qb3!? (White believes that an exchange of queens will remove some of the dyna
mism from Black's set-up; it's a reasonable claim, but 6 Qc2 followed by Nd2-c4
can hardly be a bad plan either) 6... Qh6?! (too ambitious, but note how 6 ...Qxb3 7
axb3 e6 8 Nh3! exdS?! 9 Nf4 would retain control of the position; Black must al
ways be very careful about exchanging on dS and usually he prefers to pressure
that pawn, such as with ... Na6-c7, in a bid to force White to exchange on e6) 7 e3 f4
8 exf4 Qxf4 undoubted Black's pawns in R.Palliser-C.Lim, York 2004, but his king
side weaknesses remained and White quickly gained a monstrous initiative: 9 Ne2
Qc7 10 Ng3 Bh6?! 1 1 Be2 0-0 12 NfS BgS 13 d6! exd6 14 Na3 (Diagram 39)

A dream position for White Good handling of the knights
14 . .. a6?! 15 N c4 (a classic demonstration of the potential power of the Trom-
124
The Un comp romi s i ng 2 .. . c s
powsky knight-pair) 15 ...b 5 16 Ncxd6 c4 1 7 Qc2 Nc6? 18 Qe4 Qd8 19 h4 Bf6 20

Nh6+ Kg7 21 Qg4+ Bg5 22 hxg5 1 -0.
TIP: Should Black negled for too long to play ...d6 himself, do look
out for a favourable chance to play d6 as White, stymieing Black's
position and further wrecking his strudure.
c) 4 ... Bg7 5 c3 (5 c4 followed by Nc3 is once again the alternative) 5 ... d6 (to see
why ...Qb6 is best played back at move 4, we need just examine 5 ...Qb6 6 Qc2
which leaves White's queen on a better square than b3 and unable to be hassled
by 6 ... d6 7 e3 Bf5?? on account of Black's vulnerable back rank) 6 e3 f5 7 Ne2 Nd7
8 Nf4 Nf6 9 Bc4 0-0 10 a4 saw White employ one of his main set-ups and after
10 ...b6 1 1 Nd2 a6 12 Nfl ! ? Bd7 13 Ng3 b5 14 Bb3 (Diagram 40) he was slightly for
choice in J.Hodgson-V .Kotronias, Belgrade 1993.
After that instructive interlude, we return to the main line with 4 ... Qb6:
5 QC1
White's invariable choice these days, which we will return to in our next section. I
don't wholly trust Hart's old favourite 5 Nd2!? Qxb2 6 e3 (Diagram 41),

A bold gambit Reasonable play for the pawn
but this gambit might well be a reasonable surprise weapon. After 6 .. .f5 Jansa has
suggested 7 Nh3!? (7 Bd3 Qf6 8 Ne2 e5 9 dxe6 dxe6 10 e4 attempted to blow open
the position in J.Hodgson-S.Conquest, Hastings 1986/87, but after 10 ... Nc6 1 1 exf5
exf5 12 0-0 Be6 Black was in control of the situation) 7... Bg7 8 Rb1 Qf6 9 Nf4 (Dia
gram 42), which certainly looks like decent compensation and could do with a
test. Here 8 ...Qxa2!? is a touch greedy, but far from refutable. After 9 Bc4 Qa5 10
0-0 d6 1 1 Bd3 White's knights have some good squares and a piece will land on
h5, but whether this is worth two pawns is debatable.
125
Sta rt i n g Out: The Trom powsky Atta ck
Conclusion
2 ... c5 3 Bxf6 gxf6 4 d5 is arguably the most complex variation of the Trompowsky
and as such warrants a fair amount of coverage. I just hope that the reader will
persist for this is undoubtedly a most enjoyable variation to contest. To avoid be
ing worse in the ensuing dynamic battle Black must obtain early activity and so
4 ...Qb6 is really required. As we will see, 5 Qcl is then the critical reaction; 5 Nd2
is playable, but probably good for no more than the odd surprise outing.
The 2 cs Main Line: 5 Oc1

...
-
1 d4 Nf6 2 Bgs cs 3 Bxf6 gxf6 4 d S Qb6 5 Qc1 (Diagram 43)

The prudent choice Black wants to go .. .f4
s ...fs
Black usually opens up the long diagonal thus, but he has also been known to pre
fer one of:
a) 5 ... Bg7 has often transposed to our next section after 6 c3 f5 7 g3, but 6 c4!? re
mains an important alternative. Now if Black doesn't want to transpose to our
main line, he might try 6 ... Qb4+!? and 7 Nd2 d6 (7 ... f5 8 Rb1 d6 9 a3 pushes Black
back and White should play to bring a knight to f4 here) 8 a3 Qb6 9 e3 f5 1 0 Ra2!?
Nd7 11 Ne2 Nf6 12 Nc3 led to a typically complex 3 ... gxf6 middlegame in
R.Palliser-M.Mitchell, York 2005, but with both knights further from f4 than Whi te
might like.
b) 5 ... Bh6 6 e3 fS!? (Diagram 44) provokes an immediate sharpening of the strug
gle. Specifically Black is hoping to enter the critical variation 7 c4 f4 8 exf4 Bxf4 9
126
The Uncom p ro m i s i n g 2 . . c s
.
Qxf4 Qxb2, which we will investigate in the notes to Black's 6th move. Those who
don't wish to get involved in such complications should prefer to halt ... f4 with
either 7 g3 or 7 Ne2!?. The latter was seen in J.Vigus-D.De Vreugt, Glenalmond
1996, which proceeded thematically with 7 ... d6 (more recently 7...e5 8 Nd2 d6 9
Nc4 Qc7 10 a4 b6 1 1 Nc3! a6 12 Qd1 Rg8 13 Qh5 Rg6 14 g4! fxg4 15 Bd3 revealed
another instructive handling of the white pieces in P.Vavrak-M.Hoekstra, Internet
2005; Black should always beware closing the centre with ...e5 thus, not least be
cause White is often pretty quick at bringing his pieces to good squares, such as
h5) 8 c4 Nd7 9 Nbc3 Ne5 10 Nf4 Bd7 11 Qc2 a6 12 Nh5! Qa5 13 f4 Ng6 14 Bd3 e6 15
0-0 (Diagram 45) and White had taken control.

Thwarting Black's aims Relative ambition from White
NOTE: Once again do observe White's use of the h5-square, as well

as how his battery on the b1-h7 diagonal does such a good job of
keeping the central situation under control.
6 c41? (Diagram 46)
White makes no attempt to blunt the long diagonal with a pawn. Shoring up d5
like this is useful, not least in rendering an early ... e6 rather ineffective, but it does
all come down to one's views on the critical note to Black's 6th move.
There are two important alternatives which prevent the play from becoming so
forcing:
a) 6 g3 has in mind a set-up with c3, Nd2, Nh3-f4 and usually Bg2, and is impor
tant enough to warrant its own section, which we'll come to after Game 1 7.
b) 6 e3 Bg7 7 c3 was once quite a popular approach, but is generally now consid
e red less critical than 6 g3 and 6 c4. However, developing with Nd2, Bc4 and Ne2
127
Sta rti n g Out: The Trom pows ky Atta ck
cannot be a bad set-up and 7...d6 (Black wants to hurry his knight to f6; practice
has also seen 7...e6 8 Nh3! Na6?! 9 Nf4 0-0 10 Na3 e5 1 1 Nh5 Bh6 1 2 Be2 Qg6 1 3
Qd2! d6 14 f4, which favoured White i n L.Geffroy-A.Delorme, French League
2007, and 7... Qd6!? 8 Qd2 Na6 9 Na3 Nc7 10 Rd1 e5 1 1 Nc4 Qf6 12 d6 Ne6, which
was quite unclear in M.Eiber-S.Bromberger, Bechhofen 1998) 8 Nh3 (8 Nd2 Nd7 9
a4 Qc7 10 Nc4 is the alternative, but with 10 ... Nf6 1 1 Qd1 e6!? 12 dxe6 d5 1 3 exf7+
Kxf7 14 Nd2 f4 Black obtained good play for his pawn in B.Sahl-K.Lie, Norwegian
League 2008; this illustrates why White often prefers to maintain his pawn on dS
by supporting it with his bishop from c4 or g2) 8 ... Nd7 9 Nf4 Nf6 10 Bc4 (Diagram
47) has led to a roughly balanced manoeuvring struggle in a number of games, as
we'll see in Game 1 7.

A tough struggle lies ahead Will the queen escape?
6 Bg7
...
Black's usual choice in practice, but 6... Bh6!? 7 e3 f4 8 exf4 (White is committed to
the exchange sacrifice; otherwise Black just obtains excellent play on the dark
squares) 8 ... Bxf4 9 Qxf4! Qxb2 10 Ne2 Qxa1 11 Nec3 (Diagram 48) reaches quite a
controversial position. Hodgson and Wells have been happy enough on the white
side, but Dembo has recently suggested that Black is better! At any rate it's clear
that Black must take care at this point:
a) 1 1 ...Qb2? 12 d6! throws a huge spanner in the works and after 12 ... Qc2?
(12 ... Nc6 marginally improves, but 13 Bd3 exd6?! 14 0-0 Ne5 15 Qf6 0-0 16 Nd5
was also crushing in J.Hodgson-J.Van der Wiel, Amsterdam 1994) 13 Qe3 Black
already gave up in P.Wells-A.Shirov, Gibraltar 2006! A perfectly understandable
decision when one considers the variation 13 ... Nc6 14 Bd3 Qb2 15 0-0 e6 16 Qh6 b6
1 7 Nd5! with a decisive attack.
b) Thus 1 1 ...d6! is essential, even if 12 Qd2 (Diagram 49) does cut off the black
queen's retreat. Indeed, the queen will be won, but Black can hope to obtain a suf-
128
The Uncom p ro m i s i n g 2 ... c s
fi ciently high price after 12. . .Rg8 (best; 12. . .Bf5? 13 Be2 Rg8 14 g3 Bh3 15 f3! a S 1 6
J(f2 Ra6 1 7 Rd1 Rb6 18 Qd3 Bd7 19 Nd2 Qb2 20 Rb1 Na6 21 Rxb2 Rxb2 2 2 a3 gave
Wh ite the decisive material advantage of queen and knight for the two rooks in
M.Leon Hoyos-O.Piot, Cappelle Ia Grande 2009) and then:

What can be obtained for the queen? The tactics work for Black
b1) 13 Be2!? (Gallagher's idea) 13 ... a6!? (Dembo's counter idea; instead 13... Rxg2 14
Bf3 Rg6 15 Ke2! Bg4? 16 Rcl Nc6! 17 dxc6 Re6+ 18 Kfl Bxf3 19 cxb7 Rb8 20 Na3
Qxcl + 21 Qxcl Rxb7 22 Nc2 turned out well for White in the only practical exam
p le of 13 Be2 so far, V.Weil-G.Diez, Bad Homburg 2007) 14 0-0 (in light of what
follows this may be inaccurate; instead 14 g3 bS 15 cxbS axbS 16 BxbS+ transposes
to variation 'b12') 14 ... Bh3! (Diagram 50) 15 Re1 Ooe Gallagher can certainly be
forgiven for failing to realize that 15 Bf3? fails to the highly-aesthetic sequence
15 ... Nd7 16 Na3?! Ne5 1 7 Be4 Bxg2! 18 Rxa1 Bxe4+ 19 Kfl Bd3+ 20 Ne2 Rg1+! 21
Kxg1 Nf3+) 15 ... Rxg2+ 16 Kh1 Rg6 1 7 Na3 Bg2+ 18 Kg1 BxdS+ 19 Kfl Bg2+ 20 Kg1
Qxe1 + 21 Qxe1 Nc6 and I have to agree with Dembo that Black is better due to his
fantastic activity.
b2) A critical alternative is 13 g3 a6 14 Bd3! when Dembo has pointed out that
1 4 ... Nd7?! 15 0-0 Ne5 1 6 Be2 Bh3 1 7 Rcl is promising for White. Thus she prefers
1 4 ... b5! 15 cxbS axb5 1 6 BxbS+ Bd7 1 7 Bd3 Bh3 18 Qh6 Bg2 19 Rg1 BxdS!? 20 Qxh7
Rf8 21 Nxd5 Qe5+ 22 Qe4 Qxe4+ 23 Bxe4 Rxa2 and claims an advantage for Black,
although this unbalanced position doesn't look at all clear to me.
WARNING! Few black players might have dared to grab the

exchange in practice, but those who wish to venture 6 c4 most
certainly need an idea here. I would not be surprised if an
improvement is out there over Dembo's impressive analysis, but
despite much searching I'm yet to find itl
129
Sta rti n g O ut: The Trom pows ky Attack
After that excitement, we return to 6...Bg7:

7 Nc3 d6
Standard, but perhaps Black need not be deterred: 7...e6!? 8 e3 Na6 9 Bd3 d6 10 a3
Bd7 1 1 Nge2 0-0-0 12 Qd2 Kb8 13 0-0 Rdg8 wasn't clear in P.Wells-H.Nakamura,
Port Erin 2004, although 9 Nh3!? d6 10 Nf4 might be a better try for the advantage.
8 e3 Nd7
I quite like White's prospects here and evidently so does Wells:
9 QC21 (Diagram 51)

Fairly pleasant for White Correctly opening the position
The best set-up, although the related 9 Nge2 Nf6 1 0 Qc2 Qa5 1 1 Ng3 b5 1 2 Bd3
bxc4 13 Bxc4 Rb8 14 0-0 0-0 15 Rab1 Qb4 16 Nce2 also gave White an edge in
K.Maslak-M.Kraemer, Olomouc 2006; it's quite instructive how ...b5 caused much
less disruption than one might imagine.
9 Nf6 10 Bd3 e51?
...
Recent grandmaster praxis has also seen:

a) 10 ... Ne4 1 1 Nge2 Nxc3 1 2 Nxc3 e6 13 0-0-0 Bd7 14 g3 Qa5 15 Rd2 0-0-0 16 Kb1
led to a tough, roughly balanced manoeuvring struggle in A.Rodriguez Vila
M.Villanueva, Avellaneda 2007, but here Romero Holmes's 12 bxc3! followed by
going short looks preferable.
b) 10 ... e6 1 1 Nge2 Bd7 12 0-0 0-0-0 13 Rab1 Kb8 14 b4! cxb4 15 a3 gave White goo d
attacking chances in K.Rusev-Z.Stanojoski, Sunny Beach 2008.
11 dxe61
Naturally White has no desire to allow 1 1 Nge2?? e4, and it's notable just how ef
fective opening the position can be. Indeed, after 1 1 ...fxe6 12 Nf3 Bd7 1 3 0-0 0-0
130
The U ncom p rom i s i n g 2 ... c s
(13 .. 0-0-0!? might improve) 14 e4! (Diagram 52) 1 4.. .fxe4 15 Nxe4 Nxe4 16 Bxe4 h6
.
17 Radl Black's position was rather loose despite his bishop-pair in P.Wells
S.Ledger, Southend 2003.
Conclusion
White has scored 51% from over 1,000 games with 5 Qcl . After S .. .fS both 6 e3 and
6 g3 Bg7 7 c3 (the subject of our next section) lead to a rich, manoeuvring struggle
in which the player with the better grasp of the key motifs for both sides is likely
to triumph. Specific theory is more important after 6 c4, which has been popular
of late despite the fact that it entails an exchange sacrifice, and one which can best
be described as risky if practically dangerous.
Illustrative Ga mes
Game 17
D J.Vigus J.Hodgson
British Cham pionship, Street2000
This is the only game I've manage to track down in which Julian Hodgson faced
his favourite opening. Unsurprisingly he chose one of Black's more challenging
systems, but still ran into some trouble...
1 d4 Nf6 2 Bg5 c5 3 Bxf6 gxf6 4 d5 Qb6 5 Qc1 f5 6 e3 Bg7 7 c3 d6 8 Ne2 Nd7 9 Nf4
Nf6 10 Bc4 Bd7 (Diagram 53)

White mustn't a l low bs
... Black seizes the initiative
131
Sta rt i n g Out: The Trompo wsky Attack
A popular choice and it's certainly useful for Black to be threatening ...b5 when his
queen drops back.
11 a4
White prevents ... b5 without delay. He has also been known to try 11 Nd2 when
Black doesn't have to move the queen, but he did in V.Laznicka-M.Parligras,
European Championship, Dresden 2007, which continued 1 1 ...Qc7 12 a4 0-0 13 0-0
b6 14 Bb5 e5! 15 dxe6 fxe6 16 Qc2 Rae8 1 7 c4 Re7 18 Bxd7 Qxd7 19 Rfd1 Qc6 with
comfortable play. Not hurrying with Nbd2 a Ia Vigus appears more challenging.
11 . 0-0 12 0-0
..
WARNING: White might ideally like to bring his queen's knight

across to the other flank, but there's no time for 12 Nd2 Rae8 13
Nf1? because the centre is fluid and 13 ...e5 (Diagram 54) gives Black
the advantage, as pointed out by Wells.
12 ... Kh8
Black wants to give himself some options down the g-file. That makes some sense,
but Wells prefers the immediate 12 ... Rae8, observing that 13 aS Qc7 14 Qd1 (White
lacks the time to bring his queen to b3, unlike in our main game, as 14 Qc2 e5 15
dxe6 fxe6 is obviously very comfortable for Black) 14 ....Kh8 15 Nh5 Rg8 16 Nd2
(Diagram 55) 16 ... e6! ? (Black must always beware closing the centre too early, but
here he is fully mobilized and 16 ... e5 1 7 Nxf6 Bxf6 18 f4 Rg6 19 Rf2 Qd8 20 Qb3
Bh4 21 Rf3 Reg8 22 Bfl Bc8 was also fine for him in E. Danielian-V.Saravanan,
Kavala 2008) 1 7 Nxf6 Bxf6 18 g3 Rg6 19 Re1 Kg7 gave Black respectable-enough
counterplay in J.Bellon Lopez-G.Hemandez, Las Palmas 1995.

It's a bout even How am bitious will Black be?
132
T h e U ncompromi s i n g 2 ... c s
1 3 Re11l
Far from essential, but I quite like this semi-prophylactic move. White can always
cover g2 with Bfl if needed and he might even break with e4 should Black mis
ti me his own ... e6-break.
13 ... Rg8 14 Qc2 Rae&
Now ... e5 is back on the agenda, but White is in time to hinder it.
15 aS Qd8 16 Qb3 1 Qc8 17 Qd1 (Diagram 56) 17 ... Bh61?
Despite my admiration for White's bravery (taking on the maestro in the Trom
powsky) and creativity (delaying Nbd2), I must admit that he isn't better here and
17 ... Qc7 would have left Vigus with a choice between repeating moves and allow
ing Black's ideal advance with rough equality. Of course, though, such a re
nowned fighter as Hodgson was never going to allow an early draw.
18 Na3 Rg4 19 g3 Reg81l
A bold but by this point fairly essential sacrifice. Otherwise White would have
had time for Be2-f3 followed by Nc4 with a pretty harmonious set-up.
20 Be2
Critical, although if White didn't want to take the exchange, he might have con
tinued 20 f3!? R4g7 21 Qd3 followed by arranging b4 when I slightly prefer his
chances; Kf2 and Rg1 will always hold the kingside if necessary.
20... Bxf41 21 Bxg4?
White underestimates the problems he will face on the kingside. It was better to
play 21 exf4! Rxf4 22 Bf3 (Diagram 57) when the exchange sacrifice is no longer so
convincing, although Black's rook does have a4 if necessary. Play might continue
22 ...Qd8 23 Qe2 e6 24 dxe6 fxe6 25 Bxb7 Ne4 26 Qe3, which remains quite unbal
anced but probably also about even.
21... Bxg3 22 hxg3 Rxg4
It' s time to take stock. Black has only one pawn for the exchange and no immedi
ate threats, but White's kingside is quite airy and it will take his knight time to get
back into the game.
23 f3?1
Pa nic. This isn't the best defence, but one can certainly see why Vigus didn't like
the look of 23 Kg2? Nxd5 24 f3 Bc6! with some initiative as 25 e4? fails to 25 ... Nf4+
26 Kh2 (or 26 Kf2 Nh3+ 27 Kg2 Rg6) 26... Rg6 27 gxf4? Qg8 followed by mate. Find
i n g a good defence is far from easy, though. White would like to play 23 c4!?, but
th en 23 ... Qg8 (23 ... Ne4 24 Qc2 Qg8 25 Kfl transposes) 24 Kfl (it looks a little ugly,
but perhaps radical measures in the shape of 24 Qf3!? Ne4 25 Qh1 are required)
24 . Ne4 25 Qc2 Qg6 sees Black calmly building up and preparing to infiltrate with
. .
some effect down the h-file.

23 ... Rxg3+
133
Sta rti n g Out: The Tro m p o ws ky Atta c k
Tempting, but i t seems that Black missed a strong deflection at this point in
23 ... Ba4!. The point is revealed after 24 b3 (24 Qe2? Rxg3+ 25 Kf2 f4 26 Rhl Ng4+!
27 fxg4 Rxe3 definitely doesn't help White) 24 ... Rxg3+ 25 Kf2 f4 (Diagram 58)
when Black's queen races into the attack, albeit at the cost of a further piece. Con
tinuing this line with 26 bxa4 (I suspect that grovelling with 26 Rgl is White's best
try) 26 ... Qh3 27 Ke2 Rxf3 28 Kd2 Nxd5 29 Kcl Nxe3 we find Black having picked
up no fewer than five pawns for the rook while retaining the initiative. In practice
White would certainly do well to survive this.

A solid defence A da ngerous attack
24 Kf2 Qg8 25 Rg1

Essential defence.
25 ... Nh51?
Black continues to play for the attack, rather than head for the slightly better if
unbalanced endgame which would have arisen after 25 ... Rxgl 26 Qxgl Nxd5 27
Qxg8+ Kxg8.
26 b4 f4?1
Tempting, but this allows White a handy resource and it seems that Hodgson
would have done better to prepare ... f4 with 26 ... Qg5! : for example, 27 bxc5 dxcS
28 Rxg3 Qxg3+ 29 Ke2 QeS 30 Kd2 Qxd5+ 31 Kcl Qe6 and Black emerges with a
clear advantage.
TIP: Nimzowitsch's advice isn't always correct, but sometimes 'the

threat is greater than the execution'! Attacking chess isn't all about
hurrying to deliver the killer blow - one must also take the oppo-
nent's plans and resources into account.
134
The U ncom p rom i s i n g 2 . .. c s
2 7 exf4 cxb4 2 8 Qd4+1

Further good defence.
28 ...f6 29 cxb4 Bh3 (Diagram 59)

How to defend? A sensible, fa irly popular choice
30 Ke2?
No doubt already running low on time, Vigus collapses. Instead he could have
shored up f3 with 30 Nc4 Bg2!? (30 ...Rg2+ 31 Rxg2 Qxg2+ 32 Ke3 Ng3 looks rather
scary, but the calm 33 Qb2 appears to keep everything defended - the resource
mentioned earlier) 31 Nd2 when 31...Rxf3+ 32 Nxf3 Qg3+ 33 Ke2 Qxf3+ 34 Kd2
Nxf4 sees Black retain the initiative, but he is two exchanges in arrears here and
White probably not worse after 35 Rael .
It would also have looked horribly risky over the board, but White could have
gone in for 30 Rxg3 Qxg3+ 31 Ke3 Bg2 32 Qe4, as 32 .. .f5 33 Qxe7! Qxf3+ 34 Kd2
Qxf4+ 35 Kc3 sees his king running away to safety and probably Black must force
a draw here.
30... Bg21
Clinical. Now f3 will fall and with it White's position.
31 a6 bxa6 32 Nc4 Rxf3 33 Ne3 Nxf4+ 34 Kd1 Rf2 35 Ke1 Qg3 36 Nf5 Re2+ 37 Kd1
Qb3+ 38 Kcl Qc2 mate (0-1)
The 2 cs Mai n Line: s Qc1 fs 6 g3

...
1 d 4 Nf6 2 Bg5 c 5 3 Bxf6 gxf6 4 ds Qb6 5 Qc1 fs 6 g 3 (Diagram 60)
135
Start i n g Out: T h e Trom pows ky Attack
As the reader has probably already ascertained, the play in this variation is often
quite unforcing unless 6 c4 Bh6 happens. The text has in mind a set-up with c3,
Nd2, Nh3-f4 and usually Bg2; moves which can be played in a number of orders.
6... Bg7 7 C3 d6
Sensible and best. Somewhat more ambitious is 7...Qd6?! 8 Bg2 b5, but after 9
Qc2!? (targeting f5, but 9 Nf3 Bf6 1 0 0-0 is a decent alternative) 9 .. .f4 1 0 Nh3! fxg3
1 1 hxg3 Bb7 12 e4 (I also rather like White's control of the position after 12 Na3!?
a6 13 Nf4; again note how Black's kingside remains vulnerable despite the un
doubling of his f-pawns) 12 ... Na6 13 Nf4 c4 14 Nh5 Qg6 15 Na3 Nc7 16 Qe2 White
was better in V.Bhat-C.Baginskaite, San Francisco 2005.
Black does better with 7... Na6!? 8 Bg2 Nc7 (Diagram 61), and 9 Nd2 Qh6 (9...e6 10
Nc4 Qa6 11 Qf4 0-0 12 Nh3 exd5?! 13 Ne3 Ne6?! 14 Qxf5 d4?! 15 Be4 was some
thing of a light-square disaster for Black in A.Fernandes-G. Djurovic, European
Championship, Budva 2009; a timely reminder of the fine line for both sides of the
balance between effective creativity and over-creativity in this complex variation)
10 Nc4 (10 Qc2!? d6 1 1 e3 is untested, but appears more critical, not fearing 1 1 ... b5
on account of 12 a4!) 1 0... Qxcl + 1 1 Rxcl d6 12 Nh3 Bd7 was about equal in
P.Kloninger-V.Baklan, Bad Zwesten 1999.

Black has ideas of ...e6 and ... bs Trying to destabilize the knights
8 Nd2
White can also begin with 8 Bg2 Nd7 and now 9 Nd2 transposes to our main line.
Instead 9 Nh3 h5!? 10 Nd2 (10 Qc2 Ne5! 1 1 Nd2 Bd7 12 Nf4 h4 gave Black the ini
tiative in B.Czap-A.Fernandes, European Championship, Dresden 2007) 10 ...h4
(once again 10 ... Ne5!? might be a better idea and after 1 1 Nf3 Nxf3+ 12 Bxf3 h4 13
Nf4 Bd7 both sides will probably castle long and have their trumps, with the
chances about even) 1 1 Nc4 Qa6 (Diagram 62) 1 2 Ne3 Nf6 (Dembo prefers
136
The U ncom p r o m i s i n g 2 . c s
..
1 2 . . hxg3 13 hxg3 Ne5 14 Qc2 Nc4!? 15 Nxf5 Bf6 16 e4 e6 when White's best try for
.
the advantage is probably to return the pawn with 1 7 dxe6 fxe6 18 Nh4) 1 3 Qc2 e6
1 4 dxe6! fxe6 15 g4! saw White seizing the initiative in G.Sargissian-P.Froehlich,
Lin ares 2001, but much remains to be clarified in these early ... h5 lines.
There was even a time when White experimented with 8 Bh3, but nowadays it is
known that Black is quick to pressure d5 and obtain decent counterplay after 8 ...e6
9 Nf3 Nd7.
S . Nd7 9 Bg2 Nf6
. .
Black's most natural move. Instead 9 .h5?! has little point when it can be met by
..
10 h4 or even 10 Ngf3!? Nf6 1 1 Nh4, but Czap's 9 . Ne5!? is an idea, keeping the h
..
p awn in reserve and meeting 10 Ngf3 with 10 ... Ng6. Here I believe that probably
White should counter Black's plan of ...e5 vigorously with 1 1 Qc2 e5 1 2 dxe6!? fxe6
13 e4 fxe4 14 Nxe4 d5 15 Neg5 followed by continuing in sharp vein with 0-0-0.
10 Nh3 (Diagram 63)

White aims to keep control Decent play for the pawn
10...0-0
Natural enough, but now White can castle safe in the knowledge that the h-pawn
won't trouble him. Thus Black might prefer one of:
a) 10 ... Bd7 1 1 Qc2 Qa6 12 0-0!? (bold, but 1 2 Nf4 e5 13 dxe6 fxe6 14 0-0 is hardly a
dead end for White and 14 ...h5 1 5 e4! fxe4 16 Nxe4 Nxe4 1 7 Qxe4 0-0-0 18 Qg6 saw
hi m obtaining the upper hand in K.Pettersson-J.Alvarez Sabor, correspondence
2006) 12 ... 0-0 (12 ... Qxe2!? is critical, but enables White to improve his pieces to
i deal squares with tempo: 13 Rfe1 Qa6 14 Bfl Qb6 15 Nc4 Qc7 16 Ne3, as indicated
by Wells) 13 e4!? fxe4 14 Nxe4 NxdS 15 NhgS BfS 16 Rad1 (Diagram 64) was ex
tremely unclear in V.Akopian-G.Hernandez, Merida 2000.
137
TIP: Attacking with the Trompowsky knights is always tempting, but

here a decent case might be made for 13 Nf4, especially since the
knight can retreat to d3 after 13 ...e5.
b) Even the normally-thorough Dembo appears to have been confused by all the
possible transpositions in this variation, as she doesn't deal with this specific posi
tion, but perhaps she would have recommended lO ... hS!? here. Compared to the
notes to White's 8th move, Black's knight can no longer come to eS and 1 1 Qc2 h4
(11 ... Bd7!? 1 2 Nf4 h4 should also be met by 13 Nc4-e3, whereas 13 0-0-0 Qa6 14
Kbl 0-0-0 15 e3 Ng4! 16 Nf3 hxg3 1 7 hxg3 Qa4 was about equal in Bui Vinh
O.Mihok, Budapest 2007) 1 2 Nc4 Qa6 13 Ne3 transposes to our coverage there.
11 o-o (Diagram 65)

The kn ight is useful on d2 for now White manoeuvres ca reful ly
White has obtained quite a harmonious position and is able to keep the central
situation under control after 1 1 ...e6, as we'll see in Game 18.
Conclusion
This most positionally-complex of variations has warranted a fair amount of cov
erage. White's score remains at 51% after 5 Qcl fS 6 g3 Bg7 7 c3, but he can hope to
outmanoeuvre and out-finesse Black. Knowledge of exact moves might be of less
importance than after 6 c4, but there are still some important move order issues
when Black delays castling.
138
The U n co m p rom i s i ng 2 ... c s
Il lu strative Ga mes
came 18
0 P.Wells P.Popovic
Au strian League 2002
1 d 4 Nf6 2 Bg5 c 5 3 Bxf6 gxf6 4 d 5 Qb6 5 Qc1 f5 6 g 3 Bg7 7 c 3 d 6 8 Nd2 Nd7 9 Bg2
Nf6 10 Nh3 o-o 11 o-o e6 12 dxe6
In his groundbreaking work on the Trompowsky, Wells pointed to 12 Nf4!? as the
way forwards. Once again White retains the central tension and would like to con
tinue with at least some of a4, Nc4, Qc2 and a rook to d l . Such a plan certainly
worked out well after 12 ...Bh6 1 3 e3 Bd7 14 Rd1 e5 15 Ne2 Rae8 1 6 Qc2 Qd8 1 7
Nc4 Qe7 18 a 4 b6 19 Re1 Kh8 2 0 Rad1 Rg8 2 1 Nel l Rg6 2 2 Nd3 RegS 23 b4 (Dia
gram 66) when White was slightly for choice in M.Adams-J.Lautier, Paris (rapid)
1 995. Fine, calm manoeuvring so typical of the Englishman's style, although I
would imagine that not everyone would feel so uncomfortable with Black's
chances after 14 ... e5. Indeed, I'm by no means certain that Wells's own over-the
board choice is actually inferior, and whether or not to open the centre, at least
when it is fairly favourable to do so as here, is probably a matter of taste.
12 Bxe6?1
...
Ugly, but Popovic had spotted the chance to get his central pawns rolling. Here I
agree with Wells that 12 ... fxe6 should have been preferred, after which 13 Nf4! d5
1 4 c4 (Diagram 67) gives White useful central pressure and decent squares for his
pieces in the event of 14 ... d4 15 Nd3.

Harmonious play White will play on the queenside
139
Sta rti n g Out: T h e Trom pows ky Attack
NOTE: Having exchanged on e6 White would like to open up the

centre with e4 to target e6 and h7, but here 13 e4 can be fairly well
met by Wells's 13 ...esl, playing against that knight on the rim.
13 Nf4 d5 14 c41
White wastes no time fighting for the d5-square and forcing his powerfully-placed
bishop back into play.
14...Rad8 15 Qc2 d4 16 a3 Bh6 17 Nd3 (Diagram 68) 17... Kh8?1
I quite like White's chances in any case, but suspect that Black should have pre
ferred the more solid 17 ...Qc7 18 b4 b6 19 Rab1, restricting White to just a pull.
18 Rab1
Far from terrible, but as Wells later pointed out, even better would have been the
immediate 18 b4, since 18 ... cxb4 19 axb4 Bxd2?! 20 c5! is a handy intermezzo.
18 ... Bxd2?1
One can understand Black's desire to change the nature of the struggle, but this
fails to wholly convince. Once again grim defence with 18 ... Qc7 would have been
the lesser evil.
19 Qxd2 Bxc4
Giving up one half of the bishop-pair was a big decision, as was this. The alterna
tive was 19 ... Ne4!? but in any case Black's kingside remains a long-term cause for
concern, and here 20 Qc2! (20 Qf4 Qd6! 21 Qxd6 Rxd6 suddenly becomes quite
unclear due to the weakness of c4) 20 ... Qc7 21 Rfcl b6 22 b4 Rb8 23 Nf4 would
have retained a pleasant advantage.
20 Rfc1 Bxd3 21 exd3 (Diagram 69)

Black is in trou ble An offside queen
140
The U n com p ro m i s i ng 2 . . . c s
It' s time to take stock. Black might be a pawn up, but he can't defend all his weak
pawns and is clearly in some trouble. However, Black does have a queen and
kni ght tandem, and so Popovic does his best to stir up some trouble with those
p ieces.
21 ...Qd6 22 Bxb7 Ng4 23 Qgs Qd7 24 Bg2 f6 25 Qhsl?
A little ambitious, especially when the simple 25 Qd2 ReB 26 Rc2 Rc7 27 Rbcl RfcB
was available. Admittedly it's not trivial to make further progress as White here,
not least because the knight must be kept under control, but I think that the ambi
tious 2B b4!? might be on: 2B ... cxb4 29 h3 NeS 30 axb4 Rxc2 31 Rxc2 Rxc2 32 Qxc2
and with 33 f4 followed by 34 Qc5 next up, White appears to be doing pretty well.
z s ...RgBI
Black spots that the queen can't get off the h-file so easily.
26 Rxcs?l
Now White's queen falls into some trouble. Preferable was 26 h3 Rg5 (26 ...Ne5 27
RxcS Nxd3 2B RdS Qe6 29 QxfS QxfS 30 RxfS is pretty good for White despite the
p assed d-pawn) 27 Qh4 NeS 2B RxcS Qg7 29 Rbcl!, relying on Black's weak back
rank to keep everything together.
26 ... Rg5 (Diagram 70) 27 Qh4
I dare say that time trouble had already reared its head after such a complex posi
tional struggle. The text is easy to criticize, but it would have taken a brave man to
play 27 Qh3!? Qe6 2B Bf3, especially when low on time. Moreover, it's not appar
ent that this is actually so clear: for instance, 2B ...Qa2 29 Qfl NeS 30 Bh1 f4 and
Black retains some tricky counterplay.
27 ...Qf7
A better way to tee up ... RhS would actually have been 27 ...QeB! in view of the
nasty tactic 2B RxfS! Rxf5 29 Qxg4 Rxf2!, although anything might still have hap
pened here after 30 Be4 Rc2 31 Rfl Qe7.
NOTE: This tactic exploited a loose white rook and loose pieces lie
behind many tactics. Indeed, it might well be useful to always keep
John Nunn's mnemonic LPDO in mind - 'loose pieces drop off'.
28 Rxfsl Yz-Yz
Good defence, whereas White's kingside would have been left looking a little
sh aky after 2B Rbcl RhS 29 ReB Rxh4 30 RxdB+ Kg7 31 gxh4 Qe7. After the text,
play might have continued 2B RxfS RxfS 29 Qxg4 ReS 30 Be4 when White has two
pawns and a good bishop for the exchange. He might even be slightly for choice
h ere, but after so many adventures over the past few moves, one can hardly blame
W ells for offering a draw.
141
Chapter Five
T h e Po p u l a r 2 N e4
- I ntrod uction
- The Radical 3 h4
- The Creative 3 Bh4

A/''/
I
--
T h e Pop u l a r 2 ... Ne4
I n troduction
1 d4 Nf6 2 Bg5 Ne4 (Diagram 1)

An ever-popular choice An audacious concept
We now come to Black's main response to the Trompowsky, in terms of both

theoretical attention and tournament practice (2... Ne4 accounts for 32% of all the
Trompowsky games in my database). Advancing the knight with tempo is clearly
a logical idea, although White can hope to regain a tempo by kicking back the
knight with f3.
At this important juncture, White basically has a choice between three options: the
highly-creative 3 h4, the old main line with 3 Bh4, and the almost universal choice
these days amongst grandmasters, 3 Bf4. In this chapter we will explore both 3 h4
and 3 Bh4, before moving on to 3 Bf4 in our final two chapters.
The Radical 3 h4
1 d4 Nf6 2 Bg5 Ne4 3 h41? (Diagram 2)
An ambitious idea! White is happy to swap the bishop-pair for an open h-file and
h opes to engage Black in a complex struggle right from the word go. Indeed, it is
n o surprise to learn that this was Hodgson's main weapon in the early Nineties
before a couple of decent responses were worked out for Black. Nowadays with 3
Bf4 all the rage, 3 h4 might once again make a good surprise weapon in the hands
of a creative player.
3 cs
...
143
Black's most popular response, but this has been played in just under half of all
games with 3 h4 and so we must also consider:
a) 3 ... Nxg5 4 hxg5 is a little too compliant. Black must now try to distract White
from the target on h7 with 4 ... c5 in any case, whereas 4 ...e6 5 f4!? d5 6 e3 Bb4+ 7 c3
Bd6 8 Nf3 Nd7 9 Qc2 g6 1 0 Bd3 left White with a strong grip on the position in
A.Ihaksi-R.Reijo, Finnish League 2007.
NOTE: Should Black obtain the bishop-pair on gS, White will usually
respond by targeting the black kingside at an early stage. Some
times White can play to double rooks on the h-file, other times a dis
ruptive g6-advance will cause chaos.
b) 3...f6 pushes White back and after 4 Bf4 (4 Nd2!?) 4 ... Nc6! we reach an almost
totally unexplored position. One rare reference continued 5 f3 e5 6 Bh2 f5! ? 7 Bxe5!
Nxe5 8 fxe4 Ng4 9 exf5 d5?! 10 e4 h5 1 1 Bd3 and White had seized the upper hand
in N.Eliet-J.Chabanon, French League 1996.
c) 3...h6 isn't essential at this early stage, but is quite playable if Black follows up
with 4 Bf4 d5 - compare with our next variation to which play may transpose.
d) The solid 3 ... d5 is a significant drawback to 3 h4 in my view. Black shows that
he doesn't have to play ball and simply develops his pieces along classical lines.
After 4 Nd2 (Diagram 3) we have:

Black is quite solid A small lead in development
d l ) 4 ... Nxd2 5 Qxd2 (5 Bxd2!? e5?! 6 dxe5 Nc6 7 Bg5 Be7 8 Nf3 Bg4 9 Qd2 Qd7 10
0-0-0 favoured White in J.Hodgson-M.Hebden, Candas 1992; having started sol
idly, probably Black should continue in such vein) 5 ... Bf5 leaves White ahead in
development, but it's far from easy to exploit that with Black yet to weaken his
144
The Pop u l a r 2 . . . Ne4
positi on: 6 e3 Nd7 7 Bd3 (7 Nf3 h6 8 Bf4 e6 9 Ne5 Bd6 10 h5 Nxe5 11 Bxe5 0-0 12
B x d6 Qxd6 13 Bd3 unsurprisingly soon led to a draw in J.Hodgson-V.Tukmakov,
Be rn 1995) 7... Bxd3 8 Qxd3 c6 (alternatively, 8 .. .f6!? 9 Bf4 e5 10 Bg3 c6 1 1 0-0-0
looks like the critical line with roughly balanced chances, and the solid 8 ...h6 9 Bf4
e6 transposes to a position which Rowson has been happy to defend via a 4 ... Bf5
move order) 9 Nf3 Qb6 (Diagram 4) 10 0-0-0 0-0-0?! 1 1 Bf4! e6? 12 Ng5 saw Black
wishing he'd never moved his king from e8 in A.Abdulla-S.Mishra, Mumbai 2008.
d2) 4 ... Nxg5 5 hxg5 c5!? (Black fights for the initiative; again he might do better to
p lay more solidly with 5 ...Bf5, although 6 e3 e6 7 g4 Bg6 8 f4 Bd6 9 Ngf3 Nd7 10
Nh4! Qe7 11 Nxg6 fxg6 12 Bd3 0-0-0 13 Qe2 gave White a pleasant long-term ad
van tage in A.Miles-M.Bescos Mambrona, Andorra 1995, and the more common
8 ... c5 is fairly well met by Hodgson's 9 Qf3!? when 9 ... Bxc2 10 Bb5+ Nd7 1 1 Rcl
6e4 12 Nxe4 dxe4 13 Bxd7+ Qxd7 14 Qxe4 regains the pawn with advantage) 6
dxcS e5 (Diagram 5)

White can seize the initiative A highly-respectable defence
7 e4!? (consistent with the aggressive nature of 3 h4, White fights for the initiative
rather than opt for the solid 7 Nb3) 7...Bxc5?! (too ambitious, but White also has
the initiative after 7... Nc6 8 Ngf3 Be6 9 Bb5) 8 Nb3! d4 9 Nxc5 Qa5+ 10 c3 Qxc5 1 1
cxd4 exd4 12 Rcl Qb4+ 13 Qd2 Qxd2+ 1 4 Kxd2 (with weaknesses on d4 and h7 to
aim at, White already has a large advantage) 14 ... Nc6 15 Bb5 Bd7 16 Nf3 0-0? 1 7
Rh4 Rad8 18 Rch1 i s something o f a model for White, A.Stefanova-J.Ivanov, Salou
2000.
d 3) Again a very sensible choice is best, namely 4 ... Bf5! (Diagram 6) when White
h as tried two approaches:
d 3 1 ) 5 e3 h6 6 Bf4 e6 (the Rawson-endorsed 6 ... Nxd2 7 Qxd2 e6 8 Bd3 Bxd3 9 Qxd3
Nd7 is a respectable alternative) 7 g4!? Bg6 (Black takes control of the gS-square;
145
Sta rt i n g Out: T h e Trom pows ky Attack
he also has 7... Bh7 8 Nxe4 Bxe4 9 f3 Bh7 10 Bd3 Bxd3 11 Qxd3 c5!, which gave him
enough counterplay in J.Hodgson-A.Beliavsky, Groningen 1994) 8 h5 Bh7 9 Nxe4
Bxe4 10 f3 Bh7 1 1 Bd3 Bxd3 12 Qxd3 c6 looks quite pleasant for White, but it's ac
tually hard to make any real progress: for example, 13 Nh3 Bd6 14 Qb3 Bxf4 15
Nxf4 Qc7 16 Rg1 Nd7 1 7 0-0-0 0-0-0 and Black was very close to full equality in
A.Bigg-Jo.Hodgson, British League 2002.
d32) 5 Nxe4 Bxe4 6 f3 h6! 7 fxe4 (this doesn't totally convince, but practice has
shown that 7 Bf4 Bh7 followed by ... c5 is quite comfortable for Black) 7 ... hxg5 8
Qd3 e6 9 Nf3!? (9 Qb5+? Nc6 1 0 Qxb7 Nb4 1 1 Qb5+ Qd7 saw the maestro getting
into trouble in J.Hodgson-V.Salov, Wijk aan Zee 1993) 9 ... Nc6!? 10 exd5 Qxd5
(Diagram 7) 1 1 e4 Qd6 12 0-0-0 gxh4 13 Qe3 0-0-0 14 Bb5 and White will regain his
pawn with rough equality.

Another unbalanced middlega me Keeping the centre closed
Returning to 3 ... c5:

4 dxc5
The more common choice, but I suspect that 4 d5!? (Diagram 8) might actually be a
better bet, refusing to worry about the b2-pawn. Play usually continues 4 ...Qb6
(two other options worth considering are 4 ... Qc7!? with the cheeky idea of 5 Nd2
Ng3 6 Rh3 Nxfl 7 Nxfl d6, and the sensible 4 ... g6, after which 5 Nd2 Nxg5 6 hxg5
Bg7 7 c3 d6 should be fine for Black, who can aim for Benoni-like queenside coun
terplay, while Hodgson's 5 Qd3!? can be met well enough by Emms's 5 ...Qa5+) 5
Nd2! (White prefers to sacrifice a pawn than cede the initiative after 5 Qcl e6)
5 ... Nxg5 (the other pawn grab is more risky: 5 ... Nxd2 6 Bxd2 Qxb2 7 e4 leaves
Black unable to both retreat his queen and prevent e5; indeed, 7... g6 8 Rb1 Qe5 9
Bd3 Bg7 10 f4 Qc7 1 1 c4 e6 12 h5 gave White excellent compensation in L.Large
T.Therkildsen, French League 2008) 6 hxg5 Qxb2 (declining the pawn is no longer
146
The Popu l a r 2 . . . Ne4
so good: 6...g6 7 e4 d6 8 Nc4 Qc7 9 a4 gives White a pleasant grip; he will expand
further with f4 before Nf3) 7 e4 (creating a target on g6 with the immediate sacri
fice 7 g6!? has been a popular alternative) 7... Qe5?! (Black hurries to move his
queen; this fails to convince, but to demonstrate the dangers he faces just look at
the line 7... d6? 8 Rb1 Qa3 9 Bb5+! Nd7? 10 Nc4 Qc3+ 1 1 Kfl and there's no good
defence to 1 2 Ne2; probably rushing another piece into play with 7... Na6 is best) 8
Ngf3 Qc7 9 g6! fxg6 10 e5 e6 1 1 d6 (Diagram 9) left White with fantastic compensa
tion in A.Bigg-M.Stierle, Lenzerheide 2006.

That's some pawn wedge! Everything to play for
TIP: Whenever Black neglects to follow up 3 ...c5 with ...g6, White

should keep the possibility of a sacrificial g6-advance in mind, ham
pering Black's development while creating a target for es and Bd3.
4 Qa5+
...
Bl ack often flicks this check in, but there's absolutely no compunction to. The al
ternatives are:
a) 4 ... Na6 is quite a solid choice. Following 5 Nd2 (the ever-creative Miladinovic
h as remained true to the gambit 5 Qd4!? Naxc5 6 Nc3 Nxc3 7 Qxc5 Ne4 8 Qd5
Nxg5 9 hxg5 Qb6 10 0-0-0 Qxf2 1 1 Rh3; I have my doubts that this can be thought
of as more than a reasonable surprise weapon, although 1 1 ... e6 12 Qd2 Be7 13 Nf3
Qcs 14 e4 d6 1 5 e5! dxe5 16 Rh4 0-0 17 Bd3 certainly gave White a strong attack in
LM ila dinovic-B.Chatalbashev, Reggio Emilia 2005) 5 ... Naxc5 6 Ngf3 Qb6 7 Nxe4
Nxe4 8 Qd4 Qxd4 9 Nxd4 a6 10 g3 e5 1 1 Nb3 d5 12 Bg2 (Diagram 10) an unbal
anced middlegame is reached. Gallagher slightly prefers White here, although I'm
not sure that 12 .. .f5 13 Bxe4!? fxe4 14 c3 would have been so bad for Black in
J. Hodgson-M.Hebden, Cappelle Ia Grande 1992, had he continued 14 ... a5! 15 0-0-0
147
a4 16 Na1 Be6 1 7 Nc2 Bd6.

b) 4 ... h6 5 Bf4 e6 is another sensible approach and one which just looks like rather
an easy equalizer, although the 6 Nd2 Qf6!? (highly ambitious, especially as
there's nothing whatsoever wrong with 6... Nxc5) 7 Nxe4! Qxf4 8 Qd3 Na6 9 Nd6+
Kd8 10 Nh3 Qb4+ 1 1 c3 Qxb2 12 Rb1 Nxc5 13 Nxf7+ Ke8 14 Nd6+ Kd8 15 Nf7+ and
V2-V2 of H.Gohii-P.Svidler, Mainz (rapid) 2001, was probably not what the strong
Russian Grandmaster was after.
c) 4 ... d5!? is an unexplored gambit which I wondered about some years ago and
recently saw being discussed on the excellent forum section of the ChessPublishing
website. Black certainly did well in a rare outing in R.Hemmerling-G.Hoegerl,
correspondence 2004: 5 Nd2 (Black's point is 5 cxd6 Qb6 when 6 e3 Qxb2 7 dxe7
Bxe7 8 Bxe7 Kxe7 9 Nd2 Nxd2 10 Kxd2 Qb4+ 1 1 Kcl Rd8 supplies good compensa
tion with his king far the least exposed) 5 ... Nxc5 6 e4?! (too ambitious, although
otherwise the value of h4 is beginning to look questionable) 6 ...h6 7 Be3 d4 8 Bb5+
Nc6 and White was already worse.
s Nd2 Nxgs 6 hxgs g6
Prudent. By now the sacrifice 6 ...Qxc5 7 g6!? (Diagram 11) shouldn't come as a
surprise to the reader, and 7 .. .fxg6 8 e3 Nc6 9 Ngf3 d6 10 c3 Bg4 (10 ... Bf5?! 1 1 Nb3
Qb6 12 Ng5! Ne5 13 a4 saw White seizing the initiative with some typically pow
erful play in I.Miladinovic-L.Genova, Nice 2002) 1 1 Rh4 Bxf3 (as pointed out by
Prie, Black does much better with 1 l ...h5!, intending to whisk his king away to the
queenside) 12 Nxf3 Qb6 13 Qc2 1eft White set to regain the pawn with interest in
A.Fier-N.Mayorga, Buenos Aires 2007.

A thematic blow Can White use the h-file?
7 C3
I suspect that White may do better with 7 Rh4!?, although it helps to have quite a
148
The Pop u l a r 2 . . . Ne4
creative streak if one wishes to play thus. One of Hodgson's games continued
7 ... Nc6 (and not, of course, 7... QxcS?? 8 Rc4) 8 Rc4 Bg7 (Black has also fought for
the initiative with 8 ...dS!? 9 cxd6 Be6, but then Miladinovic's 1 0 c3! looks like a
reasonable counter, intending 1 0... Bxc4?! 1 1 Nxc4 QxgS 12 d7+ Kd8 13 Nf3 with
good compensation) 9 c3 NeS 10 Nb3 Qc7 1 1 Rh4 aS!? 12 a4 b6 13 cxb6 Qxb6 14
Nf3 with a highly original and unclear position, J.Hodgson-M.Pein, London 1990.
7 ...Qxcs 8 Ngf3 Bg7 9 e3 Nc6 {Diagram 12)
Black has developed sensibly, and now the issue is whether White will be able to
do anything with the half-open h-file:
a) 10 Nb3 Qb6 1 1 a4!? d6 (Black might also pose the question with 1 1 ... hS!?; here
12 gxh6 Rxh6 13 Rxh6 Bxh6 14 Bc4 d6 1S Qe2 Bg4 16 0-0-0 Bg7 looks okay for the
second player despite the issue of f7) 12 aS Qc7 13 Ra4 saw some typically creative
play from White in J.Hodgson-J.Nunn, English Championship, London 1991, but
1 3 ... Bd7 14 Qa1 hS! 1S gxh6 Rxh6 16 Rxh6 Bxh6 {Diagram 13) saw Black refusing to
panic and retaining a rough balance.

Black defends sensibly Preparing pressure aga inst h7
b) The immediate 10 Rh4 has also been seen, after which 10 ... dS 1 1 Nb3 Qd6 12 e4
Be6 13 exdS BxdS 14 Bc4 Rd8 1S BxdS QxdS 16 QxdS RxdS led to a roughly bal
anced endgame in P.Jirovsky-M.Brodsky, German League 2006.
c) 10 Bd3!? d6 1 1 Qc2 aims to dissuade ... hS ideas and will be seen in Game 19.
Conclusion
Ju st as it once did with Hodgson, 3 h4 is likely to appeal to creative types who are
keen to escape from the beaten theoretical track. White has scored 51% from over
149
1,300 games with 3 h4, but nowadays it's unlikely to trouble a well-prepared op
ponent. That said, its shock value against the uninitiated should not be dis
counted. The solid 3 ... dS is an eminently sensible response, questioning the use
fulness of the pawn on h4, and the more dynamic 3 ... cS also offers Black a fair
share of the chances.
Illustrative Games
Game 19
D M.Kanep A.Holmsten
Fin n ish League2006
1 d4 Nf6 2 Bg5 Ne4 3 h4 c5 4 dxc5 Qa5+ 5 Nd2 Nxg5 6 hxg5 g6 7 c3 Qxc5 8 Ngf3
Bg7 9 e3 Nc6 10 Bd31? (Diagram 14) 10 ... d6
Standard development, but I suspect that Black might also be more ambitious at
this point with lO .. dS!?, after which 11 e4 (White needs to challenge either here or
.
immediately after ...eS) 1 l ...Be6 12 Qa4 a6 13 Nb3 Qd6 14 0-0-0 Bd7!? posed some
problems for the white queen in M.Arbouche-V.Akobian, Turin Olympiad 2006,
and would have been about equal after 1S exdS Nd4 16 QaS b6 1 7 Qb4 Qxb4 18
cxb4 Nxb3+ 19 axb3 aS 20 Rh4 with a messy structure; a far from atypical event
after 3 h4.
11 Qc2 b51?
Black hurries to obtain counterplay. He has also developed with:
a) 1 l ...Be6 12 Nb3 Qb6 13 Nfd4?! Bd7 14 Nxc6 bxc6 1S 0-0-0 aS! gave Black good
counterplay in H.Teske-R.Kotter, German League 2001, but I would prefer 13
Nbd4 Bd7 14 0-0-0.
b) 1 l ...Bd7 12 Rh4!? (12 0-0-0 looks like a decent alternative, with the idea of
12 ... 0-0-0 13 Ne4 Qb6 14 Bc4, probing the weaknesses fixed by that pawn on gS)
12 ... 0-0-0 13 0-0-0 hS! was fine for Black with his rooks connected in J.Sanchez-
C.Robles Garcia, La Roda 2008.
12 Rh4 Bd7
NOTE: The point of White's queen and bishop battery is revealed af-
ter 12 ... h5? 13 Bxg61 when 13 ...fxg6? 14 Qxg6+ regains the piece
with significant interest as 14 ... Kf8? can be met by 15 Rf4+.
13 Ke21 (Diagram 15)
Kanep's creative play is fully in keeping with his opening choice. White doesn't
want to castle into a strong attack, but would like to double on the h-file and so
moves his king forward to facilitate that.
150
T h e Pop u l a r 2 . . . N e4
1 3 ... b4 14 Nb3 Qb6 15 Rah11?

It' s tempting to go after h7, but Black will obtain enough for the pawn with c3
weak. Thus I suspect that White does better to first close queenside lines with 15
c4! before doubling. Moreover, it's not impossible that he might also play for a
central initiative with a timely c5-push and, of course, 15 ... h5? still fails to 1 6
Bxg6!.
15 ... bxc3 16 Rxh7 Rxh7 17 Rxh7 Nb4? (Diagram 16)

White must con nect his rooks Too a mbitious from Black
Black is enticed by some tactics, but he should never have given up his key bishop
so lightly. He had to prefer 17 ... Kf8 when 18 bxc3 Rc8 19 Rh4 would have re
mained quite unbalanced, although one would be surprised if Black could be
worse here with his bishops and strong centre.
18 Rxg71
The strong Estonian IM, arguably these days the leading exponent of the 3 h4
Trompowsky, has the tactics all worked out.
18 ... Kf8?1
Continuing his faulty plan. A better defence was 18 ... Bb5 based on the point that
19 Qxc3 Nxa2 20 Bxb5+ Qxb5+ 21 Qd3 Ncl +! is a handy resource. Following 22
Nxcl Qxb2+ 23 Nd2 Qxg7 24 Qb5+ Kf8 25 Qb7 ReB 26 Qxa7 White has decent
ch ances to eventually co-ordinate his queen and knights, but I suspect that this
radical defence would have offered Black better chances than the game.
19 Qxc3 ReS 2 0 Bc4 Rxc4
The point of Black's play. Instead 20... d5? 21 Rh7 would, of course, have been
hopeless for him.
151
Sta rt i n g Out: The T rom powsky Attack
21 Rxf7+1 Kxf7 22 Qxc4+

Thus White gets to capture the rook with check and so avoid having his queen
skewered.
22 ... Be6 23 Qf4+ Ke8 24 Kf11
TIP: It's always good to make sure that one's own king is safe before
launching an attack. Here White would like to attack with his queen
and knight-pair, but first Kanep returns one of his extra pawns to
avoid a potentially awkward counterattack.
24...Qa6+ 25 Kg1 Nxa2?1
A better defence was 25 ...Qxa2 26 Qxb4 Qxb3, although with 27 Qd4 White would
have retained a clear advantage.
26 Nbd4 (Diagram 17)

The pressure mou nts The old main li ne
26 ... Bds?l
Under pressure with his pieces uncoordinated, Holmsten collapses. He had to
cover g6 with 26 ... Bf7, even if 27 Qe4 followed by Nh4 would most likely have
won that pawn in any case.
27 e4! Bb7
Now White's pieces pour in, but 27...Bf7 28 eS would have been crushing too.
28 Ne6 Qc4? 29 Qf8+ Kd7 30 Qd8+1 1-0
A neat finish. Black's queen falls after 30 ... Kxe6 31 Qg8+ and 30 ... Kc6 31 Nfd4+
also forces him to part with her company.
152
The Pop u l a r 2 .. Ne4
.
The Creative 3 Bh4

1 d4 Nf6 2 Bg5 Ne4 3 Bh41? (Diagram 18)
This was once quite a popular handling of the main line Trompowsky, but is now
rarely seen, largely due to our main line, below. Nevertheless, a creative player
who revels in unbalanced structures may find something to work with here and
Nigel Davies was certainly of the view that 3 Bh4 was worthy of a reappraisal in
his 2005 work.
3 cs
...
Black's main move, but there are important alternatives:

a) 3 ... g5 4 f3 (with Black having weakened his kingside, 4 Bg3 cannot be so bad,
but Black will still gain good play on the dark squares, such as with 4 ... c5 5 e3 Qb6
6 Nc3!?, as in O. Danielian-M.Hebden, Cappelle Ia Grande 1993, and then Galla-
gher's sensible suggestion of 6... Nxc3!? 7 bxc3 Bg7) 4 ... gxh4 5 fxe4 usually trans-
poses to our main line after 5 ... c5 6 e3. Black can, though, also attack on the dark
squares with 5 ... e5!? (Diagram 19), after which 6 Nf3! exd4 (the 6 ...Bh6!? 7 Nxe5 d6
8 Nf3 0-0 9 Nc3 f5! 10 exf5 Bxf5 1 1 e4 Bg4 of M.Illescas Cordoba-J.Gallagher, Ali
cante 1985, is a decent, fun alternative) 7 Qxd4 Rg8 8 Qe5+ Qe7 9 Qxc7 Na6 10 Qc4
b5 was pretty unclear in L.Keitlinghaus-R.Knaak, German League 1991.
Diagram 19 {W) Diagram 20 {B)

Striking out on the dark squares Black has a l ready gone wrong
WARNING: White must appreciate just how weak his dark squares
can be in this variation. That doesn't mean that 3 Bh4 is unplayable,
but care is required. In our last variation, for instance, 6 e3?1 Qgsl is
awkward since Black will increase the pressure further with 7 Bh6....
153
b) The solid 3 ...d5 cannot, of course, be such a bad choice, but it does give White
much more hope of obtaining an advantage than the sharper options against 3
Bh4. Here 4 f3 forces the knight back without delay and then:
bl) 4 ... Nf6 5 Nc3 Bf5 (Black tries to avoid the gambit 5 ... c5 6 e4! dxe4 7 d5 with
extra space and good compensation, but he might settle for the solid, French-like
5 ...e6 6 e4 Be7 when White has various aggressive options, as well as a direct
transposition to the Classical French: 7 e5 Nfd7 8 Bxe7 Qxe7 9 f4) 6 Qd2 Nbd7?!
(after the superior 6 ... e6 I wonder about going long and advancing the g-pawn;
White might also advance in the centre, but 7 Bxf6 Qxf6 8 e4 dxe4 9 fxe4 Bg4 10
Be2 Nc6! 1 1 Bxg4 Qh4+ 12 Qf2 Qxg4 13 Nf3 Bb4 turned out okay for Black in
L.Alburt-Cu.Hansen, Reykjavik 1986) 7 g4! (Diagram 20) 7...Bg6 (7...Be6 8 Nh3
would also leave Black in a pickle) 8 g5 Nh5 9 Nxd5 netted a pawn for insufficient
compensation in S.Mamedyarov-B.Ashe, online blitz 2004.
b2) 4 ... Nd6 is the more usual retreat, but here the knight is not ideally placed and
5 Nc3 Nf5 (or 5 ... c6 6 e4 Qa5 7 Bd3 with a pleasant pull in A.Femandes-L.Galego,
Portuguese Team Championship 2002; note that 7 ... c5?! can, as often in this varia-
tion, be well met by 8 dxc5!) 6 Bf2 (Diagram 21) 6 ... e5 (the sharper 6 ... c5!? 7 dxc5 d4
can be met by either the aggressive counter 8 e4!? or more simply by 8 Ne4! Nc6 9
g4 Qa5+ 10 c3 Ne3 1 1 Bxe3 dxe3 1 2 Qd3 g6 13 Qxe3 when, like Gallagher, I don't
find Black's compensation entirely convincing, J.Hodgson-Z.Sturua, Tilburg 1993)
7 dxe5 d4 8 e4! dxc3 (practice has shown the e3-pawn to be more of a weakness
than a strength after 8 ... dxe3 9 Qxd8+ Kxd8 10 0-0-0+ Bd7 1 1 Bel !) 9 Qxd8+ Kxd8
10 0-0-0+ Ke8 1 1 exf5 cxb2+ 12 Kxb2 Bxf5 13 Ne2 leaves White with a useful lead in
development and the safer king, M.Carlsen-K.Lie, Norwegian League 2004. It's
interesting to see the then 2550-rated Carlsen giving 3 Bh4 a go.

White prepares 7 e4 Black doesn't have to retreat
154
The Popu l a r 2 . . Ne4
.
c) 3 ... c6 has a little tactical idea in mind: 4 f3 Qa5+ 5 Nd2 Qh5 and Black obtains
the bishop-pair. White should, however, prefer 4 Nd2!, after which 4 ...Qa5 5 c3
Qh5 6 Nxe4 Qxh4 7 Ng3 d5 8 Nf3 followed by 9 e4 should suffice for a pull, as did
the 4 ... Nxd2 5 Qxd2 d5 6 f3!? Bf5 7 0-0-0 of R.Vaganian-J.Moreno Camero, German
League 2003.
Returning to the critical 3 ... c5:
4 f3 (Diagram 22)
With the bishop a little sidelined, 4 d5?! Qb6 cannot be the way to go and I also
find Rausis's 4 Nd2 a little inconsistent. Admittedly the move is solid, but even
4 . Qa5 (I doubt this is quite the most accurate; instead the slightly sharper 4 ... d5!?
. .
5 Nxe4 dxe4 6 dxc5 Qa5+ 7 c3 Qxc5 is fine for Black, as must be 4 ... Nxd2 5 Qxd2
cxd4 6 Nf3 Nc6 7 Nxd4 Qb6) 5 c3 Nxd2 6 Qxd2 cxd4 7 cxd4 Qxd2+ 8 Kxd2 e6 9 a3
Nc6 10 e3 b6 1 1 b4 Bb7 12 Bg3 Be7 13 Ne2 ReS 14 Nc3 d5 15 Bd3 Kd7 16 f3 Bd6 was
hardly all that much for White to write horne about in S.Marnedyarov-A.Rornero
Holmes, Benidorrn (rapid) 2003.
4...gsl
Black pursues his quarry. This is critical, but not the only reasonable approach.
Indeed, the inclusion of 4 ... Qa5+!? 5 c3 before 5 ... g5 (or just 5 ... Nf6 6 d5 Qb6 which
should be compared with lines in our final chapter; here 7 Bel isn't available, but
Hodgson dabbled in 7 e4!? Qxb2 8 Nd2 in the mid-Eighties - a variation with
much in common with a certain highly topical gambit after 3 Bf4) 6 fxe4 gxh4 is
hard to assess. White has bolstered d4 and Nd2-c4 may come with tempo, but he
is deprived of an important Nc3 resource. Indeed, 7 e3 Qb6 (best I think, although
even 7...Bh6 8 Nd2 Qb6 9 Nc4 Qg6! 10 Qf3 d6 turned out okay for Black in
G.Chepukaitis-A.Eliseev, St Petersburg 2003) 8 Qc2 Bh6 9 Nd2!? (not ideal, but
one can understand Ward's reluctance to become involved in 9 Kf2 cxd4 10 exd4
Nc6 1 1 Nf3 e5) 9...Bxe3 10 d5 Qh6 1 1 Ngf3 d6 12 0-0-0 Bg4 failed to give White
quite enough for his pawn in C.Ward-R.Palliser, British League 2005.
5 fxe4 gxh4 (Diagram 23 ) 6 e3
White shores up his centre, but might he prefer to sacrifice? Indeed, 6 Nc3!? was
seen in no lesser game than S.Marnedyarov-L.Van Wely, Moscow 2004, which
continued 6...cxd4 (6 ... Bg7!? 7 e3 e6 is a critical alternative) 7 Qxd4 Rg8 8 e5 Nc6 9
Qe4 Qa5 1 0 Qxh7 Rg4 1 1 0-0-0 Qxe5 with something of a mess.
NOTE: The unfashionability of 3 Bh4 might make it a decent surprise

weapon. Moreover, much remains to be discovered after it and crea
tive players who pursue its various unbalanced byways may find
themselves well rewarded.
6 ... Bh61
Black takes immediate aim at e3. This is critical, whereas 6 ...Qb6?! 7 Nc3! e6
155
Sta rt i n g O ut: T h e Trom pows ky Attack
(7...Qxb2 8 NdS! is the point and after 8 ... Kd8, Davies's 9 Rb1 Qa3 1 0 Nf3!? e6 1 1
Rb3 QaS+ 1 2 Nc3 Rg8 13 Qd2 offers White good, enduring compensation) 8 Nf3
Qxb2 (greedy, but it's not easy to suggest an improvement; 8 ... Bh6 was tried in
D.Rogozenko-J.Armas, Wijk aan Zee 2007, when I would be sorely tempted to go
9 NeS!? Qxb2 1 0 NbS with some initiative) 9 NbS Qb4+ 10 c3! QaS 1 1 NeS d6?
( 1 1 ... a6! 12 Nc4! QxbS 13 Nd6+ Bxd6 14 BxbS axbS 1 S QhS is the critical line, but
like Davies, I quite fancy the queen against the three pieces here) 1 2 Nc4 (Diagram
24) 12 ... Qd8 would have been a disaster for Black in A.Tugui-R.Brajovic, Bucha
rest 2000, had White simply played 13 dxcS! dxcS 14 Qf3 followed by the crushing
1S Rdl .

A messy structure The knights dominate
7 Kf2
This has been played in more than 60% of the games to have reached this critical
tabiya, but it hardly inspires. Let me quote no less an expert than Hodgson: 'How
can White hope to gain an advantage after a move such as 7 Kf2? The position
looks bad, feels bad and is bad!' So what should White do? Well, let's look at the
alternatives:
a) 7 Bc4 is a very aggressive try, but I don' t wholly buy it: 7 ... e6 (Black is advised
to avoid 7... Bxe3? 8 Bxf7+! Kf8 9 Nh3, but Davies points out that 7... dS!? is a move,
after which 8 exdS Bxe3 9 dxcS QaS+ 10 Kfl QxcS already left Black for choice in
A.Kharitonov-S.Novikov, Moscow 2003; indeed, White is in need of an improve
ment here, which 8 BxdS e6 9 QhS Bxe3 10 Nf3!? cxd4! 1 1 Bc4 Nc6 doesn't appear
to be) 8 Qf3 (8 QhS is well met by 8 ... QgS, but perhaps White might try 8 Nf3!?
when 8 ... Qb6 9 0-0! Bxe3+ 10 Kh1 h3 11 g3 d6 1 2 Nc3!? cxd4 13 NbS is murky and
at least gives him some compensation, M.Mihalik-M.Konopka, Slovakian Cham
pionship, Topolcianky 1 994) 8 ...QgS! 9 Kf2 cxd4 10 exd4 Nc6 (Diagram 25) 1 1 Nh3?
156
The Pop u l a r 2 ... Ne4
( White had to cover both weaknesses with 1 1 Ne2, even if 1 1 ... Bg7 followed by
castling and probably .. .f5 would have left Black in the driving seat) 1 1 ...Qg7 12
c3?! Nxd4! 13 cxd4 Qxd4+ 14 Ke1 Qxb2 15 Qc3 Bg7! and 0-1 was a complete disas
ter in T.Gavriel-R.Palliser, Sutton 1997. Those dark squares again!

Strong dark-square pressure Black has sufficient control
b) 7 Qh5?! prevents ...Qg5 defences, but does leave e3 en prise: 7 ... Bxe3 (7 ... Qb6!? 8
Nd2 cxd4 is a decent alternative) 8 Bc4 (Bellon Lopez once preferred 8 Nf3, but
8 ... cxd4 doesn't really give White anything better than 9 Bc4, transposing) 8 ... e6 9
Nf3 cxd4 10 Nbd2 Nc6 1 1 Rfl Qa5 1 2 Qxh4 Ne5 left White somewhat worse in
G.Buckley- D.Gormally, British League 1998.
c) 7 d5?! was Bellon Lopez's next try. The Spanish Grandmaster is renowned for
his creative, independent approach to the game, but here the 7...Qb6 (7... d6 8 Nd2
Nd7 9 Qh5 Bxe3 1 0 Nc4 Bd4 1 1 Nf3 wasn't so clear in N.Eliet-W.Van den Brande,
Belgian League 2005) 8 Qcl d6 9 Nf3 Nd7 (Diagram 26) of J.Bellon Lopez
J.Eriksson, Stockholm 1999, already looks quite comfortable for Black to my eye;
W hite's position is a little clogged up and Black has good play on the dark squares
once again.
d ) More often of late Bellon Lopez has been happy to sacrifice his e-pawn with 7
Nd2!?; a course of action endorsed by Davies. One critical line runs 7... Bxe3 (the
vigorous counter 7... d5!? 8 Qh5 Bxe3 looks like a decent alternative: 9 0-0-0!? cxd4
1 0 Ngf3 dxe4 1 1 Bc4? exf3 12 Qxf7+ Kd7 13 Rhe1 Rf8 14 Qg7 Kc7 15 Rxe3 dxe3 16
Nxf3 Bd7 was most spirited but didn't give White enough for the rook in }.Van
Ruitenburg-A.Timman, Amsterdam 2006, and 9 exd5 cxd4 10 Nc4 Nd7 1 1 Nxe3
Qa5+ 12 c3 dxe3 13 Nf3 Nf6 has worked out well for Black in a few games, begin
ning with A.Romero Holmes-E.Sutovsky, Vendrell 1996; perhaps White must in
vestigate Davies's suggestion of 8 Qf3!? if he wants to keep 7 Nd2 afloat) 8 d5
157
Sta rt i ng O ut: The Trom pows ky Atta c k
Bxg1!? (more usual has been to maintain the bishop with 8 ...d 6 when 9 Nc4 Bd4 1 0
c3 Bf6 1 1 QhS!? Bd7 12 a4 Rf8 13 eS dxeS 14 NxeS wasn't at all clear in J.Bellon Lo
pez-P.Velicka, Figueira da Foz 2007) 9 Rxg1 Qc7, as in E.Lebraud-G.Grigore, Con
dom 2008, and now 10 Nf3 Qb6 1 1 eS!? Qxb2 12 Kf2 (Diagram 27) offers White a
strong centre and some long-term compensation for his pawns.

An i nteresting double ga mbit Both sides have serious trumps!
Returning to White's most popular choice, 7 Kf2:

7 ... cxd4 8 exd4 (Diagram 28) 8 ...Qb6
Black might very well play on the dark squares too with 8 ...e5!?. Again I find the
white position a little unpalatable, but that may just be a matter of taste. Certainly
some titled players have been happy enough here and 9 QhS Gust as in our main
line, I suspect that 9 Nc3! is best and 9 . Nc6 10 QhS exd4 11 Qxh6 dxc3 12 bxc3 dS
. .
13 exdS QxdS 14 Nf3 was about equal in G.Chepukaitis-I.Zakharevich, Saint Pe

tersburg 2000) 9 ... Qf6+! 10 Nf3 exd4 1 1 Bc4 saw White continuing in Nineteenth
Century style in the grandmaster-clash N.Sedlak-E.Vorobiov, Vrsac 2008, al
though I have my doubts that his compensation was sufficient after 1 1 ...Be3+ 12
Ke1 Nc6 13 Rfl Qg6.
9 Nc31
NOTE: This is the only move. Once again White must leave his b
pawn to its fate and attempt to seize the initiative with a rapid Nds.
9...e6
Black's usual choice, prudently preventing the aforementioned threat.
10 Nf3 Nc6 11 Bbs (Diagram 29)
White has developed very logically, but remains vulnerable on the dark squares
158
T h e Pop u l a r 2 ... N e4
and due to his king position. Black, on the other hand, is yet to develop his queen
si de and an unbalanced, roughly level struggle lies ahead. At this stage Black has
often developed with 1 l...d6, as we'll see in Game 20, but he might also try
l l . . .Bg7!?. Then 12 Bxc6 bxc6 13 e5!? (tempting, but probably too ambitious; I
w ould prefer 13 Na4, after which 13 ...Qb5 14 b3 Qh5 15 Re1 Rg8 was pretty un
clear in E.Tatlicioglu-G.Peli, correspondence 2006) 13 ... f6! 14 exf6 Bxf6 15 Ne4
Bg7! ? 16 Nd6+ Ke7 1 7 Nc4 Qb4 18 b3 Ba6 1 9 Qe1 Qxe1+ 20 Rhxel c5 (Diagram 30)
sa w Black starting to get on top with his bishops in the instructive encounter,
J.Hodgson-J.Hjartarson, Moscow 1987.

Both sides strive for the advantage The bishops are getting on to p
Conclusion
White has scored 51% from over 2,000 games with 3 Bh4, but crucially that score
falls to a rather disappointing 44% from 440-odd games after the critical line 3 ... c5
4 f3 g5 5 fxe4 gxh4 6 e3 Bh6. Then 7 Kf2 might be ugly, but is best I dare say, al
though the more speculative-minded reader might try one of the Bellon Lopez
endorsed alternatives for shock value.
Illustrative Games
Game 20
0 G.Arzumanian I.Brikov
Tula 2007
1 d4 Nf6 2 Bg5 Ne4 3 Bh4 g5 4 f3 gxh4 5 fxe4 c5 6 e3 Bh6 7 Kf2 cxd4 8 exd4 Qb6 9
159
NC3 e6 10 Nf3 Nc6 11 Bb5 d6 (Diagram 31)

Black tries to develop Trying to close the centre
12 Bxc6+1
Best in my opinion; White swaps his second bishop for a useful grip on the posi
tion. He has also delayed the exchange, but 12 Qd3 Bd7 (12 ... 0-0!? 13 Bxc6 bxc6 14
Na4 QaS is a decent alternative when it's not so clear that White's early queen de
ployment helps him) 13 a4 a6 14 Bxc6 Bxc6 15 aS Qc7 16 Rhfl 0-0 1 7 Nxh4 Qe7
gave Black enough for his pawn with ... fS high on the agenda in I.Rausis
A.Tzermiadianos, Athens 1 993.
12 ...bxc6 13 Na4 Qa5 14 b3 h31?
Black desperadoes his sickly pawn. He preferred 14 ... 0-0 in A. Villar Nieves
S.Gonzalez de Ia Torre, Ayamonte 2002, which should have been met by 15 eS!?
dxe5 1 6 NxeS when the knights control some good squares and 16 .. .f6 1 7 Qg4+!
Kh8 18 Nf3 leaves White for slightly choice.
15 gxh31?
A decent case might also be made for 15 g3, but opening the g-file is certainly in
keeping with the creative nature of 3 Bh4!
15 ...Qh5 16 e51 (Diagram 32)
C) TIP: Fixing the centre will establish clear posts for White's knights, at
l/ the same time as restricting Black's light-squared bishop.

Earlier a legendary blitz player preferred 16 Rg1 Qxh3 17 Qd3 Bb7 18 Rg3 Qh5 19
Rag1, which is also quite logical, but despite some further creativity in the shape
of 19 ...0-0-0 20 Rg4 Qa5 21 Rh4 Bf8 22 Rg5 Qc7 23 Rhh5 h6 24 Rg3 the position re
m ai ned dynamically balanced in G.Chepukaitis-S.Sivokho, St Petersburg 2000.
160
The Po p u l a r 2 . . . N e4
16 ds? J
.. .
Too compliant. Black should have preferred 16 ... Bb7! when 1 7 exd6 (it seems
White must capture; 17 Qe2 0-0-0 only helps Black introduce nasty ideas of ... c5)
17 ... 0-0-0 1 8 Nc5 Rxd6 19 Rg1 Qxh3 20 Rg3 Qh5 21 Qe1 is no more than a little bet
ter for White. The c5-knight might be a monster, but White's king position isn't
ideal and that bishop on h6 is hardly a bad piece.
11 Ncs Rf8 18 Rg1 Qxh3?
Black's last appeared to prepare 18 .. .f6 and this would have been somewhat pref
erable to grabbing the weak pawn. Following 19 exf6 Rxf6 20 Rg8+ Kf7 21 Rg3 Bf4
22 Kg2 Ke7 White's knights are strong and he will obtain promising compensation
for the exchange after either 23 Qe2 Bxg3 24 hxg3 e5 25 Qxe5+ Qxe5 26 Nxe5 or 23
Rg4!? e5 24 Nxe5 Bxg4 25 Nxg4, but at least thus way Black would have obtained
some material for his suffering.
19 Qe2 fs
Closing the centre is not ideal, but White had arranged to meet 19 .. .f6 with 20 Rg3
Qh5 21 exf6 Rxf6 22 Kg1 followed by making good use of the hole on e5.
2 0 Rg2
White is in complete control and so the strong Armenian IM simply improves his
major pieces, preparing to penetrate down the g-file.
2 o as 21 Rag1 Qhs 22 Rg8 (Diagram 33)
...

White has taken control There's no defence
2 2 Rb8?1
...
This doesn't help, but even 22 ... Ke7 23 Qel ! followed by Qc3 would have retained
a very pleasant advantage indeed for White.
161
23 Qe11
White brings his final piece into play and he is now able to improve the queen
with tempo.
23 ... Ra8 24 Qc3 Ra7?
Black was already in serious trouble, but this misses the threat. He had to try
24 ...Qf7 when 25 Rxf8+ Kxf8 26 Ng5 Bxg5 27 Rxg5 Ra7 28 Qh3 would have kept up
the pressure, and 25 R8g2!? Ra7 26 Ng5 Bxg5 27 Rxg5 might be even stronger,
aiming to invade on g7 after Qh3-h6.
25 Nxe61
I dare say this sacrifice might not have occurred to everyone; the knight looked
fantastic on c5. However, White has been thwarted down the g-file for the time
being and the introduction of his queen into the attack quickly proves decisive.
2S ... Bxe6 26 Qxc6+ (Diagram 34) 26 ... Bd7
The attack would have been completely overwhelming too after 26... Kf7 27 Rxf8+
Bxf8 28 Ng5+ Kg7 29 Qxe6 Qxh2+ 30 Kfl.
27 Qb6
The point of White's sacrifice; the queen sidesteps with tempo and now the e
pawn is all set to advance.
27 ... Ra8 28 e6
Now Black cannot avoid losing serious material and already he might have given
up.
28 ... Ke7 29 QcS+I Kxe6 30 Re1+ Kf7 31 QxdS+ Kf6 32 Qxd7 Rae8 33 RxfB+ RxfB 34
Re6 mate (1-0)
162
C h a pter Six
T h e Mod e r n P refe re n ce :
2 N e 4 3 B f4

- I ntrod uction and the Sol id 3 ds
- White Plays 4 e3
- The Critica l 4 e3 cs
- The Blackma r-Diemeresque 4 f3

, '/.
e
:r<' / /, ' I
-
St a rti n g Out: The T rom pows ky Attack
I ntroduction and the Solid 3 d s ...
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 (Diagram 1)

The most popular choice Grab- like development
White's invariable choice these days and a move played in 68% of all the games in
my database with 2 ... Ne4. While 3 h4 and 3 Bh4 remain decent surprise weapons
and might suit the creative club player, they do have their theoretical drawbacks
as we have seen. The retreat to f4 is a more sensible choice, since Black will gener
ally find it much harder to exploit the bishop's position to obtain early counter
play.
3 ... ds
Black usually chooses either the dynamic, dark-square approach 3 ... c5 (played in
53% of all games with 3 Bf4), which we'll consider in our next chapter, or the solid
text (36%), but his lesser alternatives are by no means unknown at club level:
a) 3 ... g5?! is an aggressive attempt to exploit the knight's position, but this weak
ening thrust fails to convince: 4 Bel ! (retreating to a safe square while attacking g5
is best; instead 4 Bg3?! has been worryingly common, but White should not part
with his dark-squared bishop so lightly, although doing so with 4 Be5!? f6 5 e3! h5
6 Be2 g4 7 Bg3 is not actually so bad, E.Kovalev-G.Evtushenko, Novosibirsk 2007)
4 ... h6 (White obtained an edge too after 4 ... g4 5 Nd2 Nxd2 6 Bxd2 Bg7 7 c3 d6 8 e4
e5 9 dxe5! Bxe5 10 Bc4 Qf6 1 1 Ne2 Nd7 1 2 Be3 in M.Vasilev-M.Haas, German
League 2005) 5 f3 Nf6 6 e4 (Diagram 2) 6 ... d6 7 Nc3 e5 8 Nge2 Bg7 9 dxe5! dxe5 10
Qxd8+ Kxd8 1 1 Ng3 Nc6 12 Be3 leaves White with a rather pleasant advantage,
G.Kasparov-Y.Dembo, Tel Aviv (simul) 1994.
1 64
The Modern Prefe rence: 2 . . Ne4 3 Bf4
.
TIP: Never be afraid to retreat your bishop back to c1 after 3 Bf4. It's
often the safest approach when the bishop is attacked early on and
is a decent way of defending against an early attack on b2 after ...cs
and ...Qb6, as we'll see in our next chapter.
b) White hasn't scored particularly well after 3 ... Nc6?!, but the critical line 4 f3!
(th is isn't actually the only route to the advantage; the simple 4 Nd2 should also
p romise an edge) 4 ...e5 5 dxe5 g5 (this aggressive thrust is the point of Black's
p lay) 6 Be3 (once again a good case can be made too for tucking the bishop out of
harm's way with 6 Bel !? Nc5 7 e4, as recommended by De Ia Villa) 6... Nc5 7 Qd2
Ne6 8 Bf2 Bg7 9 Nc3 (Diagram 3) 9 ... Bxe5 10 e4 looks quite pleasant for White. One
Trompowsky expert showed the further way forwards in W.Gerstner-S.Giemsa,
German League 1989: 10 ... d6 1 1 Bc4 Bd7 12 Nge2 Rb8 13 Be3 and with the king
side situation under control, the greater trumps belonged to White.

Black may well regret ... gs White takes over the centre
c) 3 ... e6 has the point 4 f3 (not essential; with Black committed to ... e6, White
might decide that this is a good time for 4 Nd2) 4 ... Bd6 5 Bxd6 Nxd6, but in prac
tice White has scored pretty well with his strong centre after 6 e4 (Diagram 4) 6 . . f5 .
(Bl ack often strikes out thus, but perhaps he does better with 6... Nc6!? when 7 Qd2
0-0 8 Nc3 Rb8 9 0-0-0 b5 10 d5 Ne5 1 1 Qd4 Qf6 was hardly clear in L.Winants
D.Hausrath, Belgian League 2007, and so White might wish to investigate 9 Nh3!?)
7 e5 (a decent case can also be made for retaining the tension with 7 Nc3!?) 7 ... Nf7
8 f4. Black's problem is that his knight on f7 isn't all that well placed and 8 ... 0-0 (or
8 ... c5 9 dxc5! Qa5+ 10 Qd2 Qxc5 11 Nc3 g5!? 12 g3 b5 13 Bg2 Nc6 14 Qf2, maintain
i n g control and an edge, M.Karttunen-E.Hintikka, Jyvaskyla 2006) 9 Nc3 d6 10 Nf3
cS 1 1 dxcS! (a deep and correct decision; White exchanges his e-pawn in return for
maintaining a grip) 1 1 ...dxe5 1 2 Qxd8 Rxd8 1 3 NxeS Nxe5 14 fxe5 Nc6 15 Bc4
165
Nxe5 16 Bb3 saw him fail to equalize in M.Karttunen-R.Nevanlinna, Finnish

League 2007; not only are there ideas of 17 0-0 and 18 Rfe1, but Nb5-d6 may cause
problems.
d) 3 ... d6!? (Diagram 5) is the best of this bunch, especially if Black has some famili
arity with either the Pirc or Philidor. Moreover, it's been the sometime defence of
Trompowsky practitioner Luke McShane. At this juncture White faces his stan
dard choice:
Diagram 5 (W} Diagram 6 (W}

Simple, but not so bad Pirc-like play
d1) 4 Nd2 is the more solid approach when the issue is whether 4 ... Nf6 (4 ... Bf5 has
been preferred by McShane, but 5 Nxe4 Bxe4 6 f3 Bg6 7 e4 e6 8 h4! h6 9 h5 Bh7 1 0
Qd2 Nd7 1 1 0-0-0 Be7 12 Kb1 gave White a pleasant spatial plus i n M.Adams
A.Kogan, online blitz 2000) 5 e4 g6 6 Ngf3 Bg7 7 Be2 gives White an improved
version of the Classical Pirc. I'm not so sure that it does after 7... Nc6 (Diagram 6),
which may explain why Hodgson has preferred 6 e5!? Nd5 7 Bg3, which is a bit
more aggressive and 7 ... Bg7 8 Ngf3 0-0 9 Bc4 Nb6 10 Bb3 Nc6 11 0-0 was perhaps
slightly better for White in J.Hodgson-G.Vescovi, Bermuda 1 997.
d2) 4 f3 Nf6 5 e4 leaves White better placed for a Pirc, especially if he is familiar
with the aggressive Be3 and Qd2 variations. Black doesn't have to fianchetto,
though, and 5 ... Nbd7! (or 5 ...e5 6 dxe5 Nh5 7 Be3 dxe5 8 Qxd8+ Kxd8, reaching the
type of queenless middlegame in which it is possible for a weaker White to be
outplayed, but 9 Nd2! Be6 10 Bc4 Bd6 1 1 Bxe6 fxe6 12 Nh3! Ke7 13 g3 Nc6 14 c3 aS
15 a4 h6 16 Nf2 Nf6 1 7 Ke2 Nd7 18 Nd3 saw White retain the upper hand with
some model play in J.Rowson-S.Gordon, British Championship, Douglas 2005) 6
Nc3 (the Samisch-like 6 Be3 e5 7 d5 Be7 8 c4 allows 8 ... Nh5 9 Nc3 Bg5 10 Bf2 g6
with sufficient counterplay for Black, R.Burnett-A.Ivanov, US Championship, Se
attle 2003) 6 ...e5 7 Be3 Be7 doesn't look like too bad a version of the Philidor for
166
The Mod e r n Preference: 2 . . . N e4 3 Bf4
him . With the aggressive 8 Qd2 (Diagram 7) 8 ...0-0 9 0-0-0 a6 10 g4! bS 1 1 gS Ne8?!
12 dxeS NxeS 13 f4 White seized the upper hand in V.Laznicka-l.Brener, online
b litz 2006, but 8 ... c6 9 0-0-0 0-0 10 g4 bS 1 1 g5 NhS looks like a more critical re
sponse.
We now return to 3 ... d5 (Diagram 8):

Aggression from White Logical play from Black
NOTE: There can't be too much wrong with this classical move, pri
marily aimed at preventing White from occupying the centre with 4
f3 and 5 e4. Solid 3 ... d5 might be, but some of the resulting lines are
among the most intriguing and complex in the whole of the Trom
powsky.
4 Nd2
At this point 4 e3, which we will cover next, just edges out the more daring 4 f3,
the subject of the final section of this chapter, in the popularity stakes. Lagging
some way behind is the text, which might appeal to a systems-orientated player,
but is too solid to really fight for the advantage.
4...csl? (Diagram 9)
This vigorous move has the support of both Gallagher and Wells, but it has been
much less common than both:
a) 4 ... Nxd2 5 Qxd2 BfS 6 e3 e6 continues the solid theme and the exchange of a
pair of knights should help Black in this London-type position: for example, 7 Bd3
(7 Nf3 might well be met by 7 ...Bd6, but with 7... Nd7!? 8 c4 dxc4! 9 Bxc4 Nb6 10
B d3 NdS! Black showed a good grasp of light-square strategy to equalize in
V.Salov-P.Van der Sterren, Biel Interzonal 1993) 7... Bxd3 (or 7... Bg6) 8 Qxd3 cS!?
167
(Diagram 10) 9 Bxb8 (9 dxc5 Na6 10 c6 Qa5+ 1 1 c3 bxc6 12 Nf3 Nc5 should be okay
too for Black) 9... Rxb8 10 Nf3 Bd6 1 1 dxc5 Qa5+ 12 c3 Qxc5 13 Rd1 and White
could find nothing better than to offer a draw in V.Salov-J.Polgar, Wijk aan Zee
1 998.
Diagram 9 {W) Diagram 10 {W)

Black fights for the initiative Further exchanges beckon
b) 4 ...Bf5 has often transposed to our last variation after 5 e3 Nxd2 6 Qxd2. White
has also tried both 5 Nf3 and 5 Nxe4 Bxe4 6 e3, but neither should trouble an ex
perienced opponent: for example, after the latter 6 ...e6 7 f3 Bg6 8 h4!? h6 9 h5 Bh7
1 0 Bd3 Bxd3 1 1 Qxd3 has been tried in a few games by Miladinovic. This is proba
bly White's best try, at least netting him some useful space, although I doubt that
Black has too much to complain about if he continues l l...Bd6 12 Ne2 Nd7!?.
5 Nxe4
Not the only try, but Black is fine too after 5 e3 Qb6 6 dxc5 Nxc5 7 Nb3 e6 8 Nf3
Nc6 9 Be2 Be7 10 0-0 0-0 1 1 c3 Ne4, V.Milov-B.Avrukh, Israeli Team Champion
ship 2000.
s ...dxe4 6 dxcs
Davies prefers 6 e3!? Qb6 7 Rb1, but I find it hard to believe that Black's free de
velopment is any less important than the dead-point on e4. Admittedly 7... Bf5 8 c3
Nc6 9 Qb3! Qxb3 1 0 axb3 cxd4 1 1 exd4 e3?! 12 Ra1 exf2+ 13 Kxf2 e5!? 14 dxe5 Bc5+
15 Kg3! didn't give Black quite enough for his pawn in A.Walton-R.Palliser, Brit
ish Championship, Scarborough 2004, but it's not so hard to improve on this se
quence, such as with 8 ... cxd4!? when 9 Qxd4 Nc6 10 Qxb6 axb6 prepares to side
line the bishop with ...e5 and looks about equal after 1 1 Bc7 Rxa2 12 Bc4 Ra4! 13 b3
Ra8 14 Ne2 e6 15 0-0 Be7 16 Ra1 0-0 17 Bxb6 Ne5.
6 .Qa5+
..
168
The Modern P refe re nce: 2 ... N e4 3 Bf4
A t this point White has tried a couple of moves, but without really coming close to
fi ghting for the advantage:
a) 7 c3 QxcS 8 Qa4+ (uninspiring, but 8 Bxb8 Rxb8 9 Qa4+?! bS 10 Qxe4 b4 gives
Bla ck excellent play for the pawn, as noted by Gallagher, while 8 Qd4?! Qxd4 9
cxd4 Nc6! 10 0-0-0 eS! already left White on the back foot in V.Mikhalevski
Ma.Tseitlin, Beersheba 1996) 8 ... Qc6 9 Qxc6+ Nxc6 10 e3 eS 1 1 Bg3 Be6 12 f3 exf3 13
Nxf3 f6 (Diagram 11) was rather comfortable for Black in G.Barbero-Y.Pelletier,
Swiss League 1998.

Easy equality for Black White prepares 5 Bd3
b) 7 Qd2 QxcS 8 Bxb8!? Rxb8 9 0-0-0 was at least a creative try in L.Keely
R.Palliser, Bradford 2008, although here 9 ... Bf5 looks like a decent alternative to
the game's 9 ... Be6!? 10 e3 g6 1 1 Qd4 Qxd4 12 Rxd4 a6, sacrificing a pawn for
queenside pressure and the initiative.
Concl usion
The critical lines of the 3 Bf4 variation are still to come, but it's useful for White to
have an inkling of what to do should Black reject both 3 ... d5 and 3 ... c5. Moreover,
3 d6 isn't such a bad approach for a black player keen to avoid theory.
...
The main subject of this chapter is, of course, 3 ... d5 when I strongly believe that
W hite should adopt either 4 e3 or 4 f3. There's nothing too wrong with 4 Nd2
When White has made 54% from over 650 games, but I'm at a loss to suggest a
good path for him after the surprisingly rare 4 ... c5 and Black's more solid defences
ap pear fine too.
169
Sta rti ng O ut: The Trom pows ky Atta ck
White Plays 4 e3
1 d4 Nf6 2 BgS Ne4 3 Bf4 dS 4 e3 (Diagram 12)
A sensible move and likely a better choice than 4 Nd2 for even the more position
ally-minded player. 4 e3 appeals to aggressive types too for the resulting play can
become quite complex if White so desires, although such players may also wish to
consider the initially-sharper 4 f3.
4... Bf5
A natural move likely to appeal to the more classically-orientated of opponents,
but we must also examine:
a) 4 ... c5 is Black's sharpest approach and by a small margin his most popular. We
will return to this active defence in our next section.
b) 4 ...g6 5 Bd3 Bg7 (Black has also avoided the exchange, but 5 ... Nf6 6 Nd2 Bg7 7
Ngf3 0-0 8 0-0 left White a handy couple of tempi up on a fairly solid line of the
London System in S.Ionov-A.Ramaswamy, Hyderabad 2000, while the more excit
ing 5 ... Nd6 6 h4!? Bg7 7 Nc3 c6 8 h5 Qb6 9 BeS! gave White an early initiative in
S.Lputian-M.Ashley, Las Vegas 2001) 6 Bxe4!? (critical, although here this unbal
ancing exchange isn't the only approach: the Schlechter Slav-like 6 Nd2 Nxd2 7
Qxd2 Nd7 8 Nf3 0-0 9 0-0 c6 1 0 h3 Nf6 1 1 c4 left White with a pleasant edge too in
M.Hebden-N. Davies, British League 1999) 6... dxe4 7 Nc3 cS!? (Black counters dy
namically in a bid to offset his inferior structure) 8 Nge2 cxd4 9 exd4 fS 10 Qd2
(Diagram 13)

A usefu l lead in development A little risky for Black
1 0 . .h6 11 h4 (with White so far ahead in development, I also wonder about the
.
more radical 1 1 0-0-0!? g5 1 2 Bxb8 Rxb8 1 3 h4 e5 14 f3, trying to blow Black away)
170
The Modern Prefe rence: 2 .. Ne4 3 Bf4
.
J l ... Bd7?! (1 l ...Nc6 1 2 0-0-0 eS! 1 3 dxeS Qxd2+ 14 Rxd2 Nxe5 15 Rd5 Nc6 would
h ave been a better idea) 12 0-0-0 Nc6 13 f3! exf3 14 gxf3 Qa5 15 Qe3 left White
50 mewhat for choice thanks to his much safer king in M.Cebalo-M.Lujan, Rijeka
2008.
NOTE: Black is rarely worse after an exchange on e4 due to White's

resulting queenside majority. Rather the problem is that White can
often attack that pawn on e4, thereby forcing a concession, such as
...fs, which can give a line-opening f3-break greater punch.
c) 4 ... c6 5 Bd3! (once again the correct approach; 5 f3 Nf6 not only forces White to
consider ... NhS ideas, but also leaves his kingside a little tangled) 5 ...Qb6!? (greedy
if critical; Black is advised to avoid the trap 5 ... Bf5? 6 f3 Nd6? 7 Bxd6 Bxd3 8 Bxb8
when a piece goes west, while 5 ... Nf6 6 c4 gives White an improved version of the
Sla v with both bishops already in play) 6 Bxe4! (Black can hardly hang on to his
pawn after this and so 6 Bxe4 can be used by even those of a solid persuasion;
they might naturally gravitate towards 6 Qcl, but then 6 ... Qa5+!? 7 c3 Bf5 8 f3 Nd6
9 Bxf5 Nxf5 1 0 e4 Nh4 11 Bg3 Ng6 12 Ne2 e6 left Black pretty solidly placed in
M.Adams-P.Leko, Groningen 1995) 6 ... Qxb2 7 Nd2 dxe4 (Diagram 14) 8 Ne2!? (the
universal choice and probably a better one than 8 Nxe4 Bf5 9 f3) 8 ... Nd7 (admit
tedly Black might cling on to the pawn, but the positionally-suspect 8 ... f5?! 9 0-0
Nd7 shouldn't worry White, who has a pleasant choice between 10 Nc4 Qb4 1 1
Ne5 Nxe5 12 Bxe5, 1 0 f3 and even 1 0 a4) 9 0-0!? (White remains relaxed about re
gaining the pawn, although by this point a perfectly good case can be made for 9
Nxe4) 9 ... Nf6 10 c4! Qa3 1 1 Ng3 g6 12 Be5! Bg7 1 3 Ngxe4 gave White a pleasant
edge in J.Hodgson-J.Hjartarson, Copenhagen 1 996; he controls the centre and may
well obtain good pressure down the b-file.
d) The solid 4 ... e6 is much more common than variations 'b' and 'c', and after 5
Bd3 Bd6 (5 ... Nf6 6 Nd2 leaves White a tempo ahead of the London System and he
should play to occupy e5, but Yermolinsky's 5 ... b6!? is more creative when White
should choose between 6 Bxe4 dxe4 7 Nbc3 and 6 Ne2 Bb7 7 c4!? dxc4 8 Qa4+ Nd7
9 Qxc4 Nd6 10 Qc2 Bxg2 11 Rg1 Bb7 12 Nbc3, which gave him good compensation
in P .Wells-A.Greet, British Championship, Edinburgh 2003) White has tried a
n umber of approaches:
d l) The simple 6 Bxe4 Bxf4 7 exf4 dxe4 8 Nc3 Nc6 9 Nge2 f5 (Diagram 15) is pretty
solid for Black, although he is a little worse due to his slightly passive bishop after
1 0 0-0 0-0 1 1 f3, J.Sisask-K.Nicholas, Hichin 1999.
d 2) Less effective is 6 Bxd6 Qxd6, after which 7 Bxe4 dxe4 8 Nc3 0-0 9 Qh5 f5 10
0-0-0 Nd7 1 1 f3?! Nb6! gave Black promising queenside play in V.Arbakov
A.Sokolov, Bern 1994.
? 3) 6 Ne2 (Diagram 16) keeps White's options open, and after 6 ... b6!? (I think this
Is best; otherwise 6 .. Nd7 7 Bxe4 dxe4 8 Nd2 f5 9 Nc4 gives White a small p ul l and
.
171
6 ... Bxf4 7 Nxf4 Nd7? 8 Bxe4 dxe4 9 Nc3 already won a pawn in Hoang Thanh
Trang-Z.IIincic, Budapest 2007) 7 c4!? (White would like to play 7 Bxe4 dxe4 8
Nd2, but then the 8 .. Bb7 9 Bxd6 of R.Palliser-R.Keely, Huddersfield 2007, can be
.
met by 9 ... cxd6! 10 Ng3 dS 1 1 0-0 0-0 12 f3 exf3 13 Nxf3 Nd7 with equality) 7 ... Bb7
8 0-0 0-0 9 cxdS exdS 10 Bxd6 Qxd6 1 1 Nbc3 Qe7 12 Qc2 Nd7 13 Nf4 he was a
touch better in J.Gustafsson-Y.Dembo, Greece 2004.

Black's bishop isn't great Black is solid but a touch worse
d4) The Londonesque 6 Bg3!? merits further attention. White hopes that Black will
open the h-file and 6 ... f5 7 Ne2 Nd7 8 c4 Ndf6 9 Bxd6 Qxd6 10 cS! Qe7 1 1 f3 NgS
12 Nbc3 supplied a small pull in M.Cebalo-H.Bajramovic, Pula 2004.
Returning to 4 ... Bf5:
5 f3
WARNING: White should not be lulled into a false sense of security

by 4... Bf5. 1t looks logical to prepare Qb3 with 5 c4?, but 5 ... e51
(Diagram 17) followed by 6 ... Bb4+ is a good reason not to.
5 ... Nf6
Less natural is 5 ... Nd6 when 6 Nd2 e6 7 c4! dxc4 8 Nxc4 Nxc4 9 Bxc4 Bb4+ 10 Kf2
0-0 1 1 Ne2 Nd7 12 Qb3 Bd6 13 e4 is quite pleasant for White, M.Wallinger-J.Ryska,
correspondence 2007; White's centre is strong, his h-pawn primed and his king
quite safe on f2.
6 c4! (Diagram 18)
Opening extra lines thus has superseded the older 6 g4 Bg6 7 h4 hS 8 gS, largely
because Nunn's 8 ... Nfd7 9 Nc3 c6 10 Bd3 Bxd3 1 1 Qxd3 g6 1 2 e4 e6 13 0-0-0 Be7
makes it hard for White to get anywhere.
172
The Modern Prefe rence: 2 ... Ne4 3 Bf4
6 e6
...
Play often transposes in the event of 6... c6 7 Nc3 e6, but it's important to be aware
tha t Black can't respond in kind with 6... c5? because of 7 cxdS NxdS 8 Bxb8!:
8 ..Nxe3 9 BbS+ Bd7 10 Bxd7+ Qxd7 11 Qe2 didn't give Black enough for his piece
.
in M.Adams-L.Van Wely, Tilburg 1996, and neither does 8 ... Rxb8 9 e4 Ne3 10
exfS!.

A most nasty centra l strike A m uch better-timed c4
7 Nc3
Those keen to avoid Black's options on his next tum should investigate the imme
diate 7 Qb3!?, although then Black might sacrifice with 7 ... Nbd7!?.
7 c6
...
Solid and common, but Black might play more sharply: 7 ... Nh5!? 8 Bxc7!? Qxc7 9
g4 Bb4 10 gxfS Bxc3+ 1 1 bxc3 exfS 12 Qd3 deserves attention but is perhaps a little
better for White, and the Chigorin-like 7... Nc6 8 g4 Bg6 9 cS! a6 10 h4 hS 1 1 gS Nd7
1 2 Bg3 gave White an edge in Bui Vinh-Z.Ilincic, Budapest 2008.
8 Qb3 (Diagram 19)
White's Queen's Gambit-like play gives him decent chances to secure an edge;
B lack must make a concession in defending b7 and those white kingside pawns
are mobile:
a) The seemingly natural 8 ...Qb6?! runs into 9 cS! Qxb3 10 axb3 Nbd7 1 1 g4!? Bc2
12 b4 a6 13 bS cxbS 14 NxbS and having undoubled his pawns, White held a
plea sant advantage in R.Palliser-M.Bush, Wantage 2003.
b) Thus Black has usually acquiesced to a slight weakening with 8 ... b6, as we'll see
in Game 21.
173
c) Moreover, 8 ... Qd7 further jams up Black's development and 9 g4! Bg6 10 h4 h6
1 1 cS Be7 1 2 NbS! (a little tactic to damage Black's structure) 1 2 ... cxb5 13 Bxb8 a6
14 Bg3 again gave White a pleasant edge in A.Miles-S.Kovacevic, Benasque 1997.
NOTE: It may appear strange to advance White's kingside pawns be

fore developing his pieces there, but their advance gains some very
useful space. Moreover, Black is a little cramped and the f4-bishop
too strong for him to effect a decent central counter.

White plays right across the board Forcing a kingside concession
Conclusion
As we have seen, 3 ... d5 doesn't have to lead to dull positions. Indeed, there are a
number of fascinating and dynamic tries in both this section and the next. Overall,
White has made an impressive 60% from almost 1,200 games with 4 e3, and even
after the solid 4 ... e6 5 Bd3 Bd6 he has a number of ways to obtain a small edge.
Illustrative Ga mes
Game 21
D L.Winants S.Mamedyarov
B led Olym piad 2002
1 d4 Nf6 2 BgS Ne4 3 Bf4 dS 4 e3 BfS 5 f3 Nf6 6 c4 e6 7 Nc3 c6 8 Qb3 b6 9 g4 Bg6 10

h4 (Diagram 20) 10.. h6
.
174
The Modern Preference: 2 . . . N e4 3 Bf4
Bl a ck plays to avoid ceding too much space, but I suspect that blocking with
10 . hS, the choice of the Czech Grandmaster Vlastimil Jansa, is more critical: 11 gS
..
N g8 (Black should avoid 1 1...Nfd7? 12 cxdS exdS 13 NxdS! cxdS 14 QxdS Bb4+ 1S
l(f2 when he loses material, but l l .. dxc4 12 Qxc4!? Nfd7 13 0-0-0 bS 14 Qb3 aS 1S
.
a4 b4 16 Ne4 Nb6 wasn't too clear in M.Archangelsky-H.Ziska, Copenhagen 2004,

a n d so I suspect that this sequence presents a good case for the less-committal 13
N h3!?) 12 Nh3 (more direct is 12 cxdS!? exdS 13 e4! Bd6! when I wonder about 14
N ge2, since 14 Nh3 Ne7! 1 S 0-0-0 0-0 16 Bd3 aS saw Black wisely rejecting the prof
fered central pawns and obtaining decent-enough counterplay in L.McShane
V.Jansa, German League 2003) 12 ... Bd6 13 0-0-0 Ne7 (Diagram 21) 14 e4 (Mik
halevski's suggestion of 14 Bxd6!? Qxd6 1S Nf4 is a simpler approach) 14 ...0-0 1S
Bd3?! (White deviates from 1S Kb1 aS! 16 cxdS cxdS 1 7 Ka1 Nbc6 18 BbS a4!?,
which gave Black good counterplay in A.Kinsman-V.Jansa, Wrexham 1998, but his
choice only makes matters worse; instead, correct is 1S Bxd6 Qxd6 16 eS Qc7 1 7
Nf4, a s pointed out b y Mikhalevski, when White won't get mated so easily, but
neither should Black be worse with the kingside pretty much blocked) 1S ... dxc4 16
Bxc4 bS! 17 Be2 aS sees Black beginning a strong queenside offensive, R.Palliser
V.Babula, European Club Cup, Rethymnon 2003.

Sensible defence from Black An aggressive set-up
11 o-o-ol? (Diagram 22)

The Belgian Grandmaster switched to this move after earlier preferring 11 Nh3,
Which has tempted a number of other strong players. The text keeps the future of
the g1-knight flexible, but going to h3 appears promising too: 1 1 ...Be7 12 0-0-0
Nbd7 (Black's a-pawn is a handy unit, but one that he shouldn't always hurry to
exploit: 12 ... a5 13 cxdS exdS 14 Bd3 Bxd3 1 5 Rxd3 Nbd7 16 Rddl prepared gS fol
lowed by e4, and left White better in L.Winants-M.Wiedenkeller, European Club
175
Cup, Neum 2000) 13 cxd5! (a well-timed exchange; White's bishops will rule the
roost after 13 ... cxd5? 14 Ba6!) 13 ... Nxd5 14 e4 Nxc3 15 Qxc3 Bxh4 16 Ba6 0-0 1 7
Qxc6 left White's diagonal-moving pieces i n charge and Black no doubt wonder
ing why he hadn't stuck with 10 ... h5 in S.Loeffler-V.Babula, Austrian League 2007.
u . . . asl?
Now this becomes justified on tactical grounds. Winants later faced the more solid
1 l . ..Be7 when 12 Bg3 introduced ideas of Nh3-f4 and 12 ... Bd6 13 Bxd6 Qxd6 14
Nh3! Qg3?! (the start of an extremely greedy and ill-judged pawn grab) 15 Be2
Qxh4? 16 Rdg1 ! dxc4 1 7 Qd1 ! Nh7 18 g5 left the errant queen in huge trouble in
L.Winants-S.Mohandesi, French League 2004.
NOTE: 4 e3 might be a fairly safe way to meet 3 ... ds, but that
doesn't mean that White shouldn't still strive for the initiative, even
at the cost of a pawn or two, if the chance presents itself.
12 cxdsl
It was too late for 12 Nh3? on account of 12 ... a4! when it would have been the tum
of Black's bishops to dominate: 13 Nxa4 dxc4 14 Qxb6 Rxa4 15 Qxd8+ Kxd8 16
Bxb8 c3! (Winants) and Black enjoys a dangerous attack.
12 ... a4?1
This asks too much of Black's position, but one can understand why such a crea
tive and energetic player as Mamedyarov chose it, rather than allow White the
initiative after 12 ...exd5 13 e4! dxe4 14 h5 Bh7 15 Bc4. Another solid choice was
12 ... Nxd5!? 13 Nge2 Nd7, but 14 e4 blunts the key bishop and leaves White pleas
antly better.
13 Nxa4 Nxds 14 Nell (Diagram 23)

Black too feels a mbitious What's White's best approach?
176
The Modern Prefe rence: 2 . . . Ne4 3 Bf4
The knight hurries to defend from c3 and now we begin to appreciate why it
d idn't go to h3.
14. Nb4
..
Th is might appear menacing, but White is in time to maintain control.

15 e4 Nxa2+1
The only real try. Otherwise 16 a3 would push Black back and leave him with no
real compensation.
1 6 Qxa2 b5 17 Nec3 bxa4
Black might be behind in development, but his queenside chances appear to be a
cause for concern. However, Black's own monarch is hardly safe either ...
18 d51
... and so Winants strikes!
18 ...a31 (Diagram 24)
Active defence. Instead 18 ... exd5? wouldn't have led to the reopening of the b1-h7
diagonal because 19 NxdS! cxdS? (19 ... Bd6 improves, although White retains a
monstrous initiative after 20 Nc7+!? Qxc7 21 Bxd6 Qb6 22 hS Bh7 23 Bc4) 20 BbS+
Nd7 21 RxdS Ra7 22 Rxd7 Rxd7 23 Rd1 would have been totally crushing.
19 dxc6?
Winants loses his way in the tactical mire. As he later pointed out in Informant 86,
19 NbS! was correct, although he didn't mention the critical line 19 ... Qf6!? (other
wise the tactical attempt 19 ... axb2+ 20 Qxb2 cxdS 21 Nc7+ Qxc7+ 22 Bxc7 Ba3
pretty much loses to 23 BbS+ Ke7 24 exdS, as pointed out by Winants, who also
gives the complex line 19 ... Na6 20 dxe6! Qf6 21 Nd6+ Bxd6 22 Bxd6 Nb4!? 23 Bxb4
axb2+ 24 Qxb2 Ra1 + 25 Kc2 Qxb2+ 26 Kxb2 Rxd1 27 hS! and by this point White's
large advantage is apparent) 20 Nc7+ Kd7 21 dxe6+ Kc8. However, it seems that
White has the answer in 22 Nxa8 Qxf4+ 23 Kb1 Qxf3 24 Bd3 Bxe4 25 Qb3 with
so me advantage as Black's king remains far from safe.
19 ..Qf61
.
Another strong resource and one which appears to give Black the initiative with
White seemingly in some trouble on b2. Instead 19 ... axb2+?? would have been
much, much weaker on account of 20 Kb1 Qxd1+ 21 Nxd1 Rxa2 22 c7 when the c
pawn decides.
20 Rd8+11 {Diagram 25)
Ole ! White understandably doesn't fancy the defensive task after 20 Bd2 Bd6 and
so sacrifices a whole rook to advance his c-pawn with tempo.
20 ..Qxd8 21 c7 axb2+
.
Essential because 21...Qc8? 22 cxb8Q Rxb8 23 Bxb8 Qxb8 would have run into 24
B bS+. Winants also points out that Black finds himself short of a good move after
177
Sta rt i ng Out: T h e Trom powsky Attack
21 ...Qxc7? 22 Bxc7 axb2+ 23 Kxb2 Rxa2+ 24 Nxa2, since 24... Nc6 25 Bb5 Kd7 26 Bg3
creates a killer pin.
22 Kb11?
Best. Instead 22 Kxb2 Rxa2+ 23 Kxa2 Qxc7 24 Bxc7 Bb4! would have enabled Black
to defend. After 25 Kb3 Bxc3 26 Kxc3 Ke7 27 Bc4 his remaining bishop is bad and
White does have the bishop-pair, but there's no way past the solid defences.
22 ... Rxa2 23 Bb5+ Nd7 24 cxdBR+ Kxd8 25 Kxa2 ( Diagram 26) 25 ...f6l

A stunning resource Black must still be accu rate
Further good defence. Instead Winants explains that he was hoping for 25 ...Nc5?
26 Na4! Kc8 (26 ... Be7? 27 Rdl+ Kc8 28 Nxc5 Bxc5 29 Ba6 and mate fully illustrates
the power of the bishop-pair on an open board) 27 h5 Bh7 28 Kxb2 Kb7 29 Rcl
Nxa4+ 30 Bxa4 when White's rook is all set to invade with a serious initiative.
26 Rd1 Be8 27 Na4 Ke71
Unpinning and now Black is in time to untangle.
28 Kxb2 Ne5 29 BxeB KxeB 30 Bxe5 fxe5
These pawns might be weak, but once again Black is very solidly placed and his
bad bishop will hold everything together.
TIP: When in trouble it can be a good idea to head for an ending with
all the remaining pawns on the same side, even at the cost of a bad
structure, poor piece or a whole pawn. That's because it's often
much harder to exploit a material or structural advantage when
there isn't a passed pawn present.
31 Rc1 Kf7 32 Rc7+ Be7 33 Nc5 Rb8+ 34 Kc2 Rb5 3 5 Nd3 Yz-Yz
178
The Mod e r n P refere nce: 2 .. N e4 3 Bf4
.
T h e Critica l 4 e 3 cs
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 d s 4 e 3 csl?
The most active defence and also the most challenging.
5 Bd31 (Diagram 27)

Simple and speedy development Black must sti ll take care
TIP: Once again White does best to waste no time introducing the
possibility of an exchange on e4.
Admittedly 5 f3 Nf6 isn't so bad for White, but his kingside has been weakened
slightly and 6 Nc3 a6!? 7 g4 (a logical-enough try to make something out of the
inclusion of f3) 7...e6 8 Bg2 Nc6 9 Nge2 Be7 10 Qd2 b5 1 1 h4 Nd7! gave Black good
cou nterplay in E.Brondum-A.Butnorius, Hockenheim 2007.
s . . . Nf6
Black sidesteps the positional threat. This has been his most common choice in
practice, although there are a number of important alternatives, which we will
consider in ascending order of popularity:
a) The solid 5 ... e6 reaches a position which might also come about via a 4 ... e6 5
Bd3 c5 move order. Following 6 Bxe4 dxe4 7 Nc3 cxd4 8 Qxd4! (ahead in devel
opment and with the better structure White is happy to trade queens, but he
mi ght also consider 8 exd4!?; V.Juergens-D.Ciuksyte, Plovdiv 2008, continued
8 . .. Bb4 9 Nge2 0-0 10 0-0 f5 and now 1 1 d5! exd5 12 Nxd5 Bd6 13 Qd2 would have
left White with a small but pleasant pull) 8 ...Qxd4 9 exd4 Bb4 10 Nge2 f5 1 1 0-0-0
(Diagra m 28) 1 1 ...0-0 (Wells's suggestion of l l . ..Nc6 12 d5! was tested in
M .Narciso Dublan-J.Fiuvia Poyatos, Badalona 2006, and turned out rather well
179
after 12 ... Ne7? 13 NbS! BaS 14 Nd6+) 12 dS Na6 13 dxe6 Bxe6 14 Bd6 Rfd8 1S Bxb4
Nxb4 16 Nd4 Kf7 Black was okay in H.Nakamura-L.Christiansen, US online
league 2007, but White might well do better with 13 a3!? BcS 14 d6.
b) S ... Nc6?! 6 Bxe4 dxe4 7 dS! (Diagram 29) enables White to advance with tempo
and seize the initiative. Now:

Calling Black's bluff That's some initiative!
b1) 7 ... eS 8 Bg3 Ne7 9 Nc3 sees Black facing ongoing difficulties untangling. The
game K.Berg-D.Navara, Marso 2002, continued 9 ... f6 (9 ... Ng6 10 QhS f6 trans
poses; instead the slightly desperate 10 ... Be7 1 1 BxeS 0-0 12 Bg3 fS 13 Qe2 Bf6 14
Qc4 failed to give Black enough for his pawn in S.Loeffler-M.Perunovic, Euro pean
Championship, Warsaw 200S) 10 QhS+! Ng6 (White wins a pawn after 10 ... g6 1 1
Qh4, because 1 1 ...NxdS? fails to 12 0-0-0 Be6 1 3 Qxe4) 1 1 0-0-0 QaS 12 f4! (wasting
no time tearing open the position) 12 ... exf3 13 Nxf3 Qc7 and now 14 Rhfl ! (Dia
gram 30) would have prepared the vicious idea of 1S BxeS fxeS 16 Nh4.
b2) 7 ... Nb4 8 Nc3 e6 (fairly essential: 8 .. .fS?! 9 a3 Na6 10 f3 opens the position to
the detriment of the black king, and 8 ... BfS?! 9 a3 Na6 10 f3!? is pretty strong too) 9
d6! has attracted some theoretical attention. The resulting play is quite sharp, but
the d-pawn is a most annoying fishbone in Black's threat. Indeed, White has
scored over 90% (!) in the limited practice so far, and we'll see one such crushing
victory in Game 22.
c) S ...Qb6?! (Diagram 31) is quite critical to a certain extent, but Black will pay for
his greed: 6 Bxe4! dxe4 (similar is 6 ... Qxb2 7 Nd2 dxe4 8 Ne2 when Black must be
extremely careful due to his lack of development; the game J.Fiores-J.Barrios Tron
coso, correspondence 1998, continued 8 ...e6?! 9 Nxe4 cxd4?! 10 Nxd4! Nc6 1 1 Rbl
Qxa2 12 NbS QaS+ 13 c3 eS 14 BxeS! NxeS lS QdS and Black was crushed) 7 Nc3
Qxb2 (Black grabs the pawn; he had to avoid the threat of 8 NdS and 7... cxd4 8
180
The Mod e r n P refere nce: 2 . . . N e4 3 Bf4
Qxd4! Qxd4 9 exd4 Bf5 10 Nge2 e6 1 1 0-0-0 with ideas of d5 and Ng3 is hardly ideal
fo r him) 8 Nge2 (White develops; 8 Nd5?! Na6 doesn't yet lead anywhere for him)
s ... Bg4 (as played in the stem game which Black is yet to improve over: 8 ... cxd4 9
Rb 1 ! Qa3 1 0 exd4 Na6 1 1 0-0 Qa5 12 d5! gave White a powerful initiative in
B.Laurent-S.Mohandesi, French League 2007, and so did 8 ... Na6 9 0-0 g6 10 Rb1
Qa3 1 1 d5! Bg7 1 2 d6 in L.Winants-I.Harovelo, French League 2004) 9 Nxe4! (the
best way to sidestep a simplifying exchange on c3) 9...Bxe2 10 Kxe2 Qb5+ (retreat
ing and so avoiding 10 ... cxd4?! 1 1 Qd3! which left White with a monstrous initia
tive in M.Adams-Xie Jun, Hastings 1996/97) 1 1 Qd3 (Diagram 32) 1 1 ...Qxd3+ at
least gets the queens off, but after Wells's 12 cxd3!? cxd4 13 Rab1! b6 14 Rhcl
White retains a powerful initiative and unsurprisingly nobody has been willing to
debate this as Black.

White should gam bit Aga in White has the initiative
NOTE: White can easily obtain the faster and more harmonious de
velopment in these lines. Indeed, he can often seize an early initia
tive with Qxd4 (in response to ...cxd4) and a rapid dS-d6 two ideas
well worth remembering.
d) 5 ... cxd4!? (Diagram 33) keeps Black's options open and is his best alternative to
our main line. Already White finds himself at a crossroads:
d1 ) 6 exd4 offers Black a number of reasonable options, including 6 ... Nc6 7 c3 (I
suspect that White should transpose to variation 'd2' with 7 Bxe4 dxe4 8 Ne2; note
that here 8 d5 is no longer anywhere near so effective on account of 8 ...e5 when 9
Bg3?! QaS+! is a rather awkward check) 7... Bf5 8 Qe2 (Black obtained a comfort
able enough version of the London System too after 8 Nf3 e6 9 Ne5 Qb6 10 Qe2
Be7 1 1 0-0 Nxe5 12 BxeS 0-0 in C.Cruzado Duenas-P.Polakovic, correspondence
181
Start i n g O u t : The Trom powsky Attack
2004) 8 ...e6 9 Nd2 Nd6!? 10 Ngf3 Bxd3 1 1 Qxd3 Nf5 with equality, J.Timman
L.Van Wely, Breda 1998.
d2) Thus 6 Bxe4!? has caught the imagination of the theoreticians, and after
6 ... dxe4 (almost exclusively played, but 6 ... Qa5+!? really deserves more than the
one test it has so far received: 7 Qd2 Qxd2+ 8 Nxd2 dxe3 9 Bxd5 exd2+ 10 Bxd2
Nc6 was prematurely agreed drawn in M.Lomineishvili-B.Jobava, Schwaebisch
Gmuend 2001; here White should either play on with 1 1 Bc3 f6 12 Ne2 e5 13 f4, as
indicated by McShane, or earlier try Wells's bold 7 c3!? dxc3 8 b4!? Qxb4 9 Bc2
Qb2 10 Ne2 Qxa1 1 1 0-0 with a strong initiative in return for the serious invest
ment) White has again investigated 7 Qxd4, but Black appears to have the answer
in 7... Nc6! 8 Qxd8+ Kxd8 9 Nc3 e5: for example, 10 0-0-0+ Ke8 1 1 Bg3 Bf5 12 f3 ex3
13 Nxf3 f6 14 e4 Be6 15 Bf2 Rc8 and Black had no problems whatsoever in
O.Bouverot-T.Hamarat, correspondence 2005. Thus 7 exd4 (Diagram 34) is proba
bly best when we reach a final divide:

A fa irly sensible defence The e4-pawn might become weak
d21) 7... Nc6 8 Ne2 (usual, but an unexplored alternative is the 8 d5!? e5 9 Bg3 Ne7
10 Nc3 Ng6 11 h4! of A.Stefanova-B.Kurajica, Solin 2007) 8 ... Bg4 (8 ...e5?! resembles
a gambit seen in certain lines of the c3 Sicilian, but here 9 dxe5 Qxd1 + 10 Kxd1 Bg4
1 1 Nbc3 0-0-0+ 12 Kcl surely didn't give Black full compensation in P.Wells
Y.Visser, London 2006) 9 Nbc3 e6 (Diagram 35) reaches an unbalanced position I
have been happy to debate with both colours. Chances are roughly equal, al
though 10 h3! (White must be alert: 10 0-0?! g5! 11 Be3 f5 gave Black the initiative
in S.Fowler-R.Palliser, Halifax (rapid) 2004) 10 ... Bh5 (and now 10 ... Bxe2 1 1 Qxe2!
Qxd4 12 0-0 gives White the initiative) 1 1 Nxe4 Bxe2 12 Qxe2 Nxd4 13 Qd3 Nc6 1 4
0-0-0 Qxd3 15 Rxd3 was slightly more pleasant for White in J.Hodgson-P.Wells,
Oxford 1998.
182
The Modern Preference: 2 . N e4 3 Bf4
. .
d22) 7...eS!? is an unexplored gambit, although I suspect that 8 dxeS (in the only
practical test thus far, 8 BxeS Nc6 9 Nc3 QaS 10 Bf4 Ba3! 1 1 Bel Bb4 gave Black
reasonable compensation in A.Stefanova-B.Kurajica, Solin 2006) 8 .. Qxd1+ 9 Kxd1
.
Nc6 10 Nc3 Bg4+ 1 1 Kcl doesn't give Black quite enough for the pawn.

Black tries to avoid .. .fs Does Black never develop?!
d23) 7...Qb6 arguably gives Black an improved version of variation 'c', although I
still quite like White's compensation after 8 Nc3 Qxb2 (if 8 ... Nc6 then one should
avoid 9 dS?! eS!, and prefer 9 Nge2 Bg4 10 0-0 Rd8 when Prie gives 1 1 Be3!? Qxb2
12 Rb1 Qa3 13 Nxe4 as better for White) 9 Nge2 Bg4 10 0-0, since even the ex
changing sequence 1 0...Qxc3 1 1 Nxc3 Bxd1 12 Rfxd1 (Diagram 36) fails to fully
help Black who should probably now return the extra pawn, rather than allow
12 .. .fS 13 Rab1 b6 14 NbS Na6 1S dS!.
We now return to the solid S ... Nf6:
6 C3
An unpretentious reply, but this is one variation in which I think that the systems
based approach is the way to go. In his 2003 work Peter Wells heavily extolled the
vir tues of the vigorous 6 dxcS Nc6 7 BbS (Diagram 37), but Prie's discovery of
7... QaS+! (stronger than the routine 7. . e6 when 8 b4!? aS 9 c3 Nd7?! 10 Qa4! Nxb4?
.
1 1 cxb4 Qf6 12 Nf3 Qxa1 13 0-0 gave White a monstrous initiative for the exchange
in A.Bigg-P.Sowray, British Championship, Scarborough 2004) 8 Nc3 a6 9 Bxc6+
bxc6 10 Qd4 e6! is a strong improvement. With 1 1 b4 (11 Bd6 Bxd6 12 cxd6 Rb8
also gives Black good play) 1 1 ...Qa3 12 Nge2 (White would like to prevent . aS, . .
but he can't: 1 2 Bc7? aS! 13 BxaS RxaS 14 bxaS BxcS leads to disaster) 1 2 ... aS (Dia
gram 38) 13 bS! (13 Rb1 ?! axb4 14 Rxb4 Nd7 regained the pawn with interest in
A.Stefanova-M.Cornette, La Roche-sur-Yon 2007) 13 ... BxcS 14 Qa4 Qb2 1S 0-0 cxb5
16 Nxb5 Bd7 1 7 Rfb1 Bxb5 18 QxbS+ Qxb5 19 Rxb5 White manages to stay out of
183
trouble, but Black is obviously fine after 19 ...Nd7 (Prie).

Relying on the bishops, but... ... Black can seize the initiative
Here White might prefer the less ambitious 7 Nf3, but then 7... e6 (sharper is
7... Bg4!? when Wells's 8 Bb5 Qa5+ 9 Nc3 e6 10 Qd4 Bxf3 1 1 gxf3 Nd7 12 Bxc6 bxc6
gives Black some compensation for the pawn) 8 Nbd2 Bxc5 9 a3 Bd6 10 Bxd6 Qxd6
1 1 c4 0-0 12 Qc2 Bd7 13 Rd1 Rac8 saw Black equalize in A.Stefanova-K.Arakhamia
Grant, Gibraltar 2009.
NOTE: After the prudent 6 c3 White finds himself a tempo up on a

variation of the London System: 1 d4 dS 2 Bf4 Nf6 3 e3 cs 4 c3 Nc6 5
Bd3. This should give him a decent chance of emerging from the
opening with a small edge.
6 ... Nc6
A critical alternative is 6...Qb6!? when, just like in the London proper, I'm not a
fan of 7 Qb3 c4! 8 Qxb6 axb6. Better appears to be 7 Qc2!, after which 7... g6 8 Nf3
Bg7 (Black might be tempted by 8 ... c4 9 Be2 Bf5, but after 10 Qcl Nc6 1 1 h3! White
has preserved his London bishop and can play for the e4-break; Prie gives the
logical line l l .. Bg7 12 Nbd2 0-0 13 0-0 Rfc8 14 Re1 Qa5 15 Bd1! Bd3 16 Bc2 Qb5 1 7
.
Qb1 which does, indeed, appear t o slightly favour White) 9 Nbd2 0-0 10 0-0 Nc6
1 1 h3 Be6 12 Rfe1 (Diagram 39) 12 ... Rac8 13 Qb3! (White's queen needed to get off
the c-file and the text does so with tempo by indirectly attacking b7) 13 ... c4 14
Qxb6 axb6 15 Bc2 b5 16 a3 Bf5 1 7 e4 Nxe4 18 Nxe4 Bxe4 19 Bxe4 dxe4 20 Rxe4
gives White a small but pleasant edge in the late middlegame, E.Prie-E.Najer, Bas
tia 2004.
7 Nd2 (Diagram 40) 7 Bg4
...
184
The Modern Prefe rence: 2 . . . Ne4 3 Bf4
Black takes the opportunity to develop his bishop outside the pawn chain. He
might also try:

It's l ike the London System ! Further London-like play
a) 7 ... e6 8 Ngf3 Be7 9 h3 (White preserves his bishop, but a more ambitious player
might opt for 9 Ne5!? followed by Qf3) 9...Qb6 10 Qc2 (not bad, but I would
p robably prefer 10 Rb1, especially as the queen might want to go to e2 or even f3)
10 ... Bd7 1 1 0-0 Rc8 12 a3 0-0?! 13 dxc5! Qxc5 14 b4 Qb6 15 c4 saw White instruc
tively transforming the structure to obtain a pleasant edge in V.Laznicka-J.Werle,
German League 2008.
b ) 7... cxd4 8 exd4 g6!? 9 Ngf3 Bg7 10 0-0 0-0 1 1 Re1 (11 h3! would enable the
bishop to control e5 by remaining on the h2-b8 diagonal) 1 l ...Nh5 12 Bg5 Qd6 13
Nfl a6 was about equal in M.Lomineishvili-K.Arakhamia Grant, Tbilisi 2005.
TIP: White should take good care to preserve his dark-squared

bishop in these London positions. Thus when Black follows up ...e6;
Ngf3 with ... Be7 White should usually flick in h3 before ... NhS hunts
down the bishop.
8 Ngf3
White continues to develop. In view of our next note he might like to play the
i m mediate 8 Qa4, but as Prie points out, this is premature when Black can still
reply 8 ... Bd7! .
8 e6 (Diagram 41)
...
Sta ndard, but one might ask why not 8 ... Nh5? A very sensible reply is 9 Be5! with
the neat point that 9. . Nxe5?? 10 Nxe5! wins a p iece and 9 ... e6 10 h3 Bxf3 1 1 Qxf3
.
also gives White the upper hand. Possible too is 9 Bg5!? when 9... h6 10 Bh4 gS 1 1
185
Sta rt i n g Out: The T ro m powsky Attack
h3! Bd7 12 Ne5 Nxe5 13 dxe5 Ng7 14 Bg3 Qb6 1 5 Rb1 was hardly clear in
K.Smallbone-M.Comette, Barcelona 2007, although I would prefer to take White
due to his queenside and central options.
9 0-0
In his column for the ChessPublishing website, Eric Pril has drawn attention to 9
Qa4!?, which is probably a more challenging idea:
a) It's worth noting the neat trap 9 ... Bxf3 10 Nxf3 Bd6? 11 Ba6!, which won mate
rial after 1 1 ...Qa5 1 2 Qxa5 Nxa5 1 3 Bxd6 bxa6 14 Bxc5 in S.lonov-E.Relange, Buda
pest 1996.
b) 9 ... Nd7 10 0-0 c4 1 1 Bbl ! Be7 1 2 e4 Bh5 13 Re1 0-0 14 exd5 exd5 15 Bf5 (Diagram
42) saw the thematic break give White an edge in M.Leon Hoyes-R.Pruijssers,
Kemer 2007.

White shou ld break the pin White has the upper hand
c) 9 ... c4!? 1 0 Bc2 Nd7 (Black prevents 1 1 Ne5 and threatens to trap the white
queen; instead 10 ...Bd6 1 1 Bxd6 Qxd6 12 b3! cxb3 13 axb3 0-0 14 0-0 Rfc8 15 Rfcl is
a line given by Prie when b4-b5 should emphasise White's pull) 1 1 Bb1 (White
must acquiesce to this loss of a tempo) 1 l ...Be7 1 2 e4 a6 13 h3 Bh5 14 0-0 0-0 15 Rel
Bg6 was quite solid for Black in E.Prie-M.Schaefer, Arvier 2007, although after the
French Grandmaster's later suggestion of 1 6 Qd1 I would still slightly prefer to
take White.
9 Be7
...
Black is happy to simply complete his development. Instead 9 . Rc8 10 Qa4! Qb6? !
. .
1 1 Ne5 saw White making full use of the pin on the a4-e8 diagonal in C.Cruzado
Duenas-B.Hens, correspondence 2005. Black might, though, prefer to bring his
bishop to d6 and 9 ... cxd4 10 exd4 (I would prefer 10 cxd4! with the point that
186
The Mode r n P refe re nce: 2 . .. N e4 3 Bf4
1 0 ... Bd6 1 1 Qb3 attacks b7 and unpins the knight) 10 ... Bd6 1 1 Bxd6 Qxd6 12 Re1
Bh5 13 Qb3 Qc7 14 Ne5 Bg6 15 Nd3 Bxd3 16 Nxd3 0-0 was extremely solid for
Bl ack and agreed drawn here in B.Predojevic-C.Balogh, European Club Cup, Kal
Ii thea 2008.
10 Qbll?
White has also played 10 Qb3, but it's tempting to prepare Ne5 while keeping up
some pressure against h7.
10 ... BhS
More critical might be 10 ... Nh5!?, although Black is still to demonstrate equality
a fter 1 1 NeS Nxf4 12 exf4.
11 Nes (Diagram 43)

The knight's idea l outpost An awkward, cramping d-pawn
Thus the knight leaps to its favourite location. Black remains quite solidly placed
here, but White has a small pull and 1 l ...Bg6 12 Nxg6 hxg6 13 Nf3 NhS 14 BeS Bd6
15 Bb5! 0-0 16 Bxc6 bxc6 1 7 Qc2 saw him preparing the c4-break while retaining an
edge in P.Wells-E.Sutovsky, Oxford 1998.
Conclusion
This has been a fairly lengthy but important section. Black can counter 4 e3 c5 5
Bct3 in a number of ways according to taste, but I suspect that he does best to ex
plore either 5 ... cxd4 or the more solid 5 ... Nf6. In response to the latter, 6 dxcS is no
longer the riposte it was once thought to be and, indeed, White has scored some
what better with the more restrained, Londonesque 6 c3.
187
Illustrative Games
Game 22
D V.lvanchuk B.Jobava
Havana 2005
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 d5 4 e3 c5 5 Bd3 Nc6 6 Bxe4 dxe4 7 d5 Nb4 8 Nc3 e6 9 d61
(Diagram 44) 9 Nc61?
...
Jobava offers a temporary pawn sacrifice in a bid to advance in the centre. The
immediate 9 ... f5?! has also been seen, but 10 Nh3!? (White preferred the more
straightforward 10 f3 Nc6 1 1 fxe4 in G.Rey-R.Wong, San Francisco 2000, and now
1 l ...e5 12 Bg3 Qxd6 13 Nf3 would have restricted him to a small advantage)
10 ... Nc6 11 0-0 (an idea of the Hungarian Grandmaster, Emil Anka) 1 l ...e5 1 2 NbS
exf4 13 QdS iooks like a lot of fun for White; the d-pawn is extremely awkward.
Black has also tried 9 ...Qa5, but 10 Nge2 Bd7 11 0-0 fS (or 1 l ...Bc6 12 a3 Na6 13
Qd2 0-0-0 14 b4!, which gave White a strong attack in K.Rusev-N.VIashki, Plovdiv
2007) 12 a3 Nc6 (Diagram 45) 13 b4! (hurrying to open the position with Black's
kingside undeveloped and his king position a cause for concern) 13 ... cxb4 14 axb4
Qb6 would have given White a large advantage in J.Hodgson-Ma.Tseitlin, Ischia
1996, had he continued 15 Qd2! Nd8 (neither does Black solve his problems with
15 ... Nxb4 16 Rfb1 aS in view of 17 Nxe4! fxe4 18 Rxb4) 16 Na4 Bxa4 17 Rxa4
(Greenfeld).

White should open further lines A tactical blow
10 Nge21?
Ivanchuk too is more interested in the initiative than material concerns. Neverthe-
188
The Mode r n Prefe rence: 2 . Ne4 3 Bf4
..
Jess, 10 Nxe4 is hardly a bad option whatsoever: 10 .. .fS (widely condemned, but
10 . eS 1 1 BgS Qb6 12 QdS! is very promising for White because, as Speelman
..
points out, 12 ...Be6 meets with 13 NxcS!; Black later preferred 1 1 ...QaS+ 12 c3 Be6
i n I.Stohl-R.Kreisl, Austrian League 2008, but White has a few decent options here,
i n cluding the game's 13 Nf3 f6 14 Bxf6!? gxf6 1S Nxf6+ Kf7 16 Ne4 h6 1 7 Nh4 with
a strong initiative for the piece) 1 1 BgS! (the point of White's play; otherwise 1 1
Nc3?! eS is, of course, very comfortable for Black) 1 1 ...Qb6?! (Black also got into
trouble after 1 1 ...QaS+ 12 Nc3 NeS?! 13 QhS+! g6 14 Qh4 Bxd6 1S 0-0-0 Qb6 16 Nf3
in M.Cebalo-P.Haener, Bern 2006, but 1 1 ...Qd7! improves; then 12 NxcS Qxd6 13
Nb3 Qxd1+ 14 Rxd1 h6 1S Bh4 gS 16 Bg3 eS gives Black decent compensation, and
12 QhS+ Qf7 13 Qxf7+ Kxf7 14 NxcS Bxd6 1S Nb3 Bc7 16 0-0-0 is only a little better
for White according to Ivanchuk) 12 Nf6+! (Diagram 46) 1 2... Kf7 13 Nxh7!? Rxh7
14 d7 saw White retain the initiative and the safer king in J.Speelman-Wang Yao
yao, Beijing 1997.
1o fs?
...
Far too ambitious. Black later improved with 10 ... eS! in D.Sadkowsky
D.Lobzhanidze, Belgian League 2006: 1 1 NbS (just as in our main game, White
must go forwards) 1 1 ...exf4 1 2 Nc7+ Kd7 13 Nxf4! Bxd6! (essential; 13 ...Rb8? 14
Qg4+ Kxd6 1S Qg3 Kd7 16 0-0-0+ Nd4 1 7 NbS would only have given White a
strong, ongoing initiative) 14 Nxa8 Ke7 1 S QhS (Diagram 47) 1S ... g6?! (Ivanchuk's
1 S .. BeS! improves, since 16 QgS+ Ke8 1 7 Qxd8+ Nxd8 18 0-0-0 b6 19 NdS Ne6
.
keeps the knight trapped) 16 Qh6 Be6 17 0-0-0 Qg8?? (badly underestimating
White's initiative, although it's not so easy for Black to regain the piece: 17 ...Qxa8
18 Qh4+ Kd7 19 Nxe6 fxe6 20 Rd2 Kc7 21 Rhd1 is hardly a panacea for him) 1 8
Rxd6! Kxd6 19 Rd1+ Nd4 20 exd4 and White won.

It's not totally clear Ripping open the kingside in style
189
Sta rt i n g Out: The T ro m pows ky Attack
Instead of moving a pawn, Ivanchuk himself offers the line 10 ... Bd7!? 1 1 Nxe4 fS
12 BgS Qb6 13 N4c3 Ne5 14 Be7 Bc6, which he feels is unclear, but 13 Nf4! im
proves: 13 ... Ne5 (Black must cover the short diagonal towards his king) 14 QhS+
Nf7 15 Nf6+ gxf6 16 Bxf6 Rg8 1 7 Qxh7 Rg4 18 0-0-0 leaves White with three pawns
for the piece and Black still badly tied up.
11 NbSI
Black's last rather asked for this. Now Jobava must move his king and already he
is in serious trouble.
11 ...Kf7 12 Nc7 RbB 13 g411 (Diagram 48)
The star move of the game. lvanchuk forces the kingside open, thereby enabling
him to go after the black king with his queen and knights.
13 ...fxg4
A little meek, but in response to 13 ... g5 Ivanchuk had prepared 14 gxfS! gxf4 15
Nxf4 (Diagram 49) when his attack would have been overwhelming, as shown by
the line 15 ... Bxd6 16 QhS+ Kf6 1 7 Qh6+! Kxf5 18 Ncxe6.
Diagram 49 (B) Diagram so (B)

There's no satisfactory defence A triumph for the knights
Indeed, I suspect that Black may already be lost by this stage because he can't
keep the kingside closed. The attempt to do so is 13 ... g6!? 14 gxfS gxf5, but here
Ivanchuk had planned another amazing blow: 15 Nd4! and after 15 ... h5 White can
even continue in style with 16 Nxf5!.
14 Ng3 Nb4
Beginning a slow manoeuvre, but I'm not sure what Black should have tried. The
alternative was 14 ... g5!?, but after 15 Nxe4! gxf4 16 Qxg4 Rg8 17 Qxf4+ Kg7 18
Rg1+ Kh8 19 Rxg8+ Kxg8 20 Ne8! White's attack would have been overwhelming.
15 a3
190
The Mod e rn P refere nce: 2 . . N e4 3 Bf4
.
As he later indicated in his notes for Informant 93, lvanchuk might well have taken
the pawn immediately: 15 Nxe4!? e5 1 6 BxeS BfS 17 0-0! and White enjoys a near
decisive advantage because Black won't survive after 17 ... Bxe4 1 8 Qxg4.
1s ... NdS 16 Nxe4 Nxf4 17 exf4 (Diagram 50)
NOTE: The Trompowsky is sometimes seen as a struggle between

knights and bishops. That's a bit of a generalization, but being a fan
of the jumping creatures does help one to play the opening well.
Here White's knights completely dominate Black's still-to-move bishops and the
end is nigh:
17 ... Kg8 18 Qxg41
Ivanchuk has made excellent use of the initiative throughout and now the threat
of 19 Ne8 forces a further concession.
18 ... h5 19 Qg3 b5 20 Rg1 Rh6 21 o-o-o KhB 22 Nxc5 b4 23 axb4 Rxb4 24 d7 Rc4 25
N5xe6 1-0
Brutal play and a real Trompowsky classic.
The Blackma r-Diemeresq ue 4 f3

1 d4 Nf6 2 Bg5 Ne4 3 Bf4 d5 4 f3!? (Diagram 51)

A rapid e4-advance is the plan French-like play beckons
A much sharper approach than 4 e3 or 4 Nd2. White wastes no time driving back
the knight and prepares to occupy the centre with e4.
4 . Nf6
..
19 1
Standard, although the slightly strange-looking 4 ... Nd6 has been seen on occasion.
After 5 Nc3 c6 6 e4 e6 Black is ready for 7 e5 Nf5 with an acceptable version of the
French. He might also meet an early Qd2 with ... Nc4, but 7 Bd3 Be7 8 Nge2 b6 9
0-0 Ba6 10 Ng3 Bxd3 11 Qxd3 0-0 12 e5 saw White timing his central advance well
to obtain an edge in N.Vitiugov-R.Wojtaszek, Warsaw (rapid) 2008.
5 e41?
The Blackmar-Diemer approach. White doesn't have to sacrifice a pawn to further
his development, though, and the older 5 Nc3 is the alternative. This approach has
been rather abandoned by theory and is fine for Black, but might make a decent
surprise weapon:
a) Black usually plays along French Defence lines with 5 ... e6 (Diagram 52) when 6
e4 leads to a further divide:
a1) Beloved of theory is the reference 6... Be7!? 7 Qd2 0-0 8 0-0-0 a6 9 exd5?! exd5 10
g4 b5! 1 1 h4 b4 12 Nce2 a5 13 Ng3 a4, which gave Black excellent counterplay in
L.Cooper-E.Gufeld, Hastings 1992, but this is all far from forced. While I suspect
that 7 e5 Nfd7 8 Be3 c5 gives Black an improved version of the Classical French,
White can improve on Cooper's play with either the less-committal 8 Nge2 or later
9 g4! b5 10 g5 Ne8 1 1 h4 b4, as White actually did in S.Lputian-E.Gufeld, Los An
geles 1998, and now 12 Nce2 a5 13 Ng3 a4 14 h5 Nc6 15 Bb5 would have been
pretty unclear.

An unusual type of Winawer An unbalanced struggle
a2) 6 ... Bb4 7 e5 (those who have read Chapter Three might not be surprised to
learn that I am quite a fan of keeping the tension with 7 Qd3!?: following 7... c5 8
dxc5 0-0 9 0-0-0! Qa5 10 exd5 exd5 1 1 Nge2 Nc6?! 12 Qb5! White was already
somewhat better in V.Vehi Bach-R.Douven, Groningen 1994, and here the 8 ... Nc6
9 0-0-0 d4 10 NbS e5 of B.Steiner-G.Danner, Vienna 2006, can be met well enough
192
The Mod e r n P reference: 2 . . . Ne4 3 Bf4
by 1 1 c3! BxcS 12 cxd4 Bxd4 13 Nxd4 exd4 14 Qa3 Be6 1S Kb1; perhaps Black
should just prefer 7...0-0) 7 ... Nfd7 8 a3 Bxc3+ (playing along Winawer lines; Black
h as also taken play back into more Classical waters with 8 ... Be7 when 9 Be3 cS l O
f4 cxd4 1 1 Qxd4 Nc6 1 2 Qd2 was perhaps a touch better for White in M.Adams
J.Smirin, New York 1994) 9 bxc3 cS (Diagram 53) leads to a complex manoeuvring
struggle: for example, 10 Be3 Nc6 1 1 f4 QaS 12 Qd2 c4!? 13 g3 Ne7 14 Bh3 Nb6 1S
Ne2 Bd7 1 6 0-0 0-0 1 7 g4 fS! and by this point Black had demonstrated that he en
j oyed a decent version of the French in Xia Yu-Zhang Jundi, Suzhou 200 1 .
b) Less effective i s S. . .BfS when 6 e4 dxe4 transposes to note 'a' to White's 6th
move in our main line, and there's also 6 g4!? Bg6 7 h4 h6 8 e3 e6 9 Bd3 Bxd3 1 0
Qxd3 Bd6 1 1 Nge2 Nc6 12 0-0-0 with a n edge, S.Drazic-R.Bulajic, Zlatibor 2007.
c) A more active approach is S ... cS!? when 6 e4 (it might seem a little strange to
combine f3 with 6 e3, but White sometimes has and here 6 ...e6 7 NbS Na6 8 g4!?
Bd7 9 a4 quickly led to a rather unclear struggle in V.Kunin-S.Feller, Le Touquet
2007) 6... cxd4 (far from forced: 6... dxe4 takes play into note 'b' to White's 6th move
in our main line, and 6 ... e6 7 NbS Na6 8 eS Nd7 9 c3 Be7 once again appears to
give Black a decent-enough version of the French) 7 Qxd4 Nc6 8 BbS Bd7 9 Bxc6
Bxc6 (Diagram 54) 10 exdS (White correctly keeps the centre fluid: 10 eS?! Nd7 1 1
0-0-0 e6 12 Be3?! Qc7 13 f4 BcS 14 Qd2 d4! 1S Bxd4 Bxd4 16 Qxd4 Bxg2 was al
ready something of a disaster for him in H.Hoeksema-A.Beliavsky, Tilburg 1992)
lO ... NxdS 11 0-0-0 e6 12 Nge2 Nxf4 13 Qxf4 QaS 14 Nd4 Be7 was about equal in
P.Cech-R.Chytilek, Bmo 2006.
We now return to the immediate S e4:
5 dxe4
...
There's nothing better. No longer can Black obtain a comfortable version of the
French with S ...e6 because of 6 eS Nfd7 7 Be3! cS 8 c3 (Diagram 55).

White frees his f-pawn Black doesn't have to accept
193
NOTE: By retreating his bishop to e3 White prepares the useful f4-

advance. Here Black struggles to undermine the pawn chain and
White enjoys a fairly pleasant version of the French Tarrasch.
The game A.Bigg-G.Pinter, British League 2006, continued 8 ... Nc6 (Black has also
manoeuvred with 8 ...b6 9 f4 Ba6, but 10 Bxa6 Nxa6 1 1 Nf3 Be7 12 0-0 0-0 13 fS! left
him worse in J.Hodgson-J.Benjamin, Las Vegas 1995) 9 f4 Qb6 (Black preferred
9 ... cxd4!? 10 cxd4 Bb4+ in F.Felecan-K.Kuderinov, Chicago 2008, and now the
positionally double-edged 1 1 Nc3!? 0-0 12 Nf3 Nb6 13 Bd3 Nc4 14 Bxc4 dxc4 15 0-0
Ne7 is a reasonable alternative to the game's 1 1 Nd2 Nb6 12 Ngf3 Nc4 13 Bxc4!
dxc4 14 0-0 Qd5 15 Qc2) 10 Qd2 aS 1 1 Nf3 a4 12 Be2 Qa5 13 0-0 with a pull; the
queen' s knight can come to a3 as well as to d2.
6 Nc3 (Diagram 56) 6 ...exf3
Critical, but in practice Black has often declined the gambit:
a) 6... Bf5 defends the pawn, but enables White to develop with tempo which is
quite useful: 7 fxe4 (De Ia Villa's 7 Bc4!?, intending 7 ... e6 8 fxe4 Nxe4 9 Nxe4 Bxe4
10 Qe2, remains untested, but does look like a good alternative) 7... Nxe4 (after
7... Bxe4 8 Nxe4 Nxe4 9 Qf3!? Qxd4 10 Rd1 Qa4 1 1 Bd3 Nc5 12 Ne2 Nbd7 13 0-0
Nxd3 14 Rxd3 White enjoyed some initiative for the two pawns in J.Hodgson
A.Kalka, German League 1994, but he might do even better with the simpler 9 c3,
as later mentioned by Hodgson) 8 Qf3 Nxc3 9 bxc3 Qc8 {Diagram 57) 10 Rb1
(Wells's 10 Bd3!? Bxd3 1 1 cxd3 avoids our next note) 10 ... c6 (Black must avoid the
overly-greedy 10...Qe6+!? 1 1 Kf2 Qxa2? 12 Qxb7 Qxc2+ 13 Kg3! Qe4 14 BbS+, but
here 1 l ...Be4! 12 Bc4 Bxf3 13 Bxe6 Be4 14 Bc4 b6 should be okay for him) 1 1 Bd3
Bxd3 12 cxd3 e6 13 Qg3! Nd7 14 Nf3 Nf6 15 BeS gave White good pressure for the
pawn in P.Mladenov-D.Kirov, Sofia 2008.

Wh ite's queen is active How to seize the initiative?
194
T h e Mod e r n Preference: 2 ... Ne4 3 Bf4
b ) 6...c5 introduces dark-squared features to the play, but here too White appears
to have decent compensation: 7 dS exf3 8 Nxf3 g6 (or 8 ... a6 9 Qe2! g6 1 0 0-0-0 Bg7
1 1 d6 when one has to agree with Wells that White has good play) 9 NbS! (a note
worthy method to prevent any expansion with ... a6 and ... bS) 9 ... Na6 10 Bc4 Bg7
1 1 0-0 0-0 (Diagram 58} 1 2 d6! (Gavrikov) gives White some initiative.
c) 6...e3 (Diagram 59) is an approach seen too in the Blackmar-Diemer proper.
Black returns the pawn in a bid to slow down White's development, although 7
Bc4!? (more often White has directly transposed to a line of the Blackmar-Diemer
both sides have lost a move - with 7 Bxe3, after which 7 ...e6 8 Qd2 Be7 9 0-0-0 0-0
10 g4 saw White wasting no time commencing his attack in A.Moiseenko
O.Romanishin, Alushta 2002, and 7...Bf5 8 Bd3!? Bxd3 9 Qxd3 e6 10 QbS+! Nbd7 1 1
Qxb7 regained the pawn with a n edge i n D.Gordon-S.Emst, Ottawa 2007) 7. . .e6
(neither is deploying the bishop outside the pawn chain a complete panacea:
7.. Bf5 8 Nge2 e6 9 g4! Bg6 10 Bxe3 and I prefer White, who was definitely better
.
after the further 10 ... Nd5?! 1 1 BxdS! exdS 12 Nf4 Bb4 13 h4 of R.Djurhuus
N.Grotnes, Namses 1 995) 8 Qd3 Nc6 9 0-0-0 Nb4 10 Qxe3 (Diagram 60} 10 ... Nbd5?!
(Prie points out that 10 ... Be7 improves, although White can still hope to attack
with 1 1 Nge2 0-0 12 g4 NbdS 13 Qd2!) 1 1 NxdS NxdS 12 BxdS! QxdS 13 Kb1 gives
White a handy lead in development, E.Tate-H.Baldursson, Reykjavik 2007.
Diagram 59 (W} Diagram 60 (B)

Black takes no chances Black is yet to eq ualize
d) 6 ... Nd5!? (Diagram 61} enables Black to fight for the initiative and is a quite a
decent try in my view: 7 NxdS (7 Bg3?! is a rather desperate attempt to sacrifice
and after 7... Nc6 8 BbS eS!? 9 dxeS Bb4 10 Nge2 exf3 1 1 gxf3 BfS Black had the ini
tia tive in A .Bigg-R.Palliser, Street 2004) 7...Qxd5 8 Bxc7 (regaining the pawn; oth
erwise Vigus's 8 c4 QaS+! 9 Bd2 QfS keeps B lack's queen annoyingly active in any
case) 8 . Nc6 9 c3 BfS 1 0 Be2 (the reader might wish to investigate Davies's idea of
. .
195
10 Qb3!? when he prefers White after 10...Qd7 1 1 Bf4 e6 12 Be2 Be7 13 fxe4 Bxe4 1 4
Nf3, but Black might prefer 1 l ...e5!? 12 dxe5 0-0-0 13 Rd1 Qxd1+ 14 Qxd1 Rxd1+ 15
Kxd1 Bc5 with sufficient play for the pawrl1 1 0 ... g5!? (Black continues creatively,
although he was fine after the simpler 10 ...e5 1 1 Bxe5 Nxe5 12 dxe5 Qxe5 13 Qa4+
Bd7 14 Qxe4 Qxe4 15 fxe4 0-0-0 in R.Djurhuus-F.Elsness, Gausdal 1995) 1 1 fxe4
Bxe4 12 Nf3 g4 13 Nh4! was rather murky in J.Berglund-T.Robertsen, Tromsoe
1992.

Quite a challenging riposte It's like a Blackmar-Diemer!
We now return to 6 ...exf3:

7 Nxf3 (Diagram 62)
NOTE: White finds himself a tempo ahead of the Blackmar-Diemer

Gambit (1 d4 d5 2 e4 dxe4 3 Nc3 Nf6 4 f3 exf3 5 Nxf3). That is often
considered a little speculative, but the inclusion of Bf4 appears to
make the gambit quite viable.
7 ..Bg4
.
The most active deployment for this bishop and Black's most popular choice, but
by no means everyone has been bold or foolhardy enough to play this:
a) Euwe's 5 ...e6 is considered a respectable defence to the Blackmar-Diemer, but
here White is a move faster castling queenside: 7... e6 8 Bc4 c6 9 Qe2 (popular, b ut
9 0-0!? has also been seen, after which 9 ... Be7 10 Qe2 0-0 1 1 Rae1 Re8?! 12 Kh1 b5
13 Bd3 b4 14 Ne4 Nd5? 15 Bxb8! Rxb8 16 Ne5 saw Black failing to untangle and
losing material in S.Sahu-T.Sanjay, Mumbai 2008) 9 ... Be7 (Black tried to blockade
the centre with 9 ... Nbd7 10 0-0-0 Nb6?! in V.Jansa-G.Sosonko, Amsterdam 1975,
but White blasted through anyway with 1 1 d5! Nbxd5 12 Bxd5 Nxd5 13 Rxd5!
cxd5 14 NbS; a powerful idea well worth remembering, as is the possibility of
196
T h e Modern Prefe re n ce: 2 ... N e4 3 Bf4
9 ... Nd5!?,
after which 1 0 Bxb8 Nxc3 1 1 bxc3 Rxb8 12 0-0 Bd6 13 Ne5 0-0 14 Bd3
g ave White enough kingside pressure for his pawn in B.Steiner-P.Steinwender,
Austrian League 2006) 10 0-0-0 0-0 1 1 Bd3 Nbd7 12 h4! (Diagram 63) 12 ... Re8 13
Be5 Nf8 14 h5 h6 15 Kb1 left White with enduring and quite dangerous compensa
tion in M.Galyas-A.Schneider, Hungarian League 2000.

Black is rather passive Spot the threat!
b) 7 ... c6 8 Bc4 (8 Bd3!? is a decent alternative when White should be especially

keen to see 8 ...Bg4 9 h3 Bxf3 10 Qxf3 Qxd4 11 0-0-0 with a healthy lead in devel
opment for the two pawns) 8 ... Bf5 (this position might also come about via a
7 . Bf5 8 Bc4 c6 move order, although there White has often preferred 8 Bd3) 9 0-0
..
e6 echoes another of Black's better defences to the Blackmar-Diemer. White might

wish that his bishop didn't block the f-file, but 10 Ne5 (Diagram 64) 10 ...Be7?
(Black had to find 10 ... Nbd7) 1 1 Nxf7! Kxf7 12 Bxb8 Bxc2 13 Qe2! Rxb8 14 Bxe6+
Kf8 15 Qxc2 saw him solving in style that particular problem in A.Almeida Saenz
S.Colli Lopez, Aguascalientes 2008.
c) It might seem sensible to shore up the kingside while not obstructing the light
squared bishop with 7...g6, but here White has promising compensation as shown
by a classic performance from Hodgson: 8 Bc4 Bg7 9 Qe2! 0-0 10 0-0-0 c6 1 1 d5!
cxd 5 12 Nxd5 Nxd5 13 Rxd5 Qb6? (Black later improved with 13 ... Nd7 14 Rhd1 e6
1 5 Rd6 Qf6 in J.Retera-F.Corrales Jimenez, Sitges 2007, but after 16 R1d4!? h6?!
w ith 17 c3! White would have retained good pressure) 14 Rb5 Qc6 15 Ne5 Qe8 16
h4! gave White a strong attack in J.Hodgson-A.Panchenko, Bern 1994.
TIP: White should always be alert to the possibility of 0-0-0 followed

by d5 in the gambit accepted, blowing open the centre to exploit his
superior development.
8 h31 Bh5
197
s t a rt i n g Out: The Trom pows ky Attack
This permits White to expand with tempo, but he also has a strong initiative after
8 . Bxf3 9 Qxf3 c6 10 0-0-0, as we'll see in Game 23.
..
g g4 Bg6 10 Ne5 (Diagram 65)

Expanding with tempo Does Black control the dS-square?
White not only has a lead in development, but much the freer development. One
has to like his compensation and 10 ... Nd5 1 1 Qf3 e6 12 0-0-0 Be7 13 Nxd5!? (keep
ing pieces on with 13 Bc4 looks quite promising too) 13 ... Qxd5 14 Qxd5 exd5 15
Nxg6 hxg6 16 Bxc7 saw him regain the pawn while netting the bishop-pair in
V.Milov-S.Battesti, Bastia (rapid) 2004.
Conclusion
White has made 54% from almost 1,150 games with 4 f3. Black certainly faces
practical difficulties whether White meets 4 ... Nf6 with 5 e4 or the slightly
forgotten 5 Nc3. White has to be happy to remain a pawn down with the former,
but I'm not surprised that he has scored 57% from 150 games with the gambit ac
cepted. His initiative is far from easy to neutralize and, indeed, I suspect that
Black might well be best advised to decline the pawn with 6... Nd5!?.
Il lustrative Games
Game 23
0 G.Kasparov M.Carneiro
S a o Pa ulo s i m u l 2004
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 d5 4 f3 Nf6 5 e41?
198
The Mode r n Prefe rence: 2 .. N e4 3 Bf4
.
It's notable to see arguably the World's strongest-ever player endorsing this gam
bit, albeit only in a simul.
s . dxe4 6 Nc3 exf3 7 Nxf3 Bg4 8 h3 Bxf3 9 Qxf3 c6 10 o-o-o e6 11 Bc4 (Diagram 66)
. .
The bishop might appear blunted here, but White has in mind a strong central
breakthrough, which we've already seen wreak havoc in similar positions. Here
1 1 g4!? is also possible and after 1 l ...Be7 12 Kb1 Nbd7 13 h4 NdS 14 Nxd5! cxdS 15
gS Black's king position remained a cause for concern in G.Meszaros-L.Pergel,
Hungarian League 1997.
11... Nbd7?
Inadvisable. Instead a famous Trompowsky predecessor continued 1 1 ... Be7 12 Kb 1
0-0 (as pointed out by Baburin, Black might have preferred 12 ... Nbd7 when 1 3 g4
NdS!? 14 Nxd5 cxdS 15 Bd3 wouldn't have been too clear, although even here I
quite like White's compensation; note that Black's king can hardly go long due to
White's raking dark-squared bishop and c4-break) 13 h4! NdS 14 Ne4 bS 15 Bd3
Nd7? (Black's position was not so easy to handle, but this is too slow; he had to
remove an attacker with 15 ... Nxf4 16 Qxf4 when Wells's 16 ... Nd7 1 7 g4 maintains
good attacking chances, but is hardly totally clear) 16 NgS! N7f6 1 7 BeS! aS 18
Rdfl a4? 19 Bxh7+! Nxh7 20 Qh5 and Black was forced to give up in S.B.Hansen
H.Olafsson, Reykjavik 1995.
NOTE: White's pawnstorm might not appear all that fast in these
lines, but don't forget that Black is a long way from creating any
problems on the queenside.
12 dSI {Diagram 67)
Diagram 67 {B) Diagram 68 {B)

The dream breakthrough Concluding in style
199
Sta rt i n g Out: T h e Trom pows ky Attack
12...cxds 13 Nxds
Rather strong too would have been 13 Bxd5!? Nxd5 (13 ...exd5? 14 Nxd5 Nxd5 15
Qxd5 Be7 16 Qxb7 leaves Black fatally pinned) 14 Nxd5 ReB 15 Rhe1 followed by
16 Qg3.
13 ... Nxds?
Not the best defence. Carneiro had to try 13 ... exd5 14 Rhe1+ Be7 when 15 Bd6 0-0
(15 ... dxc4!? is the alternative when 16 Bxe7 Qxe7 1 7 Qxb7 Rb8 18 Qxa7 Qe6 1 9 Qc7
c3 20 Rxe6+ fxe6 21 Qxc3 looks good for White with an active queen and three
connected passed pawns for the two knights and rook) 16 Bxe7 Qc7 1 7 Bxd5! NeS
18 Qa3 Rfc8 19 Bb3 Nc4 20 Bxc4 Qxc4 21 Qb3 (Prie) would have left White clearly
for choice thanks to his dominant forces.
14 Bxds as?
This radical move doesn't help, but White's initiative would have been crushing
even after 14 ... exd5 15 Rhe1+ Be7 16 Bd6 Ne5! 1 7 Bxe5 0-0 18 Rxd5 when a sacrifice
on g7 would have followed.
15 Bxb7 Ra7 16 Rxd71 (Diagram 68)
By no means the only good approach, but definitely the most attractive. How
Kasparov must have been enjoying himself by now!
16...Qf6
Black tries to save his queen, but he should really have resigned.
17 Rhd1 Be7 18 Rxe7+1 Qxe7
The diagonal-moving pieces combine to force mate in the event of 18 ... Kxe7 19
Bd6+ Kd8 20 Bc5+ Kc7 21 Qc6+ Kb8 22 Bd6 mate.
19 Qc6+ Kf8 20 Bd6
By now Kasparov must have been rather displeased by his opponent's unaccept
able manners, but he might have forced a quick mate with 20 Qc8+ Qe8 21 Bd6+.
20 ... g6 21 Bxe7+ Kxe7 22 QcS+ Kf6 23 Qxa7 Rf8 24 Qd4+ es 25 Qd6+ Kg7 26 Qxe5+
Kg8 27 Qf6 hS 28 BdS Kh7 29 Bxf7 1-0
200
Chapter Seven
T h e Ma i n Li n e :
2 N e 4 3 Bf 4 c s

m I ntrod uction a nd 4 d 5
m The 4 d 5 Mai n Li ne: 4 Qb6 5 Bc1

m Forci ng the Pace: 4 f3

m An I m portant Check: 4 f3 Qa 5+
m Retreating to c1
m Avoidi ng a Benoni: 5 c3 Nf6 6 Nd2
I ntroduction and 4 d S
1 d4 Nf6 2 Bgs Ne4 3 Bf4 cs (Diagram 1)
Diagram 1 (W} Diagram 2 (W}

A dark-square defence How best to deal with the threat?
The current main line. This dark-square approach is especially popular with
King's Indian and Modem Benoni fans, and has built up a sizeable body of theory
over the years due in no small part to its many fascinating sub-variations.
Unlike after the immediate 2 ...c5, White cannot dislocate Black's structure, but he
will be able to expand in the centre with tempo. Indeed, 4 f3 has been played in
around 75% of all the games which have reached this position and is clearly the
modern preference. Note that after 4 f3 we will consider 4 . . Nf6 in the third section
.
of this chapter, 4 .. Qa5+ 5 c3 Nf6 6 d5 Qb6 without 7 Bel in the fourth, that 7 Bel in
.
the fifth and, finally, 4 ...Qa5+ 5 c3 Nf6 6 Nd2 last but by no means least.
However, 4 d5 still very much deserves to exist and it is to this variation which we
will first turn our attention.
4 d5
White takes play straight into a Benoni structure. However, it's a moot point
whether this is the most accurate way to do so: 4 f3 Qa5+ 5 c3 Nf6 6 d5 is the alter
native and a method which certainly reduces Black's options en route. Some play
ers, though, like the nowadays fairly rare 4 d5 Qb6 5 Nd2 and 5 Qcl, both of
which are only available via a 4 d5 move order.
NOTE: Another very important reason for studying this section is if

one meets 2 cs with 3 ds. Then Black doesn't have to head into the
...
Vaganian Gambit and 3 .. Ne4 4 Bf4 is simply a transposition to our

.
coverage here.
202
The Ma i n l i n e : 2 .. . Ne4 3 Bf4 c s
4 ... Qb6 (Diagram 2)

By some margin Black's most popular move, eyeing the pawn which White's
bishop has abandoned. However, due to the unforcing nature of 4 d5, there are
alternatives and some important ones at that:
a) 4 ... d6 5 f3 Nf6 (as the reader will quickly become aware there are a number of
t ranspositions between this section and those devoted to 4 f3; here 5 ... Qa5+ 6 c3
Nf6 is our first, transposing to note 'a' to Black's 6th move in the fourth section of
this chapter - see the move order 4 f3 Qa5+ 5 c3 NJ6 6 d5 d6) 6 e4 g6 (Black shows
himself happy to head for a Benoni structure; fans of the Czech Benoni might,
th ough, prefer 6 ... e5, against which White played aggressively in A.Summerscale
B.Lalic, Coulsdon 1999: 7 Be3 Be7 8 Ne2 0-0 9 c4 NeB - Black frees his f-pawn and
p repares the thematic freeing manoeuvre ... Bg5 - 10 Qd2! g6 1 1 Bh6 Ng7 12 Nbc3
Nd7 13 0-0-0 Nf6 14 h4! and it was already clear that White's attack was the more
po tent) 7 Nc3! Bg7 8 Qd2 (Diagram 3) sees White preparing a direct attack with
8 . . . 0-0 9 Bh6, as we'll see in Game 24.

White hopes to attack A respectable a lternative to 4... Qb6
NOTE: Here White also has 7 c4, which resembles the S amisch King's
Indian (1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 f3).
White's bishop might look a little odd on f4, but after 7... Bg7 8 Nc3 0-0 9 Nge2 it
can be seen usefully restraining the . ..e6 break and 9 ... Nbd7 10 Ng3 h5 1 1 Be2 h4
12 N fl gave an edge in A.Motylev-M.Ahn, European Club Cup, Kallithea 2008.
b ) 4 ... g6 is merely a transposition to our last section after 5 f3 Nf6 6 e4 d6.
c) More original and critical is 4 ... e6!? (Diagram 4) and after 5 f3 (Michael Adams
on ce preferred 5 dxe6 fxe6 6 e3, but this is a bit too simplistic for my taste and
B la ck is surely fine after 6... d5 7 Bd3 Nf6 8 c4 Nc6) we reach a further divide:
203
Sta rt ing Out: The Trom pows ky Attack
cl) Relatively compliant is S ... Nf6, after which 6 e4 (both 6 c4 and 6 Nc3!? are
highly sensible alternatives; theory has advocated 6... Nh5 in response to both, but
in the case of the latter I quite like White if he follows up the 7 Be3 d6 8 Qd2 Be7 9
Bf2 0-0 of J.Piket-T.Shaked, Merrillville 1997, with Davies's logical suggestion of
10 0-0-0!?) 6 ...exd5 7 exdS d6 (Diagram 5) reaches a structure generally considered
slightly more pleasant for White. That assessment seems fair here too and 8 Bb5+!?
(Stefanova had earlier preferred 8 Nc3 Be7 9 Qd2, and after 9 ... a6 10 a4 0-0 1 1
Nge2 Re8 12 g4! QaS! 13 Ng3 b5 14 gS b4 1 5 Nce2 a sharp struggle arose in
A.Stefanova-A.Shirov, Drammen 2005) 8 ... Nbd7 9 Qe2+ Qe7 10 Nc3 a6 1 1 Bxd7+
Bxd7 12 0-0-0 Qxe2 13 Ngxe2 0-0-0 14 g4! h6 15 h4 bS 16 Ng3 saw White seizing
some useful kingside space in A.Stefanova-T.Kosintseva, Krasnoturinsk 2006.

White has more space A creative idea
c2) Black might fork with 5 ...Qf6?!, but White doesn't have to reply 6 Qcl . Instead
6 Bxb8! Qxb2 7 fxe4 Qxa1 8 Bg3 Qxa2 9 Nf3 is surely good for White, whose extra
minor pieces should come into their own once he has developed his kingside.
c3) A better attempt at creativity is 5 ... Bd6!? 6 Bxd6 Nxd6 (Diagram 6), which re
mains fairly rare but did worry both Hodgson and Gallagher in their respective
works. Following 7 Qd2!? (more natural and playable is 7 e4,_ but 7... Qb6! is a little
awkward; both Winants and Wells have now pointed to 8)5S1 the way forwards,
since 8 Nc3?! Qxb2 9 Nge2 NbS! didn't give White quite enough for his pawn in
S.Conquest-A.David, Mondariz 2000) 7...0-0 (Wells indicates that 7... Qh4+ 8 g3
Qd4 9 Qxd4 cxd4 10 c3! gives White good chances to emerge with an edge, but
7 ... Nc4!? is a better alternative: 8 Qc3 Qh4+ 9 g3 Qd4 10 Qxd4 cxd4 1 1 e4 Ne3 12
Na3 exdS 1 3 exdS NxdS 14 NbS Nc6 15 Ne2 0-0 16 Kd2 b6 was about equal in
A.Lambert-P.Howard, correspondence 2004) 8 e4 b5!? (more challenging than
8 ...exd5?! 9 Qxd5 Qb6 10 Qb3 when 10 ...Qa5+ 1 1 Nc3 Nc6 12 0-0-0 Nd4?! 13 QdS
204
T h e M a i n L i n e : 2 ... N e4 3 Bf4 c s
Qc7 14 Nge2 Nxe2+ 15 Bxe2 emphasised Black's problems down the d-file in
A.Stefanova-A.Van Elst, Montpellier 2000) 9 Nc3 (Diagram 7) 9 ...Qb6 10 b3 (10
dxe6!? might improve, since 1 0 .. .fxe6 11 0-0-0 Nc4 12 Bxc4 bxc4 13 Qd6 Nc6 14
Nge2 prepares 15 Na4 and keeps the queenside situation under control) 10 ...e5 11
Nh3 Na6 12 Rb1 Rb8 13 Be2 b4 14 Na4 Qc7 the position was about equal in
M.Comette-H.Hamdouchi, Nice 2003.

Both sides have their trumps Did you expect this retreat?
c4) 5 ... Qa5+ can be met by either 6 Nd2 or 6 c3 when 6... Nf6 transposes to note 'b'
to Black's 6th move in the fourth section of this chapter - see the move order 4 f3
Qa5+ 5 c3 Nf6 6 d5 e6.
d) Finally, fairly rare is 4 ...Qa5+ when 5 Nd2!? (very logical too is 5 c3, after which
S ... d6 6 f3 Nf6 transposes to note 'a' to Black's 6th move in the fourth section of
this chapter - see the move order 4 f3 Qa5+ 5 c3 Nf6 6 d5 d6 - but that was better for
Black than 5 ... Qb6?! 6 Nd2! Nxd2 7 Qxd2 d6 8 e4 e5 9 dxe6! Bxe6 10 Nf3, which
gave White a pleasant edge in T.Radjabov-A.Romero Holmes, Benidorm (rapid)
2003) 5 ... e6!? (again this challenge is more critical than 5 ... d6 when 6 c3 Nf6 7 e4 g6
8 Ngf3 Bg7 9 h3 0-0 10 Bd3 slightly favoured White in J.Hodgson-G.Morrison, Ed
inburgh 1989) 6 c3!? is an idea stemming from Hodgson which surprisingly re
mains untested. His idea was 6 ... exd5 7 Nxe4 dxe4 8 Qd5 when 8 ...Qb6 9 Qxe4+
Qe6 10 Qxe6+ 1 1 e4 (Wells) again looks a little better for White.
After that long but important interlude, we return to 4 ...Qb6:
5 Bell (Diagram 8)
A remarkable idea - White moves the bishop for the third time in five moves. This
may look ridiculous at first, but White will drive the knight back with tempo.
Moreover, Black's queen is not so ideally placed: on b6 it obstructs the b-pawn (a
key source of counterplay in Benoni structures) and can be a target for Nc4.
205
Sta rt i n g Out: T h e Trom powsky Attack
WARNING: White doesn't have to retreat his bishop, but he must

avoid the horribly weakening 5 b3?? when s Qf6 (Diagram 9) picks
...
up the exchange for clearly insufficient compensation.

Do not allow this! Quite an am bitious sacrifice
We will return to this critical retreat in our next section. It has been seen in more
than half of all games with 4 ...Qb6 and become established as the modem choice,
but that doesn't mean that the alternatives are invalid:
a) The move White would like to make work is 5 Nd2!?, gambiting the b-pawn
just as we've seen White do in the 3 h4 variation. This once had Hodgson's pa
tronage and more recently has been used by Winants, but the bishop can be tar
geted by 5 ...Qxb2! (critical, whereas 5 ... Nxd2 6 Bxd2 Qxb2 7 e4 offers White easy
play and decent compensation: for example, 7 ... Qb6 8 Nf3 d6 9 Bc3! Nd7 10 Bc4
Rg8 1 1 Rb1 Qc7 12 eS! dxeS 13 Qe2 left Black with horrendous difficulties develop
ing in J.Percze-I.Mackintosh, correspondence 2006, and 7... g6 8 Rb1 QeS 9 Bd3 c4!?
10 Nf3 Qc7 1 1 Be2 Bg7 12 0-0 d6 13 Rb4 regained the pawn with an edge in
S.Emst-C.Scholz, German League 2007) 6 Nxe4 Qb4+ 7 Qd2 Qxe4 (Diagram 10)
when White has tried a few ideas, but without demonstrating quite enough objec
tive compensation. That said, this might make a decent surprise weapon and
Black must be careful with his queen floating around:
a1) 8 f3?! Qd4! 9 Qxd4 cxd4 10 BeS (10 0-0-0!? might improve, although I still
slightly prefer Black after the messy 10 .. .f6! 1 1 Rxd4 eS 12 Rc4 Nc6) 10 ... e6 1 1 Bxd4
exdS 12 Bb2 Bb4+ was very pleasant for Black with his much less-exposed king in
W.Gerstner-J.Gallagher, Biel 1993.
a2) 8 e3 gS!? (theory used to prefer the somewhat less radical 8 ...Qb4, but after 9 c3
QaS 10 d6! Nc6 1 1 Nf3 exd6 12 Bc4 Be7 13 h4!? White enjoyed decent compensa-
206
T h e M a i n Line: 2 ... Ne4 3 Bf4 c s
ti on i n L.Winants-G.Schebler, Belgian League 2002; however, Black might prefer

Hod gson's recommendation of 8 ... e5!? when White can hardly allow the centre to
cl ose, but 9 dxe6 Qxe6! looks like it should favour Black) 9 f3 (naturally White re
fu ses to fall for 9 Bxg5? Qe5) 9...Qf5 10 Bg3 Qf6 1 1 c3! (White must prevent the
queens from coming off after 1 1 ...Bg7 and 12... Qc3) 1 1...Bg7 12 Rcl d6 13 f4 (Dia
gra m 11) 13 ... Nd7 14 Nf3 gave White some play for his pawn in L.Winants
L. Verat, French League 2002.

Not so clea r due to ... gs Neither sacrificing nor retreating
a3) Davies makes a decent case for 8 Nf3!?, although I'm not entirely convinced
that the critical 8 ...e6 (8... d6 9 e3 Qb4! 10 c3 Qa5 is a sensible alternative, after
which 1 1 e4 g6 1 2 Rc1 Bg7 13 e5! dxe5 14 Bxe5 0-0 15 Bxg7 Kxg7 16 h4 saw White
doing his utmost to attack in M.Karttunen-R.Nevanlinna, Finnish League 2008,
but Black was most certainly not worse after 16 ... h5) 9 dxe6 Qxe6 1 0 e4!? offers
White quite enough: 10 ... Be7 (10 ... Qxe4+ 1 1 Be2 Be7 12 0-0 0-0 13 Rfe1 offers decent
play for the pawns) 1 1 Be2 0-0 12 0-0 Nc6 was seen in R.Calderin-I.Blanco Sing,
Li nares 1997, and now White must probably try to keep Black bottled up with 13
e5!?.
b) 5 Qcl !? (Diagram 12) is another move White would quite like to make work. He
is ready to continue with f3, e4, c3 and Na3-c4, which should give him an edge if
Black develops slowly, but there are some critical responses:

bl ) Theory used to highly praise 5 ... c4!? 6 e3 Qa5+, but White has the answer in 7
N d2! because 7... c3 (fairly essential; Black just loses a pawn after 7... Nxd2? 8 Qxd2
Qx d2+ 9 Kxd2 b5 10 a4) 8 bxc3 Nxc3 (Alburt, Dzindzichashvili and Perelshteyn
prefer Black after 8 ...Qxd5, but unfortunately they stop their analysis here; Wells,
however, continues with 9 Nxe4 Qxe4 10 Nf3 d6 1 1 Bd3 Qc6 12 Qb2 Nd7 13 0-0
When White's superior development should offset his weakened structure) 9 d6!
207
(Diagram 13) leaves Black with ongoing developmental difficulties, as shown by a

long-time occasional Trompowsky exponent: 9 ... exd6 (Black finds himself badly
tied up after Wells's 9 ... Ne4 10 Ngf3 Nxd6 1 1 Bxd6 exd6 12 Qb2) 10 Ngf3 d5 1 1 a3!
Nc6 12 Nb3 Qb6 13 Qd2 left White better in V.Hort-F.Goldstem, Davos 2002.

A strong positiona l sacrifice Has White enough?
b2) Thus it seems that 5 ...e6! is more challenging. Then A.Stefanova

S.Maksimovic, Kusadasi 2006, quickly plunged into chaos with 6 f3 (White has
also tried 6 c4, but after 6 ...exd5 7 cxd5 c4! 8 Nh3 Bb4+ 9 Nc3, as seen in K.Kaunas
J.Czakon, Warsaw (rapid) 2005, and then 9 ...Qf6!? 10 Bd2 Nxd2 1 1 Qxd2 0-0 Black
is slightly for choice) 6... Nf6 7 e4 exd5 8 e5 Nh5 (Diagram 14) 9 Bd2!? (an attempt
to improve over 9 Ne2 d6 10 Nbc3 Be6 1 1 g4 dxe5 1 2 Bxe5 Nc6, which was messy
but good for Black in J.Hodgson-B.Lalic, St Helier 1997) 9 ...Qe6 10 g4!? (radical,
but 10 f4 can be well met by 10 ... g5) 10 ...Qxe5+ 1 1 Kf2, but does White really have
enough for the two pawns here after 1 l ...Nf6 ( 1 l ...c4? 12 gxh5 Bc5+ 13 Kg2 0-0 14
Ne2 Nc6 15 Nbc3 d6 gave Black some compensation in the game, but I don't see
any need to sacrifice a piece so) 12 Bf4 Qe6 13 Nc3 c4! 14 Bh3 Bc5+ 15 Kg2 Qc6 16
Qe1+ Kf8 1 7 Qh4?
Conclusion
4 d5 has been largely superseded by 4 f3, but it remains an important variation,
especially for those who like to meet 2 ... c5 with 3 d5. Of Black's alternatives to
4 ... Qb6, only 4 ... e6 prevents White from obtaining a pleasant set-up, but the queen
move is both critical and common. I don't particularly trust either 5 Qcl or the 5
Nd2 gambit in response, but they might appeal to those keen to get off the beaten
track.
208
The M a i n L i n e : 2 .. Ne4 3 Bf4 cS
.
Il lustrative Ga mes
Game 24
0 M.Adams K.Georgiev
B u rgas 1993
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 c5 4 f3 Nf6 5 d5 d6 6 e4 g6 7 Nc3 Bg7 8 Qd2

As Michael Adams puts it so well: 'White's strategy here is similar to that in the
legendary 150 Attack. The aim is to exchange the dark-squared bishops and
launch a mating attack.' Not everyone is quite so familiar with the 150 Attack (1 e4
d6 2 d4 Nf6 3 Nc3 g6 4 Be3 Bg7 5 Qd2), or similar direct attacking schemes, most
notably White's play in the main lines of the Sicilian Dragon, but it's not too hard
to handle the white pieces here, as we will see.
8. . 0-0?1
.
Castling into it. Black might try to keep the dark-squared bishops on the board,
but after 8 ... h6 he cannot castle in a hurry and 9 Nge2 Na6 10 Ng3 Nc7 1 1 Be2 Nd7
12 a4 a6 13 Be3 Rb8 14 Rb1 ! (Diagram 15) 14 ...h5 15 0-0 h4 16 Nh1 h3 1 7 g3 b6 18
Nf2 Bb7 19 b4 was something of a white model in J.Nogueiras Santiago-F.Gomez,
Santa Clara 2000.

Black ca n't castle just yet Beginning the attack!
0
NOTE: Such Benoni structures with the c-pawn back on c2 are often
referred to as types of Schmid Benoni. Nomenclature aside, what
matters most is that White does his utmost best to restrain ... bs
while looking to advance with either b4 or es himself.
2 09
1 suspect that Black's best course of action is to follow in Lanka's footsteps with
8 ... a6 (now castling long will be pretty risky with Black's b-pawn set to advance) 9
a4 QaS, which prepares ...bS, although 10 Nge2!? bS 1 1 Ng3 hS (White is happy to
regroup with 1 l...b4 12 Nd1; his knight will come to e3 and then quite possibly c4)
12 Be2 h4 13 Nfl Nbd7 14 Ra2! (forcing the queenside issue) 14 ...bxa4 15 Nxa4 Qc7
16 Ne3 saw White refusing to panic and retaining a pull in J.Degraeve-Z.Lanka,
Montpellier 1996.
9 Bh61 (Diagram 16)
Naturally White wastes no time exchanging the dark-squared bishops. Moreover,
Black will no longer be able to block an advance of White's h-pawn with ...h5.
g. .Bxh6
.
Black did his best to drum up counterplay with 9 ... e6!? 10 0-0-0 (and not 10 Bxg7
Kxg7 1 1 dxe6 Bxe6 12 0-0-0?! Nc6 13 Qxd6 Qa5 when Black has swapped his weak
pawn for the initiative) 10 ...exd5 11 exdS b5 in L.Aiburt-E.Gudmundsson, Reykja
vik 1984, but 1 2 Bxb5 Qb6 13 h4! Ba6 14 Ba4! Nbd7 15 Bxd7 Nxd7 1 6 Bxg7 Kxg7 17
hS Rab8 18 b3 Nf6 19 hxg6 fxg6 20 Nh3 saw White bagging a pawn while retain
ing the stronger attack.
10 Qxh6 e6 11 o-o-o exds 12 exds a6 13 Nge21
Calm. White simply wants to bring a knight to e4. Such a plan is strong, although
a more aggressive player might have got on with it: 13 h4!? b5 14 h5 Qe7 15 Kb1
b4 16 Nce2 and with two knights on the kingside, I like White's attacking chances.
13 ...b5 14 Ng3 ReB?!
Too routine. As Adams later pointed out, Black should have preferred 14 ... Ra7!,
preparing to defend along his second rank.
15 Nce41 Nxe4 16 Nxe4 (Diagram 17)

The threats mou nt How best to conclude?
2 10
The Ma i n L i n e: 2 . . . N e4 3 Bf4 c s
16 . . .ts
This weakening thrust isn't ideal, but Black had to deal with the threats of 17 Ng5
and 17 Qf4, not to mention a rapid advance of the h-pawn.
17 Ng5 Qe7?
The Bulgarian Grandmaster has endured something of a torrid time and now
misses a tactic. Black had to find 17 ... Qc7 when 18 Bd3 (18 Bxb5? axb5 19 Rhe1 no
longer comes with tempo and so Black can sidestep with 19 ... Rf8) 18 ...Qg7 19
Qxg7+ Kxg7 20 Rhe1 (Adams) would have left White set to invade down the e-file,
but Black still fighting.
1 8 Bxbsl Qe3+
This doesn't help, but after 18 ... axb5 19 Rhe1 Black would have had to part with
his queen.
19 Kb1 Re7 (Diagram 18) 20 Be81 1-0
Neat. The check on f7 will cause utter devastation, as shown by the line 20 ... Raa7
21 Rhe1 Qxe1 22 Bf7+ Rxf7 23 Rxe1 Rf8 24 h4 with a material advantage and an
ongoing attack.
The 4 d s Main Line: 4 0b6 5 Bc1 ...

-
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 cs 4 ds Qb6 5 Bc11 (Diagram 19)

A critica l retreat White begins to take control
This cool retreat very much deserves main line status after 4 d5 Qb6. White might
h ave only developed his d-pawn thus far, but he will obtain a pleasant pull after 6
f3 and 7 e4 unless Black is careful.
211
s ...g6
Fianchettoing is very natural, but Black has also been known to close the centre
with 5 ...e5, after which 6 f3 Nf6 (by analogy with our main line Black has tried
6... Nd6!?, but 7 e4 g6 8 Nc3 Bg7 9 Nh3 0-0 1 0 Be2 c4 1 1 g4!? Qc5 12 Bg5! f6 13 Bh4
Bh6 14 Bf2 Be3 15 Bxe3 Qxe3 16 Qd2 retained a comfortable edge in J.Avila
Jimenez-R.Edouard, La Massana 2008) 7 e4 d6 8 Na3 Be7 9 Nc4 Qd8 10 a4 (Dia
gram 20) is typical of the time Black must often lose with his queen if he develops
slowly. Here 10 ... 0-0 11 Bd3 NeB prepared a typical Czech-Benoni freeing device
in A.Stefanova-F.Manca, Saint Vincent 2000, but the instructive sequence 12 Ne2
Bg5 13 f4! exf4 14 Nxf4 Nd7 15 0-0 Ne5 16 Be2 Nxc4 17 Bxc4 Qe7 18 Qe2 h6 19
Nd3! Bxcl 20 Raxcl Nf6 21 e5 saw White winning the key battle for the e5-square.
Stefanova has also shown the way against 5 ... f5: 6 f3 Nf6 7 Nc3 e6 8 e4 fxe4 9 fxe4
exdS 10 NxdS! Nxd5 1 1 Qxd5 Qe6 12 Bc4 kept the central situation under control
and gave White an edge in A.Stefanova-A.Panchenko, Barbera 1999.
NOTE: Black's most popular choice after 5 Bc1 has actually been
s ...e6 when 6 f3 Nf6 transposes to the critical main line of our penul
timate section (4 f3 QaS+ 5 c3 Nf6 6 ds Qb6 7 Bc1 e6), albeit with
both sides having saved one on the move number.
Black might also follow up 5 ...e6 6 f3 with 6 ...Qa5+, as Kasparov once did, and af
ter 7 c3 Nf6 8 e4 d6 (8 ...exd5?! is well met by 9 eS!) 9 Na3 (White might well prefer
to unpin with 9 Bd2!? when 9 ...Qb6 10 c4 Qxb2 1 1 Nc3 transposes to note 'c' to
Black's 8th move in our penultimate section - see the move order 4 f3 QaS+ 5 c3 Nf6 6
d5 Qb6 7 Bel e6 8 c4 Qb4+ 9 Bd2 Qxb2 1 0 Nc3) 9 ... exd5 10 exd5 Be7 1 1 Nc4 Qd8 12
Ne3 0-0 he was fine in J.Van der Wiel-G.Kasparov, Moscow Interzonal 1982.
6 f3 (Diagram 21)
Diagram 2 1 (B) Diagram 22 (W)

Forcing the knight back It cou ld be a Modern Benoni!
212
The Ma i n l i n e : 2 . . . Ne4 3 Bf4 c s
6 Nd61?
...
The more creative retreat. Black cares not for his queenside development, prefer
ri ng a quick .. .f5 advance while keeping the fianchettoed bishop unobstructed.
However, at club level the more straightforward 6... Nf6 7 e4 d6 is likely to be pre
ferred and is by no means a bad option. At this point White must choose between
t wo main set-ups:
a) 8 c4 Bg7 9 Nc3 0-0 is very similar to a Samisch King's Indian. Black has gained a
tempo, but his queen obstructs his b-pawn and so is likely to have to move again.
Here Hodgson was twice happy to play along the lines of a Modem Benoni with 10
Nge2 (10 Bd3 e6 1 1 Nge2 is the alternative; after 1 l...exd5 White shouldn't recapture
with his c-pawn as he will then struggle to castle, and instead 12 exdS! Nbd7 13 f4
NeB 14 0-0 fS 15 Kh1 Ndf6 16 Ng1 Bd7 17 Nf3 Nc7 led to a tough manoeuvring
struggle in T.Radjabov-A.Grischuk, Bastia (rapid) 2003) 10 ...e6 1 1 Ng3 exd5 12 cxdS
h5 13 Be2 (here so that ...c4+ won't hit the bishop) 13 ...Nbd7 (Diagram 22).
NOTE: In this Samisch-like position White wants to castle and keep

Black's counterplay under control on both flanks. Once that is done
he will look to advance in true anti-Benoni style with f4 and es,
which may even be played as an attacking sacrifice followed by fs.
In a battle between two Trompowsky experts 14 a4 (Hodgson later switched to the
immediate 14 Nfl !? and after 14 ... Re8 15 Qc2 Nh7 1 6 Ne3 NeS 1 7 0-0 g5?! 1 8 Kh1
Qd8 19 Bd2 a6 20 a4 Qf6 21 NfS! Black had been outmanoeuvred in J.Hodgson
N.Berry, Grangemouth 2001) 14 ... a6 15 aS Qc7 16 0-0 ReB (wisely holding up f4) 1 7
Rf2!? Rb8 18 Nfl NeS 19 Ne3 saw White angling t o get i n 2 0 h3 and 21 f4, although
Black retained sufficient counterchances in J.Hodgson-J.Gallagher, Bern (rapid)
1 996.

Black lacks effective counterplay Unusual play even for the Tromp
213
b) Hodgson also experimented with the alternative plan, bringing a knight to c4: 8
Ne2!? Bg7 9 Nd2! 0-0 10 Nc4 Qd8 1 1 Nc3 Na6 12 Be2 Nc7 13 a4 Rb8 14 0-0 b6 15
Bg5 Ba6 16 Qe1 (Diagram 23) gave White control and a pleasant Schmid Benoni
type edge in J.Hodgson-B.Socko, Istanbul Olympiad 2000.
Which approach to prefer is a matter of taste, but unless the reader has some
King's Indian or Modem Benoni experience, he will probably find it easier to
place a knight rather than a pawn on c4.
7 e4
By some margin White's most popular response, challenging Black to come and
do his best. A respectable alternative is the more restrained 7 Nc3 Bg7 8 e3 0-0 9
Nh3 (Diagram 24), refusing to create a target for the black f-pawn. However,
9 ... Na6!? (by no means the only approach: 9 ... Qc7 10 a4 b6 1 1 Nf2 f5 12 Be2 e5 13
e4 f4 also seemed fine for Black in J.Hodgson-M.Tumer, Oxford 1998) 1 0 Be2 c4 is
a bit awkward, targeting the pawn on d5: 11 0-0 Qc5 12 Kh1 b5 13 e4 b4 14 Na4
Qa5 15 c3 bxc3 16 bxc3 Nc5 certainly left White on the defensive in M.Van Gysel
T.Schmidt, correspondence 2004.
7 ... Bg7 8 Nc3
NOTE: Here c3 is the best square for the queen's knight. With Black's
knight controlling the c4-square, bringing it instead to d2 or a3
makes less sense.
8 ...f5 (Diagram 25)

Wh ite cannot really hold e4 Wonderfu l creativity
Black has also delayed this by a move, beginning with 8 ... 0-0. Then 9 Nh3 f5 10
exfS NxfS transposes to our main line, but this way White does gain time to shore
2 14
The Ma i n L i n e : 2 ... Ne4 3 Bf4 c s
up e4 i f he prefers. After 1 0 Nf2 c4 (Black has also tried 10. . .fxe4 when White
should keep the f-file closed with 1 1 Ncxe4! Nxe4 1 2 Nxe4) 1 1 Qe2 eS 12 g4!? f4 1 3
g S Qd8 14 h4 b S a highly-complex struggle arose in M.Comette-R.Picard, French
League 2004; Black is, though, still some way from sorting out his queenside
pieces and I must admit to quite liking White's kingside chances.
9 exfs Nxfs 1o Nh31
White refuses to obstruct his king's bishop and this is quite a sensible approach.
Instead 10 Bd3 0-0 1 1 BxfS Rxf5 1 2 Nge2 c4! kept White's king in the centre and
allowed Black to stir up a fair amount of trouble in M.Adams-B.Gelfand, Cap
D' Agde (rapid) 2003, but 10 g4!? is a fascinating alternative. Notably this was
Hodgson's choice when he first met 6... Nd6 and it was later taken up by another
highly-creative player, Luke McShane: 10 ... Nh6 (otherwise 10 ... Nd4?! 1 1 Ne4! em
barrasses Black's knight, and 10 ... Nd6 1 1 h4!? Nf7 12 Ne4 d6 13 c3 Na6 14 hS! sees
White charging most admirably ahead on the kingside, as given by Wells) 1 1 d6!
e6 (Davies points out that 11 ... Nf7!? 12 dxe7 Bxc3+ 13 bxc3 Qf6 reaches quite an
irrational position, which isn't, indeed, so easy assess, although I quite like
White's dynamic potential) 12 NbS Na6 13 Bf4 Qc6 14 Bxh6!? Bxh6 15 c4 0-0 16 h4!
(Diagram 26} 16 ... Bf4 1 7 hS saw White playing most creatively for a bind right
across the board, but was hardly totally clear in L.McShane-V.Kotronias, Gibraltar
2003.
10 ... 0-0 11 Ne4 (Diagram 27)

Calm, sensible play from White White ta rgets c4
White has good central control here and may well follow up with c3, frustrating
B lack's two active minor pieces. We will see how the play might pan out in Game
25.
215
Conclusion
After 4 dS Qb6 my invariable choice in practice has been 5 Bel . Then 5 ... e6 6 f3 N f6
is pretty critical, as we will see later in this chapter. Fairly respectable too is S ... g6,
against which White has scored 52% from almost 140 games with 6 f3. The mod
ern defence 6... Nd6 is a very intriguing one, but by following our main line White
has decent chances to emerge with the upper hand.
Illustrative Games
Game 25
D A.Stefanova D.Stellwagen
Wij k aan Zee2004
1 d4 Nf6 2 Bg5 c5 3 d5 Ne4 4 Bf4 Qb6 5 Bc1 g6 6 f3 Nd6 7 e4 Bg7 8 Nc3 0-0 9 Nh3 f5
10 exf5 Nxf5 11 Ne4 Na6
It's worth noting that Black can't win a pawn here with 1 l ...Bxb2? because 12 Rb1
QaS+ 13 Bd2 Qxa2 14 c3 traps and wins the bishop. However, Black later sought to
improve with the active 1 l ...d6 12 c3 c4!? in M.Taylor-A.Corke, British Champion
ship, Swansea 2006. Following 13 Nhf2 Ne3 14 Bxe3 Qxe3+ 15 Qe2! (Diagram 28)
Black had obtained the bishop-pair, but at the cost of c4 which eventually fell leav
ing White better after 15 ...Qxe2+ (perhaps Black might consider 15 ...Qh6!?, al
though I'm not sure that there's enough for a pawn after 16 Qxc4 Nd7 1 7 Be2 NeS
18 Qb3 Qe3 19 Qc2) 16 Bxe2 bS 17 a4 Bb7 (White is for choice too after 17 ... b4!? 18
cxb4 Bxb2 19 Rb1 c3 20 Nd3, since 20... Bb7 21 Nxb2 cxb2 22 Rxb2 BxdS 23 Kd2
leaves him with a mobile queenside majority) 18 axbS BxdS 19 Nd2 ReB 20 Ra4
Nd7 21 Nxc4.
12 c3 Nc7
This manoeuvre is quite time-consuming, but 12 ... Nd6 13 Nhf2 would maintain a
knight on e4, as pointed out by Prie, and looks a little better for White too; a
timely a4-a5 might follow.
13 a41 (Diagram 29)
NOTE: White's a-pawn is an important unit in these Benoni-like

structures. Not only is it often used to prevent an early ... b5 ad
vance, but sometimes it can annoy Black's pieces too, especially if
his queenside is a little tangled.
13 ... Ne8 14 Be2 Nf6 15 o-ol
This allows Black to remove the knight from e4, but I dare say that Stefanova
216
The Ma i n Li n e : 2 . . Ne4 3 Bf4 c s
.
didn't consider 15 Nhf2 Nxe4 1 6 Nxe4 Nd6 17 0-0 Nxe4 18 fxe4 Rxfl+ 1 9 Qxfl d6
to be especially promising.
1S . d6
..
White appears to enjoy the upper hand no matter what Black tries here. Indeed,
i t's useful to note that 15 ... c4+ 16 Kh1 Nxe4?! 17 fxe4 Ne3 doesn't force White to
part with the bishop-pair, as 18 Rxf8+ Bxf8 19 Qg1 ! is much better, taking aim at c4
and meeting 19 ... Bh6 with 20 g4!.
16 Bc4 Bd7 17 Qe2
Stellwagen has managed to develop his queenside, but only at the cost of reaching
an inferior form of the Dutch Defence; Black lacks activity to compensate for his
weaknesses down the e-file.
17... Rae8
Perhaps 17 ... Nxe4!? 18 Qxe4 QaS followed by ... Rae8 was the lesser evil.
18 Negsl (Diagram 30)

Ta king control of bS Not the best of Dutch Defences
Stefanova wastes no time exploiting those weaknesses.

tB . esl
..
Rather desperate, but there was nothing better. Indeed, to appreciate just how bad
Black's position already is do consider the line 18 ... h6? 19 Ne6 Bxe6 20 Qxe6+ Kh7
21 Nf4 when a nasty accident is set to occur along the b1-h7 diagonal.
19 dxe6 ds 20 Qd1?!
This doesn't lose the whole advantage, but White would have done much better
with the bold 20 g4! . Stefanova may not have wanted to risk such an advance, but
after 20 ... Nd6 21 aS Qd8 22 Bd3 White retains the key bridgehead on e6 and some
advantage because 22 ... h6?! fails to 23 Qc2 BbS 24 BxbS NxbS 25 Nf7.
217
20... Bc6l
Returning the favour. As pointed out by Prie, Black had to remove the key pawn
with 20 ... Bxe6 when 21 Nxe6 Rxe6 22 Bxd5 Nxd5 23 Qxd5 Rfe8 would have left
Black well centralized and with some play for the pawn.
21 as Qc7 22 Bd3 h6
Easy to criticize, but Black was already rather short of a good move.
23 BxfSI gxfS 24 Nf7 Rxf71l
Stellwagen prefers to give up the exchange rather than see his king's defences de
stroyed by 24 ... Rxe6 25 Nxh6+ Bxh6 26 Bxh6 Rfe8 27 Nf4, and neither would
24 ... Kh7 have saved Black on account of the crushing 25 Bf4 Qc8 26 Qc2! when the
check on g5 indirectly protects the key e-pawn.
25 exf7+ Qxf7 26 Re1 d4 (Diagram 31)

Black is getting desperate White's main choice
A last try to obtain some activity; a quality which Black has lacked all game de
spite his creative early play.
27 cxd4 cxd4 28 Nf4 Kh7 29 Bd2 Rg8 30 Qe2 Nd7
Now White's queen invades, but it also would have done after 30 ... Nd5 31 a6 and
32 Qe6.
31 Qe7 Rf8 32 a61
TIP: It's often good to undermine the defences of your opponent's

pieces. Once less well defended, they become vulnerable to tactics
and being attacked with gain of tempo.
32 ... b6 33 Qd6 NbS 34 Re7 Qf6 35 Qxf6 Rxf6 36 Rxa7 1-0
218
The Ma i n Li ne: 2 . Ne4 3 Bf4 c s
..
Forcing the Pace: 4 f3

1 d4 Nf6 2 Bg5 Ne4 3 Bf4 cs 4 f3 (Diagram 3 2)
White wastes no time driving back the knight and now in comparison with 4 dS
Black loses options, such as that of an early ... e6 and ... Bd6. At this point Black's
main approach is 4 ... Qa5+, as we will see throughout the rest of this chapter, but
first we must consider the immediate, natural-enough retreat of the knight.
4 Nf6
...
I wouldn't go so far as to categorize this as a mistake, but it is surprising just how

many players still employ it without being fully aware of its drawbacks.
5 dxcsll (Diagram 33)

A tempting a lternative to 5 dS White has good compensation
A promising choice and one which should especially appeal to those who are try
ing to avoid Benoni structures; i.e. fans of 4 ... Qa5+ 5 c3 Nf6 6 Nd2, which will be
covered in the final section of this chapter.
Instead 5 dS is the Benoni option when we saw the position after S ... d6 in note 'a'
to Black's 4th move in the first section of this chapter and that after 5 ...e6 in note
'cl' there. Black might also flick in S ...Qb6 when 6 Bel takes play back into our
previous section. However, here I would be sorely tempted by 6 e4!? Qxb2 7 Nd2
(Diagram 34) with good compensation: 7 ... d6? 8 Nc4! Qb4+ 9 c3 Qxc3+ 10 Bd2 Qd4
1 1 Qcl ! forced Black to part with his knight to rescue his errant queen in
A.Weindl-N.Bindzus, Esbjerg 2007, and 7...Qc3 8 Bc7!? d6 9 Ne2 Qe3 10 Nc4 Qh6
1 1 BaS g6 12 Bc3 kicked the black queen offside in M.Leon Hoyos-G.Carreto Nieto,
Toluca 2009 - it's well worth comparing this gambit with the topical 4 ...Qa5+ 5 c3
N f6 6 dS Qb6 7 e4, below.
219
s Qa5+
...
Black often regains his pawn without delay thus, but he might prefer one of:
a) Just like in our main line, White obtains a pleasant form of the Open Sicilian in
the event of 5 ... Na6 6 e4 NxcS 7 Nc3: for example, 7 ... d6 (best; Black's knights were
targeted rather quickly after 7... g6?! 8 Be3! Ne6 9 eS in S.Beshukov-Y.Shulman,
Gausdal 1994) 8 Qd2 eS 9 BbS+ Ncd7 10 BgS! saw White winning the fight for dS
in J.Hodgson-D.Reinderman, Leeuwarden 1993, and after 10 ... a6 simply 1 1 Bxd7+
Bxd7 12 NdS would have been quite pleasant.
b) It's been quite rare, but I suspect that Black's best chance to reach a roughly
level middlegame is 5 ...b6! (Diagram 35), which at least has a couple of fairly
regular grandmaster adherents: 6 e4! (White refuses to accelerate Black's devel
opment with 6 cxb6 Qxb6, preferring to fight for the initiative) 6...bxc5 (Black
might limit White's options with 6... Nc6 when 7 Nc3 bxcS transposes while avoid
ing our next note) 7 Nc3 (theory has suggested that 7 eS Qc7!? is okay for Black,
but Mamedyarov was recently happy to allow this and 7... Ng8 8 Nc3 Nc6 9 BbS
Nd4 10 Nge2!? NxbS 1 1 NxbS QaS+ 12 Nec3 a6 13 Nd6+! exd6 14 exd6 Qb4 15
Qe2+ Kd8 1 6 Bd2 gave him a strong initiative for the piece in S.Mamedyarov
T.Nedev, Dresden Olympiad 2008) and now:
Diagram 3 5 (W) Diagram 36 (B)

Black too ca n ga mbit! Fighting for the initiative
b1) An idea well worth remembering is 7 ... d6?! 8 eS! which, as here, often gives
White a strong initiative. Indeed, after 8 ... dxe5 9 Qxd8+ Kxd8 10 0-0-0+ Nbd7 1 1
BxeS Bb7 1 2 Nh3 Black faced a long defence in M.Carlsen-S.Ganguly, Dubai 2004.
b2) 7... Nc6 8 Bc4 (White continues to develop, but has also tried some even more
direct ideas, including 8 NbS!? d6 9 eS NxeS! 10 BxeS dxeS 1 1 Qxd8+ Kxd8 12
0-0-0+ Nd7 13 Nh3 g6 14 Bc4 which gave him enough for his pawn in I.Schneider
J.Stocek, Pardubice 2008) 8 ... d6 (Hodgson's suggestion; instead unsurprisingly
2 20
The Ma i n L i n e : 2 . N e4 3 Bf4 c s
..
8. . .g6 9 NbS!? d 6 10 e S dxeS 1 1 Qxd8+ Kxd8 led to something of a mess in

J.Hodgson-A.Shirov, Groningen 1996, and now 12 BgS! h6! 13 Bxf6 exf6 14 Bxf7
Rb8 15 0-0-0+ Nd4 would have remained most unclear) 9 eS dxeS (and not, of
course, 9 ... Nxe5? 10 BxeS dxeS? 1 1 Bxf7+!) 10 Qxd8+ Kxd8 1 1 0-0-0+ (Diagram 36)
1 l ...Bd7 12 BbS exf4?! 13 Bxc6 Rc8 14 Bxd7 Nxd7 15 Nh3 favoured White in
G.Kasparov-D.Reinderman, Wijk aan Zee (blitz) 1999. However, Black later im
p roved with 1 1 ...Kc7!? 12 Be3 e6, which kept White at bay in V.lotov-V.Babula,
European Club Cup, Kallithea 2008, and now it's probably best to whisk the final
p ieces into play with 13 Nh3!? h6 14 Nf2.
6 Nc3
Common, but if White wishes to avoid our next note he should consider 6 Qd2!?
when 6 ... Qxc5 7 Nc3 d6 8 e4 transposes and 6...Qxd2+ 7 Nxd2 Na6 8 Nb3 e6 9 Bd6
safeguards the extra pawn.
6...Qxcs
Again standard, but Black might consider 6...e6!? 7 Bd6 NdS when 8 e4 Nxc3 9
Qd2 b6 1 0 Bxf8 Rxf8 1 1 cxb6 axb6 12 Qxc3 Qxc3+ 13 bxc3 Ba6 14 Bd3 Nc6 15 Ne2
gave him some if not quite enough compensation for the pawn in V.Kovacevic
L.Ftacnik, Hastings 1982/83.
7 e4 (Diagram 37)

An improved Sici lian A kingside adva nce beckons
Thus we reach a Sicilian structure, but that shouldn't put off those with little ex
perience of that opening.
NOTE: White is basically a tempo up on the Sicilian as his bishop has

taken two moves to reach f4, whereas Black's knight has needed
three to land up on f6.
221
Moreover, Black's queen is a little out of place and White has the easy plan of cas
tling long followed by making good use of his active pieces and/or advancing his
kingside pawns. Indeed, he has scored pretty well in practice from here.
7 ...d6 8 Qd2 a6
Black must be careful: 8 ... g6?! 9 0-0-0 fails to convince when White can easily trade
the Dragon bishop once Black has castled, but 9 ... Nc6?! 10 NbS! Qb6 1 1 eS! only
made matters worse in M.Olea Perez-R.Cenal Gutierrez, Asturias 1999; and White
also clearly won the opening duel after 8 ... Nbd7?! 9 Be3! Qa5 10 NbS in R.Palliser
R.Browne, Leeds 2005.
9 0-0-0
White can begin too with 9 g4 when 9... h6 10 h4 e5?! 1 1 Be3 Qc7 12 Bh3 Be6 13 gS
Nh5 14 0-0-0 Nd7 15 Nge2 Rc8 16 Bg4 was excellent for him in G.Sargissian-
A. Vidarte Morales, Barcelona 2000, and so Black should try to transpose with
10 ... Nbd7 1 1 0-0-0.
9 . Nbd7 (Diagram 38) 10 g41
..
White wastes no time commencing his pawnstorm; developing the kingside pieces
can wait and, besides, it's not yet entirely clear where they will be best placed.
10 h6
...
Here White has a pleasant choice between 11 Nge2 and 1 1 h4!?, and we'll see the
latter in action in Game 26.
Conclusion
4 f3 is by some margin White's most popular choice after 3 ... c5, and we've seen
plenty of evidence why in this section. How to meet 4 ... Nf6 is both a matter of
taste and affected by how one likes to meet the related 4 ... Qa5+ 5 c3 Nf6. With 5 dS
White has good chances to obtain an edge in a Benoni structure, but 5 dxc5 is
pretty tempting and has made 58% from over 400 games.
Illustrative Ga mes
Game 2 6
D J.Gallagher T.Nedev
E u ropea n Club Cup, St Vi ncent 2005
1 d4 Nf6 2 Bgs Ne4 3 Bf4 cs 4 f3 Nf6 5 dxcs QaS+ 6 Qd2 Qxcs 7 e4 d6 8 Nc3 a6 9
0-0-0 Nbd7 10 g4 h6 11 h41? bS
As White has delayed the development of his king's knight, 1 1 .. .Ne5!? may well
222
T h e M a i n L i n e : 2 ... Ne4 3 Bf4 c s
be more critical (after 1 1 Nge2 Ne5 this could just be met by 12 Nd4). The game
A.Moiseenko-E.Gersov, Kapuskasing 2003, continued 12 Kb1 Be6 13 Nd5 (13 Be3!?
Qc7 14 Qf2 b5 15 Nge2 is a reasonable alternative) 13 ...Bxd5 14 exd5 g6 15 Be3 Qc7
16 Be2 Bg7 1 7 Nh3 and while White was slightly for choice, a tough fight lay
ahead.
12 Be3 Qc7 13 Nh31 (Diagram 39)

Black is a l ready struggling A killer fork
One might have expected the knight to be en route to e2 and then perhaps later d4
or g3, but Gallagher's choice is much better: White is determined to force through
g5 and be able to recapture there if needed with his h-pawn.
13 ... Nb6?
The Macedonian Grandmaster frees the d7-square for his f6-knight, but misses a
nasty tactic. Black would have done better with 13 ... Ne5 14 Be2 e6 when Postny
proposes 15 g5 (I would be tempted too by 15 Nf4!?, intending 15 ...b4 16 Na4 Rb8
1 7 g5 again with pressure) 15 ...hxg5 16 hxg5 Nfd7 1 7 Nf4 Rxh1 18 Rxh1 Bb7, keep
ing White's advantage within check.
14 esl b4?
TIP: Often a good way to make use of a lead in development and/or

the initiative is to overload the opponent's pieces.
Here Gallagher has spotted that 14 ... dxe5? 15 Nxb5! axb5 16 Bxb6 Qxb6 17 Bxb5+
overloads the black queen, which must stay in touch with d8 and so the attack
down the d-file is utterly crushing. However, Nedev's choice is hardly ideal and I
suspect that Black should have tried 14 ... Nfd7!?, unpleasant though this would
have been after 15 exd6 Qxd6 16 Qel Qc6 17 Bd3.
223
Sta rt i n g Out: The T rom pows ky Attack
15 exf6 bxc3 16 Qd4 (Diagram 40)

It just gets worse for Black: the fork will win a piece.
16 ...Qcs 17 b41 es 18 bxcs exd4 19 Bxd4
There's no defence and Black might well have given up without torturing himself
further.
19 ... Na4 20 fxg7 Bxg7 21 Bxg7 Rg8 22 Bf6 Be6 23 Rxd6 Rg6 24 Bes Bxa2 25 Rxg6
fxg6 26 Bd3 1-0
An I m porta nt Check: 4 f3 Qa S+
1 d4 Nf6 2 BgS Ne4 3 Bf4 c s 4 f3 QaS+ (Diagram 41)

More critical than 4 ... Nf6 Again White heads for a Benoni
P layed in over 75% of the 3,500 games which have reached this position on my
database, but why does Black so often flick in this check?
NOTE: By checking on as Black reduces White's options. It may no

longer be so easy to support the dS-pawn with c4 and in the event of
... Qb6, White may think twice before sacrificing his b-pawn since the
c3-pawn will usually have to follow.
5 c3 Nf6 6 dS (Diagram 42)
Very natural. White plays by analogy with the immediate 4 d5 and angles to ob
tain a slightly favourable Benoni position. There is a very important alternative,
though, in our final section's 6 Nd2!?, and while researching this work I was
224
The Ma i n L i n e : 2 . . Ne4 3 Bf4 c s
.
amazed to find that it has actually become slightly the more popular choice (ac
counting for 53% of all games after S ...Nf6).
6 Qb6
...
Black moves his queen yet again, but the attack on b2 is not so easy for White to
meet. However, the text, while critical, is by no means the only approach:
a) 6... d6 7 e4 g6 (or 6...g6 7 e4 d6) sees Black happy to head for Benoni waters.
Unlike earlier in the chapter, White doesn't have the option of c4 and Nc3 here,
but the pawn on c3 does a useful job of blunting the fianchettoed prelate and Nc4
will come with tempo:
a1) With 8 a4 White prevents ... bS in response to Nc4 and after 8 ...Bg7 9 Na3 (Dia-
gram 43) 9...0-0 (Black's blockading plan fared better in P.Wells-A.Moen, Hastings
2002/03: 9 ... Nbd7 10 Nc4 Qc7 1 1 Nh3 0-0 12 Qd2 NeS!? 13 Nf2 Rd8 14 BgS Rb8,
although now White might have let Black do his worst with 15 NxeS!? dxeS 16 Bc4
Ne8 1 7 0-0 Nd6 and then played to arrange f4 after 18 Qe2) 10 Bd3!? Nbd7 1 1 Nc4
Qd8 12 Ne2 (a reasonable alternative to the more common Be2 and Nh3-f2 set-up)
12 ... b6 (slow, but Black lacks counterplay and 12 ... Nh5 13 BgS NeS 14 NxeS BxeS
15 0-0 would also have been fairly pleasant for White, as would have 12 ... Nb6 13
Nd2) 13 0-0 a6 14 Be3 NeS?! 15 NxeS dxeS 16 aS! broke up Black's queenside with
some advantage in M.Gurevich-V.Sanal, Antalya 2009.

The knight is en route for c4 Central play from Black
NOTE: Black's panicking 14th wasn't the best, but White already had
an ideal Schmid Benoni scenario: control of the position and options
on both flanks in b4 and f4 (after a preliminary h3).
a2) That was just one of several possible move orders: for instance, 8 Na3 Bg7 9
Nc4 will transpose to our last variation if Black's queen goes to c7. However, it
225
doesn't have to and here 9...Qd8 10 a4 0-0 11 Qd2 Re8 12 Be2 Nbd7 13 g4!? Nb6 14
Ne3 Bd7 1S aS! Nc8 16 Nh3 e6 wasn't all that clear in I.Miladinovic-D.Milanovic,
Bar 2008, but White doesn't have to play in such committal vein on the kingside
and the more restrained 1 l ...b6 12 Ne2 Na6 13 Ng3 Nc7 14 Be2 hS 1S 0-0 Rb8 16
Rfd1 a6 1 7 Ra3! h4 18 Nh1 bS?! 1 9 axbS axbS 20 NaS saw him nicely keeping con
trol in J.Speelman-B.Bratovic, Bled Olympiad 2002.
b) A respectable alternative to our main line is 6 ... e6!? when 7 e4 (Diagram 44)
leads to a further divide:
b1) It's too late for 7 ... Qb6? on account of 8 Na3! because 8 ... exdS (similar play re
sults from 8 ... Qxb2 9 NbS! when Davies points out that 9 . . Na6 10 Rb1 !? Qxa2 1 1
.
Ra1 Qb2 12 Rxa6! is strong) 9 exdS Qxb2 (Black might a s well grab the exchange:
9... NxdS 10 QxdS Qxb2 1 1 Rcl Qxa3 12 Bc4 is just plain miserable for him) 10 NbS
NxdS 1 1 QxdS Qxa1+ 12 Kf2 (Diagram 45) gives White a rather powerful initiative
for the exchange. The game I.Ivanisevic-B.Bok, Biel 2008, continued 12 ... Qb2+ 13
Kg3! Be7 and now it was time to stop being flashy and just recover some material
while retaining a clear advantage with 14 Nc7+ Kd8 1S Nxa8 Qxc3 16 Nh3.

That's a strong initiative! Ta rgeting d6
b2) Black also does well to steer clear of 7...bS?! 8 a4! b4 9 Nd2! bxc3 10 Nc4, which
gives White a pleasant advantage if the queens are exchanged and 10 ...Qb4 1 1
bxc3 ( 1 1 Kf2!? is highly tempting too) 1 1 ...Qxc3+ 1 2 Bd2 Qd4 1 3 Qcl forced Black
to part with a piece for insufficient compensation in A.Stefanova-K.Nadig, Gibral
tar 2006.
b3) Better is 7... d6 when 8 dxe6!? (White plays against the backward d-pawn, al
though he can also go 8 Qd2, intending to transpose to variation 'b4' after 8 ... exdS
9 exdS) 8 ... Bxe6 9 Na3 is a critical response. Black countered in kind in
I.Miladinovic-M.Pavlovic, Serbian Team Championship 2008, with 9... Nc6 (9 ... d5
226
T h e Ma i n L i n e : 2 . .. Ne4 3 Bf4 c s
1 0 Bxb8!? Rxb8 1 1 Bb5+ Ke7 1 2 e5 Nd7 13 f4 wasn't entirely clear i n L.Fressinet

L.Dominguez, Wijk aan Zee 2004, although I would prefer to take White here) 1 0
Nc4 (Diagram 46) 1 0. . .Qd8!? (the meek 10. . .Bxc4?! 1 1 Bxc4 Be7 12 Qb3!? had earlier
left White doing well on the light squares in P.Wells-H.Hamdouchi, Pulvermuehle
2000) 1 1 Nxd6+ Bxd6 12 Bxd6 (far from forced; indeed, I suspect that White should
consider 12 Qxd6!? Qb6 13 b3 c4 14 b4 when the onus is on Black to demonstrate
sufficient compensation) 12 ...Qb6 13 Qcl 0-0-0 14 Bg3 Rd7 15 Be2 Rhd8 and Black
had enough for his pawn with White a little tied up.
b4) Black's main choice is 7 ... exd5 8 exd5 (much sharper is 8 e5!? when 8 ... Ng8 9
Qxd5 Ne7 10 Qe4 Ng6 1 1 Bg3 Qb6 12 b3 gave White a pull in J.Hodgson-Z.Ilincic,
Istanbul Olympiad 2000, but 8 ... Nh5 9 Bel Qd8! is much more critical; White be
gan to get on top after 10 Be3!? d4 1 1 cxd4 cxd4 12 Qxd4 Nc6 13 Qe4 Qa5+ 14 Kf2!
g6 15 Nc3 in R.Palliser-C.Doran, Doncaster 2005, but both 10 ...Qb6!? and Wells's
10 ... Nc6!? may improve) 8 ... d6 (Diagram 47) when the critical test is probably 9
Qd2!? (9 Qe2+ is rather well met by 9. . . Be7 1 0 Bxd6 Nxd5, against which White has
failed to prove any advantage; indeed, he may have nothing better than 1 1 Bxe7
Nxe7 12 Qb5+, but this is hardly inspiring) 9. . .Be7 10 c4 Qxd2+ (or 10 ... Qd8 1 1 Nc3
0-0 12 Bd3 Nh5 13 Be3 Nd7 14 Nge2 Ne5 15 Ng3 Nf6 1 6 Be2 a6 1 7 a4 with an edge
in V.Moskalenko-M.Pena Gomez, Can Picafort 2008; White has managed to keep
control and later arranged f4) 1 1 Kxd2!. White enjoys more space here and it's not
so easy for Black to take control of the dark squares:

There's nothing down the e-file White may expand on the kingside
b41 ) 1 l ...Nh5 12 Be3 f5 13 Nc3 (Diagram 48) 13 ... Nd7 (Hodgson had previously
faced 13 ... 0-0?! when 14 Bd3 Nd7 15 f4! Ndf6 16 Nge2 saw him taking control in
J.Hodgson-P.Wells, York 2000) 14 Re1 Kf7 15 Nh3 (15 f4 is no longer so effective
on account of 15 ... Ndf6 intending 16 ... Ng4) 15 . . . Ne5 was the course of }.Hodgson-
227
L.Schandorff, German League 2001, and now White can secure an edge with 16
Be2! Rf8 (Black might do better with 16...h6 1 7 f4 Ng4 18 Nf2 Nhf6 19 Nxg4 Nxg4
20 Bg1, although this is still more pleasant for White) 1 7 f4 Ng4 18 Ng5+! Bxg5 19
fxg5 (Wells).
b42) White is slightly for choice too after 1 l ...b5!? 12 Nc3 bxc4 13 Re1 Kd8 14 Bxc4
Nbd7 15 b3, J.Salimanki-L.Schandorff, European Club Cup, Saint Vincent 2005.
After another large but important note, we return to 6 ...Qb6:
7 e41? (Diagram 49)

A topical ga mbit The older a pproach
A remarkable idea and one which is currently pretty fashionable.
WARNING: Unlike after 4 d5 Qb6, White cannot move his queen

without dropping material and 7 Qd2? Nxd5 8 Qxd5 Qxb2 9 Qb3
Qxa1 10 e4 Nc6 11 Nh3 e6 is unconvincing since Black is all set to
free his queen with ...c4.
However, the risky if highly interesting 7 e4 isn't forced and the reader may well
prefer one of:
a) To those who have read the coverage of 4 d5, 7 Bel !? shouldn't come as a sur
prise. Moreover, this popular retreat is important enough to warrant its own sec
tion and we will tum to it next.
b) 7 b3 (Diagram 50) hasn't been too popular of late. It's hardly a bad choice, al
though practice suggests that Black is okay in the tabiya arising after 7...e6 (Black
might also head for a Benoni formation, happy to have provoked b3: for example,
7... g6 8 e4 d6 9 Qd2 Bg7 10 Bh6 0-0 1 1 Bd3 e6! 12 Bxg7 Kxg7 was about equal in
A .Stefanova-M.Choisy, Evry 2003) 8 e4 (8 c4 exdS 9 cxdS c4! gives Black a decent
228
The Ma i n L i n e : 2 . . . Ne4 3 Bf4 c s
version of a gambit we'll see more in our next section) S...exd5 9 exd5 Bd6 1 0 Bg5:
b1 ) White is hoping for 10 ...0-0?! 11 Bxf6! ReS+ 12 Ne2 gxf6 13 Nd2 when Black's
w recked kingside turns out to be of more importance than his extra dark-squared
bishop: 13 ... Qc7?! (better to defend with 13 ... BfS) 14 Ne4 Be5 15 Qd2 KhS? 16 Qh6
Qb6 1 7 d6! led to a decisive attack while Black's queenside slept in R.Palliser
J.Rudd, British Championship, Scarborough 2004.
b2) However, Black does much better with 10 ... Be7! (Diagram 51) and then:

Black keeps h is structure intact How much com pensation?
b21 ) 1 1 c4 has been fairly common, but White lacks a particularly good answer to
1 l ...Qd6!, threatening 1 2 ... Qe5: for example, 12 Qe2 Nc6!? 13 Nc3 Nd4 14 Qd2
Qe5+ 15 Be3 Nc2+ 16 Qxc2 Qxe3+ 17 Nge2 0-0 was fine for Black in G.Evans-C.Van
Dijk, correspondence 2002.
b22) 11 Bc4 d6 12 Ne2 Nbd7 13 0-0 0-0 14 a4 QdS 15 Bf4 Nh5 16 Be3 Ne5 1 7 Na3 f5
was fine too for Black in A.Smirnov-V.Yemelin, St Petersburg 2007.
b23) Hodgson once even tried 1 1 d6!? Qxd6 12 Qxd6 Bxd6 13 Bxf6 gxf6 when the
issue is what is going on after something like 14 Bc4 Be7 15 Na3 (S.Dishman
A.Mack, British League 2001) 15 ... Nc6 (Diagram 52). Black's structure is a mess,
but he can untangle his queenside after ... a6 and while quite an interesting strug
gle lies ahead, I would be surprised if White had more than enough compensa
tion.
b3) 11 Ne2 Qd6!? (more critical than simple development with 1 1 ...0-0 12 c4 ReS 13
Nbc3 d6) 12 Na3 Qxd5 13 Qxd5! Nxd5 14 Bxe7 Kxe7 15 0-0-0 gives White some
initiative for his pawn, although the position remained unclear after 15 ... Nc7 16
Nf4 d6 1 7 Nc4 Rd8 18 Bd3 g6 19 Rhe1+ Be6 20 Be4 Nc6 21 Bxc6 bxc6 22 Na5 Rd7 23
Nxc6+ Kf6 in D.Navara-L.Ftacnik, Czech League 2008.
229
b4) The other line with some current adherents is 1 1 Na3 d6 (Black is yet to grab
and probably with good reason: 1 1 ...Qa5 12 Nc4 Qxc3+ 13 Kf2! Ne4+ 14 fxe4 Bxg5
15 Nf3 looks like excellent value for a pawn, as analysed by Davies) 12 Nc4, but
12 ... Qc7 (Mark Hebden's opening choices are always well worth keeping an eye
on and here his 12 ...Qd8!? 13 a4 0-0 14 Ne2 Na6! may be an even better way to sort
out the black queenside) 13 a4 0-0 14 Ne2 Re8 15 Qd2 Nbd7 (Diagram 53) followed
by ... Nb6 was fine for Black in A.Almeida Saenz-S.Pla, Aguascalientes 2008.

It's a bout even Chernyshov's point
Returning to the fashionable 7 e4:

7 ...Qxb2
Black doesn't have to accept, but White is happy to see 7 ... e6?! 8 Na3!, as we saw
above, and I suspect that 7... d6 can also be met by 8 Na3!?.
8 Nd2 Qxc3 9 Bc71 (Diagram 54)
It had long been thought that this gambit was much less effective once the c-pawn
had been forced forward, but then the Voronezh Grandmaster Konstantin
Chemyshov discovered this clever idea, cutting off the black queen's retreat. It is
still too early to stay whether the gambit is totally sound, but it certainly packs a
big surprise punch and isn't so easy to meet:
a) 9 ... g6 10 Rcl (White has also tried 10 Ne2 Qd3 1 1 Rcl, but 1 l ...Bh6! 12 Rxc5 Na6
13 Rc3 Qxd2+ 14 Qxd2 Bxd2+ 15 Kxd2 d6 16 Nd4 Nc5 left him struggling to dem
onstrate full compensation in X.Stockman-R.Macayo, correspondence 2007)
10 ...Qe3+ 1 1 Ne2 Na6 (hastening the knight to b4 is most certainly better than
1 1 ...Bh6??, which left the queen out of squares after 12 Nc4 Qg5 13 h4 Qh5 14 Ng3
in B.Steiner-E.Karavade, Pardubice 2005; Black later preferred 1 1 ...d6!?, but 12 Nc4
Qh6 1 3 Qb3 Nbd7 14 BaS! g5 15 Bd2 Qh4+ 16 g3 Qh5 17 Bg2 Rb8 1 8 a4 kept him
230
T h e Ma i n l i n e : 2 . . . Ne4 3 Bf4 c s
rather bottled u p and White with full compensation in H.Walsh-P.Chopin, corre

spondence 2006) 12 Nc4 Qh6 1 3 Bf4 (Diagram 55) 13 ...Qg7? (as pointed out by Prie,
Black had to try 13 ... g5! even if 14 Be3 b5 15 Na3 would have retained good com
pensation) 14 Qa4! g5 15 BeS g4 16 f4 Qg6 1 7 Ng3 Rg8 1 8 Ne3 left Black in huge
trouble both on the queenside and in terms of overall development in
K.Chemyshov-A.Grishuk, Russian Team Championship 2005.
Diagram 55 (B) Diagram 56 (W}

The queen remains a liability White controls the centre
NOTE: Rarely does one see a player of Grischuk's calibre outplayed

so quickly. Black can do better against Chernyshov's gambit, but in
pradice his task is clearly not an easy one.
b) For a while Black pinned his hopes on 9 . .b6?!, but 10 Rcl ! retains dangerous
.
compensation as we'll see in Game 28.

c) 9 . Na6!? 10 Bxa6 bxa6 1 1 Ne2 (1 1 Rb1!? d6 12 Ne2 may be a better try) 1 1 ...Qd3
. .
12 Rb1 aS! at least saved the queen in N.Sulava-A.Suarez Real, French League
2008, although 13 0-0 g6 14 Rb3 Qa6 15 Nc3 still retained reasonable compensa
tion.
d) Another rare defence in need of further testing is 9 .. e6 10 Ne2 Qa3, after which
.
1 1 Nb1 !? (unexpected and not really necessary; indeed, Komarov's idea of 1 1 Rcl
exdS 12 Nc3! Na6 13 BeS Nb4 14 exdS NbxdS 15 NxdS NxdS 16 Nc4 Qb4+ 17 Kf2
looks like a promising alternative) 1 1 ...Qb4+ 12 Nbc3 c4 13 Rb1 QcS 14 Bg3 Na6 15
Nd4 sent a knight towards bS and retained compensation in I.Miladinovic
V.Sikula, Nancy 2008.
e) 9 . Qe3+ 10 Ne2 Na6 (far from forced: 10 ... d6 will transpose to variation 'fl ',
. .
Bates's 10 ...b5 should be met by Boel's 1 1 Rb1, and 10 ...Qh6 1 1 Bf4 gS 12 Be3 Qg6
13 BxcS d6 14 Be3 Bg7 1 5 Nd4 0-0 16 Rcl .Kh8 1 7 g4! left White doing quite well in
231
E.Brondum-B.Lindberg, Helsingor 2008) 1 1 Nc4 Qh6 (Diagram 56) 12 BaS!? (Ko

marov points out that 12 Bf4 gS 13 Be3 bS 14 Na3 b4 1S Nc4 d6 16 h4 also offers
good compensation) 12 ... gS (12 ...bS!? 13 Bd2 gS 14 Ne3 e6 might well be a more
critical test) 13 Ng3 d6 14 Ne3 b6 1S Bc3 Nc7 16 eS dxeS 1 7 BxeS saw White open
ing the centre to his advantage in D.Komarov-M.Ahn, Eupen 2006.
f) Black's main choice has been 9 ... d6 (Diagram 57) when White has two options:

A critical defence Still quite unclear
f1 ) 10 Ne2 Qe3 (Black should avoid 10 ...QeS? 11 g4! Nxg4!? 12 fxg4 Bxg4 13 Rb1
Nd7 14 Qa4!, which didn't give him quite enough for his piece in F.Volkmann
J.Czakon, Aschach an der Donau 2006, but 10 ... Qd3!? deserves serious attention,
since 1 1 Rb1 b6 12 Ng3 Qe3+ 13 Be2 Bd7 14 Qc2 g6 1S Nc4 Qf4 16 Bd3 Bg7 while
messy saw White coming up a little short in A.Smimov-K.Maslak, St Petersburg
2007) 1 1 Nc4 Qh6 (White definitely had enough compensation after 1 1 ...QgS 12
BaS! Na6 13 Bd2 Qg6 14 Nf4 QgS 1S Rb1 in M.Cebalo-F.Guido, Bratto 2006) 12 BaS
(White should save his bishop: he preferred 12 Bxb8? Rxb8 13 Qa4+ Nd7 14 Qxa7
in A.Aleksandrov-A.Danin, Minsk 2006, but wouldn't have had enough had Black
untangled with Prie's 14 ... bS! 1S NaS g6) 12 ... gS 13 Bd2 Qh4+! (preventing the pos
sibility of a Ng3-fS manoeuvre) 14 g3 QhS 1S Bg2 Bh6 (Diagram 58) was the course
of I.Ivanisevic-l.Nataf, French League 2007, and now 16 0-0!? Nbd7 1 7 g4 Qg6 18
Ng3 Nb6 19 Ne3 Bg7 20 NefS hS is unclear according to Nataf.
f2) 10 Rb1!? facilitates ideas of Rb3 and Qa4+. The key line appears to run
10 ... Nfd7! (paradoxical, but Black wants to free the f6-square for his queen; alter
natively, 10 ...b6?! 11 Qa4+! Bd7 12 BbS QaS? 13 QxaS bxaS 14 Bc4 cost Black a piece
in L.Trent-V.Kotronias, Liverpool 2008, and 10 ...Qe3+ 1 1 Ne2 Na6 1 2 Nc4 Qh6 13
BaS! gS 14 Ng3 g4 15 Rxb7!! Bxb7? 16 NfS and 1-0 was the brutal course of
I.Miladinovic-Y.Gozzoli, Nancy 2008: 16 ...Qg6 17 Ncxd6+ exd6 1 8 BbS+ leads to
2 3_2
The M a i n L i n e : 2 . . . Ne4 3 Bf4 c s
mate) 1 1 Bb5! (seemingly a better try than 1 1 Qa4 g6 12 Ne2 Qd3! 1 3 BaS Bh6 14 f4
0-0 15 Rb3 b5!, which left Black doing very well in I.Miladinovic-M.Vachier La
grave, European Championship, Plovdiv 2008) 1 l ...g6 (1l ...a6!? also deserves close
scrutiny, since 12 Ne2 Qf6 13 Qa4 g6 14 Nc4 axb5! 15 Qxa8 bxc4 16 Bxb8 Bh6 1 7
Bxd6 0-0 18 Bg3 c3 gave Black enough for the exchange in K.Zhdanenia-O.Ornek,
correspondence 2006) 12 Ne2 Qf6 (Diagram 59) 13 Nc4 Bg7 14 e5! Qg5 15 0-0!?
dxe5 16 f4 Qh5 1 7 Rb3, which left White enjoying a strong initiative for the three
pawns in M.Bosiocic-N.Fercec, Croatian Team Championship 2008.

White can open the centre Will the queen escape?
Conclusion
This has been a lengthy section, but the number of rich ideas at White's disposal
illustrates why to my mind 3 Bf4 is no less fun than 3 h4 and 3 Bh4, not to mention
being much sounder! Most opponents nowadays flick in 4 ...Qa5+ when the double
gambit 5 c3 Nf6 6 d5 Qb6 7 e4 leads to fascinating complications, and might well
turn out to be a more promising path than the older 7 b3. White's most popular
option, though, remains 7 Bel to which we will turn in our next section.
Illustrative Ga mes
Game 2 7
0 M.Pavlovic A.Aieksan d rov
Vrnjacka Banja 2005
1 d4 Nf6 2 Bgs Ne4 3 Bf4 cs 4 f3 Qas+ s c3 Nf6 6 ds Qb6 7 e4 Qxb2 8 Nd2 Qxc3 9
233
Bc7 b6ll
Black regains control of the aS-square and has ideas of trapping the bishop on c7,
but White is in time to rescue the bishop while troubling the black queen.
10 Rc11 (Diagram 60)
WARNING: White must chase the queen this way: 10 Ne2ll Qas 11
1o...Qas
Bes bs enabled it to escape in C.Hanley-M.S andu, Montpellier 2006.
Black preferred to go the other way with 10 ...Qe3+ in J.Gamett-M.Ozanne, Doug

las 2005, but after 1 1 Ne2 d6 12 Nc4 Qg5 White could have seized a strong initia
tive with 13 Qa4+! b5 (13... Bd7 14 Qa3! is also awkward) 14 Qxb5+ Nfd7 15 Rb1 !
followed by 16 Qa4, as pointed out by Peter Boel in New in Chess Yearbook.
11 Bes Ba6
Relatively best: 1 l ...d6? 1 2 Bc3 Qxa2 (or 12 ...Qa3 13 Nc4 Qxa2 14 Ra1) 13 Bc4 Qa3
14 Nb1 costs Black his queen, and 1 l ...b5? 12 Rxc5 Na6 13 Rxb5 Qd8 14 Bc3 was
far from ideal either for him in S.Emst-K.Van der Weide, Netherlands 2007.
12 Bc3 Qa3
The black queen remains in trouble after this, but the desperate 12 ... Bxfl ? 13 Bxa5
Bxg2 14 Bc3 Bxh1 15 Qa4 was hardly a remedy in V.Bhat-M.Panchanathan, US
online league 2006, since the bishop was trapped on hl.
13 Nc4 Bxc4 14 Bxc4 d6 15 Bd2 Qb2ll
Black's desire to rescue his queen is quite understandable, but I suspect that he
does better with 15 ... a6!?. That said, 16 Rc3 Qb2 17 Ne2 (17 a4!? is the alternative)
17 ... b5 18 Bb3 Qa3 19 Nd4! g6! (White's lovely point is that 19 ... cxd4? 20 Rc8+ Kd7
21 Qc2 Ra7 22 e5! would have enabled him to invade the black camp) 20 e5!? still
left White with dangerous compensation in L.Trent-M.Hebden, Southend 2006.
16 Ne2 Qe5 17 BbS+ Kd8
White's position would also have been both dangerous and easy to handle after
17 ... Nbd7 18 Qa4 Rd8 19 0-0.
18 o-o g6 19 Bf41 (Diagram 61)
Yet another move with this piece, but now a central breakthrough with e5 begins
to loom large.
19...QhS 20 Ng3 Qh4 21 Qd2 h6 22 eSI
q;
TIP: When you have much the more co-ordinated forces, it's often a
good idea to look for ways to attack.
22 ... Ne8
Easy to criticize, but Black would have done extremely well to survive too after
22 dxe5!? 23 BxeS Nbd7 when White might even elect to pick up the queen with
...
234
The M a i n L i n e : 2 . . Ne4 3 Bf4 c s
.
24 Rc4 Qxc4 25 Bxc4 Nxe5 before going 26 d6!.

23 Ne4 fs 24 Bg3 QhS (Diagram 62) 25 Nxcsl
A crushing sacrifice to open lines against the black king.

No respite for Black How to breakthrough?
25 ... dxcs
White would have won more easily after 25 ...bxc5 26 Qa5+ Nc7 27 exd6 exd6 28
Rxc5 Rh7 29 Rxc7! Rxc7 30 Rcl (Boel).
26 Bc6 Nc7 27 BxaB NxaB 28 e61
There's no rest for poor Aleksandrov.
28 ... Nc71 29 Rxcsl
A neat tactic to conclude proceedings.
29 ... bxcs 30 Qb21 Nba6 31 QxhB
White could have forced mate with 31 Qb7!, but Pavlovic's choice is plenty good
enough.
3 1...gs
Neither would 3l...Ke8 have helped on account of the lethal return of the queen:
32 Bxc7 Nxc7 33 Qb2.
3 2 QxfB+ QeB 33 Qxfs 1-0
Retreating to c1
1 d 4 Nf6 2 Bgs Ne4 3 Bf4 c s 4 f 3 QaS+ 5 c 3 Nf6 6 ds Qb6 7 Bell? (Diagram 63)
235

Simply defending b2 White has control
White refuses to weaken his queenside with 7 b3 and argues, just like the closely
related 4 d5 Qb6 5 Bel, that the black queen is slightly misplaced for a Benoni
structure.
7 . e6
. .
Black's usual choice, beginning immediate counterplay, although once again 7 g6 ...
8 e4 d6 is a viable alternative when White has his usual choice of set-ups: 9 c4

(with the bishop back on c1 playing in Schmid Benoni style is no longer so effec
tive: for example, 9 a4 Bg7 10 Bb5+!? Nbd7 1 1 Nd2 0-0 12 Nc4 Qc7 13 aS a6 14 Ba4
Rb8 15 Ne2 b5 16 axb6 Nxb6 17 Nxb6 Rxb6 18 0-0 e6 saw Black equalize in
P.Wells-J.Gallagher, Hastings 2001/02) 9 ... e6 (Black doesn't have to transform the
central structure straight away, but 9 ...Bg7 1 0 Ne2 0-0 1 1 Nec3 e6 1 2 Be2 Nbd7 13
0-0 ReB 14 Kh1 a6 1 5 Nd2 Qd8 16 dxe6! fxe6 1 7 f4 b6 18 Nf3 saw him punished for
dallying too long in J.Rowson-A.Ziegler, Porto Mannu 2007) 1 0 Ne2 exd5 1 1 cxd5
Bg7 1 2 Nec3 0-0 13 Be2 a6 14 a4 Nbd7 1 5 Na3 Qd8 16 0-0 Rb8 1 7 Be3 (Diagram 64)
1 7... Ne8 18 Qd2 Nc7 19 Rab1 ! Qe7 20 b4 supplied a model Samisch handling in
H.Nakamura-K.Lie, Gjovik (rapid) 2009.
8 c4
NOTE: After 8 c4 we've reached a position which often comes about

too via a 4 d5 Qb6 5 Bel e6 6 f3 Nf6 7 c4 move order.
As well as 7 e4, the variation 8 e4!? exd5 9 exd5 (Diagram 65) has become popular
of late. In contrast to the 7 b3 variation, White hasn't weakened his queenside and
9 ... Bd6 won't come with tempo, but White is completely undeveloped! Despite
that factor, this is actually a reasonable bid for the advantage:
236
The M a i n L i n e : 2 ... N e4 3 Bf4 c s

A modern development Black lacks counterplay
a) 9 ... d6 10 c4 (a key feature; White wants to develop with Nc3, Bd3 and Nge2)
10 ... g6!? 1 1 Qe2+ KdS 1 2 Qc2 Bg7 13 Bd3 Na6 14 a3 ReS+ 15 Ne2 Ke7 16 Nbc3 KfS
17 0-0 Nb4 1S axb4 cxb4+ 19 Kh1 bxc3 20 Nxc3 gave White a small edge in
S.Movsesian-V.Kotronias, Moscow 2007.
b) 9...Be7 1 0 c4 0-0 1 1 Bd3 ReS 1 2 Ne2 d6 1 3 Nbc3 Nbd7 14 b3! (ensuring that
White can keep his bishop in the event of ...Ne5) 14 ...BfS 15 0-0 g6 16 h3! Bg7 1 7
Bd2 QdS 1S f4! (Diagram 66) 1S ...a 6 19 a 4 NfS 2 0 Rf3 b6 21 Qc2 h6 2 2 Rafl Ra7 23
g4 was a model display from White in V.Laznicka-A.Zubarev, Polanica Zdroj
2006.
NOTE: Due to the symmetrical pawn structure, Black can easily find
it hard to develop counterplay in this line. Compared to a Modern
Benoni structure (with White having d- and e-pawns), the ... bs
advance is both harder to arrange and generally carries less punch.
c) Perhaps Black does best with 9 ... Bd6!? and 10 Na3 (10 c4 0-0 1 1 Nc3 ReS+ dis
rupts White's development) 10 ... 0-0 11 Bd3 ReS+ 12 Ne2 BfS 13 Nc4 Qd8 14 Ne3 d6
15 0-0 g6 was very solid for him, albeit still not quite equal in I.Ivanisevic
V.Kotronias, Kavala 2007.
8 . exd5
. .
Black opens the centre in a bid to obtain immediate counterplay, but he has also
tried:
a) 8 .. d6 9 e4 can easily transpose to the note to Black's 7th move, such as after
.
9...exd5 10 cxd5 g6 1 1 Ne2, and the more passive 9 ... Be7 10 Nc3 exd5 1 1 exd5!? 0-0
12 Bd3 Nbd7 13 f4 Re8 14 Nge2 also gave White a pull in V.Yemelin-D.Kovachev,
237
Sta rti n g Out: The Trom powsky Atta ck
European Club Cup, Kallithea 2008, because 14 ... Ng4 15 0-0 Bf8 16 Ng3 Ne3? runs
into 1 7 Bxh7+! Kxh7 18 Qd3+.
b) Shipov's 8 ... Bd6!? (Diagram 67) is a much more independent approach. It's
never really caught on, but 9 e4 (9 Nc3 0-0 10 e4 Re8 1 1 Nge2 is a sensible alterna
tive) 9 ... Be5 (best, as 9 ...Qc7 is well met by 10 Ne2) 10 Nd2 0-0 1 1 Ne2 Na6 12 a3 d6
13 g4!? Bd7 14 Bg2 exd5 15 exd5 Bd4! 16 Nxd4 cxd4 17 0-0 d3+ 18 I<h1 Rae8 re
mained quite messy and unclear in H.Kruse-V.Grigoryev, correspondence 2005.
Diagram 67 {W) Diagram 68 {B)

Exploiting White's dark squares A critical gam bit
c) One might argue that 8 ... Qb4+!? is critical since it wins a pawn. However, Black
must be prepared to defend long and accurately after 9 Bd2! (9 Nc3 is also possi
ble, but 9 ...Qxc4 10 e4 Qb4! 1 1 Bd2 Qb6 12 Bc4 d6 13 Nge2 e5 14 f4 Nbd7 15 0-0 a6
16 Ng3 Qd8! left the black defences far from easy to break down in C.Dunworth
M.Borriss, European Club Cup, Panormo 2001; this looks like a reasonable version
of the Vaganian Gambit for Black) 9 ...Qxb2 (the correct pawn: 9...Qxc4?! 10 e4 Qd4
1 1 Nc3 exd5 12 Nge2 Qe5 13 Bf4 gives White a strong initiative) 10 Nc3 Qb6 1 1 e4
(Diagram 68).
Here White's c-pawn is further advanced than in the Vaganian Gambit and he
undoubtedly has decent compensation, but Black can keep the centre closed:
l l .. d6 (ll .. .e5 12 f4 d6 transposes and is far preferable to 12 ...exf4? 13 e5 Ng8 1 4
.
d6!, which left Black horribly tied up in L.Trent-D.Howell, British League 2004) 12
f4 e5 13 Nf3 (White might also flick in 13 Rb1 Qd8, after which 14 Nf3 exf4!? 15
Bxf4 Nh5 16 Be3 g6 saw Black threatening to take control of the dark squares in
R.Palliser-D.Howell, Hastings 2003, and now White might wish to investigate the
further sacrifice 1 7 e5!? dxe5 18 Ne4 Nd7 19 Qa4) 13 ... Nbd7 (13 ...exf4!? 14 Bxf4
Nh5 a Ia Howell might be more testing) 14 Bd3 (not the only approach: 14 f5!? g6
15 Bd3 Qd8 16 0-0 also left White with good compensation in A.Vyzmanavin-
238
The M a i n Li n e : 2 ... Ne4 3 Bf4 c s
J.Ehlvest, USSR Championship, Lvov 1984) 14. . .a6 1 5 Rb1 Qd8 16 Qa4 Qc7?! 17 0-0
Be7 18 fxeS dxe5 1 9 Nh4! retained the initiative and good compensation in
M.Comette-S.Abello, French League 2004.
9 cxds c41? (Diagram 69)

Another dark-square strategy A useful intermezzo
A highly critical course. However, Black doesn't have to offer a pawn for pressure
on the dark squares and once again 9 ... d6 10 e4 g6 1 1 Ne2 transposes to the notes
to his 7th move.
10 e3
Almost universally played, but I'm not convinced that White has to block the di
agonal. Going 10 e4!? BcS is risky, but Davies's comment that this 'should appeal
only to White players with suicidal tendencies' is rather harsh. My game with
Mark Hebden (British League, 2005) continued 1 1 Nh3 d6 12 Nd2! (this seems
stronger than 12 Bxc4; no lesser player than Kasparov once suggested that White
was better after 12 ... Bxh3 13 gxh3 0-0 14 Nc3 and, indeed, 14 ... Nbd7 15 Kfl ! a6 1 6
Na4 Qc7 1 7 Nxc5 Nxc5 18 Be2 Rae8 19 Rg1 didn't give Black enough compensa
tion in M.Gurevich-A.Kochyev, Lvov 1984, but Chemin and Gurevich point out
that Black can improve with 14 ... a6! 15 Kfl Qd8 followed by ... Nh5) 12 ... Bxh3 13
Nxc4 (Diagram 70) 13 ...Qc7 (earlier 13 ... Bf2+?! 14 Ke2 Qc5 15 gxh3 0-0 had been
seen in D.Lemer-V.Rajlich, Stratton Mountain 1999, when the indefatigable Prie
points out that 16 Ne3!? Bxe3 17 Bxe3 Qb5+ 18 Kf2 Qxb2+ 19 Be2 Qe5 20 Qd4 is
excellent for White with his unopposed bishops; I also quite like White after
13 ... Qd8!? 14 gxh3 Nh5 15 Qd2, but perhaps Black should investigate this) 14 gxh3
0-0 15 Bd2! Re8 16 b4 Bd4 17 Rei and White had definitely started to take control.
1o Bcs
...
239
Black's usual choice, but I wonder whether theory has been too quick to judge
10 ...Qa5+!?, which was the choice of a noted theoretician in J.Hodgson-l.Stohl, Isle
of Man 1995. White undoubtedly enjoyed great play for the exchange after 1 1 Nc3
b5 12 Qd4! Bb4 13 Qe5+ Kf8 (neither was 13 ... Kd8!? 14 Bd2 Bb7 15 d6 Nc6 1 6 Qg5
Rg8 1 7 0-0-0 h6 18 Qg3 Ne8 especially clear in C.Ward-V.Mikhalevski, Copenha
gen 2000) 14 a3 (Diagram 71) 14 ... Bb7?! 15 axb4! Qxa1 16 Nge2, but 14 ... Bc5! would
have been a much stiffer test. Hodgson only mentions 15 d6 Na6 16 Bd2 Qb6? 1 7
Nxb5!, but 16 ... Bb7! was a serious improvement i n D.Eckert-J.Barrance, corre
spondence 1997, which continued 17 Ne4 (17 Nd5!? Qd8 18 Nc7 Rc8 19 a4 might
be tried) 17 ...Bb4 18 Qe7+ Kg8 19 Nxf6+ gxf6 20 Bxb4 Nxb4 21 0-0-0 Kg7 and
Black's attack was good enough for a draw.

It's rather murky A fairly popular gambit
11 Kf2
Essential; the pawn must be defended.
11 0-0 12 Bxc4 (Diagram 72)
...
Trompowsky expert Rainer Knaak once tried 12 Nc3!? Re8 13 Qe2 Na6 14 Na4
Qd6 15 Nxc5 Nxc5 16 Qxc4, but Black should be okay after 16 ... Nxd5 17 Qd4 Ne6,
K.Abdel Razik-M.Makropoulou, Antalya 2009.
12 Re8
...
Black takes immediate aim at e3 and we will see how the play might pan out in
Game 28.
An important alternative is 1 2 ... d6, but 13 Nc3! (White races his knight to a4; he
has often inserted 13 Ne2 Nbd7 and only then gone 14 Nbc3 Ne5 15 Na4, but
Gormally's important novelty 15 ...Qb4 16 b3 b5! 17 a3 Qa5 18 b4 Qxa4 19 Bb3 Qa6
20 bxc5 dxc5 left Black slightly for choice in S.Buckley-D.Gormally, Portsmouth
240
The M a i n L i n e : 2 ... Ne4 3 Bf4 c s
2003) 13. . .Nbd7 ( I suspect that Black may again d o better with 1 3. . .Re8!?) 14 Na4
Qb4 15 b3 (Diagram 73) 15 ... Ne5 (Black found himself in serious trouble after
15 ... b5?! 16 Bd2 Qa3 17 Bxb5 Nb6 18 Bc6 in F.Viatge-J.Jeffrey, correspondence
2001 ) 16 a3 Qa5 17 Nxc5 Qxc5 18 Bb2 leaves White better.

Black can't keep his bishop Making e3 the ta rget
Concl usion
This has been another important section and Black certainly needs to be well pre
pared after 7 Bel, as do those who employ this slightly provocative retreat. White
has made 52% from over 250 games with 7 Bel, but much remains to be discov
ered. Indeed, both the emerging 8 e4 and the underrated 8 c4 exd5 9 cxd5 c4 10
e4!? deserve further scrutiny and testing.
Illustrative Games
Game 2 8
0 V.Milov C. Landenbergue
Swiss League 2002
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 cs 4 ds Qb6 5 Bel e6 6 f3 Nf6 7 c4 exds 8 cxd5 c4 9 e3 Bc5
10 Kf2 e-o 11 Bxc4 ReB (Diagram 74) 12 Qb3
White develops with tempo since Black doesn't really want to see the queens
come off. Stefanova has experimented with the alternative 12 Qd3 d6 13 Nc3, but
13 ... Qa5! (moving out of the way of Na4) 14 Nge2 Nbd7 saw the queen coming
241
under fire on d3 in A.Stefanova-F.Nijboer, Wijk aan Zee 2005. That continued 15

Bb3?! (better 15 BbS NeS 16 Qc2 when the queen might reach b3 after all) 15 ... Ne5
16 Qc2 Neg4+!? 1 7 fxg4 Nxg4+ 18 Ke1 Nxe3 19 Bxe3 Rxe3 when Black enjoyed a
strong initiative for the piece, since White was rather poorly co-ordinated.
12 ...Qd6
An important alternative is 12 ...Qd8!? as played by Grischuk, putting the queen in
touch with the kingside without blocking the queenside pieces. The game
K.Georgiev-A.Grischuk, Mainz (rapid) 2004, continued 13 Ne2 (White has also
tried 13 d6!? when 13 ...Rf8 14 Ne2 Bxd6 15 Rd1 BeS 16 e4 Nc6 17 Nbc3 NaS 18 Qb4
Nxc4 19 Qxc4 d6 20 Bf4 left him a touch better in H.Kruse-M.Rinesi, correspon
dence 2005, but Black might do better with Sulskis's 13 ... Kh8!?) 13 ... d6 14 Nbc3 a6
15 a4 Nbd7 (Diagram 75) 16 Ng3 Rb8! (in this important tabiya Black must find
counterplay; Wells has pointed out that 16 ... Ne5 1 7 Be2 leaves White relatively
well co-ordinated and 16 ... Qe7 1 7 Re1 Nb6 18 aS! Nxc4 19 Qxc4 QeS 20 Nge4 Ba7
21 Nxf6+ Qxf6 22 Kg1 saw him beginning to consolidate his extra pawn in
L.Tinture-P.Taboada, correspondence 2005) 17 Bd2 (I suspect that White should
explore 17 Re1 !?; now 1 7... b5?! is no longer so convincing as e3 is covered in
preparation for 18 axbS axb5 1 9 BxbS NeS 20 Qc2 RxbS 21 NxbS Qb6, but 17 ... h5 18
Bd2 h4 19 Nge4 Nxe4+ 20 Nxe4 h3! gave Black decent counterplay in R.Baumann
A.Bamsley, correspondence 2004) 17 ...b5! 18 axbS axbS 19 Be2 (prudent; Prie
points out that 19 BxbS?! NeS! 20 Qc2 RxbS! 21 NxbS Qb6 takes aim at e3 with a
rather strong initiative) 19 ...b4 20 Na4 Bb7 21 NxcS BxdS 22 Qc2 NxcS 23 Rhe1 ReS
and having regained his pawn, Black was slightly for choice.

It's hard to consolidate Two undeveloped queensides
13 Ne2 (Diagram 76}

Those who want to avoid the next note might wish to consider 13 Nc3!? followed
242
by 1 3... a6 1 4 Nge2 b5 1 5 Bd3.

13 ... Na6
This feels a bit slow and I suspect that Black does better with the active 13 ... b5!?.
Certainly 14 Bxb5 Na6 15 Nbc3 Rb8 1 6 Qd1 Nb4 saw him regaining one pawn
while maintaining compensation for the other in M.Ramesh-M.Panchanathan,
Indian Championship, Nagpur 2002.
14 Nbc3 Nc7 15 Rd1ll
I wouldn't be so keen to part company with my h-pawn and would prefer to
eliminate an annoying piece with 15 Na4!. Wells continues 15 ... b5 1 6 Nxc5 Qxc5 1 7
Bd3 Nfxd5 and now 1 8 Bd2 (18 Be4 Rxe4! 19 fxe4 Nf6 doesn't seem so clear)
18 ... Nxe3 19 Bxh7+ Kxh7 20 Bxe3 leaves White slightly for choice.
15 ... b5
Tempting, but I'm not sure what was wrong with the immediate 15 ...Qxh2!?.
White can try to trap the queen with 1 6 Ng3, but then Black has at least 16 ... Rxe3!
17 Bxe3 Bxe3+ 18 Kxe3 Qxg3 19 d6 Ne6 with a pawn and much the safer king for
the exchange.
16 Nxb5 Nxb5 17 Bxb5 Rb8 18 Qd3 Qxh2 19 Ng3l (Diagram 77)

What did the players m iss? That unmoved bishop is a killer!
It would have been somewhat better to complete development with 19 Bd2.

19...Qh6l
Returning the favour. It seems that both players underestimated Black's initiative
after 19 ... Nxd5! 20 Qxd5 Rxb5. The pin is actually not so effective and Black ap
pears to be doing quite well after 21 Rd4 Bxd4 22 Qxb5 Be5 or 21 Rd3 Re6! 22 Qa8
Rg6.
243
20 b41? Bxb4 21 Qc4 Bas 22 Nf5 Qg6 23 Nd61

The knight had to go forwards: 23 Nd4? Qh5 24 Nb3 Bb6 would have left White in
some trouble with both e3 and d5 rather weak.
23 ...Ng4+ 24 Qxg4 Qxd6 25 Rb1
White appears to have consolidated, but before Milov can generate anything on
the kingside, disaster strikes.
25 ... Bc3 26 Rb3 Bf6 27 Ba3?
A step in the wrong direction. White had to prevent the queen from returning to
h2, such as with 27 g3.
27 ... Qh21 28 d6?
Positionally well motivated, but Milov has completely underestimated the danger.
White would have done better with 28 Rdb1, although 28 ... Bh4+ 29 Ke2 Rxb5 30
Rxb5 Ba6 31 Bc5 Bxb5+ 32 Rxb5 f5 33 Qh3 Qf4 34 Rb4 Rxe3+ 35 Bxe3 Qxb4 would
still have left Black doing pretty well.
'
WARNING: Always take extra care even when one's king is just
slightly exposed in an open middlegame. Milov is going to rue his
X lack of an h-pawn and resulting shelter on g1.
2B... Bh4+ (Diagram 78)
The bishop did go to f6 for a reason!
29 Ke2 Rxbsl 30 Rxbs Ba6 31 Qxd7
The pin is fatal. 31 Rdb1 Qe5 leaves White unable to defend both his rook and e3.
3 1...Qes 32 Bcs Qb2+ o-1
Avoiding a Benoni: 5 c3 Nf6 6 Nd2

1 d4 Nf6 2 Bgs Ne4 3 Bf4 c s 4 f3 Qas+ 5 c 3 Nf6 6 Nd21? (Diagram 79)
Popularized like so much in the Trompowsky by Hodgson, this appeals to those
less keen to close the centre when it involves falling behind in development. Oth
ers may like the resulting queenless middlegames and by teaming up this varia
tion with 4 ... Nf6 5 dxc5 one can avoid having to learn too much critical theory.
6 ... cxd4
Invariably played as 7 Nb3 was threatened and 6 ... d6 7 Nc4 Qc7 8 e4 just gives
White an easy edge.
7 Nb3 Qb6
Played in two-thirds of all games to have reached this position, but if Black isn't
keen on a queenless middlegame, he will probably prefer one of the alternatives:
244
The M a i n L i n e : 2 ... Ne4 3 Bf4 c s
a) 7... Qd8 8 cxd4 (Diagram 80) gives White a lead in development, but his knight
doesn't contribute much from b3 and he must beware a timely ... Nh5 strike:

White declines to advance White is ahead in development
a1) Black should avoid 8 ... g6?! 9 e4 Bg7 1 0 Ne2!? 0-0 11 Nc3 d6 1 2 Be2 a6 13 0-0 b5
14 a3 Nbd7 15 Qd2, which gave White a comfortable edge in J.De la Villa Garcia
J.Polgar, Pamplona 2000.
a2) However, Black can tempt the e-pawn forwards and 8 ... e6!? 9 e4 Nc6 10 Ne2
d5 1 1 e5 Nd7 12 Nc3 Nb6! 13 Bd3 Nc4 14 Qe2 Nb4 gave him reasonable French
like counterplay in P.Wells-V.Kotronias, Port Erin 2007.
a3) Standard is 8 ... d5 9 e3 e6 when we reach a position in which White has tested a
number of ideas. Probably best is 10 Rcl (White wants to avoid 10 Bd3 Nc6 1 1 Ne2
Nh5! with quite an easy game for Black, but he might consider 10 g4!?, after which
10 ... Nc6 1 1 Ne2 Nd7 12 Nc3! Be7 13 Bd3 Bh4+ 14 Ke2 a6 15 Bg3 looked a little bet
ter for him in R.Rizzo-J.Helbich, correspondence 2005) 10 ... Nc6 1 1 Bb5 Bd7 12 Ne2
Bb4+ 13 Kf2! (Diagram 81) when White has control and isn't overly worried by
... Nh5 ideas. Then 13 ...Qb6 (unsurprisingly 13 ... Nh5!? has also been tried, but 14
Nc5 Nxf4 15 Nxf4 Bxc5 16 Rxc5 0-0 1 7 Qd2 ReB 18 Rhcl gave White a pleasant
edge in Hoang Thanh Trang-G.Morrison, Budapest 2006) 14 Bd3 0-0 was seen in
J.Hodgson-V.Babula, German League 2000, in which 15 g4!? was rather unclear,
but 15 a3 Be7 16 Qc2 followed by Nc5 would have been a simple route to an edge.
0
NOTE: White's king is often fairly well placed on f2 in this structure,
at least so long as he can keep the central situation under control.
Indeed Kf2 is usually the best response to an early ... Bb4+, since the
knight is quite good on e2 and White would like to play down the c
file.
245
b) Much more provocative is 7...Qf5!?, obtaining the bishop-pair in returning for

accelerating White's development with 8 Bxb8! (clearly best, whereas 8 Qxd4?!
Nc6 9 Qd2 eS 10 Bg3 dS 1 1 e3 Be7 12 Bd3 Qe6 already slightly favoured Black in
Z.Marnedjarova-P.Crarnling, Istanbul 2003) 8 ... Rxb8 9 Qxd4 b6 10 e4 (Diagram 82).

The king is quite safe on f2 Black grabs the bishop-pair
Black should now hurry his queen horne with 1 0...Qf4 (10...Qe6?! 1 1 Bc4! dS?! 12
0-0-0 already saw White making good use of his superior development in
M.Karttunen-H.Westerinen, Finnish League 2007) 1 1 Nh3 (practice has shown that
this is a better option than 1 1 Ne2; White doesn't obstruct his bishop and has ideas
of NgS, as well of Nf2-e4 after eS) 1 1 ...Qc7 when 12 eS Ng8 reaches a pretty inter
esting rniddlegarne. White is a long way ahead in development, but the black
camp doesn't contain any weaknesses and we will examine attempts to force some
in Game 29.
We now return to the main line and 7...Qb6:
8 Qxd41
The modem preference. Hodgson used to prefer 8 cxd4, but this has scarcely been
seen since Black unveiled a powerful counter in J.Hodgson-B.Gelfand, Groningen
1996: 8 ... Nc6!? (Black fights for the initiative, although it should be said that he
had hardly been doing badly with 8 ...e6, intending 9 e4 aS!) 9 e4 (White should
avoid 9 dS?! eS! and 9 e3 NdS 10 Ne2 eS! was also already somewhat more fun for
Black in G.Heisel-M.Ahn, German League 2003) 9 ... e5! 10 dxeS Bb4+ (Diagram 83)
11 Ke2 (risky and White has tended to prefer 11 Bd2 on the rare occasions that 8
cxd4 has been repeated; here too Black is very comfortable and I don't wholly
trust Vysochin's l l ... Nxe5 12 Bxb4 Qxb4+ 13 Qd2 Qe7!? 14 0-0-0!?) 1 l . ..Nxe5 12
Qd4 Ng6 13 Be3 dS! 14 exdS 0-0 and Black enjoyed a dangerous initiative.
246

Good cou nterplay A French-like middlegame
8 Nc6
...
White should be fairly happy to see 8 ...Qxd4 9 cxd4 since Black's position is
somewhat less dynamic with the queens off:
a) White enjoyed a very pleasant advantage after 9 ... d6 10 e4 g6 1 1 Rcl Nbd7 1 2
Be3 Bg7 13 Bd3 0-0 14 Ne2 a S 1S Nd2 eS?! 16 d S i n K.Georgiev-L.Gutman, Euro
pean Championship, Dresden 2007.
b) Likewise the 9 ... dS 10 e3 e6 1 1 g4! Nc6 1 2 BbS Bd7 1 3 a3! Kd8 14 Rcl NeB 1S NcS
BxcS 16 RxcS of V.Salov-V.Akopian, Wijk aan Zee 1993, was a model demonstra
tion of how to utilize White's extra space.
c) Probably Black's best is to head for a French structure with 9 ...e6 10 e4 Nc6
(play transposes in the event of 10 ... dS 1 1 Rcl Nc6 12 eS Nd7, but the independent
1 l ...Bb4+ 12 Bd2! Bxd2+ 13 Kxd2 0-0 14 eS Nfd7 1S Bd3 b6 16 Rc7! saw Black fail
ing to equalize in I.Popov-R.Pruijssers, Kerner 2007) 1 1 Rcl dS (Diagram 84) 12 eS
(12 Bd3!? deserves attention too when the critical line might well be 12 ... Bb4+ 13
Bd2 aS!?) 12 ... Nd7, after which 13 BbS aS! 14 a4 (14 Bxc6 bxc6 1S Rxc6 Ba6 sees
Black's strong light-squared bishop and long-term queenside pressure compen
sate for the pawn) 14 ... Na7 1S NxaS (Miladinovic has tried 1S Bfl ! ? against which
1S ... Nc6-b4 looks critical, since 1 S ... Nb6 16 NcS! BxcS 17 dxcS Nd7 18 BbS Nc6 19
Ne2 f6 20 exf6 gxf6 21 Bd6 favoured White in I.Miladinovic-D.Milanovic, Herceg
Novi 2007) 1S ... NxbS 16 axbS was seen in S.Lputian-A.Shirov, Belgrade 1999, and
now 1 6... NcS! 1 7 dxcS RxaS 18 b4 RxbS 19 Bd2 b6 20 cxb6 Ba6 (Hertneck) would
have been fine for Black.
9 Qxb6 axb6 (Diagram 85)
A critical position for the assessment of 6 Nd2. Black's queenside has been broken
247
up, but he enjoys some dynamic compensation not least because his central pawns
are about to advance with tempo.
10 Nd4
This has become established as White's main try, but is most certainly not his only
option:
a) 10 a3?! is directed against ideas of ...b5-b4, but is too slow and 10 ...e5 1 1 Be3 d5!
12 Bxb6 d4 13 e4 (White has also struggled after 13 cxd4 Be6) 13 ... Nd7 14 Bc7 dxc3
15 bxc3 Nc5 16 Nxc5 Bxc5 17 Bb5 f6 saw Black regaining his pawn with advantage
in N.Marcetic-S.Brenjo, Belgrade 2007.

The key tussle after 6 Nd2 An unexplored idea
WARNING: Those doubled b-pawns might appear sickly, but White

should not go after them too quickly. Pradice has frequently seen
Black obtaining a strong initiative when he tries to do so.
b) 10 e4 d5 1 1 exd5?! (Wells points out that White should try 1 1 Be3 dxe4 12 Bxb6
when 12 ... e5 transposes to variation 'd2') 1 l ...Nxd5 12 Bd2 e5 13 a3 Be6 reaches
another position in which Black's extra space and better development more than
compensate for his doubled pawns.
c) A recent development has been 10 Bg5!? (Diagram 86), which certainly deserves
further consideration. That said, 10 ... d5 (10... h6 1 1 Bxf6 gxf6 12 e4 d6 13 Bb5! Kd8
14 Ne2 f5 15 0-0 Bg7 16 Nf4 favoured White in Hoang Thanh Trang-E.Toth, Buda
pest 2007, but Black may be able to improve) 11 Nd4 Bd7!? (Black doesn't want to
stabilize the central structure and 1 l ...Nxd4 12 cxd4 Bd7 13 e3 e6 14 Bd3 gave
White a small edge in E.Torre-M.Paragua, Ho Chi Minh City 2003, although of
course Black remained rather solidly placed) 12 NbS ReB 13 e3 e6 14 Nh3 Be7 15
Rd1 h6 16 Bh4 Na5! 1 7 Nf2 Nc4 18 Bxc4 Rxc4 was no more than quite unbalanced
248
The Ma i n l i n e : 2 . . . Ne4 3 Bf4 c s
and unclear i n R.Raijmaekers-H.Buczinski, correspondence 2006.

d) White's main alternative to our main line has been 10 Be3!? when Black must
decide whether or not to sacrifice:
d1) With 10 ... b5 he keeps his front b-pawn, but 1 1 Nd4! (practice has shown that
1 1 Bd2 e5 12 e4 b4 gives Black enough counterplay; that a-file pressure should not
be underestimated) 1 l ...Nxd4 (Black might play more actively with 1 l ...Nd5!? 12
Bf2 Nxd4 13 Bxd4 e5!?, but Prie has pointed out that the forcing sequence 14 Bxe5
Ne3 15 Kf2 Nc2 16 Rei Bc5+ 17 Bd4! Nxd4 18 cxd4 Bxd4+ 19 e3 Bxb2 20 Rc2 Be5 21
Bxb5 gives White an edge; however, here Black might do better with 12 ... Nc7!?
and 13 Bg3 d6 14 e4 Nxd4 15 cxd4 g6 16 Bd3 Bg7 1 7 Ne2 Bd7 18 Bf2 Na6 gave him
enough counterplay in I.Miladinovic-L.Ftacnik, Istanbul Olympiad 2000) 12 Bxd4
(Diagram 87) 12 ... e6 (Miladinovic has shown that White has an edge too after
12 ... d6 13 e4 Bd7 14 Bxf6 gxf6 15 a3; note how White only restrains the b-pawns
having first dealt with the central situation) 13 e4 b4 14 Bb5 bxc3 15 Bxc3 Be7 16
Ne2 0-0 1 7 Nf4 left White with a small edge in I.Miladinovic-G.Grigore, Istanbul
2002.

Black lacks activity A decent ga mbit
d2) More critical is 10 ... d5!? 11 Bxb6 e5 (Diagram 88), offering a pawn for the ini
tiative. White too can fight for the initiative, though, and 12 e4! (prudent, not least
because otherwise ... d4 ideas give Black a handy initiative) 12 ... dxe4! (theory used
to prefer 12 ...Be6, but 13 Nc5! Bxc5 14 Bxc5 is problematic for Black despite his
superior development: 14 ... dxe4 15 a4 exf3 16 Nxf3 Nd7 1 7 Bd6 f6 18 Bb5 Nb6 19
Bc5 Nc4 20 b4 saw White mobilizing his queenside majority in I.Miladinovic
D.Lekic, Vrnjacka Banja 2005, and 14 ... Nd7 15 exd5 Nxc5?! 16 dxe6 Nb3?! 17 Rb1
Rxa2 18 Bc4 Nca5 19 exf7+ Ke7 20 Bd5! Rd8 21 Rd1 left Black's tactical attempts at
an end in B.Socko-P.Bobras, Polanica Zdroj 2006) 13 Bc4 Bf5 14 Nd2 Be7 1 5 fxe4
249
Sta rti n g Out: The T rom powsky Attack
Nxe4 16 Nxe4 Bxe4 1 7 Nf3 NaS! (Diagram 89) was about even in Zhang
Pengxiang-E.Sutovsky, Poikovsky 2007.

It's roughly equal A useful gai n of kingside space
We now return to 1 0 Nd4:

10 es1
...
This critical central counter is the best way to deal with the issue of 11 NbS, as we
can quickly appreciate:
a) 10 ... Nxd4 1 1 cxd4 dS is fairly solid for Black, but also leaves him unpleasantly
worse due to his queenside weaknesses: 12 Bd2!? (White should avoid 12 Bc7?! e6
13 Bxb6 Kd7!, but it's not clear that he has to take action against ... NhS and 12 e3
Bd7 13 Bd3 Bc6 14 Ne2 e6 1S Nc3 Kd7 16 g4 h6 1 7 Kf2 gave him a pleasant pull in
L.McShane-K.Lie, Drarnrnen 200S) 12 ...e6 13 e3 Bd7 14 g4! h6 1S Bd3 Bc6 16 Ne2
Kd7 1 7 Kf2 Be7 18 Nc3 (Diagram 90) left White with an edge and options right
across the board in P.Wells-H.Ziska, European Championship, Dresden 2007.
b) Neither is 10 ... RaS the complete answer because of 1 1 b4! Ra4 (I suspect that
1 l ...Nxd4 12 cxd4 Ra3!? 13 Rcl Ra8 improves, although 14 e4 e6 1S Bd2 Be7 16 a3
0-0 17 Rc3 d6 18 Bd3 Bd7 19 Ne2 Rfc8 20 Rxc8+ Bxc8 21 Bel was still a little better
for White in J.De Jong-M.Solleveld, Hilversurn 2006) 12 NbS g6 13 e4 Bg7 14 Nc7+
Kf8 1S BbS, which saw Black's rook getting into trouble in E.Torre-P.Svidler, Bad
Hornburg 1998.
c) Finally, we should note that the threat is actually worth avoiding: lO ... dS?! 1 1
NbS RaS 12 e4! dxe4 1 3 0-0-0 Bd7 14 Be3 exf3 1 S Nxf3 gives White a strong initia
tive for the pawn, R.Palliser-M.Rose, Oxford 2004.
11 Nxc6 exf4
Black accepts another set of doubled pawns, but does obtain a potentially powerful
2 50
dark-squared bishop and something of a bind on the e3-square. Hebden's sacrificial

1 l...dxc6!? is a critical alternative when the stem game continued 12 Bxe5 Be6 (the
other try is 12 ...b5, but 13 e4 Be6 14 Ne2! Nd7 15 Bg3 Rxa2 16 Rxa2 Bxa2 17 Nd4
gave White a pull in V.Milov-V.Kotronias, French League 2004) 13 a3 (White has
tried to improve with 13 e4!?, which makes a fair amount of sense; he is happy to
transpose to our last variation and 13 ... Nd7 14 Bd4 Rxa2?! 15 Rxa2 Bxa2 16 Bd3 Nc5
17 Bc2 b5 1 8 Ne2 gave him a pull too in L.Le Quang-Goh Wei Ming, Singapore 2006)
13 ...b5 (Diagram 91) 14 e4 (De Ia Villa's 14 e3!? deserves consideration too, although
14 ... Nd7 15 Bf4 Nc5 16 0-0-0 Bb3 1 7 Rd2 Ba2! gave Black enough compensation in
E.Prie-Y.Solodovnichenko, Rochefort 2005) 14 ...b4 15 cxb4 Bxb4+ 16 Ke2 Ba5 1 7 Nh3
(Prie points out that 17 Rei!? 0-0-0! 18 b4 Bb6 19 b5 Rhe8! introduces ideas of ... Nxe4
and gives Black good counterplay) 17 ...Bb3! 18 Nf4 0-0-0 19 Bc3 Bb6 and Black had
enough for his pawn in A.Miles-M.Hebden, British Championship, Street 2000.

The Hebden Gambit White must remove the f4-pawn
12 Nd4 (Diagram 92)

Yet another fascinating middlegame has arisen in which both sides have their
trumps. At this point Black has tried a number of options:
a) 12 ... Nd5 13 g3! Bc5! (Black shouldn't try to maintain a pawn on f4: 13 ... g5?! 14
e4! fxe3 15 Bc4 Nc7 16 Ke2! Ne6 1 7 Nc2 Ra5 18 Nxe3 b5 19 b4! Ra3 20 Bb3 saw
White taking control with some powerful play in A.Stefanova-Tan Zhongyi, Elista
2004) 14 Nh3!? (I much prefer this sensible move to the messy 14 gxf4 Ne3 15 Kd2
Nxfl+ 16 Rxf1 Rxa2 17 Kc2 when Prie has drawn attention to the strength of
17 ... d5!?) 14 ... Ne3?! (critical must be 14 .. .fxg3 15 hxg3 Ne3 16 Kd2 Bxd4 1 7 cxd4
Nf5 18 Nf4 Nxg3 19 Rgl Nxfl+ 20 Rgxfl when White's well-placed knight and
strong centre compensate for the pawn; at this point Prie has indicated 20... Kd8 21
Rh1 as best, simply intending to provoke a weakness after a3 and Ragl) 15 Kd2
251
sta rt ing Out: The Trom powsky Attack
Nxfl+ 16 Rhxfl fxg3 1 7 hxg3 (Diagram 93) 1 7...0-0?! 18 Nf4 Bxd4 19 cxd4 left White
doing pretty well in Hoang Thanh Trang-E.Berend, Beijing (blitz) 2008.

The knights are quite good Activity vs. structural concerns
TIP: As we have just seen, it's often a good idea to remove that
bridgehead from f4 as quickly as possible with some combination of
the g3- and e3-breaks allied to Nh3.
b) 12 ... d5 13 e3 (13 g3 Bd6! now seems okay for Black, who was fine after 14 NbS
BeS 15 gxf4 Bxf4 16 e3 BeS in M.Turner-L.Ftacnik, Hastings 2000/01) 13 .. .fxe3 14
Ke2 BcS (pinning appears best: 14 ... Bd6 15 Kxe3 0-0 16 BbS! Bd7 1 7 Nge2 BcS 18
Kd2 gave White an edge in P.Wells-J.Lappage, Oxford 2003, and so did 14 ...g6 15
Kxe3 Bh6+ 16 Kf2 0-0 17 BbS! Bd7 18 Nge2 Rfe8 19 Rhd1 in L.McShane-O.Cvitan,
Calvia Olympiad 2004) 15 Kxe3 (Diagram 94) is quite a solid option for Black.
Nevertheless, his weaknesses down the d-file promise White a small advantage
and 15 ...0-0 (I suspect that Black may do better to prevent White's next, and
15 ... Bd7!? 16 Bd3 bS 17 Nge2 b4 18 cxb4 Bxb4 19 Rhcl 0-0 didn't seem too bad for
him in P.Chakov-K.Georgiev, Bulgarian League 1995) 16 BbS!? Ne8 1 7 Nge2 Nd6
18 Kd3! Nxb5 19 NxbS BfS+ 20 Kd2 Bd7 21 Ned4 retained an edge in R.Langeveld
J.Munoyerro Ajuriagoxeascoa, correspondence 2005.
NOTE: White's bishop finds itself well placed on bs in this 12 ...ds

U sub-variation. It prevents Black from obtaining counterplay with
... ReS and White is generally happy to see a minor piece exchange.
c) 12 ... Bc5!? 13 Nh3 (Black seems to be okay after 13 g3 Bxd4! 14 cxd4 Ra4 15 gxf4
NdS) 13 ... Nd5 14 e4!, fighting for central control, is a critical line and one which
we will return to in Game 30.
252
The M a i n L i n e : 2 . . . Ne4 3 Bf4 c s
Conclusion
For those who don't like Benoni structures, 6 Nd2 is a decent approach and one
with which White has scored 52% from over 1,400 games. Both sides have a num
ber of options, but our rather unbalanced main line appears the critical test. White
has made 55% from over 200 games with 1 1 Nxc6, but that falls to just 51% after
l l ...dxc6 and it may well be that Hebden's pawn sacrifice gives Black enough play
for a draw.
Illustrative Ga mes
Game 29
D M.Galyas . Z.llil'lcic
Torokba lint 2004
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 c5 4 f3 Qa5+ 5 c3 Nf6 6 Nd2 cxd4 7 Nb3 Qf5 8 Bxb8 Rxb8
9 Qxd4 b6 10 e4 Qf4 11 Nh3 Qc7 12 e5 NgB 13 0-001 (Diagram 95)

Vigorous play is requi red Black's king isn't too ha ppy
White has tried a number of alternatives, but this obvious move, keeping his op
tions open, appears best.
TIP: It's often a good idea when developing the pieces to make the
moves you know you will definitely play before deploying those
pieces you are less sure about.
u . . . e6 14 Ngsl?
253
Sta rting Out: The Trom powsky Attack
Again this is by no means the only approach, but this direct approach is quite a
challenging one.
14... Ne71
Black must develop this way. He might like to keep the knight from e4 with
14 ... f5?, but then 15 Bb5! is simply devastating.
15 Bb5
Seizing space with 15 f4 is a very valid alternative after which, for example,
15 ... Nf5 16 Qf2! h6 (16 ... Be7?! 17 Bd3! h6 18 Nf3! h5 19 Bxf5! exf5 20 Nfd4 left
Black's structure a complete mess in J.Degraeve-V.Chuchelov, Le Touquet 1995)
17 Ne4 (Prie prefers 17 Nf3!? with the idea of 17 ... d5 18 Bd3 Ne7 19 Nbd4 followed
by kingside expansion) 1 7 ... Bb7 18 Bd3 d5 19 Ng3 Nxg3 20 hxg3 was a little better
for White in A.Miles-B.Avrukh, Biel 1995.
15 ...Ng6
Black fared better with 15 ... a6!? 16 Qf4 Nf5 17 g4!? (very direct, but I would prefer
Rybka's discovery of 17 Bd3!? with the idea of 17 ... f6 18 Nxh7! Rxh7 19 Bxf5 exf5?
20 Rhe1 Be7 21 Qg3 when White's threats are too strong, although this isn't actu
ally so clear after 19 ... Rh6!) 17 ... axb5 18 gxf5 f6 19 fxe6 dxe6 20 Ne4 Qxe5 21 Qd2
(Wells wonders about 21 Qxe5!? fxe5 22 Rhg1 when White certainly has structural
compensation, but no more than that I would say) 2l...Be7 which was unclear but
roughly balanced in G.Treppner-A.Braslavsky, German League 2004.
16 f4 Bcs 17 Qd2 (Diagram 96)
One has to quite like White's chances here. Black cannot develop his light-squared
bishop without castling, but castling is risky with White already well advanced on
the kingside.
17...0-0 18 Bd3 f6 19 h411
White would only have been a little better after 19 Nf3 Ne7 20 Rhe1, but Galyas's
choice doesn't come without risk.
19...fxgs 20 hxgs Kf71
The black king tries to flee, but h7 was too important a pawn for him to give up so
lightly. Rather than panic, Ilincic needed to stay calm and find 20... Rf7! when 21
g3 (otherwise 21 Bxg6? hxg6 22 Rh2 Bb7 23 Rdh1 Re7! only leaves White strug
gling, and 21 Rh2 Nxf4! 22 Bxh7+ Kf8 appears okay for Black, since 23 Nxc5 bxc5
24 g3 Qxe5 25 gxf4 Qxf4 still leaves him a pawn ahead) 2l...Bb7 doesn't seem to
leave White with anything better than 22 Nxc5 Qxc5! 23 Rxh7 Kxh7 24 Qh2+ Kg8
25 Bxg6 Qe3+ 26 Rd2 Qe1+ 27 Rd1 when Black has perpetual.
21 Rxh7 (Diagram 97) 21... Ne7
Probably best; 21...Bb7? would have left White doing rather well after 22 Nxc5
bxc5 23 Bxg6+! Kxg6 24 Qd3+ Kf7 25 Qxd7+ Qxd7 26 Rxd7+ Ke8 27 Rhxg7.
22 Qe2
2 54
The Ma i n l i n e : 2 . . . N e4 3 Bf4 c s
I suspect that already White might have relied on his pawn-roller with 2 2 Nxc5!?
Qxc5 23 g4 when it's far from easy to see a good defence for Black.
22 ...Rg8 23 Rf1 Nf5ll
This only eases White's task. Black's position would have been none too pleasant
after 23 ... Ke8 24 g4 Kd8 25 f5, but he had to sit tight and hope to defend thus.
24 Qh5+ Kf8 25 Rh81
TIP: When one has plenty of firepower in an attack, it's rarely a bad
idea to exchange off some key defenders.
25 ... Bb7 26 Rxg8+
Not bad, but 26 g4! would have been more clinical with the brutal idea of 26... Bg2
27 gxf5! Bxfl 28 Qh7, which forces mate.
26...Kxg8 27 g4 (Diagram 98) 27 ... Bg2

A dangerous sacrifice Black is swept away
The knight couldn't run away on account of 27... Ng3? 28 Qh7+ Kf8 29 Qh8+ Kf7 30
g6+ when g7 falls with fatal consequences, and 27... Ne7 28 Qh7+ Kf8 29 f5 would
have been totally crushing too.
28 Re1 Ne7
There's no defence and 28 ... Bf3 29 Qh3 Be3+ 30 Kbl Bxg4 31 Qxg4 would have
been but a temporary stay of execution.
29 Qh7+ Kf7 30 g6+ Kf8 31 f5 exf5 32 gxf5 Bd5 33 f6
The pawns sweep all before them like a giant tidal wave. Indeed, I dare say that
Black only played on because White's flag was hanging.
33 ...gxf6 34 g7+ Ke8 35 exf6 Qg3 36 gBQ+ BxgB 37 Rxe7+ KdB 38 Rxd7+ KcB 39
Ba6+ 1-0
255
Sta rti n g O ut: The Trom pows ky Attack
Game 30
0 S.Schneider T.Ernst
Swedish Tea m Cha mpionship2008
1 d4 Nf6 2 Bg5 Ne4 3 Bf4 c5 4 f3 Qa5+ 5 c3 Nf6 6 Nd2 cxd4 7 Nb3 Qb6 8 Qxd4 Nc6 9
Qxb6 axb6 10 Nd4 e5 11 Nxc6 exf4 12 Nd4 Bc5 13 Nh3 Nd5 14 e4 fxe3 15 Bc4 (Dia
gram 99)

Bold but good play from White Black can't hold the e3-pawn
Thus White has managed to develop his kingside and with tempo, but at the cost
of a pawn.
15 ... Bxd4
A fairly simplistic approach and Black has usually preferred to move his knight:
a) 15 ... Nc7?! 16 Nf4 0-0 17 0-0-0! leaves Black still to untangle his queenside and
White for choice: 17 ... d5!? (a radical try, but 17 ...b5 18 Nxb5 Nxb5 19 Bxb5 Rxa2 20
Kb1 Ra7 21 Bxd7 b5 22 Bxc8 Rxc8 23 Rd5 gave White a pleasant edge with two
weaknesses to aim at in J.Hodgson-E.Sutovsky, York 1999) 18 Nxd5 Nxd5 19 Bxd5
Rd8 20 Ne2 Bf5 21 Rhe1! prepared Ng3-e4 and after 2l...Rxd5?! 22 Rxd5 Rxa2 23
Kd1 ! Black didn't have enough for the exchange in L.Bruzon-D.Flores, Santa Clara
2004.
TIP: While White should keep an eye on the e-pawn in this variation,
he shouldn't hurry to round it up. Instead White should maintain
control, seek to exchange pieces and then round up the pawn at his
leisure.
b) 15 . Nf6 16 Nf4 (again White hurries his knight into play) 16 . Bxd4 17 cxd4 0-0
.. ..
256
The Ma i n L i n e : 2 ... Ne4 3 Bf4 cs
(with 1 7... g5 18 NdS Nxd5 19 BxdS RaS 20 Bb3 5 Black tried to remain a pawn up
in A.Schirbel-G.Hoegerl, correspondence 2005, but 21 4! gx4 22 Rfl Kd8 23 Rxf4
left White slightly for choice) 18 Ke2 Re8 19 g4! (Diagram 100) 19 ... g5 (easy to criti
cise, but Black was understandably keen to avoid 19 ...d6?! 20 gS Nd7 21 BbS) 20
Ng2 Ra4 21 Kd3 RaS 22 Nxe3 regained the pawn with advantage in F.Urkedal
A.Evdokimov, Bomholm 2008.
c) Probably Black is best advised to force matters with 15 ... Nb4!? 16 0-0-0 (Prie has
shown that both 16 Ke2 Bxd4! 17 cxd4 dS 18 Bb3 Bxh3 19 gxh3 Kd7 and 16 NbS
Nc2+ 17 Ke2 Nxa1 18 Nc7+ Ke7 19 Nxa8 dS 20 BxdS Nc2! are fine for Black, but
Wells's 16 cxb4!? Bxd4 1 7 0-0-0, intending 17 ... Bf6 18 Nf4 Bg5 1 9 Rd4 0-0 20 h4 Bh6
21 g4, could still do with a test) 16 ... Bxd4 1 7 Rxd4 dS! (Diagram 101).

A critical line A comfortable advantage
Then 18 Bb3!? (improving over 18 cxb4 dxc4 19 Re4+ K8 20 Re1 Bxh3 21 gxh3
Rxa2 22 Kb1 Ra8 23 R1xe3 which led to a fairly level endgame in L.McShane
A.Volokitin, Igualada 2005) 18 ... Nc6 (18 ... Nxa2+? 19 Kb1 BfS+ 20 Bc2 would trap
Black's errant knight) 19 RxdS Be6 20 Rdd1 Bxb3 21 axb3 0-0 22 Rhe1 Rfe8 23 Nf4
left White slightly better in I.Schneider-F.Beck, Bi::iblingen 2008, although Black
managed to draw.
16 Bxds
The correct recapture, since 16 cxd4?! Ne7 1 7 Ke2 d5 1 8 BbS+ Bd7 19 Bxd7+?! Kxd7
20 Kxe3 walks into 20 ... Nf5+ 21 Kd3 Ra4 when d4 must fall.
16 ... Bf6 17 Ke2
White plays to regain his pawn. He would like to improve the knight with 17 Nf4,
but here the dS-square is unfortunately unavailable after 1 7 . Bg5.
..
17...d6 18 Kxe3 (Diagram 102) 18...Be6?1
2 57
Sta rt i n g O ut: The Trom powsky Attack
Now White gets to improve his knight and obtain an edge. I suspect that Black
should seek refuge in an opposite-coloured bishop scenario and 18 ... Bxh3!? 19
gxh3 Kd7 20 f4!? Rhe8+ 21 Kf3 Ra5 would restrict White to just a pull after 22
Rad1! (22 Bxb7?! Rh5 gives Black enough counterplay). Nevertheless, White has
the better bishop here and could most certainly try to grind.
19 Nf41
Schneider has judged that he will be able to grab a pawn without anything too
dangerous happening.
19 ... Bg5 20 Bxb7 Ras 21 g3 Bxa2?
The Swedish Grandmaster's desire to regain his pawn is understandable, but he is
simply far too uncoordinated after this. It would have been better to develop the
remaining rook with 2l...Kd7.
22 b41 (Diagram 103)

Black is in serious trouble Several tempting options
Now the pin down the a-file will be killing if the rook remains on it, but neither
does checking help Black.
22 ... Re5+ 23 Kf2
Hardly bad, but even stronger would have been 23 Kd3! Be6 (23 ...Bb3 24 c4 traps
the bishop) 24 Nxe6 Rxe6 (24...Re3+ 25 Kd2 Rxe6+ 26 f4 Bd8 27 Bc6+ Ke7 28 Bd5
also leaves Black's king facing a rather nasty attack) 25 Bc6+ Ke7 26 f4 Bh6 27 Ra7+
Kf6 when White's pieces dominate and he has a number of tempting options, in
cluding just 28 Bd5 Re7 29 Rxe7 Kxe7 30 Ra1 Rd8 31 Ra6 followed by pushing his
soon-to-be-passed b-pawn.
23 ...Bc4 24 Ra8+ Bd8 25 Rd 1
258
I n dex of Va riatio n s
Now Black faces the unpleasant choice between losing his queenside pawns and
having to keep his king in the centre.
25 ... Kd7 (Diagram 104} 26 Bd5
White heads for a good knight versus bad bishop scenario, but he must also have
been tempted by 26 Ra7!? when 26 ... Bc7 27 Bd5 g5 (desperate stuff, but 27... Bxd5?
28 Nxd5 costs Black serious material) 28 Bxc4 gxf4 29 Bxf7, netting a pawn while
retaining control.
NOTE: There are opposite-coloured bishops present here, but with

rooks on and White's bishop much the more adive, White would
have enjoyed good winning chances.
26... Bxds 27 Nxds Rhea 28 Rd2 bs 29 Ra7+ Kc6 30 Ra6+ Kd7 31 Ra7+
Was the 2450-rated Swede extremely low on time or did a draw clinch an impor
tant match? I'm afraid I don't know, but even here 31 f4! would have retained
some advantage.
31 ...Kc6 32 Ra6+ Kd7 33 Nb6+ Kc7 34 Nd5+ Kd7 Y2-Y2
259
I n d ex of Va ri atio n s
The Classica l 2 ds
1 d4 Nf6 2 Bg5 d5
3 Bxf6
3 e3 - 3 1 , 35
3 exf6
.
3 gxf6
... - 24, 29
4 e3 Be6
4 ... Bd6 5 c4 - 1 2, 1 7
5 g3 c6 6 Bg2 Bd6 7 Ne2 Nd7
7 . 0-0 8 0-0 Nd7 9 Qd3 - 2 1
. .
s Nd2 ts 9 b 3 - 15
260
I n d e x of Va riations
2 g6 and Minor Lines

...
1 d4 Nf6 2 BgS g6
2 d6 - 40
...
2 c6 - 40
...
2 b6 - 4 1
...
2 Nc6 - 42
...
3 Bxf6 exf6
4 c4 - 43
4 ... Bg7 5 g3 0-0 6 Bg2 d6 7 Ne2 fS 8 c4 - 45, 47, 50
Sta rt i ng Out: The T ro m pows ky Attack
The Positional Choice, 2 ... e6

1 d4 Nf6 2 Bg5 e6
3 e4
3 e3 - 92, 95
3 ... h6
3 ... Be7 - 55
3 d5 - 56
...
3 . . c5 - 57, 63
.
4 Bxf6 Qxf6
5 Nc3
5 Nf3 - 59, 65
5 c3 dS (5 d6 - 80, 89)
...
262
I n dex of Va riatio ns
6 e5 - 81
6 Nd2 - 81, 85
s ...Bb4
5 d6 6 Qd2 - 69, 74
...
6 Qd2
6 Qd3 - 72, 77
6...cs 7 a3 Bxc3 8 bxc3 d6
9 Nf3 - 74
9 f4 - 74
The U ncom promising 2 .. cs

1 d 4 Nf6 2 Bg5 c5
Sta rti ng Out: The Trom powsky Atta ck
3 Bxf6
3 Nc3 - 1 02, 1 06
3 dS Qb6 4 Nc3 Qxb2 5 Bd2 Qb6 6 e4 - 1 1 1, 1 16, 1 19
3 ...gxf6 4 ds Qb6 s Qc1 fs
6 c4 - 127
6 e3 - 127, 131
6 g3 - 135, 139
The Popular 2 . Ne4

1 d4 Nf6 2 BgS Ne4
3 Bh4
3 h4 - 143, 150
3 cs
..
2 64
I ndex of Variations
3 ... g5 - 153
3 ... d5 - 154
4 f3 g5 5 fxe4 gxh4 6 e3 Bh6 - 155, 159
The Modern Preference: 2 .. Ne4 3 Bf4

1 d4 Nf6 2 Bg5 Ne4 3 Bf4
3 ...ds
3 ...d6 - 166
4 e3
4 Nd2 - 167
4 f3 Nf6 5 e4 (5 Nc3 - 1 92) 5 ... dxe4 6 Nc3 - 1 94, 1 99
4 cs
...
265
Sta rti ng O ut: The Trom pows ky Attack
4 Bf5 - 1 70, 1 74
...
4 e6 - 1 71
...
5 Bd3 - 1 79, 188
The Main Line: 2 ... Ne4 3 Bf4 cs

1 d4 Bf6 2 Bg5 Ne4 3 Bf4 c5
4 f3
4 dS Qb6 5 Bel
s g6 - 2 1 2, 2 1 6
...
s e6 - 2 1 2
...
4.... Qa5+
4 Nf6 - 2 1 9, 222
...
5 C3 Nf6
266
I ndex of Va riations
6 d5
6 Nd2 cxd4 7 Nb3 Qb6 (7...Qd8 - 245; 7....Qf5 - 246, 253) 8 Qxd4 - 246, 256
6...Qb6
6 d6 - 225
...
6 ...e6 - 226
7 Bel
7 b3 - 228
7 e4 Qxb2 8 Nd2 Qxc3 9 Bc7 - 230, 234
7 e6 8 c4 - 236, 241
...
8 e4 - 236
267
I n d ex of Co m p l ete
G a m es
Adams.M-Georgiev.K, Burgas 1993 209

. . . .........................................................................
Arzumanian.G-Brikov.l, Tula 2007 .

. . . . . . . . . . . . . ................... . . ..... 159 ......................................
Bigg.A-Hanley.C, Street 2004 77

. . . . . .. . . . . ................................. ......... ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bigg.A-McPhillips.K, Street 2004 . .

............................................... .............. 29 ....... ..............
Brandics.J-ott.F, Kecskemet 1990 ................................................................................ 106

Bruzon Bautista.L-Nijboer.F, Wijk aan Zee 2004 . 116
................................................. ...
Coleman.J-Skacelik.P, Brno 2004 89

............... .....................................................................
Gallagher.J-Nedev.T, E u ropean Club Cup, St Vincent 2005 . 222 ............. ....................
Galyas.M-IIincic.Z, Torokbalint 2004 .

............ . . . . . . . . . . . . . . . . 253 . . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Galyas.M-Voiteanu.G, Buda pest 2006 . . .... 47

....... . . . . . . . . . ....................... ..... ..... . . . ................
Georgiev.K-Horvath.A, European Club Cup, Fuegen 2006 . . 20 ........................... ... .......
lvanchuk.V-Jobava.B, Hava na 2005 . . . . ... .... ..... 1 88

............................................ ................
Kanep.M-Holmsten.A, Finnish league 2006 . . . . . . . . . .

....... 150 ....................... ......... ... ... ......
Kasparov.G-Carneiro.M, Sao Pa ulo simul 2004 . 198

................................. ......................
Kramnik.V-Tkachiev.V, Ta l Memoria l Bl itz, Moscow 2008 . . . . . . 16 .............. ... ... ..... . . .....
Mensch.E-Giffard.N, French league 2004 . . 50

.............. . . . . . . . . . . ..................................... .....
Miladinovic.I-Martorelli.A, Rome 2005 . 95

..................... . . . . .. . . ........................ ...................
Milov.V-Landenbergue.C, Swiss League 2002 . .. .. ..

................ 241 ........ . ............... . ... . ......
Nuesken.N-Sawatzki.F, German league 2007 .. . . ...... . . 74 ..... ............ .................. .............
Palliser.R-Taylor.P, Golders Green Rapidplay 2004 . .. . . 63

. .............. . ........................... .. .
Pavlovic.M-Aieksandrov.A, Vrnjacka Ba nja 2005 . . .. . . . 233

. . . . . .. . . . . . . . . . . . . ..... . . . . ....... . ....... . .
Popovic.M-Andersen.M, Reykjavi k 2009 . . . . . . ... ..

.... . ... ... . . . 65
....... . . ... . .............................. ...
Rahman.Z-Carlsen.M, Dresden Olym piad 2008 . . .. .

... 85
....................... ....... . ...................
268
I ndex of C o m p l ete G a mes
Schneider.S-Ernst.T, Swedish Team Championship 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . .. 256

Siewert.W-Noble.M, Correspondence 2007 ............................................................. 119
Stefanova.A-Stellwagen.D, Wij k aan Zee 2004 ........................... . . . . . . . . . . . . . ............... 216
Tu Hoang Thong-Vui.A, Ottawa 2007 ............................................................................ 35
VigusJ-HodgsonJ, British Championship, Street 2000 .................................. ....... 131
Wells.P-Popovic.P, Austrian League 2002 . . . . . . . . . . . . . . .......................................... ......... 139
Winants.L-Mamedyarov.S, Bled Olympiad 2002 ................ ............................ ......... 1 74
2 69

Starting Out The Trompowsky Attack - Palliser PDF

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Starting Out The Trompowsky Attack - Palliser PDF

Încărcat de

Drepturi de autor:

Formate disponibile

starting out:

Copyright 2009 Richard Palliser

All rights reserved. No part of this publication may be reproduced, stored in a re

British Library Cataloguing-in-Publication Data

ISBN: 9781 85744 562 6

For the chessplayers at the Minster Inn

EVERYMAN CHESS SERIES

Typeset and edited by First Rank Publishing, Brighton.

2 2 g6 and Minor Lines

3 The Positional Choice: 2 e6 ... 54

4 The Uncompromising 2 cs ... 100

5 The Popular 2 ... Ne4 142

6 The Modern Preference: 2 ... Ne4 3 Bf4 163

7 The Main Line: 2 ... Ne4 3 Bf4 cs 201

Index of Variations 260

Index of Complete Games 268

The Structure of this Book

Diagram 2 Diagram 3 (B)

3 Bh4 c 6 i s a pretty tough nut to crack.

m The Sol i d 3 exf6

m The Dyn a m ic 3 . gxf6

m Wh ite Dec l i n e s to Exc h a n ge

Diagram 1 (W) Diagram 2 {W)

The Solid 3 ...exf6

Diagram 3 (B) Diagram 4 (B)

knight on f6 to recapture on d 5 or i f White has a knight on c3), but more often he

Diagram 5 (B) Diagram 6 (B)

NOTE: Due to the omission of ...Be6, White has managed to break

Diagram 7 (B) Diagram 8 (B)

Diagram 9 (W) Diagram 10 (B)

White might, th ough, begin with 5 B d3 fS an d n ow 6 N d2 is sensible, but a lea ding

Trompowsky authority preferred to plough a creative furrow i n I.Miladinovic

Common. Inexperienced Trompowsky players have been known to worry about

Diagram 11 (B) Diagram 12 (B)

q uality, as one would expect from such a strong player.

Diagram 13 (B) Diagram 14 (W)

Diagram 15 (B) Diagram 16 (W}

12 h 5 Nf6 13 hxg6 fxg6 (Diagram 16)

Diagram 17 (B) Diagram 18 (B)

TIP: When a pawn or two ahead in an endgame, remember to con

1 d4 Nf6 2 Bg5 d5 3 Bxf6 exf6 4 e3 c6 5 g3 Bd6 6 Bg2 0-0 7 NeZ Be6

Championship, Scarborough 2004, which continued 1 0... Nd7 1 1 c4 Nf6 1 2 Nbc3 a6

Diagram 19 (W) Diagram 20 (B)

down the c-file i s also possible.

Diagram 21 (B) Diagram 22 (B)

19 a4 a6 20 f3 Nf6 21 Ra3 Qb7 22 Qd2 RaB 23 Rfa1 Reba

43 Bxg6 Qa7 44 f51

Diagram 23 (B) Diagram 24 (W)

The Dynamic 3 ... gxf6

WARNING: Despite having been used by a number of strong players,

Diagram 25 (W) Diagram 26 (B)

Diagram 27 (W) Diagram 28 ( B)

Diagram 29 (W) Diagram 30 (W)

Diagram 31 (W) Diagram 32 (W)

NOTE: High-level correspondence games are an excellent source of

after 1 7 NhS f4 18 Nxf4 e5.

Diagram 33 (B) Diagram 34 (B)

3o... Rxd2l 31 Kxd2 Kd6 32 Ke3 Kc51?

Diagram 35 (W) Diagram 36 (B)

Wh ite Declines to Exchange

Diagram 37 (W) Diagram 38 (W)

b) However, 3 ... Nbd7 4 Ngf3 g6 doesn't seem as good as against 3 e3 in view of

TIP: Never forget about transpositions when planning your reper