queen's

Indian
defence

by Jacob Aagaard

EVERYMAN CHESS
Everyman Publishers pie www.everymanbooks.com
First published in 2 002 by Everyman Publishers plc, formerly Cadogan Books plc,
Gloucester Mansions, 140A Shaftesbury Avenue, London WC2 H 8HD

Copyright© 2 002 Jacob Aagaard

The right of Jacob Aagaard 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 retrieval
system or transmitted in any form or by any means, electronic, electrostatic, magnetic
tape, photocopying, recording or otherwise, without prior permission of the publisher.

British Library Cataloguing-in-Publication Data

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

ISBN 1 85744 300 4

Distributed in North America by The Globe Pequot Press, P.O Box 480,
2 46 Goose Lane, Guilford, CT 06 437-0480.

All other sales enquiries should be directed to Everyman Chess, Gloucester Man­
sions, 140A Shaftesbury Avenue, London WC2 H 8HD
tel: 020 7539 76 00 fax: 02 0 7379 4060
email: chess@everymanbooks.com
website: www.everymanbooks.com

EVERYMAN CHESS SERIES (formerly Cadogan Chess)

Chief advisor: Garry Kasparov
Commissioning editor: Byron Jacobs

Typeset and edited by First Rank Publishing, Brighton.

Production by Book Production Services.
Printed and bound in Great Britain by The Cromwell Press Ltd., Trowbridge,
Wiltshire.
Everyman Chess

Popular opening books:

18574 4 218 0 Unusual QG Declined Chris Ward

18574 4 253 9 Alekhine's Defence Nigel Davies
18574 4 256 4 Queen's Gambit Declined Matthew Sadler
18574 4 232 6 French Classical Byron Jacobs
18574 4 2 81 4 Modem Defence Speelman & McDonald
18574 4 292 x Symmetrical English David Cummings
18574 4 290 3 c3 Sicilian Joe Gallagher
18574 4 2 42 3 Offbeat Spanish Glenn Flear
18574 4 262 8 Classical Nimzo-Indian Bogdan Lalic
18574 4 2911 Sicilian Grand Prix Attack James Plaskett
1 8574 4 252 0 Dutch Stonewall Jacob Aagaard
18574 4 257 1 Sicilian Kalashnikov Pinski & Aagaard
18574 4 276 8 French Winawer Neil McDonald

Books for players serious about improving their game:

1 8574 4 226 1 Starting Out in Chess Byron Jacobs

·
18574 4 2318 Tips for Young Players Matthew Sadler
18574 4 236 9 Improve Your Opening Play ·Chris Ward
18574 4 2 41 5 Improve Your Middlegame Play Andrew Kinsman
18574 4 2 46 6 Improve Your Endgame Play Glenn Flear
18574 4 223 7 Mastering the Opening Byron Jacobs
18574 4 228 8 Mastering the Middlegame Angus Dunnington
18574 4 233 4 Mastering the Endgame Glenn Flear
18574 4 238 5 Simple Chess John Emms

Books for the more advanced player:

18574 4 233 4 Attacking with 1 e 4 John Emms

18574 4 233 4 Attacking with 1 d 4 Angus Dunnington
1 8574 4 219 9 Meeting 1 e 4 Alexander Raetsky
18574 4 22 4 5 Meeting 1 d 4 Aagaard and Lund
18574 4 273 3 Excelling at Chess Jacob Aagaard
queen's
Indian
defence

by Jacob Aagaard

EVERYMAN CHESS
Everyman Publishers pie www.everymanbooks.com
CONTENTS I

Bibliography 6
Preface 1
Introduction 9

White plays without g2-g3

1 Opting for Nothing: 4 �f4 and 4 �gS 16

2 The Petrosian System: 4 a3 24
3 5 'ii'c2 and 5 �gS 62
4 5 'ii'b3 76
5 4 e3 87

Lines with g2-g3

6 4 g3 i.b4+ and 4 ... �a6 without 5 b3 93

7 4 g3 i.a6: Main Line with 5 b3 111
8 4 g3 i.b7 121

Index of Complete Games 143

 PREFACE I

The Queen's Indian Defence (arising after 1 d4 lLif6 2 c4 e6 3 lLi£3 b6) is one of the most
popular openings ai: both club and international level. The reason for this is obvious - ever
since its introduction at the beginning of the century, the Queen's Indian has proved to be the
most dependable defence for Black against 1 d4, together with its companion, the Nimzo­
Indian (3 lLic3 J.b4). Additionally, most players like to create some kind of imbalance and
avoid structures of a symmetrical nature, such as those found in the Queen's Gambit Declined
and the Slav, although these openings also have their followers.
When I was originally asked to write a book about the Queen's Indian I asked if I could
write a two volume work in order to delve deep into the different positional aspects of the
opening. My publisher judged that most people would be more int\:rested in a single volume,
standard work, which is what I have ultimately agreed to. In order to achieve this I had to
make some decisions concerning the structure of the book because it was evident that certain
things which, in an ideal world, merited space, would have to be omitted. First, it was impossi­
ble to provide a deep positional study in a naturally restricted book form, as I did with the
Dutch Stonewall. In this case I counted 19 different pawn structures that would have to be
discussed, and the only way this could be done would be via a multitude of arrows and short,
abstract words - something other writers have done before, but for which I have found little
use for myself and, consequently, believe is the same for others. Therefore, other than brief
discussions about the isolated pawn structures, below, this book is more a traditional work on
the Queen's Indian. Moreover, when it came to the lines to include I had to make numerous
decisions, minimising the material as much as possible. In ECO there are about 1 20 pages de­
voted to the QID and the positions that can result via transpositions which, in a book like this,
would be around 300 pages. Meanwhile, ECO includes only those games that were previously
published in Informant, so they actually exclude quite a lot of important games! I decided to
include games from the very highest level -more or less ;n the games in this book, as well as all
the sidelines, originate from Grandmaster competition. But this only got me down to 20,000
games! Then I made the obvious decision: only critical lines should be represented. But merely
eliminating mistake after a mistake was far from enough. I also had to decide that some moves
were not critical in any way and that, therefore, they had no role in the book because nobody
would try to memorise the continuation, while looking it up afterwards also appears to me to

7
Q u e e n 's In dia n D e fe n c e

b e rather indifferent. Finally, in 'normal' positions, those 'normal' moves that give Black sev­
eral ways to equalise have been omitted.
The result is a traditional work of theory, aimed at the tournament player and thus designed
for practical use. I feel that I have been good at finding the critical lines in modem play, and
picking out the most important games. I hope you will find this a useful and enjoyable book.

Jacob Aagaard
Copenhagen
July 2002

8
 INTRODUCTION
Theoretical Overview
Here, briefly, is a an overview of the status of
the various lines found in the following chap­
ters.
Chapter 1
4 .tf4 is by no means an attempt to create
an opening advantage. I have a feeling that
the bishop is not very well placed in this
variation as Black can still play ... d7-d6 in
order to limit its scope. In fact Black can
equalise in more than one way.
4 .tg5 should normally lead to transposi­
tions because the best move for White on the
next move is 5 lilc3. Ideas with 5 lilbd2 ap­
pear to be nothing other than a means of
avoiding opening theory and promise White
nothing.
Chapter 2
4 a3 is, generally, an interesting system. As
there is no line that guarantees an advantage
against the Queen's Indian White should
select his approach based on style and taste,
although I do feel that 4 a3 is not the most
annoying move order for Black. 4 lll c 3, with
the idea of meeting 4... �b7 With 5 a3!, ap­
pears to avoid some of Black's most popular,
fashionable lines.
Anyway, after 4 a3 Black has the following
I
possibilities.
4 . . d5!? leads to standard positions after
.
5 cxdS exdS; these are normally very slightly
better for White, but are ultimately a matter
of taste. Such positions are discussed
throughout the book and are examined in
some detail in Chapter 1 .
4 . c6? ! has been played only once at the
..
top level and will 'rarely be repeated. White
should secure a small plus in a symmetrical
position, which is always unpleasant for the
second player.
4 . c5!? leads to a sort of Benoni set-up
..
with the bishop on a6 after 5 dS �a6 where,
for the time being, theory promises Black a
good game (yet nothing more). I have a feel­
ing that White will eventually find a way to
prove a small plus here, although this has yet
to happen. This line has gained significantly
in popularity thanks to Kasparov's convinc­
ing victory against Gelfand.
4 . .ta6 is the most popular system these
..
days, leading to a variety of different posi­
tions, goth hedgehog and the more flexible
systems with ... cS and cxd4. This should lead
to equality and pleasant play for Black if he is
sufficiently prepared, but it is a system that is
still open for improvements and new ideas.
4 . . . .tb7 is the other main line (besides
4... �a6), and a solid reply. White has tried a
9
Q u e e n 's In dia n D e fe n c e
number o f different ways to gain an advan­
tage but it still seems that Kasparov's idea
from the 80's with 5 ltic3 dS 6 cxdS exdS 7
'ilfc2! is the only real chance of giving Black a
headache. I believe that these positions are,
perhaps, equal after correct play, but it
should be home in mind that this correct
play could often prove difficult to produce.
Chapter 3
4 lll c 3 i.b7 5 Wc2 is harmless and Black
equalises easily, but the chapter is also mainly
about 5 i.g5. If Black has played 4... i.b7,
then 5 a3! is the best way to fight for an ad­
vantage, and the move order I consider the
most dangerous for Black (what White
should play after 4... .ib4 is another question
- see below). 1bis particular move order
gives Black the chance to equalise with 5 .. h6 .
6 i.h4 i.e7!, after which there are no prob­
lems for the second player.
4 ...i.b4 is a logical response, often played
to avoid the 5 a3 transposition after 4... i.b7.
These positions can arise after 1 d4 ltif6 2 c4
e6 3 ltic3 .ib4 4 ltif3, where both 4... cS and
4... b6 are main lines, the latter obviously
transposing to the QID.
5 i.g5 and now Black has ):wo ways to
equalise. The first is 5 .. h6 6 j_h4 i.b7 7 e3
. doing so leads to a clear gain of some sort.
(the move that has been proven to give Black
most problems, unlike7 1i'c2, for example,
which does not impress) 7 ... .ixc3 8 bxc3 d6
when a Nimzo-Indian structure has emerged,
with Black being quite comfortable. The
other comes after 7 ... gS 8 .ig3 ltie4. Al­
though this has been played many times and
the 'opening' theory can continue to 30
moves or more, I still feel that there is room
for improvement and that Black should not
trust the verdict of equality completely. It
should be noted that White must play 9 'ilfc2
and that the gambit with 9 ltid2 seems rather
dubious.
Chapter 4-
4 lll c 3 i.b4 5 Wb3! - I have given this
10
move an exclamation mark simply because it
is the most unpleasant for Black to meet.
1bis hitherto offbeat line became a main line
with the match Timman-Seirawan, Hilver­
sum 1 990, where Seirawan employed it in all
his White games with convincing results.
Since then it has been played by some of the
other world's elite, with varying degrees of
success. Black has three main options.
5. . . a5 does equalise but the resulting po­
sitions seem to be a little passive for Black
because he will often have to play on against
the two bishops.
5 . . . We7 is also sufficient for equality.
However, one should always remember that
if Black first has to fight to solve various
problems before equalising, then this only
means that both players have equal chances
in the position itself, the assessment exclud­
ing other factors such as familiarity with the
position, time used on the clock and playing
strength. Consequently, in the reality of
competition, it is not necessarily always
'happiness' when one equalises.
5...c5 is the main move and, it seems,
also the best. After 6 a3 .iaS (the point)
Black is allowed to keep the two bishops for
the time being, and he will exchange only if
Then comes 7 i.gS i.b7!, as played several
times in the match between Seirawan and
Adams in Bermuda 1 999. Black should be
able to keep the game level, but this line is
still in its early childhood, and new discover­
ies will most certainly emerge with time. It
should be noted that this line is full of all
kinds of tactical tricks and options, which is a
major reason why I recommend it. Better
preparation will pay off.
Chapter 5
4 e3. 1bis chapter provides some easy to
use ways for Black to equalise against this
(from a theoretical point of view) harmless
system. 1bis might be a good place for a little
warning: If you (as Black) do not know what

you are doing against this system, you can
quickly find yourself in terrible trouble. See,
for example, Yusupov-Beliavsky (Game 33),
where top players produce moves about
which a deeper theoretical knowledge would
have made them think twice.
Chapter 6
4 g3 sees the beginning of the great saga of
White's most popular. I have always played 4
g3 when facing the QID, but I have never
mistaken the great taste I had for the posi­
tions with an advantage, and neither, I am
sure, have the majority of the players who
employ the fianchetto. Black has a consider­
able variety of choices, but I would recom­
mend playing something along the main lines
with 4 ... i.a6 while, for White, I would seri­
ously ask you to consider 4 �c3. Of course,
perhaps you play these lines to avoid fun, just
like I do on a bad day.
4 . . . .i.b4+: The lines with ....tb4+ gener­
ally benefit from the inclusion of the prepara­
tory 4... .ta6. The only exception is 4 ... i.b4+
5 i.d2 .txd2+ 6 'i'xd2 i.a6!, which trans­
poses to 4... i.a6 5 b3 i.b4+ after 7 b3, which
is the best move. Generally these lines seem
to be me to be slightly worse for Black, but
never anything less than that. They are not
very exciting and the second player rarely has
any reason to hope for a full point.
4 . . .ta6 is probably a better move than
. positions -there is no pressure.
4... i.b7, and certainly more modem. The
reason why this appears to be so clever is
that almost no matter which way White pro­
tects the c4-pawn, Black can always play
5 ... i.b7 and claim that the white pieces are
no better placed than they were the previous
move. The idea certainly appeals to many
players. Black should equalise in all the main
lines; perhaps there is a theoretical plus in the
main line, but that is only something for
players to worry about if they cannot draw
drawn endgames ... We will come to this but,
first, let us look at White's options to be pre­
sented in this chapter.
Intro du ctio n
5 'i'c2 is completely harmless. Black
probably has several ways to equalise, but the
simplest is 5 ... dS!, when there is no doubt
about the evalua�on of the position.
5 'i'b3 might look stupid and, after
5 ... �c61 (best) Black is able to exploit the
misplacement of the queen immediately to
further his own development. However,
White's move should not be mistaken for
being bad since Black only equalises - noth­
ing more.
5 �bd2 is my own choice, although it
should not give Black any problems. Black
has some different and interesting choices
here. 5 ... .tb4!? is very interesting and leads to
an unbalanced situation in which, while
White might be slightly better, the stronger
player will be in the driving seat. All the lines
with a quick ... c7-c5 seem a bit dangerous to
me, but (of course) I might be mistaken. It all
comes down to an evaluation of Game 39
and the lines featured therein. Finally 5 ... dS
should equalise without too much effort, and
that is the main reason why this is not a main
·
line.
5 'i'a4 is popular but simply lacks punch.
5 ... c6 is rather dubious, but after 5 ... cSI Black
has several ways to generate a good game.
Actually I am normally pleased when I see
this move from Black as I simply cannot see
how White can play for a win in the resulting
Chapter 7
5 b3 is the main line. And actually the only
argument I see for playing the sytem with 4
g3 is to prove that this move is an advantage
for White rather than a weakening of the
dark squares. Having investigated the rele­
vant V!lriations I must say that I still cannot
see this as anything other than a weakness,
and it is for this reason that I recommend
that everyone dump 4 ... i.b7 and choose a
line from below.
5 . . . d5!? is not in itself poor but the idea
is to answer 6 i.g2 with 6 ... dxc4, which is
11
Q u e e n 's In dia n D e fe n c e
unsound. 6 ... .i.b4+ might still be theoretically
okay, but I do not know many people who
like to play such positions.
5 c5 also looks slightly fragile, although
..•
it might work as a surprise. But if you are
looking for good positions, look elsewhere.
5 . . . b5!? is Adorjan's move, about which a
positional verdict is quite simple. If Black
finds success in ... a5-a4xb3 he can look for­
ward to a good position. But why should
White allow this? Against a logical treatment
Black should be suffering from his inferior
pawn structure and be slightly worse. How­
ever, 5 ... b5 is not a poor move and it does
offer Black chances to steer the game away
from more theoretical lines.
5 . . . i.b7 6 i.g2 i.b4+ is an interesting
option which promises something close to
equality after both 7 i.d2 a5 (Adams'
choice) and 7 . . . c5 (Korchnoi's choice) . The
struggle often hinges on deep strategic un­
derstanding, and I would not recommend
that you undertake the responsibilities of
Black if this book is the only material you will
look at. But if you work through the games
and find them interesting, then I will certainly
not discourage you from seeking further
knowledge and subsequently including these
lines in your repertoire. In fact most players
with White are unaware of what is best after
5 . . . i.b7 and can thus find themselves on
unfamiliar territory when Black refrains from
entering the long main lines.
The main lines after 1 d4 lbt6 2 c4 e6 3
lbf3 b6 4 g3 i.a6 5 b3 i.b4+ 6 .td2
i.e7 7 i.g2 c6 have been debated endlessly
for the last decade, and the verdict is - not
surprisingly - that it is very difficult for White
to gain an advantage; and if he succeeds it is
only of 'scientific' value - that is if Black is
well prepared! I have looked at 8 0-0 d5 9
'il'c2, which is the new idea at the top level,
but it seems to be harmless, and if anybody
takes any risks in the games played, it has
mainly been Black. 8 .i.c3 d5 9 llle 5 is the
main line and has been treated as something
12
which is highly respected yet no longer has
any place among the living. The story of the
main line is a funeral - I know that this is an
easy thing to say for a believer, so let us keep
it blurred for now, and see if the evaluation
should ever change.
Chapter 8
4 . . . .tb7 is still popular with some top
grandmasters. Only the absolute elite such as
Kramnik, Anand and Karpov stick to
4 ... .i.a6, while Tiviakov, Timman and
Korchnoi still have some affection for this
old main move. Personally I have no doubt
about the correctness of 4 ... .i.b7, but I do
feel that the play after 4... .i.a6 is more flow­
ing and should suit most players - at least
those who I know - better. But there are
some players with a very relaxed attitude to
chess, and these do best with the calm and
strategic 4 ... .i.b7. Anyway, let us turn to the
lines arising after 5 i.g2:
5 . . . c5 is simply positionally suspect and
Black has nothing to show for it but a very
modest - and very temporary - initiative. Stay
away from this move.
5 . . . i.e7 is the main move (5 ... .i.b4+ be­
longs to the comments given above for
4... .i.b4+). Now White has two paths. First,
after 6 0-0 0-0, there are some minor side­
lines.
7 b3 is harmless. Actually, I play this oc­
casionally in order to reach an equal endgame
(you know - hot summer days when you
would like to be at the beach ... ). Nonetheless,
do take a quick look at the easiest equalising
methods.
7 d 5 ! ? was once wildly popular. First Po­
lugaevsky used it, then Kasparov. Now -it's a
joke. Black has several ways to equalise and
can probably also play for a win without any
great risks -if he is well prepared.
7 J:e 1 was modem a few years ago. The
problem seems to have been solved with
7 ... llla 6!, after which the new try is 8 a3.
Come on guys! Call that an advantage?

7 �c3 raises nwnerous issues. First White
has 6 �c3 with some different ideas, and this
is the most popular line at the moment. In
the event of 7 �c3 I would go for the highly
exciting 7 ... �a6!? - which is quite compli­
cated - rather than the somewhat drawish
7 ... �e4. Anyway, after 6 �c3 Black usually
replies 6 ... �e4. Then 7 'ifc2 is hannless. In­
stead after 7 .td2 it seems to me that only
7 . . .tf6! guarantees Black a good game. Of
.
course this is tested again and again at the
top level, but it seems to me that those lines
that are given as leading to an advantage for
White (8 1i'c2 �xd2 9 'i'xd2 is an example)
are often the lines I would fear least, while
those given as hannless might conceal a trick
or two. There are some tricks involving de­
laying castling but after 7 ...tf6 these do not
. them appropriately before White's perma­
pay off (as they do against all the other lines).
That's all folks. Enjoy your games, no
matter what faith or choice or line you might
take.
The isolated d5-pawn
The QID is characteristic for its many differ­
ent lines and prototypical positions. It can be
compared to the Sicilian, where the struc­
tures after the opening vary from, for exam­
ple, the wildly different Sveshnikov, Dragon
and Hedgehog variations. In the QID I have
found 1 9 different prototypical variations
which, of course, might themselves contain
nwnerous variations. Only one pawn struc­
ture (along with its distinct variations) is tran­
scendental universal and appears, in some
form, in all lines. I am referring to a situation
such as the following example:
Gelfand-Leko
Cannes 2002
1 d4 lllf 6 2 c4 e6 3 lllf3 b6 4 g3 i.a6 5
"iia4 ..i.b7 6 ..i.g2 c5 7 dxc5 ..i.xc5 8 0-0
..te7 9 lll c 3 0-0 1 0 ..i.f4 lll a 6 1 1 J:fd 1
lll c 5 1 2 "ifc2 "ii c 8 1 3 J:d4 d5 1 4 cxd5
exd5 1 5 J:dd 1
Intro ductio n
This is it! Black has accepted an i solated
pawn on dS. As is usually the case this gives
Black considerable· free play for his pieces,
but he will have to keep them active or use
nent advantage of a superior pawn· structure
begins to tell. Normally exchanges tend to
favour White, as the removal of pieces serves
to enhance the significance of the static as­
pects of a given position. This does not nec­
essarily mean that endgames are bad for
Black. Typically White might have weakened
his queenside with b2-b3 at some point,
when Black will be able to gain counterplay
with a knight placed on e4, heading for c3.
Gelfand-Karpov, Game 31 is an example of
how activity can last long into the endgame.
In this game Black creates compensation
for the isolated d-pawn by advancing his g­
pawn, creating some potential threats on the
light squares around White's king. Gelfand
then decides to eliminate the g-pawn and, in
return, takes a weak pawn on e3 (which is
·
just as weak as dS).
1 5 . . . lllce4 1 6 J:ac 1 ..i.c5 1 7 e3 lllx c3 1 8
'i'xc3 llle4 1 9 °ifd 3 "iie 6 20 ..i.e5 J:ac8
21 ..i.d4 ..i.e7 22 °iff1 J:xc 1 23 J:xc1 J:c8
24 'ifd 1 i.d6 25 ..tn °iff5 26 �g2 J:xc 1
27 'ifxc 1 f6 28 ..i.d3 �7 29 'ifc2 g5 30
"ife2 . h5 3 1 llld 2 g4 32 f3 gxf3+ 33
"ifxf3 %-%
In the following game Black places all his
13
Q u e e n 's In dia n D e fe n c e
pieces in a more o r less ideal way. I recom­
mend you play through this game carefully
and note how Black's pieces generate pres­
sure.
Kramnik-Hi.ibner
Yerevan 1 996
1 lllf3 lllf 6 2 c4 b6 3 g3 i.. b 7 4 i..g 2 e6
5 0-0 .te7 6 lll c 3 0-0 7 J:e 1
1bis is Kramnik's way of playing the 7
l:tel variation. The threat of 8 e4 often
makes Black respond with 7 ... dS, but 7 ... cS,
along the lines of the hedgehog set-up, is a
sensible move.
7 . . . d5 8 cxd5 exd 5 9 d4 llla 6!
1bis is where the knight belongs. Black
has no reason to fight for d4 in these posi­
tions because it is a fight he will only very
rarely win. Rather he focuses on bringing the
knight to an active post, the ideal spots being
cS and, primarily, e6.
1 0 .tf4 c5 1 1 dxc5 lllx c5 1 2 J:c1 lllfe4
1 3 lll b 5 llle 6 1 4 i..e 5 'i'd7 1 5 lll bd4
J:ac8 1 6 l:lxc8 J:xc8 1 7 .i.h3 .tc5 1 8 a3
a5 1 9 e3 l:le8 20 .i.g4
Here a draw was agreed.
If we look at the final position we see that
all Black's pieces have a function. The knight
on e6 monitors both d4 and f4 and can, po­
tentially, jump to cS, from where an invasion
might be possible. The knight on e4 keeps a
14
good eye on f2 and also prevents White from
putting his pieces completely at ease. The
bishop on cS exerts pressure on d4 and e3
and will, in the long run, prevent White from
playing f2-f3, or at least serve to remind
White that e3 would then become a weak­
ness. The bishop on b7 looks bad but can
often come into play with ... i.a6 or, perhaps,
even the pawn sacrifice ... d5-d4 on some
occasions. Only the heavy pieces are not easy
to place. Usually the rooks should be at d8
and e8 if Black is playing for ... d5-d4, and at
c8 and e8 if he (like here) has more or less
ignored the fight for the d4-square.
White, on the other hand, tends to have
problems in finding a good square for the
dark-squared bishop. Principally it is best
placed on b2, but that, of course, cannot
happen without White weakening his queen­
side with b2-b3 at some point. As I men­
tioned above, this will give the knight, which
is usually placed on e4, a future. Additionally,
the bishop on g2 can often find that it is
nicely placed, but with the knight on e4 it
tends to be looking at a piece that cannot
realistically be removed (the exchange would
be far from ideal anyway) for fear of the re­
sulting weaknesses on the light squares.
Let us take a look at another variation of
the structures after ... d7-d5 and cxdS exdS,
where White is in no hurry to exchange on
cS.
Karpov-Spassky
Riga 1 975
1 d4 lllf 6 2 c4 e6 3 lllf3 b6 4 g3 i.. b 7 5
i.. g 2 .te7 6 lll c 3 0-0 7 'i'c2 d5 8 cxd5
lllx d5 9 0-0 llld 7 1 0 lll x d5 exd5 1 1 l:ld 1
ll:lf6
Where should White place his pieces here?
I could use this as an exercise for my pupils
because the answer is not so obvious if you
are unfamiliar with the position. Well, the
queen is fine (£5 or d2 look appropriate for
later), the queen's rook goes to cl, the other

rook is already perfectly placed, the queen's
bishop has no obvious post and the other
bishop is good. But what about White's
knight? As Black will not fight for d4 the
knight is not required on £3, but where would
we prefer it to be? The correct answer is f4,
from where it exerts considerable pressure
on dS. The following manoeuvre is standard
and will not have come as a surprise for
Spassky.
1 2 ibe5 c5 1 3 dxc5 .i.xc5 1 4 tbd3 .i.d6
1 5 .i.f4!
This bishop has no ideal square so seeking
to remove it is quite logical.
1 5 . . . .t!.eS 1 6 e3 ibe4 1 7 ..i.xd6 'iix d6 1 8
ibt4
White has an advantage. The d4-square is
a good outpost for the heavy pieces and both
the knight and the bishop are well placed.
But most importantly all the play will re­
- 1 6 ibb5!
volve around White's pressure on the dS­
pawn, so Black will not be able to generate
sufficient active play to compensate for the
weakness.
1 8 . . . l:!.acS 1 9 'ii'a4 'ii'e 7 20 'it'xa7 ibxf2
21 ibxd5 ..i.xd5 22 'it'xe7 ibxd 1 23 .l:k 1
i:tb8 24 'it'b4 .i.xg2 25 �xg2 ibxe3+ 26
'it>g 1 i:te6 27 'ii'f4 l:ld8 28 'ii'd4 l:!.de8 29
'ii'd 7 ibg4 30 l:!.c8 ibf6 31 l:!.xe8+ i:txe8
32 'ii' b 7 l:le6 33 'ii' b 8+ ibe8 34 a4 g6 35
b4 �g7 36 'ii'b 7 h5 37 h3 �f6 38 �g 2
l:d6 39 a5 bxa5 40 bxa5 l:!.e6 41 a6 ibc7
42 a7 l:!.e7 43 'ii'c 6+ �e5 44 'it>f3 1 -0
Intro du ctio n
The same manoeuvre is also possible even
if the pair of knights has yet to be exchanged.
The following example illustrates how White
handles the possibility of ... c5xd4.
Beliavsky-Chuchelov
Ohrid 2001
1 d4 ibf6 2 c4 e6 3 ibf3 b6 4 g3 .ta6 5
b3 d5 6 cxd5 exd 5 7 ..i.g2 ..i.b4+ 8 ..i.d2
..i.e7 9 0-0 0-0 1 0 ibc3 ..i.b7 1 1 ..i.c 1 iba6
1 2 ..i.b2 l:le8 1 3 l:lc 1 ..i.f8 14 ibe5 c5 1 5
ibd3 cxd4
Black's trade in the centre is probably not
the best. Clearly Black was familiar with the
following manoeuvre, but it still seems to be
a particularly unnecessary concession to give
up the d4-square when having to face a well
placed knight on f4.
The main idea. In this position the knight
transfer is rather straightforward, but some
players have a tendency to forget about this if
White has the option of recapturing on d4
with the queen. Now White has an edge.
1 6 . . .'ii'd 7 1 7 ibxd4 ibe4 1 8 ibf4 fled8 1 9
..i.h3 'i'es 20 a3 ibg5 2 1 .tg2 ibc5 22 b4
iba4 l3 ..i.a1 ibe4 24 l:!.c7 l:lab8 25 ibf5
d4 26 'iic 2 l:bc8 27 ibxd4 a5 28 b5
i:txc7 29 'ii'x c7 'it'd7 30 'ii'c 2 ibac5 31
ibd3 :ea 32 ibc6 ibxd3 33 .i.xe4 ibc5
34 .i.xh7+ �h8 35 ..i.f5 ibe6 36 l:!.d 1
'ii'e8 37 'it'e4 1 -0
15
 CHAPTER ONE I
Optin g for Nothin g :
4 i.f4 and 4 i. g 5

1 d4 lDf6 2 c4 e6 3 lDf3 b6 4 J.f4/4 J.g5 ties exist, take a look at them once or twice
There are some players, no matter what and then let your understanding of the posi­
their rating, who always try to avoid main line tion be your guide. The three games I have
theory. This can be for different reasons, of presented here should more than cover what
course - one is simply a surrender to the you need to know about the system.
modem reality of the enormous amount of
games in a standard database, while another
is the need for an independent and original
way of playing, based more on psychology
than actual playing style.
Both 4 .i.gS and 4 .i.f4 are ways to stay
clear of all established theory in the Queen's
Indian. Neither is dangerous for Black, but if
you are content to get your pieces out and
then play a game on practically alien soil,
then 4 .i.f4 in particular makes good sense,
as can be seen from :Miles' and Bareev's rela­
tive success with this approach.
4 J.b7 5 e3 J.b4 +
...

Game 1 This is an interesting choice. The simpler

Miles-Unzicker
Johannesburg 1979
1 c4 lDf6 2 lDf3 b6 3 d4 e6 4 i.f4
This move is mainly played by those want­
ing to avoid the extensively explored waters
of opening theory. It is neither good nor bad,
just a plausible option. The thing to do when
you are preparing an opening system like the
QID is to acknowledge that these possibili-
5 ... .i.e7 is presented in Game 3.
6 lLJfd2
The key reasoning behind this retreat is as
follows:
1) The queen's knight belongs on c3 and
not d2, and
2) White does not want doubled pawns.
There are two alternatives, each contra­
dicting these points. Somewhat similar to the
Nimzo-Indian and the lines in this book fea-

16
 Optin g fo r No t h in g : 4 i. f4 a n d 4 i. g 5
turing 4 l£1 £3 and 5 i.g5 are the positions
arising after 6 l£ic3 i.xc3+ 7 bxc3 d6, when
there are two set-ups for Black to choose
between:
a) 8 i.d3 l£ibd7 9 0-0 'fie7 10 l£id2 e5 1 1
i.g3 0-0 1 2 l:.e 1 c5 sees Black trying to
dominate White's centre and bishops. In fact
this seems to be rock solid, and Black had a
good game in Ostenstad-Hjartarson, Gausdal
1 985.
b) 8 h3 l£ibd7 9 i.h2 'fie7 10 l:.bl 0-0 1 1
i.e2 l£ie4 1 2 l:.b3 f5 brings about another
traditional set-up from the Nimzo-Indian. It
is not clear that i.h2 is, after all, such a great
piece, but it is easy for Black to find good
squares for his minor pieces, e.g. 13 0-0 e5!?
(limiting the scope of White's bishop, al­
though it was also possible to ignore it) 1 4 c5
i.c6 and Black is fine, Agdestein-Yrjola,
Gausdal 1 985.
Meanwhile 6 l£Jbd2 0-0 7 i.d3 d5 8 0-0
l£ibd7 9 l:.cl i.d6 10 l£ie5 l£ie4 1 1 cxd5
exd5 12 l£ixd7 'iWxd7 gave Black full equality
in Matera-Dzindzichashvili, New York 1 980.
6 . 0-0 7 a3 i.e7 8 lll c 3 d5
..
9 cxd5 lllx d5 !
This is the safe path to equality. Recaptur­
ing with the pawn would be ungrateful to the
bishop on b 7, although it has been tried.
After 9 ... exd5 10 i.d3 c5 1 1 0-0 l£ic6 1 2 l£1£3
White probably has a slight edge. In Meduna­
Vilela, Leipzig 1 980, play continued 1 2 ... a6
13 l£ie5 cxd4 14 l£ixc6 i.xc6 15 exd4 l:.e8 16
l:.el i.d6 17 'fif3 with a slightly worse posi­
tion for Black.
1 0 lll xd 5 i.xd5 1 1 ';ifc2
Here White has a serious alternative in 1 1
l:.cl . Then 1 1 ...i.d6!? 1 2 i.xd6 cxd6 looks
solid, while Miles-Hubner, England-West
Germany 1 979, continued 1 1 ...c5 12 dxc5
i.xc5 13 i.c4 i.b7! 14 0-0 i.d6 with
approximate equality. Instead 13 ... i.xc4 14
l£ixc4 'iWxdl+ 1 5 �xdl l£Jd7 1 6 �e2 l£if6 1 7
i.d6 led to a lasting ending advantage for
White in Miles-Ligterink, Amsterdam 1 978,
while 13. ..l£ic6 14 b4 i.e7 15 i.b5! favours
White as both i.xc6 and e3-e4 are threat­
ened . .
1 1 . . . c5
Equally good looks 1 1 ...i.d6 12 i.xd6
cxd6 13 i.d3 g6 14 0-0 l£Jd7 1 5 l:.acl l:.c8 1 6
1Wa4 l:.c7 wi th equality in Meduna-Unzicker,
Moscow 1 982. Here 14 e4 i.b7 1 5 0-0 l£ic6
1 6 l£1£3 'iWf6 also seems to fail to give Black
problems in finding good squares for his
pieces.
1 2 dxc5 i.xc5 1 3 llle4
13 i.d3 h6 14 0-0 l£Jd7 1 5 l:.fdl l£if6
looked fine for Black in Meduna-Brynell,
Leon 2001.
1 3 . . . llld 7!
Black chooses to complete his develop­
ment. This is more important than the two
bishops. Now, thanks to the activity of the
knight and the possibility of exchanging
dark-squared bishops, Black can equalise.
17
Q u e e n 's In dia n D e fe n c e
1 4 lid 1 'i'c8 1 5 .td3 .txe4 1 6 .txe4 ltif6
1 7 .td3
The trick was, of course, 17 .i.xa8 .i.b4+!,
and Black wins.
1 7 . . . .te7 1 8 'i'e2 Wc5 1 9 h3 ltid5 20
.tg3 llad8 21 0-0 .td6 22 llc 1 ft'a5 23
.txd6 J:lxd6 24 llfd 1 l:fd8 25 .tb5 ltie7
26 llxd6 .l:lxd6 27 J:lc3 lld5 28 a4 Wb4
29 J:lc7 a5 30 g3 g6 31 �g2 'i'd6 32
lld7 fic5 33 J:la7 lids 34 fif3 ltif5 35 e4
Yz-Yz
After ...ltid6 Black has nothing to fear.
------
Game 2
Bareev-Karpov
Moscow (blitz) 1993
._____________...
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 .tf4 .tb7 5
e3 .te 7
1bis move is more natural than 5 ... .ib4+
and should lead to a good game without too
much trouble.
6 ltic3 llih5!
If White fails to address the well-being of
the bishop then Black should, of course,
attack it. Now White will anyway have to lose
a tempo...
7 .tg3 0-0
Karpov's strategy is logical and simple.
The key idea is to develop, eliminating the
bishop only when he is somehow forced to
do so. Another possibility is to delay castling
by exchanging th e bishop immediately, e.g.
18
7 ... ltixg3!? 8 hxg3 .if6 9 'i'c2 g6 1 0 0-0-0
'i'e7 1 1 .id3 lLic6 1 2 .i.e4 0-0-0, leading to
an original position where both players had
good chances in Rodriguez-Fedorowicz,
New York 1 987 .
8 fic2
In this game Bareev elects, quite logically,
to bury the bishop on b7 by advancing d4-
d5. If he had not done so he might have
played along the following lines: 8 .ie2 g6! (a
standard reaction; now the knight won't be
hanging after surprises like ltif3-e5) 9 0-0 d6
1 0 llcl ltid7 1 1 lLid2 lLixg3 1 2 hxg3 a6 1 3
.i.£3 c6 1 4 g4 'i'b8 1 5 a4 d5 1 6 cxd5 cxd5 1 7
.i.e2 'i'd6 with equality in Agdestein-
Adorjan, Oslo 1 984.
8 . . . d6 9 d5
Essentially the same position as in the
game, with the same players, arose after 9
.i.e2 g6 1 0 0-0-0 lLid7 1 1 d5 e5 1 2 ltid2 in
Bareev-Karpov, Paris 1 992. Now Black did
not play the best move, 12 ... ltig7!, which he
must have realised later (which brings us to
the present game). Instead 1 2 ... lLidf6 13 �bl
a6 1 4 £3 lLixg3 1 5 hxg3 h5 1 6 e4 .ic8 1 7 lLifl
lle8 1 8 ltie3 .i.f8 1 9 'i'd3 'i'e7 20 llh2 .i.d7
21 lldhl .ig7 22 ltic2 .i.h6 23 ltib4 llec8 24
ltic6 'i'e8 25 g4 saw White being allowed to
create a strong attack that was enough to win
the game...
9 . . . e5 1 0 .te2 g6 1 1 lLld2
1 1 . . . ltig7 !
The correct decision. The g3-bishop is not
 Optin g for No th in g : 4 i.. f4 a n d 4 i.. g 5

impressive and White would benefit tremen­ i..d 1 J:.fe8 32 Wf2 J:.xe 1 33 tbxe 1 b5
dously from the opening of the h-file, as can
be seen in the following game: 1 1 ...ltixg3?! 12
hxg3 ltid7 1 3 0-0-0 Wg7 1 4 g4 a6 1 5 g3 c6 16
i.£3 l:tc8 1 7 dxc6 i.xc6 18 i.xc6 l:txc6 19
'ii'e4 with a clear advantage to White in
Bareev-Karpov, Paris 1 992. It should be
noted that all these games were in 'active'
chess of some form, and not played with the
usual, longer time control.
1 2 e4 tbd7 1 3 0-0
1bis looks very suspicious.
1 3 . . . i..g 5!
Compare the bishops on gS and g3. Black
has at least equalised. 34 cxb5 i..x d5 35 a4 J:.e4 36 i..f 3 .l:.e5
1 4 b4 h5 1 5 h3 'i'f6 1 6 tbf3 i..f4! 37 tbd3 IH5 38 tbt4 i..c4 39 .l:.e 1 �f7
0-1

Game 3
Nikolic-Lautier
Moscow (Fide knockout) 200 1
1 d4 ll:if6 2 tbt3 e6 3 c4 b6 4 i..f4 i..b 7 5
e3 i..e 7 6 h3
Preserving the bishop, but this costs time,
so Black decides to act at once in the centre.
6 . . . c5!

Exchanging the bishop to win the eS­

square and making way for the g-pawn.
1 7 i..xf4
17 ... gS, with an attack, was a serious
threat.
1 7 . . . exf4 1 8 ii'd2 g5 1 9 e5!?
White is worse and he knows it, so he de­
cides to sacrifice a pawn in a bid to change
the nature of the position and win some
squares for his pieces.
1 9 . . . tbxe5 20 tbxe5 'i'xe5 21 J:!.ae 1 'i'f6
22 h4 gxh4 23 tbe4 'i'f5 24 ii'd4 f6 25 Till s is the easiest way to equalise. White's
llld 2 h3! dark-squared bishop has left the queenside
Black uses his additional material to ruin (g1 -a7 and everything on the queenside as a
White's king position. Black is winning. result of e2-e3) so Black then operates on the
26 gxh3 'i'xh3 27 'ilixf4 J:.ae8 28 'i'g3 dark squares there. 1bis is perfectly logical.
'i'xg3+ 29 fxg3 .l:.e5 30 tbf3 J:.e3 31 The old way to play this was 6 . . 0-0 7 ltic3
.

19
Q u e e n 's In dia n D e fe n c e
d 5 but this might result in a modest advan­
tage for White in some cases. After 8 cxd5
ll:ixd5 9 ll:ixd5 'ii'xd5 1 0 a3 c5 1 1 dxc5 'ii'xc5
12 ..ie2 Z%d8 13 'ii'b l ..ia6 14 ..ixa6 'ii'as+ 1 5
c;ti ft 'ir'xa6+ 1 6 <it>gl ll:id7 17 <it>h2, a s in Me­
duna-Lemer, Lvov 1 981 , White's minor
pieces are well placed, although Black should
be okay. Perhaps if White plays something
slower in the opening Black will have to face
some problems.
7 d 5?
This is completely suicidal. It is difficult to
imagine a top GM playing something like
this. There are two essentially different set­
ups here.
a) 7 dxc5 bxc5 8 ll:ic3 0-0 9 ..ie2 ll:ic6 1 0
0-0 d6 1 1 'ir'c2 'ii'b6 is a universal approach
for Black in the QID and almost always
guarantees an equal position. Lputian­
Mikhalchishin, Frunze 1 979 continued 1 2
l:.adl l:tfd8 (1 2 . . .:ad8 1 3 :d2 h 6 14 l:tfdl
l:td7 1 5 ll:iel l:.fd8 was equal in Langeweg­
Timman, Leeuwarden 1 980) 1 3 l:td2 d51? (an
interesting decision, sharpening the position)
1 4 :fdl ll:ib4 1 5 'ii'b t dxc4 1 6 :xd8+ :Xd8
1 7 ll:id2 (1 7 ..ixc4 ..ixf3 1 8 gxf3 ll:ibd5)
1 7 ... ll:ibd5 1 8 ll:ixc4 ll:ixc3 19 bxc3 'ii'x bl 20
:Xbl ..ie4 21 l:tb3 ll:id5 and here it is White
who is fighting to keep the balance (which
Lputian failed to do).
b) 7 ll:ic3 cxd4 and now a possibility is 8
exd4, leading to a position similar to the
Panov/Nimzo-Indian after 8 ... 0-0 9 .td3 d5
1 0 0-0 dxc4 1 1 .ixc4 ll:ic6 1 2 l:.cl Zlc8 1 3
.td3 ll:id5, Rivas Pastor-Hubner, Linares
1 985. Here the moves h2-h3 and .if4 were
clearly not very aggressive. Black has equal­
ised. This leaves us with 8 ll:ixd4 and a fur­
ther branch:
bl) 8 ... a6 9 .ie2 0-0 and now 10 .tf3 is
logical, dealing with Black's bishop, which
has considerable influence on the centre.
Gretarsson-Stefansson, Stockholm 1 998
continued 10 ...l:ta7 11 'ii'a4 .ixf3 1 2 ll:ixf3
d6 1 3 0-0 'ir'c8 14 l:tfdl :d8 1 5 :d2 :c7 1 6
'ii'd t with an equal position, while Gretars-
20
son-Hansen, Gentofte 1 999 went 1 0 0-0 d6
1 1 ll:ib3 'flc7 12 :et ll:ibd7 with equality,
although Black's game is probably the easier
to conduct.
b2) 8 ... 0-0!? is provocative but if White
cannot exploit it then it is by far the most
logical response. Miles-Kupreichik, Reykjavik
1980 continued 9 ll:idb5 ll:ie8 1 0 'ii'd2?! (this
move is obviously just losing time; better was
1 0 ll:id6 with an unclear position) 1 0 ... a6 1 1
ll:id6 ll:ixd6 1 2 .ixd6 .txd6 1 3 'ii'xd6 b5! 14
a3 bxc4 1 5 Adl ll:ic6 1 6 .ixc4 'ii'g5 and
Black has a useful lead in development.
7 . . . exd5 8 lll c 3 dxc4!
Black is testing the pawn sacrifice. Or he
is just taking the pawn. You decide.
9 lll b 5
It is too late to play for compensation
with 9 .ixc4 because 9 ... d5 1 0 .ib5+ i..c6
gives White very little.
9 . 0-0!
..
Simple chess. White should not really take
the exchange so Black simply makes the
move from which he most benefits.
1 0 lll c 7?
The strategically decisive mistake. The
only move here is 10 .tc7! 'flc8 1 1 i..d6,
with .ixc4 to come, trying for some kind of
compensation.
1 0 . . . d5 1 1 lllx a8 .txa8
Black has two big central pawns and an
enormous lead in development. The advan­
tage is not in doubt.
 Optin g fo r No t h in g : 4 i. f4 a n d 4 i. g 5
1 2 i.e2 ©c6 1 3 0-0 b5 1 4 a4 b4 1 5
'i'b1 ©e4 1 6 :d 1 ©a5 1 7 ltie5 ©b3 1 8
i.xc4
Who cares about pieces here? The pawns
are doing the job!
1 8 . . . ©ed2 1 9 'ifc2 dxc4 0-1
...----.
Game 4
Teske-Paehtz
East Germany 1988
._______________..
(When one is confronted with a game
such as this one always wonders whether the
draw was pre-arranged, as well as the moves
themselves - often this is the case, but you
can never be really sure)
1 d4 ©f6 2 c4 e6 3 ©f3 b6 4 i.g 5
This move has less independent relevance
than 4 .i.f4, as it often transposes to 4 lLic3
.i.b7 5 .i.gS, which is treated in Chapter 3.
4 . . . i.b4+
This is just as natural as 4 ... .i.b7, although
Black now loses the option of 5 ... .i.e7 in­
stead of ... .i.b4. Of course if Black is not
interested in 5 ... .i.e7 the text is quite a funny
move, threatening to engineer a win, as hap­
pens in this game.
4... .i.b7 5 l2Jbd2?! fails to impress. After
5 ... .i.e7 6 e3 h6 7 .i.h4 0-0 8 .i.d3 cS 9 0-0
cxd4 10 exd4 dS!, as in Wuts-Matthias, Ger­
many 1 996, Black has at least equalised. The
long-term potentials in the position are in
Black's favour and the knight on d2 is not
well placed in terms of the fight for the dS­
square. Now Black will be able to take con­
trol over this quite easily and thus develop
more freely than is usual in such positions.
5 ©bd2
This move gives the line independent
value, but it is not good. 5 lLic3 .i.b7 leads us
back to Chapter 3.
5 . . . h6 6 i.h4
6 a3!? is a possibility. If you are an occa-
sional Bogo-Indian fan you can play
6 ... .i.xd2+ 7 .ixd2 .i.b7 with a good ver­
sion!? Otherwise the simple 6 ... .i.e7 should
equali�e directly.
6 . . . g5 7 i.g3 g4 8 ltle5?
Losing. The only move is 8 a31, with good
prospects for Black after 8 . . . .i.e7 9 lLieS d6
10 'i'a4+ c6 1 1 lLid3 'i'd71 followed by
... .i.b7 and ... c6-c5.
8 . . . ©e4 9 .if4 'i'f6 1 0 ©d3 'i'xd4!
21
Q u e e n 's In dia n D e fe n c e

Capturing the pawn without releasing the would take the risk!
pressure. 1 2 exd4
1 1 e3 Now we have a perpetual.
1 2 . . . �f3 + 1 3 �e2 �xd4+ 1 4 �e3
�f5 + 1 5 �e2
1 5 <i>e4 .tb7+ 16 ci>eS d6+ 1 7 <i>f6 �d7
mate(!) is a sour way to end your days.
1 5 . . . �d4+ 1 6 �e3 �f5 + 1 7 �e2 �d4+
%-%

·1 1 . �xd2??
..

What a blunder! After 1 1 ....txd2+! 1 2

'ii'xd2 �xd2 1 3 exd4 �xf1 1 4 :Xf1 .tb7
Black is clearly better. White has some com­
pensation on the dark squares but I, for one,

22
 Op tin g for No th in g : 4 1.. f4 a n d 4 1.. g 5

Summary
4 i.f4 is a rather harmless move and after 4 ... i.b7 5 e3 both 5 i.b4+ and 5 . i.e7 lead to safe
... . .

equality. Still it is a good idea to check out the most important ideas here, or White might be
able to pose Black some problems. 4 i.gS should not gain any independent significance, but
simply transpose to Chapter 3. After S lL!bd2 Black is always going to be fine.

1 d4 tL'lf6 2 c4 e6 3 tL'lf3 b6 4 .tf4 (D)

4 i.gS (D) i.b4+
S lL!c3 i.b7 - Chapter 3
5 lLi bd2 Game 4
-

4 . . . 1.b7 5 e3 1.e7 (D)

5 i.b4+ 6 lLifd2 Game 1
... -

6 tL'lc3 Game 2
-

6h3-Game3

4 1.. f4 4 .tg5 5...1.. e 7

23
I CHAPTER TWO I
The Petrosian System:
4 a3

1 d4 ltlf6 2 c4 e6 3 .!bf3 b6 4 a3 10 i.f4 tl:ihS, which currently looks like the

The Petrosian System had its heyday in most promising system for Black in the
modem times during the 1980's, when Kas­ 4 ... .ta6 line. In Game 1 3 it is time for
parov - still a boy - blew away the world's 10 ... i.cS, which seems to give White an edge.
strongest GMs. Today the variation does not
enjoy the same status and is regarded as one Game 5
of several options, not necessarily offering Portisch-Nikolic
White an advantage, b�t leading to positions I.jub§ana 1985
where many players feel comfortable. And,
of course, new ideas are constantly produced. 1 d4 .!bf6 2 c4 e6 3 .!bf3 b6 4 a3
In Game 5 we see what happens when
Black pushes with ... d7-d5 and recaptures
with the e-pawn. White might have a slightly
more comfortable game but Black should do
okay with decent play. Game 6 features an
idea, 4... c6, tried once and not again (for
good reasons), while Game 7 introduces a
very interesting, dynamic system, namely
4... c5l?, a line that resembles the Benoni,
which seems to be totally playable at the
moment. In Game 8 we just glance at what
happens if White is unambitious against
4 ... c5 - and not surprisingly, nothing happens.
Game 9 includes the minor lines against 4 d5!?
...

4 ... .ta6, which is one of Black's main replies,
while we discuss the direct assault 5 'ii'c2
.ib7 (5 ... c5 leads to Game 7) 6 tl:ic3 cS 7 d5?!
in Game 1 0. Game 1 1 features one of the
main lines, 7 e4 cxd4 8 tl:ixd4 i.cS, and
Game 1 2 investigates 8 ...tl:ic6 9 tl:ixc6 i.xc6
This move is not the standard theoretical
recommendation, but such great players as
Petrosian, Larsen, Karpov and Bronstein,
have given it a strong vote of confidence,
simply by employing it against fellow players
from the world's elite. Besides this, 4... i.b7

24

and 4... i.a6 are the standard moves. 4 ... c6
and 4 ... cS are treated in the following games,
while 4 ... .!De4 S .!Dfd21 dS 6 e3 i.b7 7 cxdS!
exdS 8 .!Dxe4 dxe4 9 .!Dc3 i.d6 10 i.bs+
c;ilf8 (1 0 ... c6 1 1 i.a41 and e4 is in trouble,
unless Black accepts something along the
lines of 1 1 ...fS 1 2 i.b3!, and the king is
caught in the middle) 1 1 'ii'c2 'ii'gS 12 g3 a6
1 3 i.ft f5 1 4 i.g2 gave White the advantage
in Glek-Rozentalis, Lvov 1 98S.
4 ... i.e7 S .!Dc3 0-0 is a strange and altema­
tive way to play. After the logical 6 dS!? (tak­
ing control over the light squares in the cen­
tre, hoping to bury the bishop on c8) 6 ... l:e8!
(directed against 7 e4) 7 g3 i.a6 8 i.g2!
i.xc4 (8 ... d6 9 dxe6 fxe6 1 0 .!Dd4 seems
risky) 9 .!Des i.xdS 10 .!DxdS exdS 1 1 i.xdS
lLJxdS 12 'ii'xdS c6 13 'ii'xf7+ 'ath8 14 'ii'h s
'ii'c 8 1S i.f4 dS 16 .l:.ct i.. £8 Black might
have been okay in Gofshtein-Vitolinsh,
Beltsy 1 979, but the general impression of
the game is that Black had to take risks with­
out ever having a genuine opportunity to
obtain more than an equal position ..
5 cxd5
White decides not to give his opponent
the chance to retum to the main lines with S
.!Dc3 i.b7, although most players would
probably stick with their choice and play
S ... i.e7, with a transposition. However, after
the usual 6 cxdS Black can remain within the
lines here with 6 ... exdS, but there are some
things with 'ii'a4+. All these nuances are
Th e Pe tro s ia n Sys tem: 4 a 3
treated under 7 °ii'a4+ in Game 1 S.
5 . . . exd5
S ... .!DxdS 6 e4 is not relevant.
6 ttlc3 i.e7 7 i.f4
This is one of the two standard ways to
place the pieces. 7 i.gS is also possible, of
course, with a position from the Queen's
Gambit Declined rather than the QID. The
real altemative, and simply a matter of taste,
is 7 g3 0-0 8 i.g2 cS 9 0-0 i.b7 10 i.f4, but
White cannot count on a significant advan­
tage after 1 0 ... .!Da6 1 1 .!Des (1 1 dxcS bxcS 1 2
°ii'b3 °ii'b6 1 3 .!Dd2 'ii'xb3 14 .!Dxb3 l:ad8 1 S
.!DaS i.a8 was equal in Lputian-Nikolic, Sara­
jevo 1 998, but 1 1 l:ct I seems to be the most
logical move) 1 1 ....!Dc7. Note that here
1 1 ... .!De4? meets with 12 .!Dxe4 dxe4 1 3 dxcS
.!DxcS 1 4 'ii'c2 .l:.c8 1 S .l:.fd1 'ii'e 8 1 6 i.h3
.!De6 17 'ii'b 3, which put Black under great
pressure in Vaganian-Gulko, Lvov 1 978. As
we shall see it is a characteristic of these posi­
tions that ... .!De4 tends to be a clear positional
error, and Black should be well aware of
when it is playable and when it is not!
Returning to 1 1 ...�c7, with 1 2 dxcS White
chooses a forced- continuation which ends
with a completely level position, but ... cxd4
was a real threat now that the knight has
control over bS. After 12 ... bxcS 13 .!Dc4 Abs
14 i.xc7 'ii'xc7 1 S .!DxdS i.xdS 1 6 i.xdS
l:fd8! (the most active, although 1 6 ... AbdS is
also good enough) 17 e4 .!Dxe4 we have the
following:
a) 1 8 'ii'hS?I g6 1 9 i.xf7+ is given by
Parma as leading to an advantage, but
1 9 ...'atg7 20 'ii'e2 .!Dd2! 21 l:.fe1 (21 .!Dxd2?1
l:.xb2 22 .l:.ad1 ti;xf7 and here the bishop
and the passed pawn must give Black a fa­
vourable position) Z1...i.f6 22 i.e6 .!Dxc4 23
'ii'xc4Jhb2 followed by ... i.d4 might even
favour Black.
b) 1 8 'ii'e2 .lhdS 19 'ii'xe4 'ii'd 7 20 Afet
i.f6 21 l:tac1 hS 22 h4 i.d4 23 l:.c2 aS 24
l:.ce2 .l:.fS with equality and a draw in
Browne-Ljubojevic, Brasilia 1981.
7 ...0-0 8 e 3
25
Q u e e n 's In dia n D e fe n c e
8 c5
. . .
The most active, but Black has more than
one reasonable move here.
a) 8 ... i.£5 seems perfectly logical because
the exchange has already been performed on
dS and, consequently, b7 is no longer a good
square for the bishop. Still, the move does
abandon the light squares on the queenside.
Van der Sterren-Short, Baku 1983, continued
9 i.e2 (9 i.d3 also looks okay) 9 ... ltie4 1 0
'iib3 c 6 1 1 0-0 i.d6 1 2 i.xd6 'ii'xd6 1 3 :ac1
ltid7 with approximate equality.
b) 8 ...i.b7. This way of holding back the
c-pawn makes sense. All the pieces can de­
velop freely, so why not? Timman-Sunye
Neto, Las Palmas 1 982, went 9 i.e2 (9
.llc 1 !?, forcing Black to make a decision, is
interesting) 9 ... ltibd7 10 .!Lies (1 0 0-0 must be
more testing) 1 0 ... ltie4 1 1 ltixe4 dxe4 1 2 l:tc1
ltixeS 13 i.xeS i.d6 14 i.xd6 cxd6 1S 'ii'a4
(this move does not impress; 1 S 0-0 must be
better) 1 S ... a6 1 6 0-0 i.dS 1 7 l:tc3 bS 1 8 'ii'c2
'iib 6 and Black looks more than fine. Bacrot­
Roder, Spain 2001 , saw positionally unsound
play from Black: 9 ... a6 10 b4 i.d6 1 1 i.g3
'iie 7 1 2 'iib3 ltibd7 1 3 0-0 i.xg3 1 4 hxg3
'iid 6 1 S :fc1 :£e8 1 6 bS c6 1 7 bxc6 i.xc6
1 8 \ib4 and Black had weaknesses in his
pawn structure, while his bishop had no
prospects.
c) 8 ... i.e6 seems illogical. The bishop is
less well placed here than on f5 and b 7. 9 h3
cS 1 0 i.d3 ltic6 1 1 0-0 a6 1 2 :cl c4 1 3 i.b1
26
bS 14 lLieS 'iib 6 1S i.gS :£e8 1 6 f4 earned
White an advantage in Wilder-Larsen, New
York 1 984.
d) 8 ... c6!? offers White some chances of a
slight advantage if he plays 9 i.d3. But after
9 .!Lies i.£5 1 0 g4!? i.e6 1 1 i.d3 ltifd7 1 2
'ii'c2, a s in A.Petrosian-Short, Yerevan 1 984,
the situation is more difficult to evaluate,
although White looks good here, too ..
9 llle 5
It may sound strange but this is the most
flexible move available to White. Note that it
is the only move that White is sure he will
make, whereas the decision to put the bishop
on e2 or d3, for example, can be made later.
However, many games have been played
using these moves directly:
a) 9 i.e2 is the most serious alternative.
After 9 ... i.b7 10 .!Lies (1 0 0-0 ltic6 or
1 0 ... ltia6, or 1 0 ... ltibd7 1 1 .!Lies transposes)
Black must make a choice - two options are
interesting, the third less good.
al) 10 ... ltia6!? is natural. It has not been
properly tested in practice, only in this one
game, Bacrot-Kurajica, Yerevan 1 996, where
White achieved nothing by playing 1 1 'iia4
ltic7 12 ltic6 i.xc6 1 3 'iixc6 with the two
bishops, being behind in development and
already fielding threats - after 1 3 ... ltie6 14
i.eS cxd4 1 S i.xd4 (1 S exd4!? is the testing
move) 1 S ... l:tc8 1 6 \ibs ltixd4 1 7 exd4 ltie4
1 8 ltixdS i.cS 1 9 dxcS :xcS 20 \ib4 'iixdS
was equal. Instead 19 :dl ! 'ii'xdS 20 dxcS

'ifxc5 21 'ifxc5 llxc5 22 0-0 lies 23 llfel
gives White a tiny lead with bishop against
knight in the endgame, although this will , in
99.9% of all cases, not influence the result.
a2) 1 0 ... tLlbd7 seems to do less well with
... c7-c5 than the other knight moves. After
1 1 0-0 l:r.cS 1 2 dxc5 tLlxc5? (1 2 ... bxc5!?
should keep the balance) 13 tLlb5! a6 14 tLld4
White had the advantage in Ribli-Gligoric,
Lucerne 1 982. 1 1 ...tLlxe5? is almost always
bad when White's bishop is on e2: 1 2 dxe5
tLle4 13 tLlxe4 dxe4 1 4 Wa4 a6 15 llfdl 'ife8
1 6 1i'h3 b5 17 e6 with advantage to White,
Lputian-Zagrebelny, Lucerne 1 993. The e4-
square will always be weak in the endgame.
a3) 1 0 .. . tLlc6 1 1 0-0 cxd4 1 2 exd4 tLle4 (as
always with the bishop on e2 Black ends up
in trouble after 1 2...tLlxe5?! 13 dxe5 tLle4 14
tLlxe4 dxe4 1 5 'ifa4 a6 1 6 l:lfdl We8 1 7 Wb3,
as in Timman-Larsen, Tilburg 1 9S2) 13
tLlxc6 i.xc6 14 llcl 'ifd7 15 i.a6 l:lad8 1 6
Wd3 i.d6 and now Farago-Matanovic, Vi­
enna 1 9S6 continued 1 7 tLle2?! i.xf4 1 S
tLlxf4 'ii'd 6, when Black appeared to have
survived the opening. With 1 7 tLle2?! White
completely lacked a plan; better was 1 7 i.xd6
Wxd6 1 8 llc2! to use the c-file in the near
future!
b) 9 i.d3 was played in Plaskett­
Stefansson, Reykjavik 1 992. There followed
9 ... tLlc6 1 0 llet (1 0 tLle5 transposes to the
main line) 1 0 ... i.b7 1 1 0-0 h6 1 2 tLle5 tLlxe5
13 dxe5 (1 3 .lxe5, with equality, is most
likely superior) 13 ... tLle4 14 'ii'h 5 Wes!, pre­
paring ... f7-£5 in the case of a sacrifice. After
1 5 tLlxe4 dxe4 1 6 i.c4 b5 17 i.a2 c4 Black
was better.
c) 9 h3 tLlc6 1 0 i.d3 should not be dan­
gerous for Black, but after 1 0 ... cxd4 (a pre­
mature decision; 10 ... i.b7 offers good
chances of equality) 1 1 exd4 i.d6 1 2 i.xd6
'ifxd6 1 3 0-0 lies 14 i.b5! i.d7 1 5 i.xc6
i.xc6 1 6 tLle5 White had an edge in Schan­
dorff-Larsen, Denmark 1 9S9.
d) Interesting is 9 llet tLlc6 10 i.e2 c4
(1 0 ... i.b7 is more natural). Now White can
Th e Pe tro s ia n Sys tem: 4 a 3
choose a very interesting plan in 1 1 b4!?, with
the idea that c3 will be a fine blockading
square, and later e3-e4 will undermine Black's
centre. Ostojic-Abramovic, Banja Vrucica
1 991 was unclear after 1 1 ...a6 1 2 0-0 b5 13
tLle5 .tb7 14 i.f.3 .
9 .i.b7
. . .
Once the pawn has arrived on c5 this is
the appropriate square for the bishop. After
9 ... i.e6 the c6-square is left weakened, some­
thing well illustrated in the following game:
1 0 i.d3 'ii'c 8 1 1 0-0 c4 12 i.c2 tLlbd7 1 3
i.a4 a 6 1 4 tLlc6 l:leS 1 5 tLlxe7+ :Xe7 1 6 f.3
b5 1 7 i.c2 Wc6 1 S Wd2 and the two bishops
gave White good chances for an advantage in
Bisguier-Villarroe� Caracas 1 970.
1 O .i.d3 lLic6
10 ... tLlbd7 is quite passive -as I mentioned
earlier, ... c7-c5 and ...tLlbd7 do not combine .
well. 1 1 'ii'f.3 lies 1 2 0-0 a6 1 3 Wh3 created
promising attacking chances for White in
Tai-Bronstein, Tbilisi 1 982.
1 1 0-0
11 cxd4
. . .
This is the main line in ECO and other
places, but I recommend 1 1 .. .tLlxe51 12 dxe5
(1 2 .i.xe5, with equality, might be the best
move) 12 ... tLle4!, the difference between this
and other situations being that the bishop is
now on d3. Browne-Frias, Santiago 1 981
continued 13 'ii'c2 tLlxc3! 14 bxc3 (1 4 i.xh7+
�hS 1 5 bxc3 g6 1 6 i.xg6 fxg6 1 7 'ii'xg6 'ii'e S
and, as Black has control over all the light
27
Q u e e n 's In dia n D e fe n c e

squares, there will be no attack, only three in the endgame, but only if he manages to
scattered pawns for the piece), and now not penetrate to the 7th rank will he be able to
14 ...g6? 15 i.h6 :e8 16 i.b5, when White win.
was winning, but 14 ... h6, with an unclear 20 'ifd3 g6 21 J.. b 3 'iPg7 22 :tae1 :ea
position that Black should not fear. The al­ 23 f3 l0d6 24 'ife3 l0e4? !
ternative 1 2...ltJe8?! was tried in Ligterink­ Forcing this exchange only benefits White.
Nikolic, Wijk aan Zee 1 984, when 1 3 'ii'g4 g6 After a more careful move Black should be
1 4 :ad1 h5 1 5 'ii' £3 ltJc7 1 6 i.h6 was slightly only slightly worse.
better for White. 25 J.. xe4 dxe4 26 :ted 1 'ifg5!?
1 2 l0xe6 ..be6 1 3 exd4 J..d 6? !
After this move Black will always face
some problems. The exchange is illogical as it
highlights the difference between the bishops
on d3 and c6. An improvement is 13 ...'ii'd 7!
14 '1'£3 l':.ae8! (the argument for using this
rook lies in the following line: 14 ....l:[fe8 1 5
l:.fe1 i.d6 1 6 i.g5! ltJe4 1 7 ltJxe4 dxe4 1 8
i.xe4 :Xe4? 1 9 :Xe4 'ii'b7 20 :ae1, where
.. f7-f5 has no protection - with the rook on
.

f8 this is not the case) 1 5 .l:[fe1 i.d6 1 6 i.e5

(1 6 i.g5? ltJe4! 17 �xe4? dxe4 18 i.xe4
:Xe4! 1 9 :Xe4 'ii'b7 and Black wins)
1 6 ... i.xe5 17 dxe5 d4 1 8 ltJe4 was Petursson­ Introducing a tactical endgame in which
Larsen, N:estved 1 988, and now Larsen Black is on the receiving end.
writes that Black can keep the balance with 27 'ifxg5 hxg5 2a d5 J.. b 7 29 :te5 :te5?
1 8 ... �xe4 1 9 i.xe4 i.xe4 20 'i'xe4 :e6!, Here 29 ... �f6 is the appropriate move, af­
which looks correct as far as I can see. 21 ter which White maintains some pressure.
:ad1 l':.d8 22 f4 d3 23 f5 l':.ee8! is an impor­ 30 :txg5 f5 31 h4 'iPf6 32 d6 :te5 33
tant line, when the e-pawn is getting as weak 'iPf2 :te6 34 l0e2 :tdxd6 35 l0f4 :txd 1
as the d-pawn, at least. Black should avoid 36 :txg6+ 'iPf7 37 :txe6 :td4
23 ...l':.c6?? 24 e6 fxe6 25 fxe6 'ii'e 8 26 e7
.l:[dc8 27 'i'xd3 and White wins.
1 4 J..g 5
1 4 'i'£3!? i s also natural, preparing a future
i.e5!.
1 4 .. .. .. h6 1 5 J.. h 4 :tea 1 6 J.. e 2 J..e 7 1 7
:te1 l0e4
This is the only move that, in the long run,
frees Black from having to push with ... g7-g5.
1 a J.. x e7 :txe7 1 9 'ife2! :td7? !
Very passive. Now White gets time to
strengthen his position without any worries.
1 9 ...l':.e6 20 i.xe4 dxe4 21 d5 i.xd5 22 :ad1
·
.l:[d6 23 'i'd2 'i'h4 24 �xd5 l':.ad8 25 g3 'ii'g4 3a :th6 :txf4 39 'iPg3 :td4 40 :th7 + 'iPg6
26 ltJe7+ �f8 27 'ii'xd6 :Xd6 28 :xd6 <lhe7 41 :txb7 :td2 42 :txa7 :txb2 43 :te7 b5
29 l':.d3 might give White a small advantage 44 :te6 + 'iPg7 45 :te5 'iPg6 46 :tee+

28
 Th e Pe tro s i a n Sys tem: 4 a 3

<iL'g7 47 .J:.c5 <iL'g6 48 h5+ <iPxh5 49 7 . . .a 6 8 l:.c 1 ! .tb7

l:xf5+ <iL'g6 50 .J:.c5 <iL'f6 5 1 a4 bxa4 52 8 ... i.e7? is very bad in view of 9 llia4!!,
.J:.xc4 a3 53 l:.a4 l:.a2 54 l:.a5 <iL'g6 55 f4 when Black is forced into 9 ..i.d7 1 0 i.c7
.

'iPf6 56 .J:.a6 + <iL'e7 57 �f3 <iL'f7 58 g4
<iL'e7 59 g5 1 -0
The lesson provided by this game, from a
theoretical point of view, is that when White
puts his bishop on d3 Black should consider
... llixeS followed by ... llie4, but with the
bishop on e2 Black will have to unravel more
slowly, although he should be able to main­
tain the balance.
i.xa4 1 1 'ii'xa4+ 'ii'd7 12 'i'b3 bS 13 llieS
with a significant advantage to White. After
e2-e3 comes a3-a4.
9 e3 ll)c6? !
Failing to appreciate the immediate danger
on the c-file. With this in mind 9 ... llibd7
followed by ... l:.c8 is absolutely necessary.
Now White springs a wonderful combina­
tion.

Game 6
D reev-Seirawan
Wijk aan Zee 1995

1 d4 ll)f6 2 ll)f3 e6 3 c4 b6 4 a3 c6? !

One of those ideas that is tried once for
the sake of originality, and never tried again
for the sake of results ...
5 ll)c3 d5

1 0 ll)b5! .J:.c8
10 ... axbS 1 1 i.xbS :cs 12 llieS sees
White win the piece back with interest.
1 1 'ii'a4 ll)d7
1 1 ...llihS 1 2 llic7+! l:r.xc7 13 1.xc7 'il'xc7
14 i.xa6 i.xa6 1 5 :xc6! and White wins.
1 2 ll)c7 + .J:.xc7 1 3 .txa6!

6 cxd5!
Letting Black choose at once.
6 . . . cxd5
This was the reasoning behind 4 .. c6, of
.

course. Otherwise the move would have

contributed nothing to Black's cause. 6 ... exdS
meets with 7 i.gS, when White has the free­
dom to play as he likes in the centre.
7 .tf4!
The bishop takes up residence on what is
now the most dangerous diagonal for Black. Precisely played. After 13 !i..x c7? 'i'xc7 1 4

29
Q u e e n 's In dia n D e fe n c e

i.xa6 ltJdb8 1 5 i.xb7 11i'xb7 Black is ready exd5 i.e7 Black is starting to unravel.
with ... b6-b5, breaking the pin. And after 1 6 22 exd4 i::tta 23 a4!
11i'b5 i.d6! Black prevents ltJe5.
1 3 . . . i..x a6
There is no alternative. 13 ... ltJdb8 1 4
i.xb7 l:txb7 1 5 i.xb8 b 5 1 6 11i'a6 and White
regains his piece, or 1 3 ... e5 1 4 i.xb7 :Xb7
1 5 11i'xc6 and White wins.
1 4 i..x c7 'i'xc7 1 5 J:!.xc6 'i'b7 1 6 lbe5
i..e 7 1 7 b4! f6
17 ... i.c4 would lead to serious trouble af­
ter 1 8 :Xb6!, but not 1 8 ltJxc4 dxc4 1 9 b5 as
indicated by Dreev. Then 19 ... ltJb8! 20 l:txc4
11i'xg2! by no means looks bad for Black!
1 8 b5?
Now Black can do nothing to stop the a­
pawn.
23 . . . i..g 5 24 J: 1 c2 J:!.f4 25 a5 J:!.xd4 26
g3 J:d 1 + 27 �g2 :a1 28 axb6 lbb8
28 ...11i'a6 29 11i'xa6 l:.xa6 30 b7 l:txc6 31
l:txc6 i.f6 32 l:td6 and White wins.
29 J:d6 + 1 -0

Game l
Gelfand':'Kasparov
Novgorod 1997
1 d4 lbf6 2 c4 e6 3 lbf3 b6 4 a3 c5!?
This is simply a mistake. After 1 S l:txb6! The alternative continuation 4... i.a6 5
11i'xb6 1 9 11i'xd7+ 'iii> f8 20 ltJc6 11i'b7 21 11i'xe6 11i'c2 c5 6 d5 is seen just as often.
11i'c7 22 �d2! 11i'd6 23 11i'xe7+ 11i'xe7 24 ltJxe7 5 d5
'iii>xe7 25 a4 White has very good chances in 5 e3 is considered in Game 8.
the endgame. 5 ... i..a 6 6 'i'c2
1 8 . . . i..x b5 1 9 'i'xb5 fxe5 20 0-0
20 l:txe6 exd4 �1 0-0 'iii> fl 22 l:tc6 ltJf6 is
okay for Black.
20 . . . exd4?
The decisive mistake. Dreev must have
not seen that Black is absolutely fine after
20 ... 0-0 21 l:tfcl ! i.d8 22 l:txe6 e4, instead
giving 21 l:txe6 as a simple explanation, but
after 21...i.xa3 22 l:tal l:.c8! all Black's pieces
are playing.
2 1 J:tc 1 ! �dB? !
Another grave error. After 21.. .i.d8 22
:Xe6+ �fl 23 l:.cc6 ltJf6 24 e4!? 11i'd7 25

30

6 b3 has been played a few times but the
pawn structure is not really suited for a
bishop on b2 as d6 is the weak spot in
Black's camp.
6 exd5
. . .
Black can try to prevent e3-e4 with
6 ...'ii'e 7?! but experience shows that this
move, which hinders development, is rather
risky if White reacts with determination.
Miles-Kudrin, London 1 982, went 7 i.g5
exd5 8 lll c 3! i..xc4 (8 ... i.b7 9 cxd5 h6 1 0
i..xf6 'ii'xf6 1 1 e4 was a shade better for
White in Vyzmanavin-Rozentalis, Lvov
1987) 9 e4! (Kasparov's contribution) 9 ... h6
10 i.xf6 'ii'xf6 1 1 exd5 i.xf1 1 2 <t>xf1 d6 1 3
l:.e1+ i.e7 1 4 'ii'a4+ � £8 1 5 'ii'g4 and Black
was under a lot of pressure. Returning to the
position after 9 e4!, Black has a couple of
alternatives to 9 ... h6. 9 ...'ii'e 6 1 0 i..xc4 dxc4
1 1 0-0-0 lll c6 12 e5 lllh 5 1 3 llld S .l:c8 14
'ii'xc4 i.e7 1 5 g4 was terrible for Black in
Hernandez-Ortega, Cuba 1 982, but 9 ... dxe4!
10 0-0-0 i.e6 1 1 lllxe4 'ii'dB 1 2 i.c4 i.e7 1 3
i.xf6 i.xf6 1 4 lll d6+ ct> f8 1 5 i.d5 i.xd5 1 6
:Xd5 gives White compensation fo r th e ma­
terial rather than an enormous advantage.
7 cxd 5 g6
This only serious move. The bishop never
belongs on e7 in a Benoni set-up and win­
ning a pawn with 7 ... i.b7?! 8 e4 'ilie7 9 i.d3
ll:lxd5 looks unhealthy. After 10 0-0 lllc 7 1 1
i.g5! f6 1 2 i.f4 d 6 1 3 lll c3 ll:ld7 1 4 b41
Black was under pressure in Sakaev-Ruban,
Th e Pe tro s ia n Sys t em: 4 a 3
Elista 1 995. And 7 ... ll:\xd5 8 'i!ie4+ llle7 9
'ii'xa8 lllec6 1 0 llle5 'ii'c7 1 1 i.f4 lllxeS 1 2
'ii'e4 simply wins fo r White, a s in Einarsson­
Thorhallsson, Reykjavik 1 996.
8 t2Jc3
There is an alternative set-up both here
and on the following move, designed to take
immediate advantage of the weaknesses of
the d6-square and the a4-e8 diagonal. Luckily
for Black, he also has his resources. 8 i.f4!?
d6 9 ll:\c3 i..g7 10 'i!ia4+ 'ii'd 7! 1 1 i.xd6
'ii'xa4 12 ll:lxa4 lllxd5 13 0-0-0 llle 7! is very
important, for if the knight goes to f6 it
blocks the bishop and does not possess any
mobility. About 6-7 serious games have been
played with 1 1 i.xd6, and in all of them
where the knight went to f6 (mainly after 1 3
e4) Black did badly, while i n th e rest, where
the knight went to e7, Black equalised.
Browne-Timman, Las Palmas 1 982 contin­
ued 14 e4 i..xf1 1 5 :men lllb c6 16 lll c3
i.xc3! 1 7 bxc3 l:td8 1 8 i.. c7 .l:d7 1 9 i.f4 f6
with an equal ending which Black went on to
win. Less ambitious is 1 1 'ilixd7+ �d7,
when Black should , be doing fine. In Garcia
Trobat-Rodriguez, Malaga 1 991 White even
got into trouble after 1 2 g3 �e7 1 3 i.g2
lllbd7 1 4 0-0 .l:he8 1 5 h3 lllh 5 1 6 i.e3 �£81
etc. Finding a good move for White here is
by no means easy.
8 .i.g7 9 g3
. . .
White can also give up the right to castle
in order to gain immediate control over the
centre with 9 e4 i.xf1 1 0 �f1 , but this ne­
glects the fact that in Benoni positions Black
is normally thrilled to exchange the light­
squared bishop for any piece. Consequently,
without any weak light squares, this makes
little sense from a positional point of view.
Of course this is my way of thinking. A
number of the world's top players think dif­
ferently, and the variation has been tried by
better players than I. Arencibia-Almasi, Elista
1 998, for example, continued 1 0... d6 1 1 i.f4
0-0 1 2 h3 l:te8 1 3 g4 b51 14 lllxb5 lllxe4 1 5
l:t e1 a 6 1 6 ll:\c3 lllxc3 1 7 :Xe8+ 'ilixe8 1 8
31
 Q u e e n 's In dia n D e fe n c e
bxc3 �d7 1 9 �g2 (White should avoid 1 9
i.xd6 c4 2 0 i.g3 �f6 2 1 d6 �e4, where the
. importance of the inactive rook is all too
evident) 1 9 ... c4 20 .l:tel 'ii°d 8 21 i.xd6 �f6
22 i.eS 'i'xdS 23 'i'd2 'i'xd2 1/2-1/2. Cebalo­
Tukmakov, Pula 1994 saw Black enjoy a
pleasant position after 14 litg2 'i'b6, with all
his pieces being able to find good squares.
Who knows - Black might eTren be a little bit
better here.
9 . 0-0 1 0 i.g2
. .
The automatic square for the bishop, and
also the best. However GM Neverov has
experimented with 10 i.h3 bS!, but now
Black does not rush with ... d7-d6 as there is
no pressure on the g2-a8 diagonal. Play
might continue 1 1 0-0 l::te 8 (Also good looks
1 1 .. .b4 12 axb4 cxb4 13 �dl �xdS 14 i.gS
'i'c7 1 5 'ii'd2 i.b7 1 6 l::tc l 'Was 17 �e3
�xe3 1 8 i.xe3 i.x£3 1 9 ex£3 �c6 with a
very complex position, which ended with a
draw in Neverov-Onischuk, Donetsk 1 998)
1 2 b4!? (the e2-pawn needs protection, and
ruling out ... b5-b4 safeguards the dS-pawn)
1 2... d6 1 3 i.f4 �hS 1 4 i.gS 'i'c7 1 5 bxcS
'i'xcS 1 6 l::t fcl i.c8 1 7 i.g2 and now instead
of 17 ... �a6, as in Neverov-Nisipeanu,
Koszalin 1 998, 1 7 ... i.b7!? looks best, with at
least equality for Black.
1 0 . d6
. .
1 1 0-0
An understandable mistake would be 1 1
a4?. The problem here compared to normal
32
Benoni set-ups is that Black has not yet
played ... a7-a6 or ...�d7. Now the knight
comes to b4 in a second, and White experi­
ences problems with the a6-f1 diagonal (d3)
and the dS-pawn. 1 1 ...i.b71 1 2 �d2 �a6 1 3
0-0 �b4 14 'i'b3 'i'e7 1 5 l::t d l i.a6 gave
Black a clear advantage in Matjanovic­
Timman, Bled 1 979.
1 1 . . .:es
This is the exact developing move. It is a
question of timing and the deeper details in
the position. After both possible knight
moves Black comes under pressure.
1 1 ...�bd7?1 12 'iia4! i.b7 (1 2 ...'iic 8?1 1 3
i.h3! and Black i s under irritating pressure,
with i.f4! coming soon) 1 3 'iWh4! and Black
has some problems, as has been demon­
strated in many games, one such being
Grooten-Riemersma, Holland 1 993.
1 3 ...�xdS? loses at once to 14 �xdS i.xdS
15 �gSI, for example.
Nor does 1 1 ...�g4?! 1 2 .l:.el improve
Black's situation because after the natural
1 2 ... �d7?? White might consider 1 3 'i'a4,
when Black resigned in Korotylev-Kiselev,
Moscow 1 996. Sometimes it is nice to see
2545 players blunder like we might do in 1 -
minute games on the Internet Chess Club! It
makes me feel that the game is still alive for
humanity...
1 2 J:e 1
Clearly the most natural move, and by far
the most popular. Others:
a) 12 i.f4 �hS 13 i.gS 'ii'c8 14 .l:.fel
�d7 15 g4?1 �hf6 16 i.f4 (1 6 h3 hS! is also
bad for White) 16 ... �xg4 17 �e4 :t.xe4! 1 8
'ii'xe4 i.xb2 and Black was better in Zhu
Chen-Ehlvest, Beijing 1 998. Look at the
prospects for the queenside pawns!
b) 12 b41? was original and not unsuccess­
ful in the following game but, as we shall see,
Black had some promising improvements:
1 2 ...�bd7 (1 2 ... �xdS 1 3 i.gSI, or 12 ... �e4
13 �xe4 i.xal 14 i.gS f6 1 5 :Xal fxgS 1 6
�fxgS, followed by �e6 o r 'i'c3, with good
compensation) 1 3 i.b2 i.c4 14 .l:tfdl cxb4

(1 4...b51?) 1 5 axb4 l::tc 8 (1 5 ... a5, gaining c5 for
the knight) 1 6 'ii'd2? (1 6 e4 lLJxe4 17 lLJxe4
i.xb2 1 8 lLJxd6 with a mess) 1 6 ... i.xe2?
(1 6 ... lLJxd5! wins a pawn� 1 7 lLJxe2 lLJe4 1 8
'ii'd3 i.xb2 1 9 :xa7 and the position was
rather unclear in Krasenkow-Emms, Hast­
ings 1 990.
c) 12 l:bl lLJbd7 13 :e1 b5 14 b4 l:c8 1 5
°ii'd3 lLJg4 1 6 lLJe4 "ii'e7 saw White already
experiencing problems in Zhu Chen­
Alterman, Beijing 1 997.
1 2 lL!bd7
. . .
1bis is the standard move. There are some
transpositions with 12 ... b5, but they are not
so relevant. Yet one idea does look interest­
ing: 1 2...h61?, playing against the cl -bishop.
13 e4 i.b7 14 i.f4 lLJhS 1 5 i.e3 lLJa6 1 6
l:adl °ii'd7 led to an interesting position in
Touzane-Bauer, France 1 996. However, I
have a feeling the bishop's value on b7 has
declined here.
1 3 h3!
By far the most logical option. White
wants to play e2-e4 without allowing ... lLJg4-
e5-d3. Others:
a) 1 3 i.f4 "ii'e 7, and now:
a 1) 1 4 e4 lLJg4! 1 5 i.g5 f6 1 6 i.f4 lLJge5
17 lLJxe5 lLJxe5 1 8 l:adl °ii'd 7 and Black has
the better prospects, Kramnik-Timman,
Linares 1 993. White's pieces are nicely
placed, but they are not really doing anything.
a2) 1 4 "ii'a4 i.b7 1 5 lLJb5 lLJxd5 1 6 i.xd6
(1 6 lLJxd6 i.xb2 17 lLJxb7 i.xal 1 8 :Xa1
Th e Pe tro s ia n Sys tem: 4 a 3
lLJc3 looked better for Black in Riazantsev­
Mitenkov, Moscow 1 998) 1 6 ...°ii'f6 1 7 l:ad1
i.c6 and White's pieces are all tangled up,
Szymczak-Panczyk, Poland 1 982.
a3) 14 l:ad1 lLJe4! 15 "ii' a4 (1 5 lLJxe4
'ii'xe4 1 6 °ii'd2 "ii'e7 with equality) 1 5 ... i.xc3!
16 bxc3 i.b7 17 l:d3 5 and Black is by no
means worse, Jusupov-Timman, Tilburg
1986.
b) 13 e4 lLJg4 14 i.g5 "ii'c7 15 i.f4 lLJde5
16 l:ad1 b5 gave Black excellent prospects in
Popov-Kastanieda, St. Petersburg 1 998.
c) 1 3 'ii'a4?! i.b7 1 4 'ii'h4 does not work
now as the rook is no longer on ffi! After
1 4...lLJxd5! 1 5 lLJxd5 i.xd5 1 6 lLJg5 i.xg2 1 7
°ii'xh7+ � ffi 1 8 �g2 lLJf6 1 9 'fi'b4 d 5 Black
has the advantage (... d5-d4 and ...°ii'd 5 are
coming, with total domination).
1 3 b5
. . .
Clearly the best move. After 1 3 ...lLJe5?1 14
lLJxe5 :xe5 1 5 e4 l:e8 16 i.e3 lLJd7 17 f4
Black's forces had problems finding squares
in Yusupov-Timman, Tilburg 1 986.
1 4 e4
1bis was supposed to be the move for
White until the present game was played.
Now nobody really knows. Just as frequently
played has been 1 4 i.f4, but Black should be
fine if he continues 14 ...'i'b6 1 5 e4 l:ac8 1 6
i.e3 i.b7!?, remembering to push with . . .a7-
a5 and then ... i.a6 in order to fight for con­
trol of the c4-square. Yevseev-lonov, St.
Petersburg 2000, went 17 lLJd2 a6 1 8 a4 b4
19 lLJc4 (1 9 a5 "ii'c 7 20 lLJcbl c4 and Black is
okay) 1 9 ...'ii'c7 20 lLJbl a5! 21 lLJbdZ i.a6 22
i.f4 i.ffi 23 l:ad 1 lLJh5 with a sound posi­
tion for Blaek. In Moreno-Emms, Mondariz
2000, 1 6 ...'i'b7 1 7 i.fl lLJb8! also proved to
be fine, e.g. 1 8 i.f4?! b4 1 9 lLJdl b3 20 'i'b1
i.xfl .f 1 �fl 'fi'a6+ 22 �g2 lLJbd7 23 lLJc3
c4 and Black was better.
1 4 'ifcS ! !
. . .
I have used this position a s an exercise in
my positional training program on the Inter­
net. The key idea is that the bishop should be
supported so that Black can play ... b5-b4 as
33
 Q u e e n 's In dia n D e fe n c e

soon a s possible. Kasparov's move i s the need to see variations to see that 1 7 'fib 1? is
only one that does this. The automatic bad. Just look at the white pieces!
14 ... :c8 1 5 ..ie3 'fic7 (1 5 ...b4 1 6 axb4 cxb4 1 7 lL!xe4 1 8 'it'c2
•••

. 1 7 :Xa6 bxc3 1 8 li)d41 followed by li)c6 has 1 8 li)c3? l::tb 8 19 'ii'c2 .l:.xb2! pockets a
long been known to give White a good posi­ pawn for Black.
tion) 1 6 ..tn 'fib7 17 ..if4 gave White a good 1 8 . . . lLidf6
position in Van Wely-Kamsky, Amsterdam
Donner 1 996. 14 ...'fib6 1 5 ..ie3! also slows
down the ... b5-b4 idea.

Here White needs to do something to

challenge the power of the e4-knight. The
way he does it is terrible, as it serves only to
1 5 .tf4 weaken his king.
The aggressive approach, but after this it 1 9 g4? !
is White who must try to keep the balance. 1 9 li)g5! was the only move.
However, other games suggest that Black is 1 9 'it'd7 20 g5? !
• • •

absolutely fine. 15 ..ifl!? b4 1 6 li)b5 ..ixb5!? Still 20 li)g5!? i s necessary.

17 ..ixb5 l::tb 8 18 ..ic6 b3 19 'fid3 li)e5 20 20 lLih5 21 .th2
• • •

li)xe5 l::txe5 21 <li>g2 l::te 7 led to an interesting 21 l::txe4 .l:.xe4 22 'ii'xe4 l::te 8 23 li)xc5
position (which Black went on to win) in l::txe4 24 li)xd7 .l:.xf4!? leads to a very clear
Markos-Nisipeanu, Cappelle la Grande. The advantage for Black. White's pieces could
question is what is the most important here - hardly be worse.
the bishop on c6 or the weaknesses of the 21 . . . f51
b2-pawn? The debate is most likely not over.
1 5 ..id2? b4 1 6 li)dl b3 1 7 'fibl c4 was
plainly good for Black in Tunik-Ionov, St.
Petersburg 2001 .
1 5 b4 1 6 lba4
• • •

1 6 li)dl b3 1 7 'fibl 'fic7 1 8 li)c3 l:.ab8

did not look impressive in Zhu Chen­
Zatonskih, Istanbul 2000.
1 6 b3!
• • •

This is the idea - now White loses control

over e4.
1 7 'it'xb3
I am one of those people who does not

34

With complete domination.
22 �c3 l:.ab8 23 l:.ab 1 i.xc3!
Securing the difference between Black's
majestic knight and the bishop on h2.
24 bxc3 l:.xb 1 25 l:.xb 1 i.c4 26 �d 2
lpxd2 27 1"xd2 f4! 28 l:.e 1 l:.e5!? 29
l:.e4? l:.xe4 30 i.xe4 'ifxh3 31 i.g2 1"g4
32 'ife 1 �g7 ! 33 f3 'ifxg5 34 'ifb 1 �f5
35 1"b8 + �g7 36 'ifxa7+ �h6 37 'iff7
i.f 1 ! 38 �xf 1 �e3 + 39 �e1 'ifh4+ ! 40
�e2 'ifxh2 41 �d3 �f5 0-1
Game 8
Silman-Gheorghiu
Palo Alto 198 1
1 d4 �f6 2 c4 e 6 3 �f3 b 6 4 a 3 c 5 5 e3
Tbis move is less ambitious than 5 d5, and
Black has no problems equalizing.
5 . . . i.b7
5 ... g6!? 6 liJc3 .tg7 7 .te2 0-0 8 0-0 .tb 7
9 b4 (9 d5?1 exd5 1 0 cxd5 l:te8 gives Black a
good version of the Benoni - where does
White's queen's bishop belong and how is
the other bishop contributing?) 9 ... cxd4 1 0
llixd4 (10 exd4!? d 5 1 1 c 5 bxc5 1 2 bxc5 ltie4
should not certainly not be worse for Black)
1 0 ... ltic6 1 1 llixc6 i.xc6 1 2 i.b2 d5 1 3 'iib3
(1 3 cxd5 llixd5 with at least equality,
Browne-Timman, London 1 980) 1 3 ... dxc4 1 4
i.xc4 'ike7 1 5 l:tadl l:tfd8 with an equal posi­
tion (but also somewhat boring, with chances
only for a draw) in Fanno-Andersson, Bue-
Th e Pe tro s i a n Sys tem: 4 a 3
nos Aires 1 980.
6 �c3 cxd4 7 exd4 d5 8 cxd5
There has been an attempt to finish Black
off immediately by weakening his light
squares with 8 i.g5 i.e7 9 i.xf61? i.xf6 10
cxd5 .txd5 1 1 llixd5 'ikxd5 1 2 'ii'a4+ ri;e7 1 3
l:tcl , but rime has shown that thi s i s not pos­
sible. Christiansen-Gurevich, Parsippany
1 996 went 13 ...l:td8 14 .td3?! a6! 1 5 0-0 l:ta7
1 6 l:tfel .txd4! 17 i.xh7 Wb5 1 8 'ikxb5 axb5
19 llixd4 :Xd4, when Black was better
placed for the endgame, while 14 i.e2 a6 1 5
l:tc7+ l:.d7 1 6 l:tc8 J:ld8 1 7 :Xd8 ri;xd8 1 8
0-0 llic6 results in equality according to
Gurevich.
8 . . . �xd5
Tbis is clearly the most flexible move, and
as White has no way of troubling Black
straight away there is no need to resort to
8 ... exd5?! 9 i.bs+ llibd7 1 0 ltie5 i.e7 1 1
ltic6 .txc6 1 2 .txc6, when White stood bet­
ter in the game Eingom-Lauber, Bad Wor­
ishofen 1 997.
9 i.b5+ i.c6 10 i.d3 �d7 !
The bishop appears t o b e exposed o n c6
but White has no way to exploit it, so Black
maintains the balance.
1 1 o:o i.e7 1 2 i.d2? !
Tbis clearly shows that White lacks suffi­
cient uriderstanding of the positions arising
with the isolated pawn. 1 2 'ike2 is an im­
provement.
1 2 . . . 0-0 1 3 l:.c 1 �xc3 1 4 l:.xc3 i.b7 1 5
35
Q u e e n 's In dia n D e fe n c e
�f4 ltif6 1 6 �e5 :ea 1 7 .J:xea Wxea 1 a
•b 1 g6 1 9 We1
This looks just like a bad endgame, and
guess what it is . . . ?
1 9 . . .Wxc 1 20 .J:xc 1 tlld 5 21 �e4 f6 22
�g3 .J:ea 23 .J:xca+ .ixe8 24 .ib8 a6 25
�xd5 exd5 26 .ic7 b5 27 b4 .if5
Here the bishop is better than the knight.
Black has good practical chances, although
White might have been able to hold it with
best play.
28 ltie 1 �f7 29 � 1 h5 30 �e2 �e6 3 1
.if4 g5 3 2 .i c 1 �d6 3 3 h 3 g 4 34 h 4 g3
35 f3 .ib1 36 tlld 3 �f5 37 ltie5 .ia2 38
tllx a6 .ic4 + 39 �e 1 .if4 40 .ixf4 �xf4
41 tll c 5 �e3 42 a4 bxa4 43 tllx a4 �xd4
44 ltic5 �c3 0-1
..------..
Game 9
Barlov-Sakaev
Budva 1996
_______________..
1 d4 tllf 6 2 c4 e6 3 tllf3 b6 4 a3 �a6
The ... i.. a6 system has gained much popu­
larity over the last decade and is now the
primary system for Black among top players.
I think this has less to do with the actual
worth of the move (both 4...i.. b7 and
4... i.. a 6 probably lead to equal play) and
more to do with the double-edged positions
that we find in the critical lines. There are
simply more chances to outplay your oppo­
nents in this variation.
36
5 •b3
The queen is quite obviously less well
placed here than on c2. Other moves have
also been tried.
a) 5 e3 is really harmless, and often trans­
poses to the 4 e3 line after 5 ... dS, where
White has the extra move a2-a3, which
hardly does him any good (Black can also
play 5 ... cS 6 �c3 i.. e7 7 i..e 2 cxd4 8 �xd4
i..b 7 9 i.. £3 �c6 with total equality, as in
Hort-Ljubojevic, OHRA 1 988). And now:
at) 6 b3 i.. e7 7 i..d3 0-0 8 0-0 cS 9 i..b2
dxc4 10 bxc4 �c6 1 1 �eS �aS 12 �d2 l:tc8
13 l:tct �d7 and White's queenside weak­
nesses began to tell in Smolej-Siegel. Ger­
many 1 985.
a2) 6 �c3 i.e7 7 �eS r 'ii' a4+ c6 8 cxdS
i.xft 9 �xf1 exdS, Flesh-Romanishin, Lvov
1 981, is simply better for Black - look at
White's bishop) 7...0-0 8 i..e 2 c6 9 0-0 �fd7!
and Black has solved all his opening prob-
lems, Browne-Benjamin, USA 1 984.
a3) 6 �bd2 is the only move for people
who want to play. Kholmov-Naumkin,
Voskresensk 1 990 went 6 ... i..d6 7 b3 0-0 8
i.. d3 i..b7 9 'ii'c2 �bd7 1 0 i..b2 cS 1 1 0-0
l:lc8 12 l:lfel bS 13 dxcS i..xcS 14 'ii'b l with
a draw, although I must say that as Black I
would have considered playing on here.
b) After 5 �bd2 Black plays 5 ... i..b 7! as
the new arrival on d2 is poorly placed.
Rohde-Miles, USA 1 989 continued 6 'ii'c2 dS
7 cxdS exdS 8 g3 i.. d 6! (controlling eS and

better than the old 8 ... i.e7) 9 i.g2 tiJbd7 1 0
0-0 0-0 1 1 lLlh4!? l:.e8 1 2 lLlc4?! (not good;
12 tiJdf3 gives both players chances) 12 ... i.f8
13 i.f4? lLle4! and Black was much better
due to the threat of ... g7-g5.
5 . . . g6! 6 tl\c3 i.g7 7 e4 0-0 8 i.e2 c5 9
d5 d6
9 ... exd5 would be too soon. After 1 0 exd5
l:.e8 1 1 'it'c2 White might be better as the
bishop is misplaced on a6.
1 0 0-0 tl\bd7
Black is just developing and, as White has
no active plans, the position is equal. Now
White goes over the edge in an attempt to do
something.
1 1 dxe6? ! fxe6 1 2 e5 dxe5 1 3 tl\g5 'fle7
1 4 �f3 e4!
Black will not allow White to occupy e4
with a knight without first opening up for his
bishop on the long diagonal.
1 5 tl\cxe4?
1bis seems to be bad, but 1 5 ltJgxe4
lLlxe4 1 6 i.xe4 lLle5 1 7 lLlb5 l:.ad8 would
also benefit Black.
1 5 . . . tl\e5 1 6 tbxf6+ i.xf6 1 7 i.xa8
�xc4 1 8 'ilg3 i.xf 1 1 9 �e4 i.d3!
A very important simplification. Now
White's hopes of an attack against the weak­
ened king disappear together with the minor
pieces.
20 i.xd3 tl\xd3 21 'ilxd3
The last finesse is that 21 lLlxh7 'i'xh7! 22
'iixd3 l:.d8! 23 'ii' e2 'ii'h 5 gives White a terri-
Th e Pe tro s i a n Sys tem: 4 a 3
ble problem on the first rank and on b2.
21 . . . �xg5 22 �xg5 'ilxg5 23 J:l.d 1 'fle5
Black has a pawn and very good winning
chances, which proved enough...
24 b3 �g7 25 a4 'flf4 26 'ild7 + .U.f7 27
'ild2 'ilf6 28 h3 e5 29 J:l.e 1 'i/f4 30 'ile2
e4 31 'ilb2+ �h6 32 'ilc2 J:l.e7 33 'ilc4
'ile5 34 'ilg8 'ilf5 35 'ild8 'ilf7 36 'ild2+
�g7 37 'ile3 'ild 5 38 'ilg3 'ild4 39 'ilh4
.U.e5 40 'ilg3 :ts 41 'flh4 :e5 42 'ilg3
a6 43 �h2 :e7 44 'ilh4 'fle5 + 45 �g 1
b5 46 axb5 axb5 4 7 'i'g4 c4 48 bxc4
bxc4 49 'ile2 c3 50 'ilc2 :d7 51 :xe4
:d 1 +
52 'ilxd 1 'flxe4 53 'ild7 + �h6 54 'ilea
c2 55 'ilf8+ �g 5 56 h4+ �h5 57 'i'c5+
�xh4 58 �h2 g5 0-1
Game 10
Dzhandzhava-Chernin
Lvov, 1987
1 d4 tl\f6 2 c4 e6 3 tl\f3 b6 4 a3 �a6 5
'ilc2 �b7 6 tl\c3 c5 7 d5?
1bis should not be considered as a serious
alternative to 7 e4, and nor should 7 dxc5
bxc5, e.g. 8 i.f4 lLlh5 9 i.g5 i.e7 10 i.xe7
'iixe7 l 1 e3 0-0 1 2 i.e2 lLlf6 13 0-0 d6 14
l:.fd1 l:.d8 1 5 l:.d2 liJbd7 1 6 l:.ad1 and White
realised that he had no advantage at all in
Gavrikov-Yemelin, Tallinn 2000. Another
game went 8 i.g5 i.e7 9 e3 d6 10 l:.d1 0-0
1 1 i.e2 tiJbd7 1 2 0-0 'it'b6 1 3 l:.d2 l:.ad8 14
37
Q u e e n 's In dian D e fe n c e
l:tfdl l:tfe8 1 S h3 lLif8! with the idea of
... lL!g6 and ... h7-h6, winning the two bishops.
After 16 .i.f4 lL!g6 17 .i.h2 l:.d7 18 'ii'b l
:ed8 Black was no worse in Portisch­
Timman, Reykjavik 1 988.
7 . . . exd5 8 cxd5 tbxd5 9 i..g 5
9 'ife4+ 'ife7 10 'ifxdS is a silly blitz line.
Black just takes the money and runs.
9 . . . i..e 7 1 0 ltib5 0-0 1 1 ltid6
1 1 . . . ltie3! !
This was a serious new move at the time.
Now Black takes over the initiative.
1 2 fxe3
Practically forced as 12 'ifd3 .i.xf.3 1 3
.i.xe3 .i.g4 1 4 'ii'e4 .i.xd6 1 S 'ii'xa8 'iff6
looks extremely dangerous for White.
1 2 . . . i..xf3 1 3 exf3
13 .i.xe7 'ifxe7 14 lL!fS 'it'e4 and Black
wins.
1 3 . . . i..xg5 1 4 i..c4
38
White must get going. The alternatives are
hopeless. After 14 f4 .i.e71 1 S lL!f5 g6 there is
no compensation for the sacrificed material,
while 14 'ii'e4 'if e7! 1 S 'ifxa8 'ii'xe3+ 1 6 .i.e2
lL!c6 17 'iWb 7 'ifd2+ 1 8 'ifi>fl 'ii'xd6 1 9 l:.d 1
lL!d4 is very close to winning for Black. The
domination of the dark squares and the mis­
placement of White's king will decide the
game.
1 4 . . . ltic6? !
Here Black could have buried 7 dS once
and for all by simply playing 14 ...'i!i'e7!, e.g. 1 S
'ife4 'i!i'xd6 1 6 'ifxa8 lL!c6 1 7 'i!i'b 7 .i.xe3, or
1 S lLifS 'it'eS 16 e4 g6, when Black has a
pawn plus an overwhelming positional ad­
vantage.
1 5 f4?
Here White could have fought back a little
bit with 1 S h4! he31 1 6 lLixfl :.Xfl 1 7
.i.xfl+ �h8 1 8 .i.dS, and Black has no clear
way to conduct the attack, although he has
good chances. Now the struggle is effectively
over.. .
1 5 . . . i..xf4! 1 6 ltixf7
16 exf4 'ife7+ 17 lL!e4 lL!d4 1 8 'ifd3 l:tae8
19 .i.dS 'it'h4+ 20 'ifi>fl (20 g3 'it'hS 21 .i.b7
dS) 20 ...'ii'x f4+ 21 'iti>gl 'if fS 22 .i.b7 dS 23
lLif2 'it'eS 24 g3 c4 2S 'ii'd l f5 and the attack
wins by itsel£
1 6 . . . 1fh4+ 1 7 g3
No alternatives: 1 7 �e2 'ifhs+ 1 8 g4 (1 8
'iltel dS� 1 8 ... 'ii'xg4+ 1 9 �d2 .i.xe3+! 20
�xe3 'ifd4+ 21 'ifi>e2 l:tae8+ 22 �ft dS and

1 7 cJ.?d 1 i.. x e3! 1 8 tllh6+ cJ.?h8 1 9 tll f7+ .l:t.xf7
20 i.. x f7 tt::l d4 21 'iid3 'ii'g4+ 22 �el 'tixg2
both win for Black.
1 7 . . . .i.xg3+ 1 8 �d2
Black would also win after 18 �e2 'ii'h5+
1 9 cJ.?d2 i.. £2 ! (funnily enough Chemin makes
the mistake of mentioning 1 9 ... d5 as winning,
but 20 hxg3! is worth a try!� 20 .l:t.afl i.xe3+!
21 �xe3 d5 and Black wins in the attack.
1 8 . . . .i.f2 1 9 �d 1
1 9 'ii'd3 .l:t.xf7 20 i..xf7+ �xf7 21 'ii'xd7+
tt::l e7 22 .l:t.afl .l:t.£8 and Black wins.
1 9 . . . d 5 0-1
--------..
Game 1 1
Bacrot-Adams
Cannes (rapid) 200 1
1 d4 ll'lf6 2 c4 e6 3 ll'lf3 b6 4 a3 .i.a6 5
'iWc2 .ll b 7 6 ll'lc3 c5 7 e4 cxd4 8 ll'lxd4
Th e Pe tro s ia n Sys tem: 4 a 3
8 . . . .i.c5!?
Here the main line is 8 ... tt::l c 6, but the text
move and 8 ... d6!? have gained in popularity.
After 9 i.. e2 i.. e7 White has played:
a) 1 0 i.. e3 a6 1 1 g4?! has been tried occa­
sionally but it seems to be too dangerous for
White in view of the weakening of his queen­
side with a2-a3. Dreev-Lemer, Vienna 1 996
continued 1 1 ...tll fd7 12 0-0-0 tllc 6 13 tt::lxc6
i..xc6 14 f4 'ii'b 8 15 g5 0-0 16 h4 b5 17 h5
b4 18 axb4 'ii'xb4 19 g6 i.. f6 20 gxf7+ (20
gxh7+ �h8 21 i.. d4 i.xd4 22 .l:t.xd4 tllc5
would also give Black a serious attack)
20 ... .l:t.xf7 21 i..d3 i.xc3 22 'ii'xc3 (22 bxc3
'ii'a3+ 23 cJ.?d2 i.. a4 24 'ii'c l 'ii'a2+ 25 i.c2
tllc 5 and Black is better) 22 ...'ii'xc3+ 23 bxc3
tll c5 24 i..xc5 dxc5 and Black was much
better and went on to win.
b) 10 0-0 0-0 1 1 i.. e3 a6 1 2 .l:t.fdl 'ii'c7 13
.l:t.acl tt::lbd7 14 £3 .l:t.fc8! 1 5 i.. fl 'ii'd 8 1 6 'ii'f2
.l:t.ab8 17 �hl i.. a8 1 8 'ii'g3 .l:t.c7 1 9 b4 'ii'e 8!
with equality in Ljubojevic-Piket, Monaco. In
these hedgehog positions a2-a3 tends to have
pros and cons. It provides White with the
opportunity to advance with a3-a4-a5 later to
generate an attack against a6, but it also
weakens the defence of a2-c4, and any pawns
that might be placed on that diagonal.
9 ll'lb3 ll'lc6 1 o .i.g5
1bis game illustrates quite clearly that this
line is okay for Black. Here are the altema-
rives:
a) 10 i.. d3 0-0 1 1 0-0 h6!? 12 tt::lxc5 bxc5
13 i.. e3 d6 14 £3! .l:t.b8 15 .l:t.fdl 'ii'e7 16 i.fl
.l:t.fd8 17 .l:t.ab l ! with a very small advantage
to White in Sakaev-1\:filos, Sao Paulo 1 991 . In
Lputian-Adams, Wijk aan Zee 2000, 1 1 i.. f4
e5 12 i..g5 h6 1 3 i..h4 i.e7 14 0-0 tt::lh5 1 5
i...xe7 'ii'xe7 1 6 .l:t.adl tt::l f4 1 7 tt::l d5 'tig5 1 8
tt::lxf4.exf4 left Black slightly better.
b) 10 tt::lxc5 bxc5 1 1 i.d3 d6 12 0-0 0-0 13
h3 e5 1 4 'ii'd l h6 1 5 .l:t.bl a5 1 6 f4 exf4 1 7
i..xf4 tt::l e5 and Black seems to b e fine,
Khalifman-Salov, Amsterdam 1 995.
c) 1 0 i.. f4! is the real test of 8 ... i.. c 5, when
wild and gruesome is 1 0 ... 0-0 1 1 tt::lxc5 bxc5
39
Q u e e n 's In dia n D e fe n c e
1 2 i.. d6 o!iJd4 1 3 9d3 e5 14 i..x c5! (the ex­
change does not appear to be so interesting
when all the remaining pieces are good for
Black) 14 ...l:te8 15 i..xd4 exd4 1 6 'ir'xd4
ltJxe4 1 7 0-0-0 and White had the advantage
in Piket-Salov, Wijk aan Zee 1 997. In Kram­
nik-Psakhis, Debrecen 1 992, 1 0 ... i.. e7 1 1
i.. e2 d6 1 2 l:tdl Wb8 was too slow, and
White had 13 c5 e5 14 cxd6 exf4 1 5 dxe7
ltJxe7 1 6 J.b5+ c;t>f8 17 0-0 with a small ad­
vantage. 1bis leaves 1 0 ... e5 1 1 i.. g5 h6 1 2
i..h4, and now Black has the following op­
tions at his disposal.
cl) 12 ... 0-0 13 f3 i.. e7 14 i.. £2 with advan­
tage to White according to Dautov. And trust
me, he is right, although Black should hurry
with 1 4... a5!, with the idea of ... a5-a4 and
... ltJd4 and, perhaps, ... d7-d5 to use his lead
in development to blow the centre apart. 1 3
o!tJxc5?! bxc5 1 4 i.. d3 ltid4 1 5 'ir'dl a 5 1 6 0-0
g5 17 i.. g3 d6 18 :et i..c 6 was better for
Black in Christiansen-Seiravan, USA 1984.
c2) 12 ... i.. e7 13 0-0-0!? 0-0 14 f3 was bet­
ter for White in Notkin-Kiselev, Cappelle la
Grande.
c3) 12 ... d6 13 0-0-0 a6 (1 3 ... 0-0 14 f3 a5 1 5
'ir'd3! gave White a serious advantage in
Akopian-Salov, Wijk aan Zee 1 993) 14 f3
l:tc8 1 5 @bt ltia5, Gruenenwald-Bischoff,
Bundesliga 1 988, and here White should
·
have played 1 6 ltixc5! bxc5 1 7 ltid5 with an
advantage.
1 0 . . . �d4!
40
1 1 �xd4 J.xd4 1 2 �b5
1bis is the test, but as we shall see it seems
to be in Black's favour. 12 i.. d3 'itb8 and the
control over the dark squares guarantees
Black equality.
1 2 . . . .tc5!
12 ... i.. e5 1 3 £4, with unclear play, was the
main line before this game. But the bishop is
better placed on c5.
1 3 e5 h6! 14 exf6
Also fine for Black is 14 i.. e 3 a6! 15 exf6
axb5 1 6 fxg7 .l:.g8, e.g. 1 7 'ir'h7 �e7 1 8
'ii'xh6 i.. xe3 1 9 'ir'xe3 bxc4 and Black ap­
pears to be better as g2 is weak. 17 i..xc5
bxc5 18 'ii'h7 c;t>e7 19 'ir'xh6 'ii'a5+ looks
good for Black. White is severely behind in
development and the potential pressure
against g2 is strong.
1 4 . . . hxg5 1 5 fxg7 l:.g8 1 6 'ifh7 </;e7 1 7
b4 a6 1 8 bxc5 axb5 1 9 'i'h6 f6
White has attacked with his lone queen
and now his king is stuck in the centre. Black
now wins almost by force from here.
20 J.d3 J.xg 2 21 l:.g 1 'i'c7 ! 22 cxb5
22 .:r.xg2 'ii'es+ is easy.
22 . . . 'i'xc5 23 'i'h7 J.f3 24 l:.b1 l:.xa3 0-1
Game 12
Bareev-Eingorn
Kiev 1986
1 d4 lbt6 2 c4 e6 3 lbt3 b6 4 a3 J.a6 5
'ifc2 J.b7 6 �c3 c5 7 e4 cxd4 8 �xd4

lDc6 9 lDxc6
In practice this is the only move played by
the top players; there are no serious alterna­
tives. After 9 .i.e3 lDg4! Black gains the two
bishops without any concessions.
9 txc6
. . ..
1 0 i.f4
This is the main move, but not the only
serious one.
a) 10 .1e2 does nothing about the impor­
tant dark squares. In fact White should now
be careful if he is to maintain the balance:
al) After 1 0 ... 'ifb8 we have 1 1 0-0?! .1c51
12 �h1 (1 2 b4 .i.d4 13 .ib2 h5 and the at­
tack continues) 12 ... h5 1 3 'ifd3 lLJg4, when
Black was already much better in Fe­
dorowicz-Miles, USA 1 989, while 1 1 .1e3
was equal in Christiansen-Miles, Linares
1 985.
a2) 10 ... .1c5! looks best. Then 11 0-0?! is
too dangerous. Now Black attacks on the
dark squares (but what should White do?):
1 1 ...'ifc7 1 2 �h1 h5! 1 3 f3 h4 14 b4 .i.d4 1 5
.id2 lDh5! with a decisive attack in prospect,
Farago-Grooten, Sas van Gent 1 988.
b) 10 .id3 'ii'h 8 1 1 'ife2! (no castling
here!) 1 1 ... .id6 12 .id2 0-0 13 h3 .1f4 1 4
.1xf4 'ifxf4 1 5 0-0 d6, a s in Sak.aev­
Poluljahov, Vrnjacka Banja 1 996, should be
equal.
c) 1 0 .1g51? might give White something.
It all comes down what happens after
10 ...'ifb8. While 1 1 g3?! is not very good, the
Th e Pe tro s ia n Sys t em: 4 a 3
refutatioi:i is not 1 1 ...lDxe4?1, as played in
Lputian-Psakhis, Sochi 1 987, but 1 1 ....ie7!
12 .1g2 lDxe4!1 13 .1xe7 lDxc3 14 .1xc6
Wes+ 15 'iit f1 dxc6 16 Ae1 'iid4 1 7 .ih4 h61!
18 g4 (1 8 'ifxc3 'i'xc3 19 bxc3 g5 20 Ae4
l:.d8! and Black will have a better endgame)
1 8 ...'ii'xg4 1 9 .1g3 "ii'xc4+ 20 �g1 l:td8 21
'ii'xc3 'ifxc3 22 bxc3 l:r.d3 23 l:r.e3 :Xe3 24
fxe3 rj;e7, and the three pawns look stronger
than the bishop. Instead 1 1 i.d3 .i.e7 1 2
'iie2 h 6 1 3 .id2, Klimov-Yemelin, St. Pe­
tersburg 2000, looks like a tiny edge for
White.
1 0 lDh5
. . .
Whether this is the best move or not is
very difficult to tell. The alternative, 10 ... i.c5,
is treated in Game 13.
1 1 i.e3
1 1 .td2 'ii'h 8 12 g3 f5 13 .td3 .i. d6 14
0-0-0 f4 gave Black good control over the
dark squares in Polugaevsky-Arnason, Reyk­
javik 1 987. Super-GM Curt Hansen evaluates
the position as unclear, but I feel that Black
should not fear such unclear situations.
1 1 i.c5?!
. . .
This is not the appropriate strategy. Kar­
pov �veloped the correct approach over a
few tries in the early 1 990's. The first was
1 1 ....id6??, which ended quickly with 1 2
'ii'd 1 !, when Black resigned in Christiansen­
Karpov, Wijk aan Zee 1 993. But the second
attempt, 1 1 ...'ii'h 8 !, is here to stay. The strat­
egy involves a fight for f4 with .. f7-f5 and
.
41
Q u e e n 's In dia n D e. fe n c e
. . .i.f8-d6 coming. Black should equalise:
a) 12 g3 fS! 13 0-0-0 (1 3 l:r.gl !? fxe4 1 4
tDxe4 i.xe4 1 5 'ii'xe4 lDf6 1 6 'ii' f3 i.cS 1 7
i.f4 eS 1 8 i.gS 0-0 with equality in San Se­
gundo-Langeweg, Zaragoza 1 995) 13 ... tDf6
1 4 i.d3 �7 1 5 f3 (1 5 tlJdS fxe4 1 6 i.xe4
exdS 17 cxdS tlJxdS does not work) 1 5 ... fxe4
1 6 tDxe4 tDxe4 1 7 i.xe4! (1 7 fxe4?! hinders
the bishop on d3; after 1 7 ... i.d6 Black was
better in Lutz-Karpov, Dortmund 1 993)
17 ... i.xe4 1 8 fxe4 i.e7 19 l:r.hfl 0-0-0 with
an unclear game according to Karpov.
b) 12 l:r.dl validates 12 ... i.cS! 1 3 i.xcS
bxcS 14 tlJbS 'ii'eS 15 tlJd6+ rJl;e7, when
White has the most problems with his king.
Bykhovsky-Soffer, Tel Aviv 1 994 went 1 6 g3
l:r.ab8 1 7 :d2 gS 1 8 i.g2 f5 1 9 0-0 f4, with
preference for Black.
c) 12 0-0-0 with a further branch:
cl) 1 2... i.cS 13 i.xcS bxcS 14 g3 0-0 1 5
f4 d 6 1 6 i.g2 tlJf6 1 7 l:r.d3 l:r.c8 1 8 l:r.hdl
tDe8 19 g4!, Gurevich-Korchnoi, Biel 1 993,
with aggressive ideas, as g4-g5 and l:r.h3 look
very good for White.
c2) 12 ... i.d6 13 g3 i.eS (1 3 ... fS 14 i.e2
tDf6 15 exfS! i.xhl 16 l:r.xhl 'ii'b7 17 l:r.dl
i.e7 18 g4 0-0 19 gS tDe8 20 i.d3 illustrates
how the light squares can suddenly become
so important; White is clearly better) 14 i.d3
'iib7 (1 4... tlJf6 1 5 f4! i.xc3 1 6 'ii'xc3 'ii'b 7
[or 16 ... i.xe4 17 i.xe4 tDxe4 1 8 'ii'xg7 with a
clear advantage, and the same goes for
1 6 ... 0-0 1 7 i.d4U 17 l:r.hel and White looks
much better) 1 5 l:r.hel tDf6 and now, instead
of 1 6 f4, as in Kramnik-Lutz, Dortmund
1 993, White could have claimed a large ad­
vantage with 16 i.f4! d6 (1 6 ... i.xf4+ 17 gxf4
d6 1 8 eS dxeS 1 9 fxeS tlJd7 20 tDe4 and
Black is in deep trouble) 1 7 i.xeS dxeS 1 8
i. fl 0-0 1 9 f3 l:r.fd8 20 'ii' £2 . This structure is
simply very bad for Black. Once all the heavy
pieces are traded off the endgame will give
White a strong passed pawn on the queen­
side and, potentially, pressure against the eS­
pawn, or e6 if the first is protected with . . f7-.
f6.
42
c3) 12 ... tDf61 is probably the only good
move here. Then 13 f4 tlJg4! does not work,
so Arencibia-Bischoff, Havana 1 998 contin­
ued 13 i.e2 i.d6 14 g3 i.eS 1 5 l:r.hel 0-0 1 6
f3 a6 17 rJi;bl :c8 1 8 i.d4 i.xd4 19 l:r.xd4
dS with equality.
1 2 ..txc5 bxc5 1 3 g3 0-0
13 ... fS 14 0-0-0 f4 worked out well after
1 5 g4?1 in Dreev-Supatashvili, Moscow 2001 ,
but White could have used hi s lead in devel­
opment more aggressively with 1 5 i.e2! 'i'gS
(1 5 ... lDf6? 16 gxf4 'fic7 1 7 eS! i.xhl 1 8
l:r.xh l followed by tDe4-d6 and White is prac­
tically winning) 16 h4 'ii'eS 17 g4 tDf6 1 8 gS
tDxe4 1 9 i.hs+ g6 20 l:r.hel gxhS 21 tDxe4
i.xe4 22 l:r.xe4 'ii' fS 23 'ii'c 3, with an initia­
tive and the safer king position.
1 4 ..td3 lLlf6 1 5 0-0-0 d6 1 6 f4 l:!.b8 1 7
l:hg 1 e5 1 8 g4
Here White's attack seems to be the most
promising, as none of Black's minor pieces
seems able to help on the kingside.
1 8 . . . lbd7 1 9 f5 'ifa5 20 i..e 2 J:l.b7 21
J:l.xd6 J:l.fb8 22 J:gd 1 J:l.xb2 23 Vxb2
J:l.xb2 24 'iPxb2 'ifc7 25 g5 'iPf8 26 f6
gxf6 27 gxf6 'iPe8 28 'iPa2 a6 29 i..h 5
'ifb7 30 lbd5 1 -0
Game 13
Lautier-Timman
Amsterdam 1996
1 d4 lLlf6 2 lbt3 e6 3 c4 b6 4 a3 ..ta6 5

1"c2 .i.b7 6 l0c3 c5 7 e4 cxd4 8 l0xd4
l0c6 9 l0xc6 .i.xc6 1 0 .i.f4 .i.c5 1 1 .i.e2
The best move. White does not want to
allow 11 0-0-0 leih51, when he has no advan­
tage after 1 2 i.d6 i.xd6 1 3 !txd6 0-0 14 g3
£5 1 5 i.g2 (1 5 f4?! fxe4 1 6 leixe4 i.xe4 1 7
'i'xe4 ltc8 i s even more dangerous for
White) 1 5 .. .f4 1 6 !thd1 'i'g5 17 'i'd2 'i'e5, as
in Greenfeld-Kindermann, Pardubice 1 994.
1 1 . . 0-0 1 2 0-0-0
. f6 15 b4 i.e7 16 i.g41 leic7 17 b5 i.a8 1 8
This is by far the most promising con­
tinuation. After 12 .:td1 a5 Black appears to
be doing all right. Play might develop as fol­
lows: 13 i.g5!? (1 3 i.g3!? 'i'e7!? 14 e5 leie8
15 leie4 £5 1 6 exf6 lDxf6 1 7 i.d3 i.xe4 1 8
i.xe4 leixe4 1 9 'i'xe4 and Black cannot be
that much worse, if at all, Gurevich­
Chuchelov, Germany 1 995) 13 ... h6 14 i.h4
e5? (Black is okay after 1 4 ...'ii'c7 1 5 i.xf6
gxf6 1 6 ltd3 'it>h7) 1 5 0-0 'ii'e7 1 6 �h1 g5 17
i.g3 l:.ab8 1 8 i.f3 i.d4 1 9 leie2 i.c5 20
leic3 i.d4 21 ltfe1 g4 and now instead of 22
i.e2?, as in Dreev-Adams, London (rapid)
1995, White could have won with 22 leid511
i.xd5 23 exd5 'ii'c5 (23 ...gxf3 24 ltxd4 and
Black is on the wrong side of a strong attack)
24 'ii' f5 gxf3 25 'it'xf6 fxg2+ 26 �xg2 'iixc4
27 !txd4 'iixd4 28 i.xe5 'ii'g4+ 29 �h1 �h7
30 ltg1 etc. (Vaisser).
12 e5 meets with 12 ... leih5! 13 i.xh5 'i'h4
14 i.xf7+ l:r.xf7 1 5 i.g3 'ii'xc4, while in the
vent of 1 2 0-0 Black has 1 2 ... i.d4!? with the
wild idea of 13 i.d6 Lc3 1 4 i.xf8 leixe4 1 5
Th e P e tro sian Sys tem: 4 a 3
i.b4 i.e5 1 6 i.f3 £5 , with strong compensa­
tion. For more cautious players I recommend
13 ... .l:.e81 14 e5 (1 4 i.d3 e5 looks even better
for Black) 14 ... i.xc3 1 5 'ii'xc3 leie4 with
equality.
1 2 . . . lOeB
Also possible are one decent alternative
and two mistakes:
a) 1 2 ... ltc8 13 'it>bl a5 (1 3 ... leie8 14 i.g3!?
'i'd3 i.xa3 1 9 'ii'xd7 'i'xd7 20 !txd7 !t f7 21
!thd1 with a clear advantage for White) 1 4
i.g3 leie8 1 5 !thel 'iie7 1 6 i.d3 f6 17 f4
�h8 1 8 i.£2 should supposedly give White a
small advantage, as in Greenfeld-Yu, Beijing
1 996, but it does not appear to be that clear,
does it?
b) 1 2... e5? 13 i.xe5 leig4 14 i.xg4 'ii'gs+
1 5 'it>bl 'ii'xg4 1 6 leid5 i.xd5 1 7 cxd5 ltac8
1 8 f3 'iig6 19 i.c3 left White with an extra
pawn in Krasenkov-Hellsten, Malmi:i 1 995.
c) 1 2 ... i.xf2?! 1 3 i.d6 l:.e8 14 e5 i.xg2 1 5
lthg1 i.xg1 1 6 !txg1 i.b7 1 7 °iid3 l:r.c8 1 8
'iig3 g6 1 9 'ii f4, with a very dangerous posi­
tion for Black (Piket).
1 3 .i.g3
1his appears to be the strongest. The al­
ternative 13 �b1!? is probably not best met
with 13 ...f6 14 i.g3 e5 1 5 f4 i.d4 1 6 leib5!
i.xb5 1°7 cxb5 ltc8 18 'iia4 leic7 19 ltc1 ,
with an advantage to White in Greenfeld­
Tunk, Beersheva 1 996. Instead Black has
1 3 ... e5 1 4 i.xe5 i.xf2 1 5 i.d3 f61? 1 6 i.f4
43
 Q u e e n 's In dia n D e fe n c e
i.d4 t 7 :lift with close to equality, or even
t3 ... i.xf2! t4 i.d3 i.c5 t 5 e5 f5 and the
. pawn sacrifice gives White compensation,
but not too much.
1 3 . . . e5
Here you can also take the direct route to
the insane asylum with t 3 ... f5!? t4 exf5 exf5
t 5 £3 'ii'g5+ t 6 'itiiib t f4 t 7 i.d3 fxg3 t 8
i.xh7+ rif fl t 9 f4 1Wf6 20 hxg3 it)d6 2 1 g4
'iiiie 7, and who can tell what the hell is going
on here without scanning all the other games
in the database and spending two weeks with
this position?! Piket-Timman, The Hague
t 995 ended in a draw.
1 4 �b1
t4 i.xe5 'ifgs+ t 5 f4 i.e3+ t 6 ritbt i.xf4
t 7 i.d4 i.e51? with equality according to
Tiviakov.
1 4 . . . .1d4 1 5 ltib5 .1xb5 1 6 cxb5 1ke7
Here I prefer t6 ... :c8 t 7 1Wa4, when
Black has a couple of options. t 7 ... it)f6 t 8
i.h4 h 6 ( t 8. . .d 5 t 9 f4 .l:lc5 was played in
Watson-Browne, USA t 996, and now White
could probably have gained an advantage by
bringing his last piece into play with 20
:lift !) t 9 f4 g5 was Van Wely-Adams, Wijk
aan Zee t 998, and now White would have
the advantage after 20 i.g3! it)xe4 (20 ...gxf4
2t hf4 it)xe4 22 i.xe5 i.xe5 23 'ii'xe4
appears even more dangerous) 21 fxe5 it)xg3
22 hxg3 i.xe5 23 l:txh6. Alternatively
t 7 ...1We7 t 8 Act it)d6 t 9 i.d3 h5 20 h4 led
to equality in Van Wely-Timman, Breda
44
1 998.
1 7 l:.xd4! ? exd4 1 8 l:.d 1 d 5 ! ?
Black i s trying to complicate the position.
t 8 ...1Wc5 t9 b4 'ifxc2+ 20 'ifiiixc2 :tc8+ 2t
'ifiiib 2 it)f6 22 .l:lxd4 .l:lfe8 23 i.d3 gives White
a small advantage in the endgame due to the
two bishops.
1 9 exd5 ltid6
t 9 ... it)f6 20 d6 'ifd7 2t .l:xd4 is just very
good for White.
20 .1d3! h6?
A grave error. After 20 ... :tfc8! 21 'i'e2 (2t
'i'a4 .l:lc5!) 21...1i'xe2 22 i.xe2 the chances
are even (Lauri.er) .
21 1ka4 1kf6?!
Black continues to hesitate about seizing
the open file, and soon it is too late. After
2t ...:ac8 22 'iixd4 .l:lfe8 23 'i'f4 l:ed8 24 h4
:c5 25 a4 White is slightly better.
22 l:.c 1 ! ltif5?
Still not doing too well.
23 .1c7! ltih4? 24 f4! ltixg 2 25 .1e5 1kh4
25 ... 'ifd8 26 'iixd4 f6 27 'iie4 wins for
White immediately.
26 1kxd4 ltie 1
26 ... f6 27 'ife4 fxe5 28 'i'h7+ riffl 29
.l:lc7+ 'ifiiie 8 30 'ifg6+ and White mates.
2 7 1kg 1 !
27 i.xg7?! meets with 27 ... it)£3!, when
matters are less clear.
27 . . .f6 28 1kxe 1 ? !
Here Lauri.er later felt his queen was better
placed and that 28 :Xet fxe5 29 fxe5 would

have won easier as both the pawns march
forward and the idea of .l:tel -e2 followed by
'Wgl -g6-h7 seems to have deadly powers.
28 . . .'i'xe 1 29 J:l.xe1 fxe5 30 fxe5 �f7 3 1
h4! g 5
31 ....l:tad8 32 d6 �e6 33 i.c4+ �d7 34
e6+ �xd6 35 e7 and White wins.
32 hxg5 hxg5 33 .�.f5 J:l.ad8
33 ...�g7 34 i.d7! followed by the e­
pawn's march forward wins for White.
34 d6 �g7 35 .i.g4 .1:1.hS
35 ... .l:tf4 36 e6! .l:txg4 37 e7 and wins.
36 �c2 J:l.h4 37 J:l.e4 J:.xg4
37 ... �fB 38 �c3 .l:th2 39 b3 �e8 40 a4
and the king marches to the centre to assist
the pawns forward.
38 J:l.xg4 �f7 39 J:l.xg5 �e6 40 �d3 J:l.h8
41 �e3 J:l.h 1 42 J:l.g8 1 -0
Game 14
Gershon-Anastasian
Saint Vincent 2000
1 d4 lbf6 2 c4 e6 3 lbf3 b6 4 a3 .i.b 7 5
lbc3
5 . . . g6
This is the most serious alternative to
5 ... d5, although Black hardly ever equalises.
The other option worth mentioning is
5 ... liJe4!?, a variation I have tried without any
pleasure. Black has no real chance to fight for
equality. A good, solid way for White to gain
an advantage is 6 liJxe4 i.xe4 7 e3 i.e7 8
Th e P e tro s ia n Sys tem: 4 a 3
i.d3, e.g. 8 ... d5!? (8 ... i.xd3 9 'Wxd3 d6 1 0 b4
c6 1 1 0-0 0-0 12 e4 ltJd7 13 i.f4 gave White
a safe plus in Browne-Trois, Buenos Aires
1 979) 9 i.xe4 dxe4 10 liJd2 £5 1 1 £3 i.d6!?
(1 1 ...c5?! 12 fxe4 cxd4 13 'Wa4+! �fl 14 liJ£3
favours White, as does 1 1 ...e5 12 fxe4 exd4
1 3 exf5 0-0 14 e4) 1 2 'Wa4+ c6?! (1 2 ... liJd7!
13 fxe4 'Wh4+ 14 �dl 'Wg4+ 15 �c2 'Wxg2
1 6 .l:tfl also gives White a grip on the light
squares, but still it was a fighting chance -
now White is just much better) 1 3 0-0 'Wh4
1 4 f4 0-0 1 5 c5! bxc5 (1 5 ... i.c7 1 6 cxb6
i.xb6 17 ltJc4 is also clearly better for White)
1 6 liJc4 'We7 (1 6 ... .ie7 1 7 dxc5 liJd7 1 8 b4
and Black has no compensation for his struc­
tural weaknesses) 17 dxc5 i.xc5 1 8 b4 i.d6
19 i.b2 with a clear advantage for White in
Kramnik-Vaganian, Horgen 1 995.
6 'i'c2 .i.xf3
This is the only move that makes sense.
After 6 ... .ig7?! 7 e4 The bishop is not very
good on b7. 7 ... d5 (7 . . . 0-0 8 .ig5!? h6 9 i.e3
d6 1 0 h3 liJbd7 1 1 .ie2 was very good for
White in Dreev-Sorokin, St.Petersburg 1 993)
8 cxd5 exd5 9 e5 ltJe4 10 .id3 £5 (White also
had a solid advantage after 10 ...liJxc3 1 1 bxc3
0-0 1 2 0-0 c5 13 .ig5 'i'c8 14 'Wd2 i.a6 1 5
i.xa6 liJxa6 1 6 a4 .l:te8 1 7 i.h6 in Farago­
Podlesnik, Bled 1 996) 1 1 exf6 liJxf6 1 2 'iie2+
'iie7 1 3 'ii'xe7+ �xe7 1 4 0-0 .l:te8 1 5 i.f4
�d8 1 6 liJb5 and Black is in trouble,
Christiansen-Schroll, Vienna 1 99 1 .
7 exf3
Also good is 7 gxf3 liJc6 8 'Wdl !? i.g7 9
.ig5 0-0 10 f4 d5 1 1 cxd5 exd5 1 2 i.g2 liJe7
1 3 b4, when White had the advantage in
Garcia Ilundain-Epishin, Manresa 1 995. Ro­
mero Holmes writes that White is better in
view of the superior bishop and the weak­
ness Qf c6.
7 . . . .i.g7
This move has to be played, so just get it
out. After 7 ... liJc6 8 i.e3 .ig7 9 0-0-0 ltJe7
(9 ... 0-0 10 d5 liJa5 1 1 h4 a6 12 g4 was good
for White in Dreev-Korchnoi, Yalta 1 995) 1 0
g4 d 5 1 1 h4 h6 White had the upper hand in
45
Q u e e n 's In dia n D e fe n c e
Slobodjan-Speelman, Lippstadt 2000.
8 .i.e3
Tbis is the most nonnal move. The alter­
natives just seem to be too premature. 8 g4
0-0 9 i.e3 d5 1 0 g5 ltle8 1 1 cxd5 exd5 12 f4
ltlc6 13 ltle2 ltle7 14 ltlg3 c5 1 5 dxc5 d4
gave Black good counterplay in Kozul­
Romanishin, Yerevan 1 996, while 8 d5 0-0 9
i.e2 exd5 1 0 cxd5 c6 1 1 dxc6 ltlxc6 1 2 'iia4
a6 1 3 0-0 b5 14 'ii'h4 ltlh5 1 5 i.g5 i.f6 was
certainly not worse for Black in Lalic­
Romanishin, Germany 1 996.
8 . . . 0-0 9 h4
Here it is actually most natural to castle
first. In the following game Black chooses a
good set-up compared to the main game: 9
0-0-0 d5 1 0 g4 dxc4 1 1 i.xc4 c6 1 2 h4 ltld5
1 3 ltle4 ltld7 14 h5 'iic7 1 5 hxg6 hxg6 1 6
i.xd5 exd5 1 7 ltlc3 'iid6 with a good game
for Black, Huzman-Van der Wiel, Pula 1 997.
1 0 cxd5 ltlxd5 11 ltlxd5 'iixd5 1 2 i.c4 'iid6
13 d5 e5 would be the critical position. I
prefer White due to the control over the light
squares.
9 . . . �c6
Here Black has 9 ... d5!? 10 cxd5 ltlxd5 1 1
ltlxd5 'iixd5 1 2 i.c4 'iid6, with a slight edge
for White.
1 0 0-0-0 d5
1 1 cxd5! exd5
1 1 ...ltlxd5 1 2 ltlxd5 'iixd5 1 3 i.c4 'iid6
14 d5 and now the knight is misplaced on c6.
1 2 g4 �a5 1 3 h5 :ea
46
Black is trying to defend. Active counter­
play with 1 3 ... c5 14 hxg6 hxg6 1 5 dxc5 l:r.c8
invites 16 g5, undermining the defence of the
centre.
14 hxg6 hxg6
14 ... fxg6 15 g5 ltlh5 16 f4 and White has
the advantage due to the weak light squares
in the centre.
1 5 .i.d3
15 i.e21? makes more sense in some ways
because it monitors the h5-square. But from
d3 the bishop might be sacrificed on g6 later!
1 5 . . . c6
Before he can think of any counterplay
Black has to protect the d5-pawn
1 6 l::r. h 3!
16 i.h6? i.xh6+ 1 7 l:.xh6 ltlxg4! gives
Black some interesting opportunities. For
this there is no reason.
1 6 . . . lbc4 1 7 .i.g5 'ii'd 6
18 .txg6! ?
Tbis sacrifice does present Black with
more problems than most players can solve
at the board. But it does seem that as White
does not have anything directly winning that
it is not the objectively best continuation. 1 8
l:tdhl ltld7 1 9 ltle2 followed by ltlf4 is also
very dangerous for Black.
1 8 . . . fxg6 1 9 'ii'x g6 �f8
Better is 1 9 ... ltld71 20 'ii'b.7+ <i;fl 21 l:r.h6
l:r.e6 22 'i!ifS+ ltlf6 23 'iig6+ �£8 24 l:r.dh1
l:r.ae8 25 l:th8+ ltlg8 26 'i!ifS+ ltf6 27 i.xf6
'iixf6 28 l:t8h5, when White has some advan-

tage.
20 :dh 1 l:.e6
20 ...liJg8 21 'iWf5+ liJf6 22 i.h6 l:te7 23 g5
and Black is just dead lost. 20 ...'iWe6 seems to
lose quite quickly to pressure on the 6th rank
with 21 l:th6! etc.
21 .bf6 'iif4+
21...l:txf6 22 l:th8+ i.xh8 23 l:txh8+ <Ji;e7
24 'iWg7+ l:tf7 25 'iWg5+ 'iWf6 26 'iWxf6+ l:txf6
27 l:txa8 l:txf3 28 l:txa7+ <Ji;e6 and Black has
some good chances of a defence, although
White has an obvious material advantage.
22 cJi>b1 :xt6
22 ... 'iWxf6 -see the note to 21 st move.
23 l:.hS + St.xh8 24 1.txhS+ @e7 25
'iih 7 + ! l:.f7 26 'iih 4+ @d6?
26 ...'iWf6! -see note to move 21 .
27 l:.xa8 'iid 2 28 'iid S+
Here moves were made in order to win
time, as the game was being played with time
increments.
28 . . . �e6 29 'ii e S+ �f6 30 'iih S+ 'iit e6
31 :ea + �d7 32 l:.dS + <ot>e6 33 'iie S+
<ot>t6 34 'iih S+ �e7 35 l:.e8 + ?
Still not finding the right path. Luckily the
win did not go away.
35 . . .�d7 36 l:.dS + 'iite 6 37 'iie S + <ot>t6
38 llle4+ !
Addressing the ltJc4 problem and there­
fore winning the game.
38 . . . dxe4 39 'iix c6 + 'iitg 7 40 'iixc4 exf3
41 'ii c8 'iixf2 42 l:.gS+ 'iith 7 43 l:.hS+
�g7 44 'iig S+ 'iitf 6 45 :th6 + 1 -0
Th e Pe tro sian Sys tem: 4 a 3
With a mating attack. Black resigned.
Game 15
Neverov-Stefanova
Rry�javik 2002
1 c4 e6 2 d4 lllf 6 3 lllt3 b6 4 a3 i.b 7 5
lll c 3 d5
1bis is, of course, the main move here.
Now White has a variety of choices, which
will be discussed in the following seven
games. The first o( them is the popular...
6 St.g5
...which should give White an advantage.
6 . . . St.e7
6 ... dxc4!? also appears to be fully playable.
After 7 e4 i.e7 8 i.xf6 i.xf6 9 e5 i.e7 1 0
i.xc4 0-0 1 1 0-0 c 5 1 2 d 5 exd5 1 3 i.xd5
i.xd5 14 ltJxd5 ltJa6 15 'iWb3 ltJc7 1 6 l:.fdl
ltJxd5 17 l:.xd5 'iWc8 1 8 l:.adl 'iWe6, as in Van
Wely-Korchnoi, Wijk aan Zee 1 997, Black
cannot be much worse, although White still
dictates events.
7 'iia4+ ! ?
1bis was the major line in the 1 990's. The
key idea is that after 7 ... c6 8 cxd5 exd5 9 g3
0-0 1 0., i.g2 White has gained a slightly im­
proved version of the 4... d5 line. Here White
has had many successes in practical play, but
experience has shown that Black completely
equalises after...
7 . . . 'iid 7! 8 'iic 2
1bis is what Gelfand invented when Black
47
 Q u e e n 's In dia n D e fe n c e
found sufficient defences after 8 'i'xd7+
tDbxd7 9 lDb5 .id8! (the most logical, and
. also the height of fashion), e.g. 1 0 cxd5 (10
.if4? dxc4 11 tDxc7+ j_xc7 1 2 j_xc7 0-0 13
llc1 .l:.fc8 1 4 .if4 b5 and Black is better. The
control over the centre and the queenside
majority is far more important than the dark
squares, Van Wely-Gelfand, Monaco 2001)
10 ... lDxdS 11 e4 lD5f6 12 .l:.c1 with a draw
agreed in Dreev-Anand, Moscow Knockout
2001. Then 12 ... 0-0?1 13 lDxc7 .l:.c8 14 lDb5
.l:.xc1+ 15 .ixc1 leads to the win of a pawn
for White, but Bacrot-Gelfand, Leon 2001
continued 12 ... .ixe41 13 tDxc7+ .ixc7 1 4
:Xc7 lDd5 1 5 l:l.c1 h6 1 6 .id2 .ixf3 1 7 gxf3
cJi;e7, and the question here is whether
White's two bishops can compensate for the
poor pawn structure. Perhaps the answer is
yes, but Black appears to have full control as
c7 is his for keeps, meaning White canno t
threaten him.
8 . . . dxc4
9 e4? !
This is an opening experiment we will not
see Gelfand try again. White permanently
sacrifices the c4-pawn and hopes to generate
an initiative in the centre. I feel that it is
doubtful whether he can prove any substan­
tial compensation for the pawn. But the main
line here has, by now, been more or less es­
tablished as equalising for Black. After 9 e3
Black needs to go for the pawn, as 9 ... 0-0?!
10 .ixc4 c5 1 1 dxc5 was somewhat better for
48
White in Gelfand-Karpov, Dortmund 1997,
when Black was still far from finishing his
development. Here 11 0-0-0?! meets with
1 1 ...cxd4 1 2 :Xd4 1i'c8 13 llh4 h6 etc. In­
stead 9 ... .ixf3! 10 gxf3 b5 is best, when 1 1
.ixf6 hf6 leaves White to make a choice.
1 2 'i'e4 0-0! 13 'i'xa8 tDc6 14 'i'x£8+ (14
'if'b7 .l:.b8 15 'i'a6 .l:.b6 and the queen does
not escape) 14 ...cJi;xfB 15 lDxb5 'i'd5 sees
Black win the £3-pawn, with dangerous
threats on the dark squares.
This leaves 12 a4 c6 1 3 axb5 cxb5 14 'ii'e4
0-0 1 5 f4! (White is playing for structure;
after 15 .l:.a5 b4 16 tDa4 .id8! 1 7 lDcS 'i'd5
1 8 l:l.a4 tDc6 1 9 .ig2 j,, b6 20 f4 .l:.fd8 Black
is ideally placed according to Gelfand), and
now time has validated 1 5 ... .l:.c8! (1 5 ...tDc6,
as in Gelfand-Lautier, Biel 1 997, has also
been played). Now White has more than one
possibility:
a) 1 6 'ii'xa8 tDc6 1 7 1i'xc8+ 1i'xc8 1 8 .ig2 a6
19 0-0 g6 and the primary factor in the posi­
tion is Black's future on the queenside.
b) 1 6 .ig2 tDc6 1 7 0-0 (1 7 lDxb5? lDb4 1 8
'i'b7 'i'd8 and White i s not well co­
ordinated, nor fully developed) 17 ... .l:.ab8 1 8
.ih3 shows that White can eam no advan­
tage from 1 6 i..g2. Bacrot-Gershon, Ber­
muda 1 999 continued 1 8 ... b4 1 9 lDa4 lld8 20
lDc5 'i'd5 21 .ig2 tDxd4 22 exd4 .ixd4 23
lDa4 1/2-1/2.
c) 1 6 .ih3!? attacks on the light squares,
where White is better. All other moves are
harmless. Piket-Khalifman, Wijk aan Zee
2002, went 16 ... lDc6 17 tDxb5 .l:.ab8 18 tDc3
.l:.xb2 1 9 0-0 .l:.b3 with a very promising posi­
tion for Black, but best seems 19 ... lDxd4!
with the idea of 20 exd4 'i'xd4 21 'i'xd4
.ixd4 22 .l:.ac1 ? (otherwise three pawns and
the better structure should be important, too)
22 ....l:.b3 and Black wins. 1 7 d5 lDd8 1 8 0-0 is
a suggestion - without an evaluation - of
Gelfand. Black should be fine if he chooses
to play 18 ...b4 1 9 lDa4 exd5 20 .ixd7 dxe4
21 .ixc8 l:l.xc8, when the passed pawns are
of paramount importance, and White is

struggling for survival.
9 . . . b5 1 0 l:l.d 1 a6 1 1 .ie2 h6 1 2 .th4
Wd8 1 3 0-0 0-0 14 ltie5
1 4 . . . c5?
Here Black loses the chance to gain a clear
opening advantage. After 1 4... lbc6! White
has no better move than the exchange of
knights, as d4 is hanging. After the trade
Black needs slightly less space in which to
manoeuvre, and the pawn grows in stature.
1 5 d5! exd5 1 6 exd5 ltifd 7 ? !
1 6 ...lbxdS!? 1 7 .i.g3 lbxc3 1 8 l:txd8
lbxe2+ 1 9 'ii'xe2 i.xd8 with a situation that
is very difficult to assess. Black is, in theory,
better placed, with the large majority on the
queenside, but there will be considerable
problems completing development.
1 7 .txe7 Wxe7 1 8 ltig4
Here White finally has a very interesting
position. For the pawn he has a strong
Th e Pe tro sian Sys te m : 4 a 3
passed pawn on dS, while Black's develop­
ment difficulties remain.
1 8 . . . f5 1 9 ltie3 f4 20 ltlt5 Wg5 2 1 ltid6
ltie5 22 f3 l:l.a7 23 a4 Wd8 24 ltice4
Wb6 25 g3 ltied7 26 �h 1 ltif6 27 gxf4
ltibd7 28 ltif5 ltixd5? !
28 ...lDhS looks better, with a complex
struggle ahead. Now White slowly takes over.
29 l:l.xd5 .ixd5 30 ltie7 + �h8 3 1 ltixd5
We6 32 l:l.d 1 l:l.f5 33 ltidc3 b4 34 l:l.d6
Wf7 35 ltid5 c3 36 i.c4 Wh5 37 �g2
l:l.f8 38 bxc3 ltie5 39 We2 ltixc4 40
Wxc4 bxc3 41 Wxc3 l:l.b 7 42 ltig3 Wh4
43 l:l.e6 J:l.d7 44 We5 Wd8 45 J:l.xh6+
4 5 ... �g8 46 We6 + J:l.ff7 47 J:l.h5 Ad6 48
Wh3 J:l.h6 49 J:l.xh6 gxh6 50 We6 'i'f8 5 1
f 5 c 4 5 2 f 6 �h7 5 3 We4+ �h8 54 'i'xc4
'i'b8 55 We4 a5 56 'i'g6 J:l.b7 57 ltie7
J:l.b2+ 58 �h3 Wts 59 ltlgt5 1 -0
Game 16
Piket-Anand
Monte Carlo (blindfold) 1997
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 ltic3 .ib7
5 a3 d 5 6 Wc2!?
1hls move has been known ever since
1 950, but it was only when Gelfand used it
against Karpov in 1 992 that it became fash­
ionable. As with the 6 i.gS line its span in
the great circulation of modem theory was
short, as Black quickly found the right way to
respond. In this game we shall see two such
49
Q u e e n 's In dia n D e fe n c e
examples.
6 . . . dxc4
This is the current trend, but I prefer
6 ... c5, although it is a matter of taste. How­
ever, not good for Black is 6 ...lbbd7?! 7 cxd5
tbxd5 8 tbxd5. Then 8 ... exd5 9 i.g5 gives
White a small edge, but 8 ... i.xd5? 9 e4 i.b7
10 i.b5! c6 (otherwise i.c6 with a great ad­
vantage) 1 1 i.xc6 l:.c8 1 2 d5 'ii'c7 is a differ­
ent story. Now instead of 1 3 tbd4? as in Gel­
fand-Karpov, Moscow 1 992, White should
play 1 3 i.e3! i.xc6 14 l:.cl exd5 (1 4 ...lbb8
15 dxc6! tbxc6 16 0-0 e5 17 b4 and Black is
still struggling to complete his development)
1 5 exd5 lbb8 1 6 dxc6! 'ii'xc6 1 7 'ii'd l 'ii'e6 1 8
'ii'a4+! with a lasting initiative.
Returning to 6 ... c5!, this is the tactician's
preference. After 7 cxd5 cxd4 8 'ii'a4+ lbbd7
White has three logical options (of different
value):
a) 9 'ii'xd4?! i.c5 1 0 'i'a4 exd5! 1 1 i.g5
0-0 1 2 e3 i.e7! and White needs to develop
quickly, but chose not to in Gofshtein­
Altennan, Israel 1 997.
b) 9 dxe6 dxc3 10 exd7+ 'ii'xd7 11 'ii'xd7+
tbxd7 12 bxc3 i.e7 13 i.e3! l:.c8 14 i.d4
0-0 1 5 e3 tbc5 16 tbd2 (1 6 i.e2? lbb3 1 7
l:.bl [17 l:.a2? i.d5 and Black is getting some
action] 1 7 ... tbxd4 1 8 cxd4 i.xa3 and Black is
better due to the pawns and bishops)
1 6 ...lba4! 17 c4 i.c5 with good compensa­
tion for the pawn, Lputian-Dautov, Budapest
1 996.
50
c) 9 lDxd4 has become the critical move,
although Black has nothing to fear after
9 ... tbxd5 1 0 lDxd5 i.xd5 1 1 e4 i.b7, e.g. 1 2
i.e3 a 6 1 3 f3 b5 1 4 'ii'd l i.e7 1 5 i.e2 0-0 1 6
0-0 l:.c8 1 7 'ii'e 1 i.g5 1 8 i.£2 tbe5 19 l:. d1
'iie 7 and Black had equalised so effortlessly
that he became too ambitious in Dreev­
Karpov, Cap d'Agde 2000. Cramling-Xu Jun,
Yerevan 1 996 went 12 i.a6 i.xa6 13 'ii'xa6
i.c5 14 i.e3 0-0 15 0-0 tbf6 1 6 l:.adl 'ii'c 8!
with equality.
6 ... i.e7 is dealt with in the next game.
7 e4 c5 8 d5
8 dxc5 seems to be rather harmless. After
8 ... i.xc5 9 i.xc4 lbbd7 10 0-0 'iic7 1 1 i.d3
a6 12 i.g5 tbe5! Black already has a pleasant
position. In fact in Sokolov-Polugaevsky,
Holland 1 994 the sequence 1 3 tbxe5 'i'xe5
14 i.h4 lDh5! 1 5 'i'a4+! �f8 1 6 'ii'd7 g5!
gave him a great position.
8 . exd5 9 exd 5 a6!
. .
This is the way they are doing it these
days. After the passive 9 ... i.d6 10 i.g5 0-0
1 1 0-0-0 lbbd7 12 i.xc4 'iib 8 13 �bl a6 14
tbe4 tbxe4 1 5 'ii'xe4 l:.e8 1 6 'ii'g4 tbf8 17
i.d3 i.c8 18 'ii'h 5 'ii'c 7 1 9 l:.cl White had a
little pressure in Lautier-Karpov, Monte
Carlo 1 996.
1 0 .i.g5! ?
The beginning of a wild stream of compli­
cations that ends with an equal position.
10 i.xc4 b5 1 1 i.a2 i.d6 1 2 i.g5 0-0 1 3
0-0 lbbd7 1 4 l:.adl 'iic 7 1 5 i.bl l:.fe8 and

Black was in no way worse in Avrukh­
Anand, Haifa 2000. Instead 1 1 ...i.e7? is poor
since the bishop belongs in front of the d­
pawn. Sakaev-Veingold, Moscow 1 994 went
1 2 0-0 0-0 1 3 i.g5 lL!bd7 1 4 l:!ad1 'ii'b6 1 5
lL!h4 wi th an advantage to White in view of
15 ...g6? 16 lL!xg6 hxg6 1 7 d6 i.xd6 1 8
'ii'xg6+, when White wins.
1 o . . . b5 1 1 0-0-0 i.d6 1 2 ll:\e4 i.e7 1 3
i.xc4? !
This attack seems incorrect. 1 3 i.xf6
i.xf6 14 lLixc5 0-0 (14... i.c8 1 5 'ii'e4+! with
an attack) 1 5 lL!xb7 'ii'c7 1 6 g4!? was a try for
an advantage.
1 3 . . . bxc4 1 4 l:.he 1 0-0 1 5 d6 i.xe4 1 6
l:.xe4 i.xd6 1 7 ll:\e5 ll:\bd 7 ? !
1 7 ... c3! would have disturbed White con­
siderably. Now we are heading for a draw.
1 8 :xd6 ll:\xe4
1 9 i.xd8 liJxd6 20 lbxd7 .=.txd8 2 1 lbb6
:ab8 22 lbxc4 lbxc4 %-%
Game 1 7
C ramling-Al masi
Horgen 1995
1 d4 ll:\t6 2 c4 e6 3 lbt3 b6 4 a3 i.b 7 5
ll:\c3 d5 6 'ii' c 2
Another way to lead to the positions in
the game (in fact the most common) is the
following: 6 cxdS lLixd5 7 i.. d2 (7 'ii'c2 i.. e7
8 i.d2 has been seen, but the most popular
replies to 7 'ii'c2 are 7 ... c5 and 7 ...lL!xc3,
Th e Pe tro s ia n Sys tem: 4 a 3
while White often prefers 8 e4, transposing
to 7 ... lL!xc3) 7 ... lLid7 8 'ii'c2. Now 8 ... cS
seems safe enough, albeit a little boring.
Akopian-Gurevich, Haifa 1 995 continued 9
e4 lL!xc3 1 0 i.xc3 cxd4 1 1 lL!xd4 a6 1 2 g3
'ii'c7 13 i.g2 i.. e 7!? 14 0-0 0-0 1 5 l:!ac1 l:!ac8
1 6 'ii'e2 l:!fd8! (the x-ray finds no practical
use, so Black ignores it) 1 7 lL!b3 lL!e5 1 8
i.b4 'ii'd7 1 9 i..xe7 'ii'xe7 with equality. In­
stead 8 ... i.e7 9 e4 lLixc3 10 i.xc3 0-0 would
be a very natural transposition to our main
game, which seems more interesting and
equally sound.
6 . . . i.e7 7 cxd 5 lbxd 5
7 ... exdS leads to positions similar to the
4 ... d5 lines. There is no need for exact theo­
retical knowledge there.
8 i.d2 0-0 9 e4 lbxc3 1 0 i.xc3 lbd7 1 1
0-0-0
1 1 . . . c6!
1 1 ...'ii'c 8 1 2 h4 1ld8 13 l:!h3 h6!? 14 d5 1 5
ltJgS! (threatening e4-e5) 1 5. . .hxg5 1 6 hxgS
lL!g4 (1 6 ... lL!e8?! 1 7 f4 gives White a very
strong attack as the black pieces are not play­
ing) 17 f4 i.cS! 1 8 i.xg7!? �xg7 1 9 'ii'c3+
and White had a very interesting attack in
Akog.ian-Granda Zuniga, Groningen 1 993.
Best for Black here is 19 ... �£8! 20 l:!h7 �e7,
bringing the king to safety, when 21 i.e2
exdS 22 f5 is messy indeed.
·
1 2 h4 b5 1 3 �b1
More direct is 13 l:!h3!? a5 1 4 d5 cxdS 1 5
i.xb5, as in Kamsky-Anand, Sanghi Nagar
51
Q u e e n 's In dia n D e fe n c e
1 994. Black defended well by taking the cen­
tral squares with 1 5 ...lll f6 1 6 ltlg5 'ii'b 6 1 7
exd5 l:ac8! 1 8 .id7 .ixd5! 1 9 .ixc8 :xc8 20
l:xd5 (the bishop seems to be stronger than
the rook, so this exchange is quite natural)
20 ... exd5 21 'ii' f5 'ii'a 6, and now instead of
the adventurous 22 c,t>d2?! White should have
settled for the repetition after 22 c,t>b1 'ii'f1+
23 c.t>a2 'i'c4+.
1 3 . . . a5 1 4 .i.e 1 ?
1bis move hurts my eyes. I cannot see
how the bishop will ever be a useful piece on
e1 .
Better, in my opinion, is 1 4 .id2 b4 1 5 a4
c5 1 6 d5?! (this pawn sac seems wrong; 1 6
.ig5! i s obviously superior, getting all the
pieces into action) 16 ... exd5 17 exd5 .ixd5
1 8 ltlg5 ltlf6, Mikhalchishin-Petrosian, Lvov
1 994, when Black has the initiative and a
pawn and, therefore, the better prospects (his
pieces are only temporarily awkwardly placed
in the centre).
14 d5 cxd5 1 5 lllg5 (1 5 .ixb5 l::tc 8 1 6
exd5 .ixd5 1 7 .id3 'ii'b 6! looks good for
Black) 1 5 ...dxe4 16 .ixb5 .id5 17 lllxe4 'fllc 7
was good enough for Black in Tregubov­
Akesson, Cap d'Agde 1 994, although White
has quite a bit of compensation for the pawn.
1 4 . . . b4 1 5 a4 :ea 1 6 ltlg5 c5 1 7 d 5
.i.xg5!
1 7 ...e5? fails to 1 8 ltle6! fxe6 1 9 dxe6,
when White is better.
1 8 hxg 5
52
No improvement is 1 8 dxe6 fxe6 1 9 .ib5
(1 9 hxg5 'i'e7 and Black's position is com­
fortably superior, with, among others, ... ltlb6
and ... c5-c4 coming) 19 ... .ic6! 20 .ixc6 :Xc6
21 hxg5 'i'e7 and White is very weak on the
highly important light squares.
1 8 . . . exd5 1 9 exd5 g6 20 .i.b5
20 'ii'b3 'fllxg5 21 'iih3 ltlf6 and Black is
close to winning. 20 f4 is met by 20 ... ltlb6.
20 . . . lLlb6
Black is clearly better according to Almasi.
21 d6 c4?
21 ... .ixg2! 22 :h2 .if3 and White is fin­
ished.
22 f3! c3! 23 b3
Forced. 23 d7 b3! wins on the spot for
Black.
23 . . . ltld5 24 .i.f2!
More tricks: 24 d7 ltle3! and Black wins.
24 . . .'ilxg5 25 d7 .z:lb8 26 .i.h4?
Here White had the chance to gain good
counterplay with 26 .id4! lll e3 27 .ixe3
'ii'xe3 28 .l:.he1 1i'g5 29 l:e8, although
Black's chances are preferable after 29 ... l:d8!
etc.
26 . . .'ilf4 27 .z:lde 1 ?
White is on the way, not to sacrifice an
exchange; but to win a piece. The only prob­
lem is that a bishop on d5 together with the
strong c3-pawn combine to provide Black
with decisive compensation. 27 .if2 was
better according to Almasi.
27 . . . ltle3 28 .z:lxe3 'ilxe3 29 .z:le 1 'i'b6 30

J:e8?! .i.d5! 3 1 �a2 J:bxe8 32 dxe8'i'
J:xe8 33 i.xe8 'i'd4!
White has a piece but, as promised, Black
is very close to winning.
34 i.g3 'i'd2 35 �b 1 'i'e3! 36 i.b5
i.xb31 37 'i'xb3 'i'g 1 + 38 �a2 'i'xg2 +
3 9 �a1 'i'xg3 4 0 'i'd 1 'i'f2 4 1 i.e2 b3!
42 'i'xb3 'i'xe2 43 'i'xc3 'i'd 1 + 44 �b2
'i'd8 0-1
White has had enough.
-------
Game 18
Timman-Polgar
Bali 2000
1 ltlf3 lLlf6 2 c4 b6 3 d4 e6 4 a3 i.b7 5
ltlc3 d5 6 cxd5 ltlxd5 7 'i'c2!
This is the most popular move in this po­
sition with GMs today, and for good reason.
As we shall see Black has no clear-cut path to
Th e Pe tro sian Sys tem: 4 a 3
development and, in fact, this is the main
reason why I cannot fully recommend the
system with 4... i.b7 followed by 6 ... i!llx d5. I
just do not feel comfortable about the result­
ing positions.
7 . . . ltlxc3
7 ... c5 is dealt with in the following game.
8 bxc3
This has become the only serious move.
In the late 1980's 8 1Wxc3!? was also played
quite a lot, but several methods to equalise
were quickly found. The simplest is 8 ... i!lld7 9
i.g5 (9 i.f4 i.d6 is no problem at all for
Black) 9 ... i.e7 10 i.xe7 cJilxe7! (the king is
not exposed, so this yields no problems) 1 1
e3 :c8 1 2 i.e2 c5 1 3 dxc5? (this starts a
kamikaze journey for the queen which leads
only to problems; 1 3 0-0 is better) 1 3 ... llxc5!
14 'ilxg7 ltg8 15 'ii'xh7 l:lxg2 (the white king
is more awkward in the centre as Black is
well developed) 1 6 1Wh4+ lll f6 1 7 ltd1 'ikc7
1 8 llld4 a51 and Black is clearly better accord­
ing to Petursson. Portisch-Karpov, Biel 1 996
saw White blunder with 1 9 i!ll b 5?? (1 9 h3!?)
1 9. . .l:.xb51 2 0 i.xb 5 ltg4 0-1 .
a . . . i.e7 9 e4 0-0 10 i.d3 c5 1 1 0-0
Here there is a very interesting alternative.
Of course this can also be played at move 13,
but it is for some reason seen more often
here - namely 1 1 i.b2!, when Piket-Rau,
Rotterdam 1 988 continued 1 1 ...'ilc8 1 2 1We2
i.a6 13 0-0 i.xd3 14 'ii'xd3 'ila6 (14 ... cxd4
1 5 cxd4 'i'a6 1 6 'i'e3 llld7 1 7 d5 gives White
the initiative, while 14 ...i!ll c 6? 1 5 d5 llla5 1 6
c4 is overwhelmingly better fo r White) 1 5 c4!
cxd4 1 6 lllxd4 'ii'b7 (1 6 ... ltc8?! 1 7 lllb 5 lllc 6
18 'iig3 e5 19 f4! was very promising for
White in Piket-Polugaevsky, Aruba 1 994) 1 7
'ii'g3 ( 1 7 i!llb 51? also looks good) 1 7 . . .i.f6 1 8
e 5 i.e 7 1 9 l:lad1 ltd8 20 lllb 5 llxd1 21
:Xd1 lll a6 with an unclear game. I think that
this is more unclear for White than for Black,
if you get my drift. White has some strong
plans involving the advance of the h-pawn.
Note that 1 5 c4! is far stronger than 1 5 1We3
i!ll d7 1 6 llad1 ltac8 1 7 Ad2 b5, where the
53
Q u e e n 's In dia n D e fe n c e
bishop is poor on b2. Black has at least
equalised, and won quite soon in Flear­
Hjartarson, Szirak 1987. Nor does 1 5 'ii'xa6
�xa6 1 6 �e5 l:tfc8 17 l:tfdl cxd4 1 8 cxd4
.l:r.c2 present Black with anything but pros­
pects, as in Malaniuk-Ehlvest, USSR 1 987.
1 1 . . . 'ifc8 1 2 'ife2 .ta6
1 3 J:d 1
Others:
a) 13 a4 i.xd3 14 'ii'xd3 ltd8 15 i.f4 (1 5
a5 bxa5 1 6 :Xa5 �c6 1 7 l:ta4 cxd4 1 8 cxd4
°ii'd7 with equality - Dautov) 1 5 ...'ii'b7 1 6 a5
�c6 (1 6 ... cxd4!? 17 cxd4 bxa5 1 8 :xa5 �c6
would probably be fine) 17 axb6 cxd4 1 8
cxd4 axb6 1 9 .:Xa8 :Xa8 2 0 .:r.d1 .:r.d8 with
equality, Azmaiparashvili-Epishin, Reggio
Emilia 1 995/96. Dautov has suggested the
following alternatives: 14 ...\i'a6 1 5 'ii'b 5 'ii'h7
1 6 ltel a6 1 7 'i'e2 .:r.d8 1 8 d5 with good
prospects for White, and 14 ... cxd4!? 1 5 cxd4
�c6, which possibly equalises. I am not
completely sure on this one, although it looks
natural.
b) 13 i.f4 gives White the advantage ac­
cording to ECO, but I will remain sceptical
on that one, too. Black should be fine after
1 3 ... i.xd3 1 4 'ii'xd3 l:td8 (more careful is
1 4... �d7 1 5 a4 .:r.d8 1 6 l:tfe1 cxd4 1 7 cxd4
'ii'b 7 1 8 l:lab1 :ac8 1 9 a5, when White might
have a slight positional edge but it will be
very difficult to prove, and he was nowhere
even close in Piket-Polugaevsky, Aruba 1 994)
1 5 'iii'e 3 (White has no advantage after 1 5
54
d5!? c4 1 6 'ii'e2 exd5 17 exd5 i.f6 1 8 d6 'iWfS
1 9 i.g3 �d7 20 'i'xc4, when a draw was
agreed in Piket-Van der Wiel, Holland 1 996)
1 5 ... �c6 16 ltfdl �a5!?, Arbakov-Levin,
Berlin 1 994. Now White had sufficient com­
pensation for the pawn after 1 7 h4 �c4 1 8
'i'e2 cxd4 1 9 cxd4 i.xa3 20 �d2 �xd2 21
l:txa3! �c4 22 l:tg3, but nothing has been
decided. Personally I would not like to be
Black here, but I think it is more a question
of style than anything else ..
13 i.b2! leads to the note to White's 1 1 th
move.
1 3 . . . cxd4 1 4 cxd4 .txd3 1 5 'ii'x d3
White has no chance for an advantage
here, it seems. The alternative looks a little
clumsy and Black equalises easily: 1 5 l:txd3
�d7 1 6 i.b2 (1 6 i.g5?! i.xg5 1 7 �xg5 h6
1 8 �h3 'ii'c4, as in Peshina-Gurevich, Eger
1 987, is already better for Black; White's
forces are misplaced) 16 ...'i'a6 1 7 l:tadl l:tfe8
1 8 'iii'e3 l:tac8 1 9 d5 exd5 20 l:txd5 �£8 21 h4
'iii'a4 22 l:t1d4 and a draw was agreed in
Bareev-Rodriguez, Sochi 1 988.
1 5 .. .'ifa6 1 6 .tb2
16 d5 'iii'xd3! 17 ltxd3 exd5 18 exd5 i.d6
1 9 .i.e3 .:r.c8 20 l:te1 �d7 is equal according
to Knaak. 1 6 'iii'e31? �d7 1 7 i.b2 ltac8 is
more adventurous, but Black should be able
to gain sufficient counterplay on the c-file.
1 6 . . .'ifxd3 1 7 J:xd3 �d 7 1 8 J:c 1
White gains the c-file but Black has the
possibility of a distant passed pawn and. of
creating . a good square for the knight on c4.
Also equal is 1 8 d5 �c5 1 9 l:te3 .:r.fd8 20
.:r.d1 :ac8.
1 8 . . .J:fc8 1 9 J:dc3 J:xc3 20 J:xc3 �f6
21 �d2
21 e5? �d5 helps only Black.
21 . . . b5!
Black would never allow 22 a4, where the
pawn is less weak and when he would not
have c4 waiting for the knight.
22 J:c6
22 ltc7 'iii f8 23 .:r.b7 a6 has been assessed
as unclear. I think White should be very care-

ful here that ... l:tc8 does not suddenly win
something. All in all it seems that improving
the king should be a priority.
22 . . . 'iPfs 23 f3 �ea 24 ltlb3
24 'iifi' f2 'iitt'd7 shows that White somehow
gained nothing by 22 l:tc6, only to have to
waste time keeping it there.
24 . . . ltld7
24 ... 'iifi' d7 25 l:tc1 l:tc8 26 ltic5+ and White
maintains the balance.
25 �f1 ?
Creating problems later. Actually it is dif­
ficult to why 25 'iifi' f2 ! was not played.
25 . . . ltlb6 26 ltlc5?
White is drifting. The knight does less
here than it seems to while the knight on c4
is really annoying. 26 ltia5 'iifi'd 7 27 l:tc2 l:tc8
28 l:r.xc8 ltixc8 29 'iifi' e2 results in an endgame
where Black has very slightly better chances
due to the prospects of creating a passed
pawn and the weakness of the a3-pawn.
However, this should be a drawable option
for White.
26 . . . ltlc4! 27 .i.c 1 :I.dB
Now White just loses something. If the
king had been on f2 the game continuation
would have worked better because there
would be no check on dl .
28 a4
28 d5 ltie5 29 l:r.c7 ..td6 and Black should
win.
28 . . . Axd4! 29 ltlb3 Ad3 30 axb5 ltle5 3 1
:ea + �d7 3 2 :as Axb3 3 3 :xa7+ �ea
Th e Pe tro s ia n Sys tem: 4 a 3
0-1
Game 19
Kasparov-van der Wiel
Amsterdam 1988
1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 a3 .i.b 7 5
ltlc3 d5 6 'i'c2 c5 7 cxd5 ltlxd5
8 dxc5!
This is the only move that is dangerous
for Black. The alternatives both promise
Black equality.
a) 8 e4 ltixc3 9 bxc3 and now:
al) After 9 ... ltid7 White reaches another
crossroads.
al 1) In reply to 10 ..td3 the best move is
1 0 ... 'iVc7! according to Kasparov, a draw
being agreed · in Kasparov-Sosonko, Lucerne
1 982 after 1 1 'iVd2 g6 1 2 0-0 ..tg7. Kasparov
was, in fact, so uncomfortable with his posi­
tion here that he claimed in Chess Informant
that Black already has a slight advantage.
Whether or not this is too strong an assess­
ment I shall not say, but the fact that Black is
not unhappy with the outcome of the open­
ing is quite clear to me. Black is also fine after
the alj:ernatives, e.g. 1 1 'iVbl g6 1 2 0-0 ..tg7
1 3 :a2 0-0 14 l:tel a6 1 5 a4 l:r.fc8 1 6 h3 ..tc6
17 ..te3 c4 1 8 ..tc2 b5 with equality, as in
Ostermeyer-Sosonko, Hannover 1 983.
Meanwhile 11 ..tb2 cxd4 12 cxd4 'i'xc2 1 3
..txc2 ..ta6 led to a carefree endgame for
Black in Yusupov-Miles, Linares 1 983, and
55
Q u e e n 's In dia n D e fe n c e
1 1 0-0 cxd4 12 cxd4 'i'xc2 13 j,xc2 l:lc8
even gave Black some initiative in Franco
Ocampos-Van der Sterren, Wijk aan Zee
1 983.
a1 2) 10 i.f4. This bishop needs to be de­
veloped immediately. 10 ... cxd4 1 1 cxd4 l:tc8
12 'i'b3 'ii' f6!? 13 j,g3 'ii'g6 (1 3 ... j,xe4 1 4
j,bS i s very dangerous fo r Black) 1 4 i.d3
j,e7 1S 0-0 0-0 16 l:tfel l:tfd8 and Black was
well placed in Van der Sterren-Van der Wiel,
Wijk aan Zee 1 986. There are both good
things and bad things to say about the queen
on g6. The rather futuristic 1 2 ... 'ii'f6!? seems
to be the best, while more solid is 12 ... j,e7
13 j,d3 �f6 14 'libs+ (the only serious
move - after 14 'i'bl 0-0 1 S j,d2 �e8 1 6 0-0
�d6 17 eS �£5 1 8 j,xf5 exf5 1 9 'ii'x f5 j,x£3
20 'i'x£3 'ii'xd4 Black was already slightly
better in Crarnling-Karpov, Oropesa del Mar
1 996, and 1 4 dS?! exdS 1 S l:td1 0-0 1 6 0-0
dxe4! was better for Black in Miles­
Polugaevsky, Sarajevo 1 987) 14 ... 'i'd7
(14... j,c6! 1S 'i'bl 0-0, with good play for
Black, is probably what Karpov would have
played against Cramling) 1 S �eS 'i'xbS 1 6
j,xbs+ 'it>f8 1 7 £3 �e8! 1 8 i.d7 %:td8 1 9 j,c6
j,c8! and Black was only slightly worse in
Timman-Karpov, Jakarta 1 993. Also possible
is 10. .. j,e7, when 1 1 dS?! is too early:
1 1 ...exdS 12 exdS 0-0 1 3 :d1 ?! gS! 1 4 j,cl
g4 1S �d2 j,xdS and Black is close to win­
ning. Instead White should develop first with
1 1 j,d3, e.g. 1 1 ...Ac8 1 2 'i'e2 0-0 1 3 0-0
cxd4 14 cxd4 �f6 (not 14 ... l:tc3? 1 S %:tfc1 !
l:lxcl+ 1 6 l:lxc1, Gulko-Timman, Amster­
dam 1 987, when White had the advantage
due to the threat of l:tc7) 1 S a4 Ac3! 1 6 eS
�dS 1 7 j,d2 l:tc7 1 8 l:tfc1 with equality in
Bareev-Chernin, Lvov 1 987.
a2) Black can also try 9 ... j,e7 10 j,bs+!
j,c6 1 1 j,d3 and the bishop is badly placed
on c6. No need for further variations ...
a3) Or 9 ... �c6, which also puts Black in
trouble after 1 0 j,b2!, e.g. 1 0 ... i.e7 1 1 dS!
exdS 12 .rtd1 0-0 (worse is 1 2. .. j,f6 13 exdS
56
�eS 14 �xeS j,xeS 1S j,bS+ 'it>f8 16 0-0
'i'd6 17 f4 j,f6 18 c4 a6 19 i.eS and White
is already winning, Tukmakov-011, Kujby­
shev 1 986 continuing 1 9 ... j,xeS 20 fxeS
'i'xeS 21 'ii' f2 f6 22 :£e1 'ti'hs 23 j,d7 'i'f7
24 j,e6 'ii'c7 2S d6 'ii'd8 26 i.h3 j,c6 27
l:te7 1 -0) 1 3 exdS �aS 14 i.d3 j,f6 1 S
j,xh7+?! (1 S 0-0! gives White the advantage
due to 1 S ... g6 1 6 c4) 1 S ... 'it>h8 1 6 j,e4 (1 6
0-0 g6 17 j,xg6 fxg6 1 8 'i'xg6 'i'e7 and
White's queen is alone in attack) 1 6 ... j,a6 17
j,d3 %:te8+ 1 8 'it>fl �c4 1 9 j,xc4 j,xc4+ 20
c;t>g1 :e2 and Black had enough for the
pawn (to put it mildly) in Portisch-Sosonko,
Tilburg 1 982. 10 ...g6 1 1 :d1 i.g7 1 2 dS exdS
1 3 exdS 'i'e7+ 14 j,e2 �eS 1 S 'ti'a4+ �d7
16 0-0 0-0 17 l:tfel left White better devel­
oped in Petrosian-Sosonko, Tilburg 1 982,
while 10 ... .rtc8? 1 1 dS!? exdS 12 exdS 'i'xdS
1 3 i.d3 �eS 14 0-0-0 was just a killer in
Agdestein-Lau, Dortmund 1 987. I would say
that Black's king is not completely safe ...
b) 8 j,gS is also not so dangerous. After
8 ... i.e7 9 i.xe7 'ii'xe7 1 0 �xdS exdS 1 1 e3
0-0 1 2 j,d3 g6 (taking control over the king­
side light squares) 1 3 dxcS bxcS 1 4 0-0 �d7,
as in Seirawan-Timman, Amsterdam 1 992,
Black secured equality. Vilela-Rodriguez,
Cienfuegos 1 98S seemed okay for Black after
1 1 g3 0-0 1 2 i.g2 �d7 1 3 0-0 :ac8.
8 i..x c5 9 i.. g 5!
. . .

The most energetic. There are two differ­
ences between this and S i.g5 - Black no
longer has a c-pawn, so the d4-square is
available to White, and Black has already
moved his bishop. Now there are three dif­
ferent paths to choose from. I would rec­
ommend 9 ... f6, although none of them
promises equality.
9 'i'c8
. . .
Others:
a) 9 .. .£6 10 i.d2 0-0 (Black should not
hesitate over this logical move - after
10 ... lDd7 1 1 e4 lDxc3 1 2 i.xc3 'i'cS 1 3 0-0-0
0-0 14 b4 i.e7 15 'i'b3 %:tf7 16 i.c4 lDf8 1 7
lDd4 White had strong pressure l_his king will
be fine on bl] in Plaskett-Short, Plovdiv
1984) and now White should play 1 1 e3!
(intending to use the b1 -h7 diagonal), e.g.
1 1 ...q.,hS?! 1 2 i.c4! 'ile7 (1 2 ... lDxc3?! 1 3
i.xc3 'We7 1 4 h41 wi th th e idea of h4-h5 with
a very dangerous attack) 1 3 %:td1 !? and White
has the better prospects (Piket). An im­
provement is 1 1 ... a5! - there is no need to go
to the h-file before being forced to do so.
Gelfand-Lautier, Manila 1 990 continued 1 2
i.d3 c:;ihs! 1 3 l:.d1 (13 h4!? 'Wes 1 4 lDg5 £5
1 5 e4 lDf6 1 6 0-0-0 lDc6! offers Black coun­
terplay according to Gelfand, while 1 3
i.xh7? £5 1 4 i.g6 'i'f6 1 5 i.h5 'ilh6 benefits
only Black) 13 ...lDd7 14 0-0 'We7 15 i.e4!
lDxc3 1 6 i.xc3 (1 6 i.xb7? lDxd1 17 i.xaS
lDxe3 and Black wins) 1 6 ...:a7 17 lDd4?! (17
l:.d21? and White is perhaps a little bit better ­
Gelfand) 1 7 ... i.xd4 1 S exd4 %:tdS! 1 9 .l:fel
i.xe4 20 'Wxe4 lLJf8 21 d5 'i'd6! with equal­
ity. Note that after 1 1 e4 there is no good
place for the light-squared bishop, e.g.
1 1 ...lDxc3 12 i.xc3 a5 13 %:td1 'i'cS 14 i.d3
lDa6 1 5 0-0 �hS 1 6 %:tfe1 e5 with equality in
Vyzmanavin-Lautier, Sochi 1 9S9, or 1 3 i.c4
'Wes 1 4 'i'e2 i.a6 1 5 ..L:a6 'ilxa6 1 6 'i'xa6
lDxa6 1 7 q.,e2 e5 1 S :bd1 %:tfdS with a good
position for Black in Benjamin-Korchnoi,
Jerusalem 1 9S6.
b) 9 ... i.e7 is slightly passive. After 1 0
i.xe7 'Wxe7 1 1 lDxd5 exd5 1 2 e 3 0-0 1 3
Th e Pe tro sian Sys t em: 4 a 3
i.d3 %:tcS 1 4 'We2 lDd7 1 5 0-0 lDc5 1 6 lDd4
lDxd3 1 7 'Wxd3 Ac5 Black had the c-file to
compensate for his structural deficiencies in
Ehlvest-Korchnoi, Zagreb 1 9S7. In Gure­
vich-Ionescu, Moscow 1 9S7 White played 1 3
i.b5! to be ready to remove Black's knight.
Play continued 1 3 ... d4 14 lDxd4 i.xg2 1 5
:g1 i.e4 1 6 lDfS! i.xfS (1 6 ...'i'e5? 1 7 lDh6+
q.,hS 1 S lDxf7+! and White wins) 1 7 'WxfS a6
1 S i.d3 g6 1 9 'Wd5! with a clear advantage to
White. Black's 1 3th move seems a bit too
optimistic. White is just better developed for
this kind of stuff. Best was 13 ... :cs, when
Black is still fighting to equalise.
1 0 :c1
1 0 e3 0-0 11 lDxd5 hd5 1 2 i.d3 h6 13
i.h4 lDd7 14 0-0 a5 1 5 l:.acl a4 was harm­
less in Farago-Horvath, Hungary 1 995.
1 o h& 1 1 .i.h4!
...
The only serious move - the bishop must
remain active. For example after 1 1 i.d2
lDf6! the bishop looks stupid on d2, e.g. 1 2
e 3 0-0 1 3 i.e2 lDbd7 14 0-0 'i'bs 1 5 b 4 i.e7
16 lDd4 l:.cS 17 'i'b3 a5 1 S £3 axb4 19 axb4
i.d6 and Black was clearly better in Chek­
hov-Timoshchenko, Berlin 1 9S6, or 1 2 b4
i.e7 1 3 lDb5 lDc6 1 4 'ila4 a5 1 5 bxa5 0-0
with unclear play in Petursson-Amason,
Reykjavik 1 9S5. Meanwhile 1 1 lDxd5?1 hxg5
1 2 b4 g41 1 3 lDe5 l:.h5 favoured Black in
Petursson-C.Hansen, Borgarnes 1 9S5, and
Ligterink-Beliavsky, Wijk aan Zee 1 9S5 went
1 1 lDe4 lDd7 (1 1 ...hxg5?? 1 2 lDd6+1) 12
lDxc5 lDxc5 1 3 e4 (1 3 i.d2 lDf6!, to take
control of e4, guarantees Black a good game)
1 3 ... hxg5 1 4 exd5 exd5 1 5 lDxg5 'it>f8 1 6
i.e2 d4! 1 7 0-0 d3 1 S i.xd3 'i'g4 1 9 lDe4
i.xe4 20 i.xe4 'i'xe4 and White resigned.
1 1 a5
. . .
Pe'1iaps the altematives are better:
a) 1 1 ...lDc6 1 2 lDxd5 exd5 13 e3 0-0 14
i.d3 a5 1 5 0-0 d4 1 6 e4 :es 1 7 i.g3 with a
slight advantage to White in Ftacnik­
Hjartarson, Esbjerg 1 9S5.
b) 1 1 ...0-0 12 lDxd5 exd5 13 e3 lDd7 14
i.e2 i.d6 15 'i'dl 'i'eS 16 0-0 lDc5 17 lDd4
57
Q u e e n 's In dia n D e fe n c e
'ii'e4 1 8 .1g3 with only a tiny edge for White,
Yusupov-Chemin, Tunis 1 985.
1 2 l2Ja4!
The only serious option. 1 2 �e4 �d7 1 3
'ibl 0-0 14 �d6 'ii'c6 1 5 �xb7 looks like
White has never heard about development,
but only about the two bishops (well, the
king's bishop is yet to wake up). Ftacnik­
C.Hansen, Esbjerg 1 985 led to Black eaming
an enormous lead after 1 5 ... 'ii'xb7 1 6 e4? (1 6
e3 was still equal) 16 ...�f4 1 7 .1g3 £5! etc.
12 e3 0-0 13 .1e2 �xc3 14 'ii'xc3 �c6 1 5
0-0 a4 was equal in Spraggett-Portisch,
Montpellier 1 985.
1 2 . . . l2Jd7 1 3 e4!
White needs to get going before Black fin­
ishes his development. After 1 3 e3· 0-0 1 4
.ib5 .1a6 1 5 .1xa6 (1 5 'i'd3 .1b4+ 1 6 <t>e2
�c5 1 7 .1xa6 'ii'xa6 1 8 'i'xa6 l:.xa6 1 9 �xc5
.1xc5 with equality) 15 ... 'ii'xa6 Black was no
worse in Salov-Timman, Belgrade 1 987.
1 3 . . . l2Jc7
Black's position is beyond repair, e.g.
13 ... �f4 14 .1g3 �h5 15 �xc5! �xc5?!,
while 15 ...bxc5 is horrible too in view of 1 6
b4! axb4 1 7 axb4 �xg3 1 8 hxg3 �b3 1 9
.ibs+ <t> f8 2 0 'ii'xc8+ l:.xc8 2 1 l:.xc8+ .1xc8
22 .1c4 and White wins. After 13 ... �5f6 1 4
.i.xf6 �xf6 1 5 .ib5+ �e7 1 6 e 5 �e4 1 7 0-0
White has a powerful initiative.
1 4 l2Jxc5 bxc5
14 ... �xc5? 15 b4 and Black drops a piece.
1 5 .i.e2 .i.a6
58
1 5 ... e5 1 6 .1c4! with a large lead.
1 6 0-0 0-0 1 7 :td 1 f6!
17 ... l:.e8?! 18 .1xa6 l:.xa6 19 'i'd2! and
White wins.
1 8 .i.c4! .i.xc4 1 9 'it'xc4 :t7 20 .i.g3 e5
21 l2Jh4!
With a clear advantage for White.
21 . . . 'it'e8
21. ..'i'a6? 22 'ii'xf7+! <t>xf7 23 l:.xd7+ and
White wins.
22 l2Jf5 'it'e6 23 'it'e2! l:l.b8 24 l:td6 'it'e8
On 24... 'i'b3 Kasparov gives the following
winning line: 25 'ii'g4 <t>h7 26 �xg7 (26
�xh6!? gxh6 27 :Xd7 �h8 28 l:.cd1 also
looks good) 26 ...'i'xb2 27 l:.£1 l:.xg7 28 l:.xd7
�e8 29 f4 and here Black has nothing better
than 29 ...'i'd4+ 30 l:.xd4 l:.xg4 31 l:ld7+ l:.g7
32 l:.xg7+ @xg7 33 fxe5 l:.b6 with a terrible
endgame.
25 :cd 1 ! l2Jf8 26 f4! l2Jb5
26 ... exf4 27 .1xf4 �fe6 28 .1g3 'Wb5 29
'i'g4 @h7 30 l:.xe6 �xe6 3 1 �d6 and White
wins.
27 fxe5 l2Jxd6?!
Apparently 27 ... fxe5 28 l:.6d5 �d7 29
'i'd2!, with a decisive advantage to White, is
better.
28 l2Jxd6 'it'a4 29 l2Jxf7 :xb2! 30 l2Jxh6 +
lfi>h7
3 1 'it'h 5
31 'ii'£3! is even stronger.
31 . . . g6 32 'it'f3 <li>xh6 33 :n 'it'd4 + 34
lfi>h 1 l2Jh7 35 exf6 l2Jxf6 36 .i.f4+

36 e5 �h7 37 i.£4+ '1tg7 38 i.cl and
Black has nothing left to do but resign.
36 . . . r;1;>g7 37 i.g5 l:.b6 38 'i'h3?!
Again 38 'iWf4! was simpler.
38 . . . "1i>g8 ?!
38 ...'ifxe4! 39 i.h6+ rt>h71 40 i.£8+ '1tg8
41 i.xc5 gives White a pawn but at least puts
up some resistance.
39 'ii'c 8+ �g7 40 'i'c7 + li:ld7 41 'ii'f4
'i'c4 42 h3 'i'e6 43 l:.d 1 ! .:l.c6 44 i.d8!
.!Lib& 45 :n 1 -0
------..
Game 20
Khalifman-Short
Paernu 1996
1 d4 li:lf6 2 c4 e6 3 li:lf3 b6 4 .!Llc3 i.b 7
5 a3 d5 6 cxd5 li:lxd5 7 e3
This is the main move in ECO, but in re­
cent times it has been less popular among
those who know. This is due to 7 ... g6, and
for this reason I will not consider other
moves here, or give 7 e3 serious attention.
Anyway, White plays the text with the idea of
e3-e4. It does not seem to be so dangerous.
7 . . . g6
8 i.b5+
Here there are other possibilities:
a) 8 h4 _ag7 9 h5 �d7 10 i.d3 0-0 1 1 e4
(this is just not very goodQ 1 1 ...�xc3 1 2
bxc3 �c5! 1 3 hxg6 hxg6 1 4 dxc5 �osing, but
Black is already doing very well) 14 ... i.xc3+
1 5 rt>e2 .axal 1 6 'ii'gl 'ii'd7 1 7 'ii'h2 l:lfd8 1 8
Th e Pe tro sian Sys t em: 4 a 3
'i'h7+ '1tf8 1 9 .i.h6+ '1te8 and the attack
never went further in Vokac-Sax, Lazne
Bohdanec 1 995. 10 ...�xc3 1 1 bxc3 .:r.c8 ap­
pears to be equally good, when White should
start thinking of castling. After 1 2 e4?! �c5!
13 hxg6 �d3+ 14 'ii'xd3 hxg6 15 Axhs+
.i.xh8 1 6 .i.f4 .i.g7 17 '1td2 c5 Black was
better in Maksimovic-Groszpeter, Berlin
1 988. 9 ... 0-0 is also okay, although it seems
natural to wait. Here after 10 hxg6 hxg6 1 1
�xd5 exd5 1 2 .ad2 �d7 1 3 j,, d3 l:le8 1 4
.i.c3 a5 15 'ili'c2 .i.a6 1 6 .i.xa6 Black has no
problems, Razuvaev-Rodriguez, Moscow
1 985 already coming to a peaceful conclu­
sion.
b) 8 �xd5 exd5 9 b4 is supposed to give
White an advantage. But due to the plan illus­
trated in this game (... c7-c6 and ... a7-a5) no­
body adopts this approach any more. Black is
just fine. Shirov-Karpov, Monte Carlo 1 995
went 9 ... i.g7 10 i.b2 c6 11 i.e2 0-0 12 0-0
�d7 13 'ili'b3 Ae8 14 Afcl a5 1 5 bxa5 Axa5
16 a4 c5 with equality.
8 . . . c6 9 i.d3 i.g7 1 0 .!Lla4!?
Short wrote that this move was an inven­
tion of Lobron and his long-term companion
] ack Daniels. Surely Mr. Daniels is an inven­
tive guy. After 1 0 e4 �xc3 1 1 bxc3 c5! 1 2
.ag5 'ii'd6 1 3 e5 1i'd7 Black seems to d o well
as White's centre appears rather fragile. And
after 14 dxc5?! 0-0! 1 5 cxb6 axb6 Black was
even better in Kasparov-Korchnoi, London
1 983. White has three weak pawns and is
lacking in co-ordination and development. .
1 0 . . . li:ld7 1 1 e4 .!Lle7 1 2 0-0
Here (and on the next move) White has a
serious alternative in 12 i.f4 0-0 13 0-0.
Then Black needs to accept a weak pawn and
he will be okay: 13 ... c51 14 dxc5 �xc5 1 5
�xc5 l;>xc5 1 6 'ife2 (1 6 'ii'c2 �c6! and Black
is fine ; while 16 Acl c4! also equalises be­
cause after 1 7 Axc4? i.a6 1 8 .ac7 'iWd7 1 9
�e5 i.xe5 20 i.xe5 i.xc4 2 1 i.xc4 'ii'x dl
22 .i::r.x dl Afd8 Black has all the chances)
1 6 ... �c6 1 7 l:lacl �d4 1 8 �d4 i.xd4 with
equality in Yermolinsky-Ivanov, Parsippany
59
Q u e e n 's In dia n D e fe n c e
1 996.
1 2 . . . 0-0 1 a i.g5
13 i.f4-see the previous note..
1 a . . . h6!
1 3 ...l:.e8?1 (Khalifman) 14 l:.cl °i'b 8 15
i.h4! (going to the h2-b8 diagonal) 15 ...h6 16
b4 bS 1 7 .lbcS .lbxcS 18 Les with a strong
positional advantage for White in Khalifman­
Sivokho, St Petersburg 1 996.
14 i.ea �h7 1 5 :.c1 f5!
Black has equalised (Short).
1 6 exf5 exf5
16 ... .lbxfS 17 i.e4 is just wrong.
1 7 i.f4 lL!d5 1 8 i.d6?!
18 i.g3 improves.
1 8 . . . :.e8 1 9 tLica °i'f6 20 i.ga lL!xca 21
bxca c5
Here Black could have rewarded himself
for his strong opening play with 21...f4! 22
i.h4 'ikf7, when all White's pieces are un­
prepared for ... c6-c5.
22 i.b5 :.e7 2a lL!e5 lL!xe5 24 dxe5 °i'f7
25 f4? !
60
Here White could have kept the balance
with 25 'ii'd 61 f41 26 i.xf4 i.xg2 27 'iii'xg2
'ir'xf4 28 'ii'xe7 'ii'g4+ with a perpetual.
25 . . . a6 26 i.e2 :.d7 27 "i'e1 "i'e6 28
i.ta i.xf3 29 :.xta i.f8 30 i.f2? !
30 i.h4! is forced. Now Black will be able
to play ... g6-g5 and prove a clear advantage.
ao . . . i.e7 a 1 c4 :.ad8 a2 :.tea g5!
The e-pawn is now in trouble. From here
Short's fantastic technique took over.
aa i.ea gxf4 a4 i.xf4 :.d4 as °i'f1 i.g5
a6 i.xg 5 hxg5 a7 "i'e2 �g6! a8 :.ea
:.d2 a9 "i'ta g4 40 "i'n :.8d4 41 9a :.b2
42 :.e2 :.xe2 4a "i'xe2 :.e4 44 °i'd2
"i'xe5 45 �2 :.d4 46 "i'e2 Wxe2+ 47
�xe2 f4! 48 gxf4 �f5 49 :.b1 :.d6 50
:.n :.h6 51 :.t2 :.ha s2 �n :.ba! sa
�g2 a5 54 :.e2 a4 55 :.e5 + �xf4 56
:.e6 :.xaa 57 :.t6 + �e5 58 :.xb6 :.ba
59 :.a6 aa 60 :.as ltf4 0-1
 Th e Pe tro s ia n Sys t em: 4 a 3

Summary
The Petrosian system still holds great dangers for Black as it did 15 years ago when Kasparov
was championing it. The move order with 4 a3 gives Black some extra lines to choose from,
most notably 4... i.a6, therefore it can be a good idea for White players to play 4 liJc3 with the
idea of 4 ... i.b7 5 a3! with transposition to the 4 a3 i.b7 lines. After 4 a3 then 4 . . dS gives a .

typical d-pawn position, as discussed in the introduction. After 4... cS!? we get some interesting
and unusual positions. Black has so far been able to keep the game level, but perhaps White
can make his better structure count for something in the future. 4... c6 is just not very good,
while 4 ... i.a6 leads to a whole series of variations which are basically okay for Black although
precise play is needed. 4... i.b7 is still the main line, and after 5 liJc3 Black only has one good
plan: 5 ... d5 6 cxd5 liJxd5 and now 7 'iilf c2! is perhaps very slightly better for White.

1 d4 lLlf6 2 c4 e6 3 lLlt3 b6 4 a3 (DJ .i.b7

4 ... d5 5 cxd5 exd5 6 liJc3 i.e7 Game 5; 4 ... c6 - Game 6
-

4 . . . c5
5 e3 Game 8; 5 d5 i.a6 6 'iiic2 Game 7
- -

4 ... i.a6
5 'tib3 Game 9
-

5 'iiic2
5 ... c5 6 d5 - 4 ... c5
5 . . . i.b7 6 liJc3 c5
7 d5 - Game 10
7 e4 cxd4 8 liJxd4
8 . . i.cS - Game 1 1; 8 . . liJc6 9 liJxc6 i.xc6 1 0 i.f4
. . (DJ
·

1 0 . . liJh5 - Game 12; 1 0 . i.c5 Game 13

. .. -

5 lLlc3 d 5
5 ...g 6 Game 14
-

6 cxd5
6 i.g5 - Game 15,· 6 'iii c2: 6 ... dxc4 Game 16; 6 ... i.. e7 Game 1 7
- -

6 lLlxd 5 (DJ 7 'ii' c 2

. . .

7 e3 - Game 20
7 c5
. . .

7 ... liJxc3 - Game 1 8

8 dxc5 - Game 19

4 a3 1 0 .i.f4 6 . . . lLlxd5

61
 CHAPTER THREE I
5 ii'c2 and 5 i. g 5

1 d4 itif6 2 c4 e6 3 itif3 b6 4 ltic3
i.b7/i.b4 5 'i'c2/5 i.g5
In this chapter you will see White try to
put Black under pressure and fight for the
control of the e4-square - and you will see
him fail. These lines do not offer White an
advantage from the opening, yet they can
nevertheless be difficult to play for Black,
and if they suit you, as they used to suit
Kamsky, there is a definite possibility that a
level opening might later lead to a superior
middlegame and, ultimately, a full point.
e4, is the best way to play with this structure.
After the not so clever 10 ... dS 1 1 0-0 liJbd7
(1 1 ...dxc4 1 2 :fdl 'i'c8 13 e4 gave White an
initiative in Razuvaev-Kruszynski, Polanica
Zdroj 1 979) 1 2 :fdl :cs 1 3 :acl White
had an edge in Sokolov-Stohl, Burgas 1 992.
1 1 0-0 ltic5 1 2 l:fd 1 ltife4 1 3 i.xe7
13 i.g3 liJxc3 14 'ii'xc3 liJe4 favours
Black thanks to the nvo bishops.
1 3 . . . 'i'xe7

Game 21
Tolnai-Adorjan
Hungary 1992
1 ltif3 itif6 2 c4 b6 3 ltic3 i.b7 4 d4 e6
5 'i'c2
A rather harmless option, the game con­
tinuation demonstrating the easiest way to
equalise.
5 . . . c5! 6 dxc5 i.xc5 7 i.g5
7 e4 h6 8 i.f4 i.b4 9 i.d3 0-0 10 0-0 1 4 itid2?1
i.xc3 1 1 bxc3 dS 12 cxdS exdS gave Black at Presenting Black with the opportunity to
least equality in Szekely-Daroczy, Debrecen take full control of the centre. 14 liJxe4, with
1 956. equality, improves.
7 . . . h6 8 i.h4 i.e7 9 e3 0-0 10 i.e2 14 . . . ltixc3 1 5 'i'xc3 d5 1 6 i.f3
ltia6! 1 6 b4 liJd7 17 cxdS :£c8 is nice for Black.
This move, aimed at planting a knight on 1 6 . . ..Z:.fdS 1 7 h3

62

1 7 b4!? tlia6 1 8 a3 .l:.ac8 1 9 'i'b3 dxc4 20
tlixc4 i.xf3 21 gxf3 tlic7 is a shade prefer­
able for Black.
1 7 . . . dxc4! 1 8 i.xb7 l'Lixb7 1 9 l'Lixc4
:ac8 20 'ila3 'i'xa3 21 l'Lixa3 lbc5 22
l'Lib5 a6
·
23 l'Lic3?!
White could have equalised with 23 tlid6
:lc6 24 tlic4 tlid3 25 tlie5 .l:.cd6 26 tlic4,
with a repetition.
23 . . . l'Lid3 24 :d2 l'Lie5 25 :xd8+
25 .l:.ad1 .l:.xd2 26 .l:.xd2 @f8 is only
slightly better for Black. Now Adorjan's fan­
tastic will to win comes into play
25 . . . :xdB 26 :d 1 :ea 27 Ad4 ..tits 28
..tin rt;e7 29 rt;e2 l'bc4 30 l'Lid 1 b5 3 1
:d3 b 4 32 f4?
32 .l:.b3 was forced.
32 . . . a5 33 a3?
33 .l:.d4, with problems but nothing more,
5 'il c 2 . a n d 5 i. g 5
was necessary.
33 . . . l'Lixb2! 34 l'Lixb 2 :c2+ 35 q;f3 :xb2
36 axb4 :xb4 37 ltc3 :b7 38 :a3 Aa7
39 q;e4 q;d6 40 q;d4 a4 41 q;c4 f5 42
rt;b5 :b7 + 43 ..t>c4 lta7 44 ..t>b5 :b7 +
45 ..t>c4 :b2 46 :xa4 ltxg2 47 .f:.a7 g5
48 :as+ q;e7 49 :a1 + q;f6 50 .:l.h7
q;g6 51 Ae7 ..t>f6 52 Ah7 gxf4 53 :xh6 +
'1Pg 7! 54 :xe6 f3 0-1
Game 22
Kamsky-Yudasin
Biel 1993
1 d4 l'Lif6 2 c4 e6 3 l'Lif3 b6 4 l'Lic3 .tb7
4 ... i.b4 is treated below.
5 .tg5
TIUs variation is known to be quite harm­
less but, as this game clearly illustrates, it is a
genuine struggle so long as the pieces remain
on the board. The most normal move here
must be 5 a3, transposing to the Petrosian
System. Actually, this is the way I would play
with White, as the lines with 4... i.b7 seem to
be the most fragile for Black in the Petrosian.
5 . . . h6 6 .th4 i.e7
No}V Black has this possibility, and there­
fore equalises easily.
7 e3
Standard. 7 'iic2 has also been tried, when
7 ... c5! is another standard reaction, White
having neglected the dark squares. Sokolov­
Salov, Wijk aan Zee 1 997 went 8 dxc5 bxc5 9
63
Q u e e n 's In dia n D e fe n c e
e3 d6 1 0 i.e2 �bd7 1 1 .l:dl 0-0 12 0-0 a6 1 3
:d2 ii'c7 14 .:tfdl :fd8 1 5 h 3 �IB 1 6 �h2
:d7 17 i.g3 l:ad8 1 8 i.£3 �g6 19 i.xb7
'ii'xb7 20 �£3 d5 and, if anyone is better, it is
Black.
7 . . . lt:le4
This is a good way to equalise. Also possi­
ble is 7 .. 0-0, and now:
.
a) 8 i.d3 c5 9 0-0 cxd4 (9 .. d6 1 0 a3 �bd7
1 1 b4 d5 12 cxd5 �xd5 13 i.xe7 ii'xe7 1 4
�xd5 i.xd5 1 5 e 4 and a draw was agreed in
Onischuk-Dautov, Gennany 2000) 1 0 exd4
d5 (also possible is 1 0 ...i.x£3!? 1 1 ii'x£3 �c6
12 �e2 d5 13 .:.adl �b4 14 i.xf6 i.xf6 1 5
cxd5 �xd5 1 6 i.e4 with a draw in Seirawan­
Schussler, New York 1 985) and now the
most testing is 1 1 i.xf6!, unlike the hannless
1 1 cxd5 �xd5 12 i.g3 �xc3 1 3 bxc3 �c6
14 :e1 i.f6 1 5 .l:cl .:r.c8 16 ii'e2 �a5 1 7
�e5 'ii'd 5 1 8 f4 i.xe5 1 9 fxe5 :c7 with
complete equality in Spassky-Hjartarson,
Reykjavik 1 988. In Sokolov-Rivas Pastor,
Leon 1 995 the position after (1 1 i.xf6)
1 1 ...i.xf6 1 2 cxd5 exd5 was reminiscent of
the Tartakower where the bishop on b7 is
somewhat silly, and where there might be a
weakness on c6 occasionally. But the ex­
change of the dark-squared bishop does also
influence White's position. After 1 3 .:.e1
�c6 14 i.c2 ii'd6 15 ii'd3 g6 16 a3 :re8 1 7
h 3 i. g7 th e game was equal, Black having
the manoeuvre ... �a5-c4.
b) 8 i.e2 c5 9 0-0 d6 (9 ... cxd4 10 exd4 d5
is, of course, also possible) 10 dxc5 bxc5 1 1
'ii'c2 �bd7 1 2 :fdl 'ii'b6 1 3 .:.d2 :fd8 with
equality in Karpov-Polugaevsky, Bugojno
1 980. 1 0 :cl �bd7 1 1 a3 :c8 1 2 i.g3 a6 1 3
b 3 cxd4 1 4 �xd4 ii'c7, Groszpeter-Tal, So­
chi 1984, was at least equal for Black. The
move a2-a3 does not improve White's posi­
tion.
8 lt:lxe4 ! ?
This should not be dangerous for Black at
all, but neither is 8 i.g3 i.b4! 9 ii'c2 d6 1 0
i.d3 i.xc3+ 1 1 bxc3 £5 (1 1 ...�xg3?! is too
early: after 1 2 hxg3 �d7 13 e4 ii'e7 1 4 0-0
64
0-0 1 5 �h4! White had a space advantage in
Korchnoi-Ligterink, Wijk aan Zee 1 984) 12
h4 �d7 1 3 i.h2 ii'e7 1 4 0-0-0 0-0-0 1 5 :dn
g5 with dynamic equality in 11ikhalevski­
Yudasin, Tel Aviv 1 994.
8 ..txh4? 1
. . .
This is risky. A safe path to equality is
.8 ... i.xe4 9 i.g3 0-0, e.g. 10 �d2 i.b7 1 1
i.d3 c5! (standard when White weakens the
dark squares) 1 2 dxc5 bxc5 1 3 0-0 d5 14
cxd5 exd5 1 5 e4 dxe4 16 i.xe4 i.xe4 17
�xe4 �c6 18 'ii'a4 �d4 with complete
equality in Uhlmann-Padevsky, Havana 1 964,
or 1 0 i.d3 i.xd3 1 1 ii'xd3 d6 12 0-0 �d7 1 3
b 4 a 5 1 4 b5 £5 1 5 ii'b3 a 4 1 6 ii'c2 'ii'e 8 1 7
�el e5, Uhlmann-Antoshin, Leipzig 1 965,
also equal.
9 ..td3 ..te7 1 0 'ii'b 3!
White has a territorial advantage to add to
his superior development, which he uses to
generate an attack.
1 o d6 1 1 h41? ltid7 1 2 0-0-0 c6? !
. . .
Slightly passive. After 1 2 ... c5!? 13 dxc5
dxc5 14 �c3 wliite has a small edge (Kam­
sky) .
1 3 g4 'ii'c 7 1 4 'itrb1 a6 1 5 g 5 c5 1 6
'ii'c 3!? .:gs 1 7 :hg 1 hxg5?!
Now the g5-square becomes available for
White, with attacking possibilities on e6 and
f7. Better is 1 7 ... 0-0-0 with a messy position
where anything could happen.
1 8 lt:lexg5!
1 8 �fxg5?! allows 1 8 ...d5! etc.

.
1 8 . . . cxd4 ! ? 1 9 exd4 iJ..f 6 20 'il*'b3!
Shooting at e6 and f7 while taking control
of the light squares on the queenside.
20 . . . il.. x f3?
From here on Black is defenceless on the
light squares. A lesser evil is 20 ... lll f8 21
i.e4!? 0-0-0 with a poor - but not losing -
position.
21 ltJxf3 l:!.h8 22 l:!.g4!
A strong, active post.
22 . . . �e7 23 :e 1 b5!? 24 cxb5 axb5 25
lLlg5! il..x g5 26 hxg5 g6
White was considering g5-g6 himself.
27 l:!.ge4 �dB 28 il..x b5 d5 29 l:!.4e3 l:!.b8
30 J:tc 1 iib6 31 a4 'ilt'xd4 32 'ifc2 �b6
33 l:!.c3 l:!.h4 34 l:!.c6 1 -0
Black resigned due to 34 .l:f.c6 'i'a7 35
.l:f.c8+ rl;e7 36 i.xd7 etc.
Game 23
Portisch -Popovic
Ljubgana 1985
1 lLlf3 lLlf6 2 c4 e6 3 lLlc3 iJ.. b4 4 d4 b6
5 il..g 5 iJ.. b 7
5 ...h6 6 i.h4 i.b7 and now 7 'i'c2, along
with 7 llld2, is aimed at gaining control of e4.
Both moves weaken the d4-square and
should therefore be countered by ... c7-c5.
6 �c2
6 lll d2 h6 7 i.h4 c5! 8 a3 cxd4 9 axb4
dxc3 1 0 bxc3 0-0 1 1 f3 d5 12 e3 a5 13 bxa5
bxa5 14 i.e2 li:Jbd7 1 5 cxd5 exd5 16 0-0
5 fl c 2 a n d 5 iJ.. g 5
'i'e7 with a good position for Black in
Moutousis-Nikolaidis, Glyfada 2001 . An­
other suggestion for Black is 7 ... i.e7. Then,
after 8 e4, 8 ... d6 is too passive - 9 i.g3 li:Jbd7
1 0 i.e2 e5 1 1 d5 lll f8 1 2 b4 lllg6 1 3 li:Jb3 h5
14 h3 h4 15 i.h2 0-0 16 0-0 gave White the
advantage in Ibragimov-Bischoff, Pulver­
muehle 2000. Black's position is very solid
but too passive, and White can hope to bene­
fit from his space advantage on the queen­
side. Instead 8 ...lllxe4!? is another prospect
altogether: 9 i.xe7 lllxc3 10 i.xd8 lllx d1 1 1
i.xc7 lllxb2 1 2 a4 d5 1 3 cxd5 exd5 1 4 .l:ta2
lll c4 1 5 lllxc4 dxc4 16 i.xc4 with equality.
However, this position looks a little nasty for
Black, so I would feel less inclined to test this
line ...
6 . . . h6 7 il..h 4 g5 8 il..g 3 lLle4
This is the simple way to equalise against
this system. White now has no path to an
advantage.
9 e3
White can also go for complications, but
Black is okay after 9 i.e5 f6 1 0 d5 exd5 1 1
cxd5 i.xc3+ 1 2 i.xc3 i.xd5 1 3 0-0-0 lllxc3
14 'i'xc3 i.f?! 15 h4 g4 16 lll e5 i.h5 17 'i'c2
fxe5 1 8. 'i'f5 'i'e7 1 9 'i'xh5+ rl;d8 with dy­
namic �quality in Portisch-Timman, Hilver­
sum 1 984.
9 . . . iJ..x c3+ 1 0 bxc3 d6
A natural move. Black can also choose to
exchange the bishop immediately, which
should lead to the same type of positions:
65
Q u e e n 's In dia n D e fe n c e
1 0 ... ll'lxg3 1 1 hxg3 ll'lc6!? (Black decides that
c4-c5 is the only White way to make sense of
the position, so he avoids creating a 'hook'
with d6 - 1 1 ...d6 12 c5 ll'ld7 13 cxd6 cxd6 1 4
a 4 'itif8 1 5 i.d3 Ac8 1 6 a 5 bxa5 1 7 e 4 cl;g7
1 8 0-0 ended peacefully in Ehlvest-Adams,
Pula 1 997) 12 Abl (in Langeweg-Karpov,
Amsterdam 1 981 , the following manoeuvre
did not favour White: 1 2 ll'ld2 'flle7 1 3 'f!lb2
g4 14 ll'lb3 'ii'g5 1 5 c5 Abs 16 'ii'a3 h5 1 7
Ah4 i.aS 1 8 i.e2 f5 and Black i s already
better) 1 2...'f!le7 1 3 c5 h5 1 4 i.e2 ll'la5 1 5
ll'ld2 i.xg2 1 6 l:.xh5 0-0-0 1 7 cxb6 axb6 1 8
Axh8 Axh8 1 9 i.. £3 i.xf3 20 ll'lxf3 'ii'f6 fa­
voured Black in Gheorghiu-Miles, London
1 980.
1 1 .i.d3 tl:lxg3 1 2 fxg3! ?
1bis decision is quite common in these
positions. White wants to build up his artil­
lery with the rooks on f1 and £2, aiming at f7.
Black should be able to meet this plan with­
out fear, but some attention is required, of
course.
1 2 . . . tl:ld7 1 3 0-0 'fke7 1 4 .i.e4
Other, equally good, options are:
a) 14 ll'ld2 0-0 15 g4!? seems rather exotic,
and Black gets a good position after 1 5 ... ll'lf6
1 6 h3 h5! 1 7 gxh5 ll'lxh5 1 8 g4 ll'lg3 1 9 Af2
f5 20 <ith2 ll'le4!, the pawn sacrifice provid­
ing enough compensation to draw, but not
more. Van Wely-Anand, Tilburg 1 998 con­
tinued 21 ll'lxe4 fxe4 22 :.Xf8+ Axf8 23
i.xe4 i.xe4 24 'fllxe4 l:.£2+ 25 cl;gl 'f!lf7 26
66
Ael cl;g7 27 a4 1/2-1/2 And Black cannot use
•.
his activity for anything, as the rook must
stay on f2 (otherwise White has Afl), while
White cannot try to win as Black is too active
to make a single pawn important.
b) 14 Af2 is the most natural.. 1bis does
not mean that White immediately wants to
play Aafl , but that he can do so whenever he
pleases, without spending time in prepara­
tion. Quite simply, his position becomes
more flexible. We are following Zvjaginsev­
Timman, Biel 1 995, which continued
1 4 ... 0-0-0 1 5 i.e4 l:.hf8!? 1 6 l:.el (intending
to meet 1 6 .. fS with i.xb7 and e3-e4 with a
.
good game) 1 6 ... i.xe4 1 7 'ii'xe4 cl;bS 1 8
'ii'h7 Ah8 1 9 'ii'c2 h 5 with unclear play. Now
Black slightly misplayed his position with 20
e4 h4 21 gxh4 gxh4 (21 . ..g4 22 ll'ld2 'ii'xh4
23 g3 'f!lh6 is better, with chances for both
sides) 22 e5 with a modest lead for White. 1 6
a4!? was suggested by Timman a s a possible
improvement. It is hard to say anything
about that...
14 ... .i.xe4 1 5 'fkxe4 0-0 16 tl:ld2 c5 1 7
g4 :ac8 1 8 :f2 �g7 1 9 :an :ce8
The position is more or less equal. Now
the longest arms won the fight.
20 ii'b7 tl:lbS 21 'fkf3 f5 22 ii'g3 tl:ld7 23
�h 1 fxg4 24 'fkxg4 :xf2 25 :xf2 :ta 26
:xf8 �xf8 27 �g 1 tl:lf6 28 'fkf3 �g7 29
e4 cxd4 30 cxd4 e5 31 'ikd3 exd4 32
'fkxd4 co1>f7 33 �f 1 'fie& 34 �e2 'ikg4+
35 �d3 'ikxg2 36 e5 'flh3 + 37 �c2
 5 'ii' c 2 a n d 5 i.. g 5

'ii'f 5+ 38 c,ilb3 dxe5 0-1 queenside and is overrun completely. 13 £3?

------.. is how this happens: 13 ...h4 14 i.£2 0-0-0 1 5
Game 24 h3 :dg8 1 6 i.e2 lCih5 1 7 'i'a4 <ittb 8 1 8 c5
Belozerov-Nikole nko dxc5 1 9 i.a6 i.a8 20 lCib3 c4! 21 i.xc4 £5 22
Moscow 1999 i.e2 g4! (the final breakthrough; it is clear
now that Black is the quicker) 23 fxg4 fxg4
1 d4 e6 2 lbf3 lLif6 3 c4 b6 4 lL!c3 i.. b4 24 i.xg4 lCihf6 25 lCic5 lCixc5 26 dxc5 lCixg4
5 i..g 5 i.. b 7 6 e3 h6 7 i.. h 4 i.. x c3 + 8 27 hxg4 'iWxc5 28 l:tfdl h3 29 e4 h2+ and
bxc3 d6 9 i..d 3 White resigned in Borik-Speelman, Dort­
This move is less good than the theory's mund 1 981 . The alternatives are no better,
favourite, 9 lCid2!, for obvious reasons. After e.g. 15 a4 a5 1 6 l:tbl g4 17 fxg4 h3 1 8 g3
the bishop is developed White is forced to :lhg8 1 9 i.e2 lCie4 20 i.f3 lCixd2 21 i.xb7+
hurry with kingside castling as the knight <itfxb7 22 'i'xd2 :Xg4 23 'ii'e2 £5 with a clear
cannot move, and White cannot build his advantage for Black in Rohde-Benjamin,
usual impressive centre. USA 1 986, or 1 5 e4 :ldg8 (1 5 ... lCih5!? 1 6 c5!?
9 'ii'e 7
• . . dxc5 17 'iWa4 lCif4! 18 i.a6 lCib8! gave Black
a good position in Hjartarson-Stoica, Taflfe­
lag-Politehnica 1 987) 1 6 h3 lCih5 1 7 l:tel lCif4
1 8 i.c2 c5 1 9 lCifl £5 20 ex£5 g4!! 21 fxg4
i.xg2 with an overwhelming attack in Ne­
nashev-Tiviakov, Groningen 1 997.
1 3 . . . .C.gS
13 ... lCig4 has also been tried, but it seems
that White gets better control of the kingside
after 14 lCif3 gxh4 1 5 lCixh4 0-0-0 1 6 e4
(Beckmann-Steil, Germany 1 990) than he
normally does.
14 f3 0-0-0 1 5 hxg5 .C.xg5 1 6 i.. h 4 .C.g7

1 0 0-0
Practically the only alternative is 10 'ii'c2,
which does not pose Black problems after
1 0 ...gS 1 1 i.g3 h5 1 2 h3 (now this is forced
because 12 �gS? drops a piece to 12 ... h4 1 3
i.f4 e5, while 1 2 h 4 g4 1 3 tDd2 i.xg2 drops
a pawn) 1 2 ... h4 13 i.h2 l:tg8 1 4 e4 lCic6 1 5
lCid2 e5 1 6 d 5 and the bishop i s now buried
on h2, and White will have to do some ma­
noeuvring to get it out. First the other bishop
needs to get outside the pawn chain:
1 6 ... lCib8 1 7 i.e2 tDbd7 1 8 i.g4 lCixg4 1 9
·
hxg4 lCic5 2 0 f3 i.c8 and a draw was agreed 1 1 .C.f2
in Beliavsky-Gulko, Munich 1 990. The alternative is very sharp, but perhaps
1 o . . lL!bd7 1 1 lL!d2 9 s 1 2 i.. g 3 h5 1 3 h4!
. the best way for White to place his pieces
This is absolutely forced, otherwise White might be 1 7 tDe4. If not Black has the fol­
has no control over the dark squares on the lowing active continuation: 17 ... :dg8 1 8 l:lf2

67
Q u e e n 's In dia n D e fe n c e
lL!xe4! (Black will have to do something
about the pin eventually, and this is a good
way - a spectacular queen sacrifice) 19 i..xe7
lLlxf2 20 �xf2 :Xg2+ 21 �e1 l::th2 22 i.. fl
l:.g1 23 e4 (White has nothing better since t'3
must be protected; after 23 l::tc 1 h4! 24 l:.c2
:Xfl+ 25 �xfl l:ht+ 26 �e2 LO+ Black
should win). Now Black should choose be­
tween three moves:
a) 23 ... fS? 24 exfS exfS 25 'ii'c 1! (address­
ing the unwelcome rook on g1) 25 ... i.. a6 26
'ii'e3 l:.xfl+ (26...l:hg2 27 'ii'e 61 and Black
cannot improve his position easily) 27 �fl
l:.h1+ 28 �£2 l::tx a1 29 'ii'e 6! and White is
winning. Black severely lacks co-ordination.
b) 23 ... d5!? seems to be the safest policy.
After 24 exd5 (24 'ii'd3 i.. a6 25 exd5 exd5 26
'ii'e3 l::tgh 1 and Black wins) 24 ... exd5 25 'ii'a4
dxc4 26 l:.c1 b5! 27 'ii'xb5 i..x t'3 28 'ii'xc4
l::tg4 29 'ii'a6+ �b8 30 i.. d3 l:tg1+ 31 i.. fl
l:tg4 a draw results.
c) 23 ... h4 24 i..xh4 l:.xh4 25 �£2 l:tg5 with a
complicated position in Barsov-Schoenthier,
Germany 1 996. I am a little sceptical about
Black's position as I cannot find a good
square for the knight, but perhaps most GMs
would see great options in ... f7-f5 and ... lLlf6.
1 7 . . .l:ldg8 1 8 Wa4
1 8 lL!e4 transposes to the previous note.
1 8 . . .*b8
1bis seems more healthy than 18 ... e5, al­
lowing 1 9 i.. fS �b8 20 e4 when White looks
good. After 20 ... lL!f8?? 21 i..xf6 he had al­
ready won(!) in Sideif Sade-Guedon, Cappelle
1 995.
1 9 l:le1 ! ?
I have most faith in closing the long di­
agonal with 1 9 e4. Then Farago-Rechlis,
Beersheba 1 987 continued 19 ... 'ii'e 8! 20 c5
dxc5 21 i.. b 5 'ii'c 8 22 e5 a6 23 exf6 axb5 24
'ii'xb5 l:tg6 25 a4 e5 26 a5 exd4 27 axb6?
lLlxb6 (now Black has a clear advantage) 28
�h2 dxc3 29 lLlfl 'ii'fS 30 g3 l:tg4 31 'ii'e2
l::tb 4 32 lL!e3 'i'e5 33 f4 'ii'e4 34 lLlc2 l:tb2 35
l:.e1 'ii'xe2 and White resigned. Instead 27 a6!
i.. d 5 (27 ... i.. a 8 28 a7+ �b7 29 lL!e4 gives
68
White a winning attack) 28 cxd4 would have
kept the struggle unclear.
1 9 . . . We8 20 Wc2 lt:lg4!
A cute tactic that allows Black to break
through to g2, albeit at a price.
21 l:lfe2
21 fxg4 l:txg4 22 l::t f4 :Xg2+ 23 �fl e5 is
inferior for White since Black's attack is even
stronger and White's forces more exposed.
21 . . . lt:lxe3! 22 l:lxe3 l:lxg2+ 23 *t 1 ?
Losing by force. It is understandable that
White did not want to go on to the long
diagonal with 23 �h1 , but at least this brings
with it the possibility of putting something
on e4. It is much harder to do something
about g1 . Anyway, Black is probably still
better here, and it is not easy to recommend
since the comer is no place to be during a
raging attack. White does avoid having to
worry about a rampaging h-pawn, though,
and has slightly better piece co-ordination
here than in the game. 23 ...l:.2g4 24 i.. £2
'i'e7 25 lL!e4 l::t4g7! 26 i.. fl h4 looks good,
with the two pawns plus good co-ordination.
23 . . . l:lh2 24 .tf2 'ile7 25 lt:le4 l:lgg2 26
: 1 e2 .txe4!
Removing a defensive piece for one not
taking part in the attack on White's second
rank. Pure logic.
27 l:l.xe4
There is nothing left to do. 27 fxe4 'i'g5
28 i..g3 (28 l:td2 llh1+ 29 �e2 :Xf2+ 30
�£2 'ii'g 1+ 31 �8 l:.h3+ 32 �f4 'ii'xe3

mate) 28 ... l:.xe2! 29 i.xe2 l:.hl+ 30 �£2 h4
and Black wins.
27 . . .iig5
... l:.h1 is now a deadly threat.
28 .l%h4
The battle is over. 28 :ld2 l:.hl+ 29 �e2
l:.xf2+ 30 �xf2 •gl+ 31 �e2 •e1 mate or
28 i.g3 •xg3 29 l:.xg2 l:.hl+ and wins.
28 . . . .1%xf2+ 29 .1%xf2 .1%xh4 30 .1%g2 iie3
31 .1%g8 + <iii b 7 32 �g2 lbf6 33 .1%g7 iih6
34 .1%g3 l::U4 35 c5 h4 36 i.. a 6+ ! <iitb 8 37
.1%g8+ lbxg8 38 'i'a4 'i'g5 + 39 �h 1 lLlf6
40 'i'c6 0-1
White resigned before Black could play
40 ...•dS.
Game 25
Piket-Tiviakov
Wyk aan Zee 1994
1 d4 lLlf6 2 c4 e6 3 lbc3 i.. b4 4 lLlf3 b6
5 i.. g 5 i.. b 7
Normally the moves 5 ...h6 6 i.h4 i.b7
are played here, but sometimes they are de­
layed. It does not really change anything, as
long as they are played eventually.
6 e3 i.. x c3+ 7 bxc3 d6
This is the move I would recommend for
Black. With sensible play from Black equality
is possible.
8 lbd2
This is the serious move here. After £2-£3
and e3-e4 White has a powerful centre.
5 ii c 2 a n d 5 i.. g 5
8 . . . lbbd7 9 f3 h6 1 0 i.. h 4 iie7 1 1 i..d 3
There is no difference between this and 1 1
e4 in practice. After 1 1 ...eS there are two
alternatives to 12 i.d3, which transposes to
the game:
a) 1 2 i.e2 lt:lf8! (this is the right way to hit
the bishop) with a further branch:
al) 13 lt:\£1 lllg6 14 i.£2 lt:\£4 1 5 ll'le3 and
now in Khalifman-Hulak, Bled, 1 991 the
natural 1 5 ... 0-0 was, for some reason, new at
the time. There followed 1 6 i.f1 ?! (planning
g2-g3 to chase away the knight, but this gives
Black a chance to generate a substantial lead
in development) 1 6 ... l:.fe8 1 7 g3 (1 7 dS!? was
probably becoming necessary) 17 ... i.xe4!!
1 8 fxe4 (1 8 gxf4 exf4 1 9 fxe4 f:xe3 20 i.g3
lllxe4 gives Black a completely winning at­
tack; the e3-pawn is worth a piece) 1 8 ... exd4!
(1 8 ... lllxe4 1 9 dS! keeps the position closed)
1 9 •xd4 (1 9 cxd4 lllxe4 20 i.g1 dS! with a
powerful attack) 1 9 ... cS 20 'i'd2 lllxe4 21
•c2 'i'f6!! (a brilliant quiet move) 22 i.g1
(22 gxf4 lllxf2 and White loses) 22...lllxc3 23
gxf4 dS! with a decisive attack .. According to
Khalifman 16 0-0, with equality, is an im­
provement. After seeing the game one tends
to agree.
a2) 1 3 •a4+ •d7 14 'iib3 does not seem
too smart. Now Black has 14 ...ll'lhS! 1 5 cS
lllg6 1 6 i.£2 dxcS 1 7 dxcS ll'lhf4 1 8 i.f1 0-0
and White was on the way to trouble in
Campos Moreno-Polugaevsky, Oviedo 1 991.
a3) 13 i.£2!? seems more to the point, but
69
Q u e e n 's In dian D e fe n c e
after 1 3. . .�g6 14 g3 0-0 1 5 � fl c6!, with the
idea of ... d6-d5, Black is able to use his lead
in development constructively. In Dokhoian­
Djuric, Philidelphia 1 989, 1 6 d5 l:.ac8 1 7
1Wd2 b5! 1 8 dxc6 i..xc6 1 9 �e3! bxc4 2 0 0-0
1Wd7! 21 i..xc4 i.. b 5! saw Black completely
equalise.
Note that here 1 2 ... g5?! 1 3 i.. f'2 �h5 1 4
g3! seems to guamtee White an opening ad­
vantage, e.g. 1 4... �g7 1 5 0-0 0-0-0 (1 5 ... h5 1 6
c 5 dxc5 1 7 i.. b 5 0-0-0 1 8 1Wa4 �b8 1 9 �b3
with an attack - Epishin) 16 l:.e1 c5 1 7 a4 a5
1 8 llb1 l:.hf8 (1 8 ... £5 19 exf5 �xf5 20 i.. d3
l:.hf8 21 i.. e4 is also better for White - Ep­
ishin) 1 9 �fl I and White had a structural
advantage in Epishin-Shneider, USSR 1 990.
When White's knight comes to d5 Black
cannot really take it but, unfortunately, nor
can he leave it!
b) 1 2 'ir'a4 is not really dangerous. The
only thing White has in this position is the
pin from h4-e7, and Black can unravel with­
out allowing �d2-f1 -e3-f5, which at times is
very annoying. Flear-Salov, Szirak 1 987 con­
tinued 1 2 ... 0-0 13 i.. e2 1We8! 14 l:tbl �h5
with equality. After 15 0-0 �f4 1 6 l:.fe1 £5 1 7
i.. fl �h8 Black's position looks like it is
more pleasant to play...
1 1 e5 1 2 e4 g5( ? ! )
•••
I a m a little suspicious about thi s move.
The hole on £5 is big enough to consume all
the pieces I have blundered during the years.
The alternatives are:
a) 12 ... 0-0-0!? is quite a normal move that
has been suggested in numerous places but
never tried.
b) 12 ... �£8! begins a logical manoeuvre.
Piket-Ljubojevic, Monaco 1 994 continued 1 3
� fl �g6 1 4 i.. f'2 (I think thi s i s the best
move because it keeps the option of g2-g3
open; in van der Wiel-Ljubojevic, Amster­
dam 1 986 Black was never really worse: 1 4
i..g3 �h5 1 5 �e3 �gf4 1 6 'ir'a4+ � £8 17
l:.dl �xg3 1 8 hxg3 �e6 1 9 i.. e2 g6 20 �f'2
q;g7 21 1Wa3 1Wg5 22 1Wc1 l:.ad8 23 1td2 h5
24 �d5 1txd2 25 :Xd2 f6 26 �e3 �fl 27
70
i.. d l c5) 14 ...�f4 15 �e3 �xd3+!? (directed
against i.. c2 and g2-g3) 1 6 'i'xd3 �h5! (not
1 6 ... 1td7 1 7 :d1 1Wa4 1 8 0-0 0-0 1 9 i..g3
l:.fe8 20 l:.f'2 �h5 21 l:.b2, when White had
an enduring advantage in Sokolov­
Govedarica, Belgrade 1 987; 1 6 ...g6? runs into
1 7 i..h4Q 1 7 0-0-0 g6 1 8 1tc2 0-0-0 1 9 l:.d2
f6 and the position is unclear, with chances
for both sides.
1 3 .i..f 2 lLih5 1 4 lL!f 1 exd4?
Handing White a large advantage, but
nothing leads to equality.
a) 1 4... £5 1 5 �e3 f4 1 6 �£5 'ir'f6 1 7 g4 has
been assessed as unclear, but I find that silly.
With the knight on £5 and a strong structure
in the centre White appears to be better.
b) 14 ... �f4 does not work here. After 1 5
�e3 g4 1 6 0-0! White already has an enor­
mous advantage: 1 6 ...gxf3 1 7 'i'xf3! exd4
(1 7 ... �xd3 1 8 ti:Jf5! 'i'g5 1 9 i..h4 leads to a
substantial plus due to 1 9 ...1tf4?! 20 �g7+
�£8 21 �e6+! fxe6 22 1Wxd3, when White
wins) 1 8 �d5! �xd5 (1 8 ... i..xd5 1 9 'i'xf4
�c5 20 exd5 �xd3 21 1txd4 �e5 22 i..h4
also wins for White) . 1 9 exd5 and Black did
not make it to move 30 in Khalifman­
yemelin, Russia 1 996. The dark squares are
simply too weak.
1 5 cxd4 f5 1 6 lL!e3! fxe4
1 7 lLif5!
An important trick to remember.
1 7 Wf7 1 8 fxe4
• • .
White has a very strong position. Now

'ii'xh5 is a threat.
1 8 . . . ltlf4 1 9 0-0 ltlxg2!?
Black is trying to create counterplay as the
'normal' 1 9 ... lLixd3 20 'ii'xd3 0-0-0 21 �g3
'ii'h7 looks very dangerous for Black. Here
White can try 22 �xd6!? cxd6 23 lLixd6+
'it>b8 24 'ii'g3 'it>a8 25 :n 'ii'g6 26 :an with
a strong attack against the king (Piket).
20 j_g3 ltlf4 2 1 j.xf4 gxf4 22 :xf4
'ilg6+ 23 :g4 'ilf6 24 ltlg7 + �dB 25
�h 1 ?
'This is unnecessary. Preferable is the logi­
cal 25 �c2! with an advantage after 25 ... hS
26 l:tg2 l:tg8 27 lLixh5 ::r.xg2+ 28 �xg2 'ii'g5+
29 lLig3, when Black has nothing concrete
for the pawn.
25 . . . �cS 26 j.c2 j.a6?
A terrible mistake, simply overlooking the
reply. Better are 26 ... c5! and 26 ... l:tfB!?, both
with counterplay.
27 j.a4!
Eyeing c6. 'This is much more important
than the c-pawn.
27 . . . j.b7 28 'ild3 ltlc5
29 :f 1 ! ltlxe4
29 ...'ii'xg7 30 'ii'h3 !.
30 :xf6 lL'lf2+ 3 1 �g 1 ltlxd3 32 ltle6
ltlb2?
Losing immediately, but after 32 ... 'it>b8 33
l:tg7 l:tc8 34 ::r.ff7 lLib2 35 �b3 a5 36 lLixc7
White also has a winning position.
33 :xh6!
No squares left for the pilgrim!
5 'il c 2 a n d 5 j_ g 5
3 3 . . . ltlxa4 34 :xh8+ �d7 35 l:.xa8
j.xa8 36 d5 1 -0
Game 26
Miles-Timm an
Ti/burg 1986
1 d4 ltlf6 2 c4 e6 3 ltlc3 i.b4 4 ltlf3 b6
5 .tg5 j.b7 6 e3 h6 7 j.h4 g5! ?
'This used to be the main line, championed
by Timman, but these days it is seen less and
less.
8 .tg3 ltle4 9 'ilc2!
9 lLid2 is dealt with in the following game.
9 . . . j.xc3+ 1 0 bxc3 d6 1 1 j.d3 f5
1 1 ...lLixg3 1 2 fxg3 leads to Game 23 by
transposition. 'This variation is a good option
for Black.
1 2 d5 ltlc5
'This is the old move, and probably good
enough for some kind of equality, but the
new 12 ... lLid7!? is certainly more logical.
Then 13 �xe4?1 gives away all the light
squares without any concessions. After
1 3 ... fxe4 1 4 'ii'xe4 'i'f6 1 5 0-0 0-0-0 Black
seems to have a good position despite the
pawn qeficit. 1 6 'i'xe6 1i'xe6 1 7 dxe6 lLic5 1 8
lLid4 l:tde8 saw Black enjoy a good ending in
a few games, among them Hort-Bellon,
Hastings 1 975. The score so far is two wins
for Black and one draw. Instead White
should play 1 3 lLid4 lLidc5 1 4 dxe6 and now:
a) 14 ... l:tfB 1 5 �e2 g4! 1 6 f3 lLixg3 1 7
71
Q u e e n 's In dia n D e fe n c e
hxg3 'i'g5 1 B 0�0-0! 0-0-0 1 9 i.d3 gxf.3 20
gxf.3 c!l)xe6 21 i.xfS l:txfS 22 'i'xfS 'i'xfS 23
c!l)xfS i.xf.3 24 l:txh6 i.xd1 25 �xd1 c!l)g5 26
g4 l:tfB?! (26 ... 'ili>b71 with compensation was
better according to Tisdall, but it still feels as
if Black is fighting for equality here) 27 'ili>e2
with an advantage to White in Zvjaginsev­
Hiibner, Elista 1 99B.
b) 1 4...'i'f6 appears to be the most natural
move. After 1 5 f.3 f4! 1 6 i.xe41 (forced - 1 6
exf4? c!l)xd3+ 1 7 'i'xd3 c!l)c5 1 B 'ii'e2 gxf4
was very dangerous for White in Komarov­
Mantovani, Reggio Emilia 1 996) 1 6 ... i.xe4
1 7 fxe4 fxg3 1 B hxg3 0-0-0 the situation is
unclear, although I feel a bit more comfort­
able about White's position. The pawns
might look ugly, but they control many inva­
sion squares.
1 3 h4 g4 1 4 ltld4 'i'f6 1 5 0-0 ltlba6
This is the only move these days since
1 5 ... c!l)xd3 16 'ii'xd3 e5 17 c!l)xfS i.cB
was met with 1B f4!! in Miles-Beliavsky,
Tilburg 1 9B6, practically winning on the spot.
There followed 1B ... 'ii'xf5 19 e4 'ii'h5 20 fxe5
dxe5 21 c5 �dB 22 d6 'i'eB 23 dxc7+ 'ili>xc7
24 'i'd5 ll:ic6 25 l:lt7+ i.d7 26 :taf1 l:ldB 27
: 1 f6 �cB 2B cxb6 axb6 29 'ii'h 5 1 -0.
1 6 ltlxe6! ?
Thi s sacrifice i s the only way fo r White to
prevent his centre from collapsing. After 1 6
dxe6 Black's bishop would be a truly power­
ful piece and Black would rock n' roll on the
light squares.
72
1 6 . . . ltlxe6 1 7 .txf5 ltlg7 1 8 .i.g6 + 'it>d7
1 9 f3 l:af8!
Preparing a possible getaway to the
queenside for the king.
20 fxg4 'i'e7 2 1 e4 'iPc8
Also possible is 21...c!l)c51? 22 'ii'e2 'ili>cB
23 e5 i.a6 24 e6 c6 25 :Xm+ :XfB 26 l:tf1
l:txfl+ 27 'i'xf1 'ili>dB 2B 'iWf4 i.b7 29 'ii'xh6
cxd5 30 i.xd6 1i'xd6 3 1 'ii'h s+ �c7 32
'ii'xg7+ �dB 33 'i'hs+, with a perpetual in
Ivanchuk-Anand, Monaco 1 993.
22 'ifd 2
2 2 . . . ltlc5?!
This game was a sad episode for Black
and, later, play was improved with 22... 'ili>bB!,
e.g. 23 l:txfB+ :txfB 24 'ii'xh6 i.cB (this is the
idea behind ... 'ili>bB - the bishop is badly
needed on the kingside to stop the pawns) 25
l:te1 i.xg4 26 c5 'ii' f6 27 cxd6 i.h5 2B e5
'i'xg6 29 'i'xg6 i.xg6 30 e6 c!l)c5 31 d7
c!l)xd7 32 exd7 l:ldB 33 l:le6 i.h5 34 i.e5
:Xd7 35 l:lh6 i.t7 36 i.xg7 i.xd5 37 i.e5
i.xa2 V2-V2, Kasparov-Timman, Hilversum
1 9B5, or 23 'ii'd4 c!l)eB 24 l:lt7 :Xt7 25 'i'xhB
:m1 26 'ii'xh6 ll:ic5 27 h5 c!l)d7 2B i.£5 i.cB
29 l:lft ll:ie5, which left Black okay in Salov­
Timman, Saint John 1 9BB.
23 l:xf8 + l:xf8 24 'i'xh6 'i'f6 25 .tf5 +
ltlxf5 26 'ifxf6 l:xf6 27 exf5 .ta& 28 .tf2
.txc4 29 .id4 l:f7 30 f6 l:h7 3 1 .txc5
dxc5 32 l:d 1 !
Keeping the bishop out of the game.
32 . . . .tb5
 5 W c 2 a n d 5 .i. g 5

32 ...:xh4 33 f7 l:h8 34 g5 and White Take the money and run.

wins. 1 0 bxc3 .bc3 1 1 :c 1 i.a5
33 g5 .i.e8 34 :e 1 ! This appears to be the safest path here. It
And now the king. is doubtful that White has gained much for
34 . . . �d7 35 g3 b5 36 �g 2 a5 37 �h3 his pawn. But other moves have also brought
b4 38 cxb4 axb4 39 g4 :ha 40 h5 @d6 good results.
41 J:l.xe8! 1 -0 a) 1 t . ...i.xd2+ 12 'ii'xd2 d6 13 c5 (13 £3? is
too slow: 1 3 ... lbd7 14 .i.d3 'ii'e7 1 5 0-0 f5 1 6
c5 dxc5 17 i..xc7 0-0 1 8 dxcS lbxcS 1 9 .i.e5
l::t ad8 and White resigned in Priehoda­
Rozentalis, Tmava 1 988) 1 3 ... dxcS 1 4 'irh21?
(harmless is 14 dxc5 'ii'xd2+ 15 'iii>xd2 i.. a6
1 6 .i.xc7 .i.xfl 1 7 l::thx ft lba6 1 8 .i.d6 lbxc5
1 9 .i.xc5 bxc5 20 l::txc5 l::td8+ 21 'iii>e2 and a
draw was agreed in Smirin-Rozentalis, New
York 1 997) 14 ... 0-0 15 h4 g4 1 6 dxc5 'ii'dS 1 7
.i.e2 and now Black wa s too careful with
17 ... lbc6? in Stocek-Cvek, Plzen 1 997, after
which the positional aspects of the position
begin to count over material. Correct was
17 ...'ii'xg21 1 8 l::th2 'ii'd5 to ruin White's set­
Game 27 up. After 1 9 'ii' f6 lbd7 20 'ii'xh6 lbxc5 21
Crouch-Harikrishna 'ii'gS+ 'i'xg5 22 hxgS we have a big mess.
undon 200 1 b) 1 t ....i.b4 1 2 h4 and now the piece sacri­
fice 1 2 ...lbc6!? should be the reason to play
1 d4 e6 2 c4 ll:lf6 3 ll:lf3 b6 4 ll:lc3 i.b 7 this system. 12 ...gxh4? is still the main line
5 i.g5 h6 6 i.h4 i.b4 7 e3 g5 8 i.g3 given in ECO, but after 13 l::txh4 .i.d6 1 4
ll:le4 9 itid2?! 'ii'g4 'ile7 1 5 i..xd6 cxd6 1 6 'i'g3! they sug­
This is a gambit that might have been gest to improve on 16 ... £5 with 1 6 ... d51?.
playable back in the days when people were About 10 years ago I told Steffen Pedersen
afraid of taking the pawns and defending an the refutation and he put it in his book 1 d4!:
awkward position. But these days - no. 1 7 cxd5 exd5 (what else? 17 ... lba6 is just
9 . . . ll:lxc3! plain bad) 1 8 :c 7 i..a6 1 9 :xa7 and White
wins. But even stronger is 1 9 .i.c4!!, winning
everything after 1 9 ... dxc4 (1 9 ... 'i'a3 20 .i.xa6
'ii'xa6 21 'ile5+ �d8 22 l::tc21 and White wins
material or mates) 20 l::txa7! :xa7 21 'ilxb8+
with a mate in 8 according to Junior. After
1 6 ... £5 White has 1 7 c5 d5 1 8 1i'e5! with a
strong., initiative, as in Agdestein-Hellers,
Gausdal 1 987. Returning to 13 d5, Schussler­
Amason, Reykjavik 1 986 went 1 3 ... 1i'e7! 14
dxc6 dxc6 1 5 .i.eS 0-0-0 1 6 .i.c3 (1 6 l::tc2
l::txd2 1 7 l::txd2 l::t d8 and Black wins)
1 6 ... .i.xc3 1 7 :Xc3 c5 1 8 hxg5 hxg5 1 9 l::txh8
:xh8 20 e4 f5 21 l::th3 :d8 22 :d3 :us 23

73
Q u e e n 's In dia n D e fe n c e
'ir'c2 'ikfl and Black actually won this posi­
tion. Here I will just say that White has no
easy way to co-ordinate his pieces, and
Black's bishop does look terribly strong. Still,
my choice would be 1 1 . ..aas.
. .
1 2 h4 .:lg8 1 3 'i'c2 'iie 7!
Black lives with the knowledge that c7
falls. So what? White is still trapped in a ter­
rible pin and g2 is still exposed. All in all it
looks as if White does not have any chances
to gain an advantage in this line, as he does
not have the fluent development you usually
get when you sacrifice a pawn.
1 4 hxg5 hxg5 1 5 i.xc7 ltia6 1 6 i.e5
74
'ifa3 1 7 .:Id 1 ltib4 1 8 'i'b3 'i'xa2 1 9
Wxa2 ltixa2 20 f3 i.b4 2 1 .:la1 ltic3 22
'it>f2 a5 23 i.f6 g4 24 i.d3 g3+ 25 'it>e 1
b5 26 i.h7 :ta 27 i.g7 bxc4 28 i.xf8
'it>xf8 29 �xc4 f5 30 i.g6 'it>g7 31 ltie5
ltie4+ 32 'it>f1 i.a6+ 33 'it>g 1 lDg5 34 f4
34 . . . i.d2!
The winning move. Once the e3-square
caves in the game is up.
35 .:lh4 i.xe3+ 36 'it>h 1 i.xd4 37 J:[d 1
i.xe5 38 i.h5 i.xf4 39 :xd7 + 'iitf6 40
:xt4 :ha 41 .:lh4 i.e2 42 :a1 :xh5 43
:xh5 i.xh5 44 :xa5 f4 45 'it>g 1 i.e2 46
:as f3 4 7 gxf3 i.xf3 0-1
 5 'il c 2 a n d 5 i.. g 5

Summary
5 'iic2 i s not ambitious and Black equalises without trouble. 4 lLic3 i.b4 5 i.gS i.b7 6 e3 h 6 7
i.h4 i.xc3 8 bxc3 d6 is a good reliable system for Black. Now 9 lLid2 is the only good move,
but 9 i.d3, which has led to disaster for many White players, is still seen occasionally. 7 ...gS 8
i.g3 lLie4 is also seen at the top level, and after 9 'iicz the game seems to be more or less
equal, even though this is a complex line. The gambit after 9 lLidZ is simply not to be recom­
mended.

1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 lll c 3 (D) i.b4

4 ... i.b7
5 'iic 2 - Game 21; 5 i.gS Game 22
-

5 i.g5
5 'iib 3 - Chapter 4
5 . . . .ltb7 6 e3
6 'iic 2 Game 23
-

6 . . . h6 7 i.h4 i.xc3 +
. 7 . . gS 8 i.g3 lLie4 (D)
9 'iic2 - Game 26; 9 lLidZ Game 27
-

8 bxc3 d6 (D)
9 i.d3 - Game 24; 9 lLidZ Game 25
-

4 t'i:Jc3 8 . . . t'i:Je4 8 . . . d6

75
 CHAPTER FOUR
5 'ifb3
1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 lllc 3 i.b4
5 'i'b3
In this chapter we will investigate Seira­
wan's pet line which, in my opinion, is one of
the most difficult for Black to face. Look out
particularly for Alterman's nice strategic idea
in the first main game, and that of Miles in
one of the sub-variations.
Game 28
Alterman-Liss
Israel 1999
1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 lll c 3 i.b4
.
5 'i'b3!
I give this move an exclamation mark not
because it is better than 5 e3, leading to the
76
I·
Nimzo-Indian, or 5 .tg5, but because this is
what I would play myself with White in this
position, as it is clearly the most problematic
line for Black to face.
5 . . . a5
Tbis is one of three ways to protect the
bishop. 5 .. .'ii'e 7 is considered in the following
game and 5 ... c5 in Game 30.
6 i.g5
The main line. Black has no problems af­
ter the alternatives:
a) In reply to 6 g3, 6 ... .tb7 7 .i.g2 0-0 8
0-0 .txc3 9 1i'xc3 d6 1 0 b3 c!Dbd7 1 1 .tb2
leads us to a standard position where White
seems to have a tiny edge thanks to the po­
tential pressure on the long diagonal, as in
Dreev-Kiselev, Podolsk 1992. Instead
6 ... c!Dc6! is the aggressive approach, e.g. 7
.i.g2 a4! (now this is possible) 8 'ii'c2 .i.a6 9
a3 (9 0-0 .i.xc4 10 c!De5 c!Dxe5 1 1 dxe5 c!Dd5
12 lDxa4 b5 13 c!Dc3 .i.xc3 1 4 bxc3 l:r.a4 was
fine for Black in Labunsky-Tsesarsky, USSR
1986) 9 ....txc3+ 10 'ii'xc3 lDa5 1 1 lDe5?! (1 1
lDd2 d5 1 2 cxd5 lDxd5 1 3 'ii'c2 is unclear
according to Giplis) 1 1 ...d5! 1 2 cxd5 c!Dxd5
and it is already clear that Black has consider­
able power in his pieces. That he used it to
win a superb game in Polugaevsky­
Ljubojevic, Linares 1 985, is no surprise: 13
'ii'c2 c!Db3 14 'ii'c6+ �f8 1 5 .i.g5 'ii'e 8 1 6

lDd7+ �g8 1 7 e4 lDxd4 1 8 'ii'xa4 cS 1 9 exdS
i.bS 20 'ii'd 1 W'xd7 21 i.e3 i.e2 22 dxe6
'i'xe6 23 'iixe2 lDxe2 24 i.xa8 °iic4 25 :td1
lDd4 26 i.xd4 cxd4 27 i.e4 g6 28 i.d3 Waz
29 .l:.b1 W'b3 30 �d2 �g7 and Black went
on to win.
b) 6 a3 i.xc3+ (6 ... a4 7 'ii'c2 i.xc3+ 8
bxc3! reminds me of the main game. 8 bxc3
has never been played before) 7 W'xc3 i.b7 8
g3 0-0 9 i.g2 a4! and Black had at least
equalised in Ahner-Bischoff, Germany 1 997
(c4 is a potential problem for White and the
queen's bishop has no good squares).
6 .tb7 7 e3
. . .
White is developing. 7 a3 i.xc3+ 8 "iixc3
d6 9 e3 lDbd7 10 lDd2 'iie7 1 1 £3 eS 12 i.e2
exd4 13 W'xd4 a4 was good enough for
equality in Zaja-Baklan, New Delhi 2000.
7 h6
. . .
There is only one really independent way
to play this position, namely 7 ...'iie7 8 "iic2
a4!? 9 a3 i.xc3+, and now 10 "iixc3 lDe4! 1 1
i.xe7 lDxc3 1 2 i.h4 lDe4 seems to give
Black equality due to the weakness of c4.
Instead 10 bxc3! is best, the idea being that
Black will have some commitment to the a4-
pawn and that White will benefit from a
strong centre, as is the case in the 5 i.gS
lines. Miles-Arkhipov, Miinster 1 993 contin­
ued 10 ... i.e4 (it is not easy to decide where
this bishop bclongs, as White will certainly
play t2-8 and e3-e4 soon) 1 1 'iib z h6 1 2
i.h4 0-0 1 3 lDd2 i.h7 1 4 £3 e S 1 5 e4 lDc6
16 i.e2 (planning to re-route the knight to
dS) 16 ... d6 17 tDn gS 18 i.t2 lDhS 19 lDe3
lLJf4 20 lDdS! W'd8 21 g4! and White had a
very promising position.
8 .th4
The only serious move. After 8 i.xf6
'iixf6 9 i.e2 0-0 1 0 0-0 i.xc3 1 1 "iixc3 d6 1 2
lDd2 lDd7 the position was even in Ivanov­
Farago, Philadelphia 1 984.
s d6
...
Black should develop smoothly. After
8 ...gS 9 i.g3 lDe4 1 0 i.d3 d6 1 1 i.xe4!
i.xe4 1 2 0-0-0 i.xc3 (1 2 ... i.b7 1 3 dS! gives
5 'il b 3
White the advantage due to his influence in
the centre) 1 3 W'xc3 lDd7 14 lDe1 i.b7 1 5 cS
White seemed to have some initiative in
Miles-Budnikov, Beijing 1991, although
Black might be fine.
9 a3! ?
Thi s builds on a very interesting concept,
similar to the one Miles used in the note to
Black's 7th move. Others:
a) 9 i.e2?! - it is not easy to see what the
bishop is doing here. Black is doing well after
9 ... gS 10 i.g3 lDe4, e.g. 1 1 0-0-0 i.xc3 1 2
bxc3 lDd7 1 3 lDe1 h S 14 h 4 g4 1 5 lDd3
lDxg3 16 fxg3 i.xg2 17 .l:.h2 i.e4, and White
did not have enough compensation for the
pawn in Conquest-Onischuk, Germany 1 997,
or 1 1 d5 lDd7 12 lDd4 lDdc5 13 'iic2 i.xc3+
14 bxc3 eS, and Black was satisfied in Stohl­
Polak, Olomouc 1 998.
b) 9 i.d3 is the main line here. After
9 ... lDbd7 1 0 0-0-0 i.xc3 1 1 "iixc3 Black has
a variety of valid options:
b1) 1 1 ...a4!? 12 'iic2 l:taS 13 lDd2 gS 1 4
i.g3 i.xg2 1 5 .l:.hg1 i.b7 1 6 £3 "iia8 1 7
:tdf1 rJi; e7 1 8 i. e1 c S 1 9 lDe4 .l:. a7 2 0 lDc3
i.a6 21 f4, Miles-Bischoff, Havana 1 998, and
White; had compensation for the pawn but it
was still a battle.
b2) 1 1 .. .W'c8?! 1 2 'iic2! cS 1 3 dS eS 1 4
lDd2 gS 1 5 i.g3 'iic7 1 6 f4 and Black experi­
enced serious problems on the light squares
in Khalifman-Tunik, Maikop 1 998.
b3) 1 1 ...'ii'e7 1 2 lDe1 eS?! 1 3 i.5 0-0-0 14
77
Q u e e n 's In dia n D e fe n c e

c5! put Black under fire in Miles-Singh, Cal­ with the a4-pawn.
cutta 1 994. After 14 ... exd4 1 5 cxd6 'iixd6 1 6 1 6 0-0 ltih5 1 7 :te 1 lLlf4 1 8 .i.f 1 'ilg6
:Xd4 lDd5 1 7 'iic2 c5 1 8 lDd3 .tc6 1 9 .i.g3 19 :ad 1 f5 20 exf5 :xf5 21 .i.g3!
'iif6 20 lDe5 cxd4 21 lDxc6 he was ready to
resign.
b4) 1 1 ...g5 12 .i.g3 lDe4 1 3 1i'c2 lDdf6 1 4
l:lhe1 'ile7 with an even position in Khalif­
man-Granda Zuniga, Ulcinj 1 998.
9 . . . a4
9 ... J..xc3+!, with the idea of 1 0 'ilxc3 a4!
with equality, is probably a better line. Now
White gets something.
1 0 'i'c2 .bc3+ 1 1 bxc3!
I would like to quote Alterman on this po­
sition because he says it all: '1bis was my
idea. It looks as if Black gets a tempo with
... a5-a4 but, as I understand it, the pawn on All White's pieces are well organised here
a4 only presents problems for Black because and Black has managed to do nothing other
it requires constant protection.' than open up the position for them. White is
1 1 . . . ltibd7 1 2 lLld2 'ile7 1 3 f3 e5 1 4 e4 clearly better.
0-0 1 5 .i.e2 21 . . . 'ilf7 22 ltie4! <iilih 8 23 ltif2 :tS!?
Leaving the pawn to its own devices
seems a rational thing to do. The pieces need
to play before everything goes wrong.
24 'ilxa4 ltif6 25 . .i.xf4 :xf4 26 dxe5
dxe5 27 ltid3 :t5 28 :xe5 ltih5
Here Black has only one minute left with
which to reach the 40th move, but White
completely overlooks his threat, and the
game takes a sudden tum.
29 l:l.de 1 ? ? :xf3!

Here Black should find a way to organise

his pieces so that he gets out of the pin and
activates his forces. I have a feeling that Liss
does not find the appropriate route.
1 5 . . . 'i'e6?!
1 5 ...'iie 8! improves, with indirect protec­
tion of a4 and the manoeuvre ...lDh5-f4 com­
ing up, perhaps even ... c7-c6 and ... d6-d5 in
some situations. Liss leaves the e8-square
free for a rook, which seems sensible
enough, but the queen appears to be slightly 1bis rook is obviously taboo.
awkward on e6 and he still has problems 30 'ild 1 'i'g6 31 'ild2? !

78

31 'i'e2! is better according to Alterman.
31 . . . lLlf4 32 lLlxf4 l:3xf4 33 g3 l:4f6 34
i.d3 flf7 35 .:e7 flh 5 36 fle2 flc5+ 37
fle3 .:t2 38 i.f1 ?
Presenting Black with the opportunity to
finish the game with a nice tactical blow.
After 38 'i'xcS .l:.g2+ 39 @hl bxcS 40 i.e4
i.xe4 41 l:r.1xe4 l:r.ff2 42 .l:.h4 .l:.a2 43 a4
:tgc2 44 .:e1 .l:.xc3 Black is close to winning,
but White retains some drawing chances.
38 . . . flc6!
0-1
Now the control ofhl is deadly.
Game 29
Timman-Karpov
Hoogeveen 1999
1 d4 lLlf6 2 c4 e6 3 lLlf3 b6 4 lbc3 i.b4
5 'ilb3 fle7
5 'il b 3
I like this move quite a lot. Th e next note
demonstrates the main reason why (... �e4) .
6 a3!
The most challenging set-up. At least it
earns the bishop pair. Others:
a) After 6 i.gS i.b7 7 e3 aS 8 a3 i.xc3+ 9
'i'xc3 �e4! Black steers the game to an end­
ing where c4 is a slight weakness. Seirewan­
Sokolov, Candidates 1 985, went 1 0 i.xe7
�xc3 1 1 i.h4 �e4 12 �d2 �xd2 13 @xd2
a4! 14 £3 �c6 15 i.d3 �aS 16 @c3 i.a6 17
e4 0-0 1 8 i.c2 �xc4 1 9 i.xa4 c6 20 i.b3 dS
21 i.xc4 i.xc4 22 b3 with complete equality.
b) 6 g3 i.b7 7 i.g2 cS 8 a3 i.xc3+ 9
'i'xc3 cxd4 10 'i'xd4 �c6 1 1 'i'h4 �as 12
i.gS h6 1 3 0-0 0-0 14 i.xf6 'i'xf6 1 5 'i'xf6
gxf6 1 6 �d2 J:tfc8 and Black was completely
fine in Dreev-K.arpov, Dortmund 1 994.
However, I think White might be slightly
better after 7 ... 0-0 8 0-0 i.xc3 9 'i'xc3 d6 10
b3 �bd7 1 1 i.b2 i.e4 12 l:.fel . At least
Black was in trouble after 12 ... cS 1 3 l:r.adl
l:r.fd8 14 dS! exdS 15 cxdS bS 1 6 �h4! i.xg2
17 �xg2 b4 1 8 'i'£3 g6 1 9 e4 in Van Wely­
Psakhis, Internet 1 9'95, the bishop making its
presence felt.
c) 6 i.f4. The bishop is nice here, but not
very useful. 6 ... dS 7 e3 0-0 8 a3 i.xc3+ 9
'i'xc3 i.a6! 1 0 i.gS l:r.c8 1 1 cxdS i.xfl 1 2
.l:.xfl exdS 1 3 .l:.cl 'i'e6 left Black n o worse
in Alterman-Hracek, Bad Homburg 1 997.
6 . . . i.xc3 + 7 flxc3 i.b7 8 e3
Here I prefer 8 g3 0-0 9 i.g2 as it puts
less restraint on the queen's bishop. Now
9 ... d6 10 0-0 �bd7 1 1 b4 is perhaps a touch
better for White, but this is difficult to prove,
e.g. 1 1 ...aS 12 i.b2 axb4 1 3 axb4 l:tfc8 14
.l:.fdl , Gurevich-K.arpov, Reggio Emilia
1991, when 14...'i'fB!, with the intention of
... .l:.xal ,and ... l:r.a8, puts Black as close to
equality as he can possibly come in positions
like this (without a mistake from White).
Indeed he should not fear such positions.
Gurevich calls it equal, but this is because
White's advantage is so minor that, at his
level, it is insignificant. 1 1 ...�e4 12 'i'c2 f5 is
79
Q u e e n 's In dia n D e fe n c e
less good fo r Black. Smyslov-Chandler, Hast­
ings 1988, went 13 i..b2 liJdf6 14 a4 a5 15 b5
l:tae8 16 ll\e 1 c6 17 bxc6 i..xc6 18 £3 lbg5 1 9
lbd3, and White was better. Look a t all the
weak pawns on the dark squares!
8 i..g5 lbe4! sees Black equalise immedi­
ately.
8 0-0
. . .
8 ... d6 9 i.. e2 lbe4 10 "iic2 lbd7 1 1 0-0 0-0
12 liJd2 f5 13 £3 ltJg5 14 .l:f.£2 e5 1 5 liJfl
exd4 16 exd4 lbe6 gave Black sufficient
counterplay in Seirawan-Karpov, Roque­
brune 1 992, but there is still a lot of play in
the position for White, and the bishop might
later become strong on b2, so I would not
recommend playing like this.
9 i.e2 d6
I think this is the most natural way to play
the position, but also possible is 9 ... c5!? 1 0
0-0 .l:f.c8 1 1 dxc5 "iixc5 (1 1 ...bxc5 1 2 b 3 fol­
lowed by 1 3 i..b2 is slightly better for White)
1 2 b4 "iie7 13 i.. b2 d5 14 °iid4 dxc4 1 5
i.. xc4 a5! (1 5. . .lbc6?! 1 6 °iih4 and Black
needs to be careful) 16 lbe5 lbc6 1 7 lbxc6
.l:f.xc6 1 8 bxa5 (1 8 b5 .l:f.d6 19 °iih4 e5 20 a4
.l:f.d2 21 i.. a3 °iid8 22 'i*'g3 i.. d 5 is no worse
for Black at all) 1 8 ... e5! 1 9 °iih4 l:txa5?! was
Sokolov-Karpov, Groningen 1 995. Better is
1 9 ... bxa5, when Black is not worse in any
way.
1 0 0-0 ltJbd7 1 1 b4 c5!
Black should seek counterplay, and the
only way to do so is to put pressure on a
80
weakness on the light squares, namely the c4-
pawn. 1 1 ...ll\e4 12 'i*'d3 f5 tries to put pres­
sure on g2, but Black is not in position to
justify this: 1 3 liJd2 'it'h4 1 4 £3 ll\g5 1 5 d5!
gave White a slight advantage in Van Wely­
Lobron, London 1 995. The point is, of
course, 1 5 ... e5 1 6 f4!, with a fight for a giant
bishop on b2.
1 2 i.b2 a5
Here Black has a great range of alterna­
tives.
a) 12 ... l:tac8 13 dxc5 dxc5 14 b5 .l:f.fd8 1 5
a4, Becker-Pitschak, Bad Liebenwerda 1 934.
This kind of position is typical for the line.
White has a very minor plus due to the
bishop's superiority over the knight.
b) 1 2... ll\e4 1 3 'i*'c2 and now 1 3 ... fS?I does
not sit well with ... c7-c5. Now White's dark­
squared bishop assumes great influence that
could otherwise be neutralised later with ... f7-
f6 if necessary. Moreover Black's second
rank is slightly more exposed, and this is felt
on g7. In Shamkovich-Kholmov, Baku 1 972,
14 dxc5 bxc5 15 b5 liJdf6 16 liJd2 l:tad8 1 7
£3 lbg5 1 8 .l:f.fdl gave White the advantage .
Better is 1 3 . .. .l:f.ac8, e.g. 1 4 dxc5 dxc5 15 b5
h6 16 .l:f.fd1 .l:f.fd8 17 a4 lbg5 1 8 'i*'c3 lbx£3+
19 i..x £3 lbf6 20 i..xb7 'i*'xb7 with an edge
for White in Cebalo-Hohler, Saint Vincent
1 998. Often Black should not fear such posi­
tions as White's bishop tends to have diffi­
culties causing any harm all by itself.
c) 12 ... .l:f.fd8 13 .l:f.fdl l:tac8 14 'i*'h3 i.. e4
1 5 ll\d2 i.. a8 1 6 dxc5 dxc5 1 7 b5 lbe4 1 8
lbxe4 i..xe4 1 9 a4 f6 2 0 a 5 ltJ£8 2 1 £3 i..g6
and, because Black's bishop looks so good
and the al -h8 diagonal is limited, Black had
equalised in Gurevich-Rozentalis, Belfort
1 997.
d) 12 ... .l:f.fe8 13 lbd2 l:tac8 14 dxc5 dxc5
1 5 b5 e5 1 6 £3 ltJ£8 17 a4 was slightly better
for White in Burmakin-Zayac, St. Petersburg
1 998. White's set-up here is probably the
most promising.
1 3 :td 1
1 3 dxc5 dxc5 14 b5 .l:f.fe8 1 5 .l:f.fdl l:tad8

1 6 tlld2 e5 equalises. Perhaps 1 6 tlle 5!?
Might offer some kind of an advantage.
1 3 . . . axb4 1 4 axb4 l:!.fb8!
Clearing space for the queen in order to
trade major pieces on the a-file.
1 5 tt:ld2 'iid 8 1 6 f3 l:l.xa 1 1 7 ..txa 1 l:!.a8
1 8 ..td3 'ii c 7 1 9 ..tc2 l:!.c8? !
A slight positional mistake. After 1 9 ... e5!
Black has neutralised the pressure on the
long diagonal and thus limited the scope of
the bishop. Now White has a chance to es­
tablish a small edge.
20 dxc5 dxc5 21 b5!
The pawn structure is now clearly favour­
able for White. His bishops enjoy open di­
agonals, but Black's knights have no good
squares.
21 . . . tt:iea
21 ...e5 22 i.f5! is uncomfortable for
White. Thanks to the misplacement of
Black's pieces, e3-e4 followed by tll fl -e3 will
soon be a relevant option.
22 tt:le4 f6 23 'iid 3! ..txe4
This move is based on the following tac­
tics: 23 ... tll f8 24 tlld 6! tllx d6 (24 ... l:f.d8 25
tllxe8 l:txe8 26 'ifd6 gives White a great ad­
vantage; b6 falls) 25 'ifxd6 (with the threat of
'ifxc7 and l:td6) 25 ... 'ifxd6 26 l:f.xd6 l:r.a8 27
i.c3 l:f.a2 28 i.d3 i.c8 29 i.fl l:ta3 30 i.e 1
and White wins b6 for e3, which gives him a
passed pawn on the queenside and estab­
lishes c5 as a weakness.
24 Wxe4 tt:lf8 25 °iWd3 <l;f7 26 ..tc3 <l;e7
5 'ii b 3
26 ... l:ta8 27 .iel ! shows the strength the
bishops have even in this semi-closed posi­
tion.
27 l:!.a 1 l:!.d8 28 'i'e2 tt:ld6
28 ... e5 29 f4! and the black king's defences
crack.
29 f4 �f7 30 e4 tt:lca 31 e5!
White is trying to open the position.
31 . . . f5
31 ...tlle7 32 'ifhs+ tlleg6 33 exf6 gxf6 34
l:tfl ! and White is ready to launch a direct
assault on f6 with f4-f5.
32 °iWf3 tt:le7 33 h3 <l;g8 34 <l;h2!
Improving the king before the attack.
34 . . .tt:lfg6 35 g3 cj;f7 36 h4! 'it>g8 37 h5
tt:lh8
37 ...tll f8 38 g4! - now this is possible as
Black cannot organise counterplay against f4.
38 g4 l:l.f8
38 ... fxg4 39 'ifxg4 'ifc8 40 l:r.a7 l:r.e8 41 h6
with a winning attack.
39 gxf5 tt:lxf5
39: .. exfS 40 e6 and the bishop comes to
life.
40 ..txf5 l:!.xf5 41 l:ta8 + l:tf8 42 l:txf8 +
c.t>xf8 43 Was + <3;17 44 Wxh8 'iib 7 45
h6! 1 -0
"
Game 30
Seirawan-Timman
Hifversum Match (5th game) 1990
1 d4 tt:lf6 2 tt:lf3 e6 3 c4 b6 4 tt:lc3 ..tb4
81
Q u e e n 's In dia n D e fe n c e
5 'ifb3 c5
1bis is the main line. Actually it was this
match between Timman and Seirawan that
put the 5 'ii'b3 line on the map as a serious
main line, instead of a decent sub-variation.
Seirawan used it in all three of his games with
White -all of which he won.
6 a3
1bis is the most serious move. Alterna­
tives:
a) 6 .Jlf4 0-0 and here White has two ways
he can go:
al) 7 a3 .Jla5 8 l:.dl (8 dxc5 llie4 9 cxb6
axb6 1 0 g3 llic6 1 1 1'.g2 .ixc3+ 1 2 bxc3
1'.a6 13 0-0 llic5 14 Vc2 .Jlxc4 15 llie5 1'.b3
16 Vd2 llixe5 17 1'.xe5 d5 was nice for
Black in Dreev-Timman, Moscow 1 993, until
he blundered a few moves later...) 8 ... 1'.a6!
and the bishop is much better placed here
than on b7. In Garcia Palermo-Eingom,
Cienfuegos 1 986 Black organised a quick
attack on White's centre: 9 e3 cxd4 1 0 llixd4
d5 1 1 1'.e2 1'.xc4 12 .Jlxc4 dxc4 13 Vxc4
Vd5 and Black had equalised, if not more ...
a2) 7 e3. 1bis must be the way to play
with 1'.f4 but it fails to eam White an advan­
tage. Dreev-lvanchuk, Linares 1 995 contin­
ued 7 ... d5 8 :d1 llibd7 9 cxd5? (9 dxc5 bxc5
10 1'.d3, with an unclear position, is better,
while 9 1'.d6 :e8 10 1'.d3 1'.a6 seems dan­
gerous for White - what is the bishop really
doing on d6?) 9 ... llixd5 1 0 1'.g5 (10 1'.d6? c4!
might have been what Dreev overlooked)
82
1 0 .. .'.c7 1 1 1'.c4 cxd4 1 2 1'.xd5 .Jlxc3+! 1 3
bxc3 exd5 14 cxd4 1'.a6 and Black had the
advantage.
· b) 6 .Jlg5 with a further branch:
bl) Now after 6 ... llic6 there has been
much debate as to whether 7 d5 gives White
an advantage. But the basic fact that the
knight looks misplaced is possibly best illus­
trated by the less ambitious, but smooth 7
e3!? h6 8 .ih4 g5 9 1'.g3 llie4 10 1'.d3
.ixc3+ 1 1 bxc3 llixg3 1 2 hxg3 d6 13 .Jle4
.id7 14 .ixc6 1'.xc6 1 5 d5, when White had
an advantage in Condie-Cullip, Edinburgh
1 989.
b2) 6 ...1'.b7
b21) 7 0-0-0 1'.xc3! 8 'iixc3 llie4 9 .ixd8
llixc3 10 bxc3 �d8 1 1 d5 ri;e7 gives Black
an even endgame.
b22) 7 e3 0-0 (7 .. .ixf3? 8 gxf3 llic6 9 d5!
.
exd5?! 10 cxd5 llie5 11 f4! llif3+ 12 ri;eZ
llixg5 13 fxg5 1'.xc3 14 'i'xc3 llih5 15 'i'e5+
'i'e7 1 6 'i'xe7+ rJ:i;xe7 1 7 1'.g2 rJ:i;d6 1 8 1'.£3
gave White a strategically won position in
Van Wely-Miles, London (rapid) 1 995) 8
.ie2 cxd4 9 exd4 llic6 1 0 0-0 1'.xc3! 1 1 1'.xf6
(1 1 Vxc3 llie4 and Black appears to have an
edge after 1 2 1'.xd8 llixc3 1 3 bxc3 l:.fxd8,
when c4 is exposed) 1 1 ...'iixf6 12 'i'xc3
l:tac8 13 'i'd2 llie7! 14 l:.acl llig6 1 5 l:tc3
llif4 and Black was doing fine in Timo­
scenko-Pelletier, Leon 2001 . Actually it might
be White who should be careful here.
b23) 7 l:tdl !? 0-0 8 e3 cxd4 9 exd4 and
now 9 ... 1'.xc3+! is a strong theoretical novelty
of Boris Gelfand which gives Black equality.
1 0 bxc3 Vc71 1 1 .ixf6 1'.xf31 1 2 gxf3 gxf6
1 3 'iic2 'i'f4 14 Ve4 'iixe4+ led to a draw in
Alterman-Gelfand, Tel Aviv 1 999. 1bis end­
ing is equal.
c) 6 e3!? 1'.a6 (6 ... llic6!? looks so natural
here that one wonders why it has not been
played - when there is no 1'.g5, then d4-d5
also becomes less of a real threat) 7 a3 1'.a5 8
1'.d2 0-0 and now:
cl) 9 l:tdl cxd4 10 llixd4 d5 1 1 llidb5
1'.b7! 12 cxd5 a6! 1 3 llid4 llixd5 14 llixd5

'i'xdS 1 5 'i'xdS .txd2+ 1 6 l:xd2 .txdS 17 f3
l:a7! 1 8 e4 .ta8! was even in Seirawan­
Browne, USA 1 987. A decent alternative is
9 ... dS!? with the following idea: 10 dxcS bxcS
1 1 cxdS exdS 12 .txa6 �xa6 1 3 0-0 l:b8 14
'i'a2 c4! and Black has good play on the light
squares.
c2) 9 0-0-0 and now after 9 ...'i'e7 10 dS
exdS 1 1 �xdS �xdS 1 2 cxdS .txf1 1 3
:hx f1 .txd2+ 14 �xd2 d 6 1 5 e 4 �d7 1 6
l:fe1 �eS, Browne-Korchnoi, Chicago 1 982,
I prefer White (slightly). 9 ... dS!? looks like an
obvious improvement. White has no way of
turning the tension in the centre to his ad­
vantage: 10 dxcS?! �bd7! 1 1 cxb6 �cS 1 2
'i'a2 axb6 1 3 cxdS ..ixf1 1 4 .l:thxf1 .txc3 1 5
.txc3 �xdS and Black i s a pawn down but
certainly has the advantage!
6 . . . .i.a5! ?
Also possible is 6. . ..txc3+ 7 1"xc3 0-0,
when 8 dxcS is the best try for an advantage,
e.g. 8 ... bxcS 9 .tgS �c6 1 0 e3 h6 1 1 .th4 gS
1 2 ..ig3 �e4 1 3 1"c2 f5 1 4 0-0-0 1"f6 1 5
.td3 �xg3 1 6 hxg3 .tb7, and now in Seira­
wan-Arnason, Manila 1 992, White played 1 7
.te2?. Instead 1 7 g4! gives White a powerful
attack. When the kings are castled on oppo­
site flanks and there are lots of open files,
time is an important aspect of the game. The
alternatives are 8 .tgS .tb7 9 e3 d6 10 dxcS
bxcS 1 1 l:d1 'i'e7 1 2 .td3 �bd7 1 3 0-0
l:fd8, which appeared to secure Black equal­
ity in Malaniuk-Lerner, Donetsk 1 998, 8 e3
.tb7 9 b4 d6 1 0 .tb2 �e4 1 1 1"c2 f5 1 2
l:d1 �d7 1 3 .te2 a S 1 4 bxcS bxcS 1 5 0-0
l:f7 16 .ta1 �b6 17 dxcS �cS 1 8 �d4
'i'gS 19 .tf3 .txf3 20 �xf3 with equality in
Seirawan-Amason, Reykjavik 1 986 (although
this set-up might be more dangerous than it
appeared here) and, finally, 8 g3? cxd4! 9
'i'xd4 �c6 1 0 'flh.4 .tb7 1 1 .tg2 l:lc8 1 2
.td2 �aS! 1 3 .txaS bxaS, when Black had a
lead in development and very active pieces in
Gurevich-Kasparov, Linares 1 991 .
7 .i.g5
The aggressive move. After 7 e3 0-0 8
5 'il b 3
.te2 �e4! 9 dS! .txc3+! 1 0 bxc3 .ta6!? 1 1
.tb2 1"e7 1 2 lld 1 l:e8! 1 3 .td3 (1 3 0-0?
exdS 14 l:.xdS �c6 with ... �aS to follow)
1 3 ...�d6!? Black was doing very well in Sei­
rawan-Timman, Hilversum 1 990. 7 .tf4 ..ib7
8 l:d1 0-0 9 e3 cxd4 10 �xd4 �e4 1 1 1"c2
..ixc3+ 12 bxc3 was fine for Black in Por­
tisch-Polugaevsky, Linares 1 985. It seems
that the bishop is rather harmlessly placed on
f4 in these lines, as the attack on d6 and c7 is
not supported by other pieces.
7 lbc6?
. . .
This move is refuted in the meanest pos­
sible way. Others:
a) 7 ... h6?! 8 .th4 gS (8 ... �c6? 9 0-0-0!
..ixc3 10 'ii'xc3 cxd4 [1 0 ...gS 1 1 dSI and
Black is clearly not developed to meet this]
1 1 �xd4 �e4 1 2 'i'h3! and White won a
quick victory in Seirawan-Timman, Hilver­
sum (3) 1 990, while 8 ... .tb7 is, of course,
best here, but because the bishop is now on
h4 Black will have less effect with the
... .txc3+ and ... �e4 trick than without ... h7-
h6). After 8 ... gS 9 ..ig3 Black has played:
al) 9 ... �c6 10 0-0-0 .txc3 1 1 dS! exdS 12
cxdS �xdS 1 3 l:.xdS ..ig7 1 4 h4 and White
had ;a powerful initiative in Nielsen-Hracek,
Germany 2000.
a2) 9 ...�e4 10 e3 �c6 1 1 .td3 �xg3
(1 1 ....txc3+ 1 2 bxc3 �xg3 1 3 fxg3! gives
White some prospects on the f-file) 1 2 hxg3
g4 13 dS! gxf3 14 dxc6 fxg2 1 5 l:g1 with a
powerful initiative for White in Psakhis-
83
Q u e e n 's In dia n D e fe n c e
Gruenfeld, Israel 1 99 1 .
a3) 9 ... g4 10 lDd2 cxd4 1 1 tiJb5 i.xd2+ 1 2
�xd2 lDe4+ 1 3 �e1 with a clear advantage
for White in Malaniuk-Lendwai, Kesckement
1 991 .
b) 7 ... .tb7! and now White has a flood of
different opportunities:
b1) 8 0-0-0!? i.xc31 9 W'xc3 lDe4 10 i.xd8
lDxc3 1 1 bxc3 �dB 12 d5 1;e7 13 e4 d6 1 4
i.d3 lDd7 1 5 l:.he1 l:lad8 with an equal end­
ing in Gretarsson-Hjartarson, Leeuwarden
1 995.
b2) There is no time for 8 dxc5?!. Black
now develops a lot of threats quickly:
8 ... lDa61 9 cxb6 (9 'iic2 .txc3+ 1 0 'iixc3
lDxc5 1 1 tt.Jd2 a5 12 b4 axb4 1 3 axb4 l:.xa1+
1 4 'iix a1 tDa6 15 'iib 2 'iie7 gave Black a
development advantage in Campos Moreno­
Adams, Cala Galdana 2001 -who cares about
two bishops when the rook is still on h1 with
no immediate prospects?) 9 ... lDc5 10 'ii'c2
.te4 1 1 'ii'd 1 i.xc3+ 1 2 bxc3 1i'xb6 and
Black has a strong initiative, Van Wely­
Seirawan, Wijk an Zee 1 995.
b3) 8 l:.d1 .txc3+! 9 bxc3 'iie7 10 d5 d6
1 1 e3 lDbd7 12 i.e2 h6 13 .th4 e5 with a
good position for Black in Gunawan­
Timman, Bali 2000.
b4) 8 e3 and now:
b41) 8 ... 0-0 is the most normal. 9 .:d1 d5
(again the most logical, but also possible is
9 ... .txc3+ 1 0 bxc3 ii'e7 1 1 i.e2 d6 1 2 0-0
lDbd7 1 3 lDd2 llfc8 14 f3 h6 15 .th4 with a
complicated struggle ahead in Seirawan­
Adams, Bermuda 1 999) 1 0 i.e2 tiJbd7 1 1
0-0 cxd4 1 2 lDb5!? (although neither o f the
players could have foreseen it, this move
practically leads to a forced draw; anyway,
Black appears to be doing well) 1 2 ...h6 1 3
.th4 lDc5 1 4 'ii'c2 d31 1 5 i.xd3 tiJxd3 1 6
l:.xd3 a6 1 7 cxd5 axb5 1 8 dxe6 ii'e 8 1 9 i.xf6
fxe6 20 .te5 .te4 21 .:d7! (21 b4?? l:.xf3 22
gxf3 'ii'g6+) 21 . ..i.xc2 (21 . . .'iixd7 22 'iixe4
gives White powerful compensation for the
exchange as his minor pieces are very
strongly placed in the centre) 22 :Xg7+ <itih8
84
and a draw was agreed in Bareev-Adams,
Wijk aan Zee 2002. Note that 9 ... cxd4?!
quickly leads to trouble: 10 l:.xd41 lDc6 1 1
l:.f4! tt.Je7 1 2 :Xf6! and White had an initia­
tive in Seirawan-Adams, Bermuda 1 999.
b42) 8 ... cxd4!? 9 lDxd4 0-0 10 'iic2 i.xc3+
1 1 'iixc3 d5 looked fine for Black in Soko­
lov-Adams, Dortmund 1 999.
b43) 8 ... h6?! 9 .th4 g5 (9 ... 0-0 10 l:.dl
cxd4 11 :Xd4 i.xf3 12 gxf3 lDc6 1 3 :d2
lDe5 14 .te2 lDg6 1 5 .tg3 lDe8 1 6 'ii'c2
i.xc3 1 7 'ii'xc3 £5 1 8 'iid3 left White in a
strong attacking position in Sokolov-Janssen,
Netherlands 1 999) 10 i.g3 lDe4 1 1 .td3
i.xc3+ 12 bxc3 lDxg3 13 hxg3 d6 14 'iic2
lDd7 and a draw was soon agreed in Nielsen­
C.Hansen, Aars 1 995, but the position does
not feel completely safe for Black.
b44) 8 . . ..txm 9 gxf3 cxd4 1 0 exd4 lDc6
1 1 0-0-0 i.xc3 12 'ii'xc3 l:.cB 13 �b1 gave
White a storming advantage in Korchnoi­
Polugaevsky, Reykjavik 1 987.
b45) 8 ... 'ii'e 7 9 .te2 d6 1 0 dxc5 dxc5 1 1
0-0-0 tiJbd7?! (1 1 ...i.xc3 1 2 'ii'xc3 is a little
better for White due to 1 2 ... lDe4 1 3 'iixg7!
with advantage) 12 lDb5 0-0 13 tiJd6 .tc6 14
e4 was critical for Black in Shabalov­
yemelin, Moscow 2002.
8 0-0-0 .txc3
9 d5!!
This i s the main point o f the fight against
... lDc6 in this system, and why I suggest that
Black should avoid ... tt.Jc6 altogether. 9 dS is

available to White because the lead in devel­
opment over-rides a possible loss of a pawn.
9 . . . exd5?
Opening up the position when behind in
development tends to be a poor idea.
9 ... i.e5! 10 dxc6 i.c7! is the correct defence
(weaker is 1 0 ...'it'c7?, e.g. 1 1 g3! i.d6 1 2 .i.g2
dxc6 1 3 nxd6!! 'iixd6 14 nd1 'it'c7 1 5 i.f4
'it'b7 1 6 tlle 5 i.d7 1 7 nxd7! lllxd7 1 8 i.xc6,
which led to a swift victory in Seirawan­
Zarnicki, Buenos Aires 1 993). Then with 1 1
e3! White maintains a lead in development
for as long as possible in order to create the
greatest possible disharmony in the oppo­
nent's camp. Soppe-Debamot, Buenos Aires
1 991 continued 1 1 ...h6 1 2 i.h4 'it'e7 1 3 'it'a4
d6 14 .i.d3 (14 i.e2!? appears logical) 14 ...g5
1 5 i.g3 tllh 5 1 6 i.e2 f5 1 7 tll d 2 tll f6 1 8 h4
g4 1 9 f3 h5 20 tllb 1 ! a6 21 tllc3 0-0 22 i.f4
:b8 23 i.h6 :n 24 i.g5 'it'f8 25 e4 with
excellent prospects for White. Note that the
bishop is yet to be of use on c8 and, there­
fore, the c6-pawn remains a constant source
of discomfort for Black.
1 0 cxd5 i.e5 1 1 dxc6 'iie 7
1 2 cxd 7 + ?
This move wins the game but, as a pupil
5 'fi b 3
o f mine illustrated when given thi s game as
an exercise, 12 i.xf6! is more straightfor­
ward: 1 2 ... i.xf6 (1 2...'i'xf6 1 3 cxd7+ i.xd7
14 'i'e3) 1 3 'iid5 0-0 14 cxd7 i.a6 1 5 e4 and
White wins.
1 2 . . . .i.xd7 1 3 e3!
With the idea of :Xd7 and i.b5+. Note
that Black cannot get the king into safety as
he will lose instantly.
1 3 . . JldB
13 ... 0-0 ? 14 lllxe5 'ii'xe5 15 i.xf6 'if'xf6 1 6
nxd7 and White wins a piece, or 1 3. . .0-0-0
14 i.a6+ 'it>b8 1 5 tllxe5 and Black can only
resign.
1 4 J:xd7! J:xd7
14 ... 'it>xd7 1 5 'it'a4+! 'it>e6 1 6 i.c4+ �fS 1 7
'ii' c2+ 'it>g4 1 8 h3+ 'it>h5 1 9 g4+ and White
wins.
1 5 .i.b5 .i.d6 1 6 l:td 1 0-0 1 7 i.xd7 'iixd7
1 8 .i.f4!
And Black has no defence.
1 8 . . . c4
1 8 ...tll e4 1 9 'ii'd5 'i'a4 20 i.xd6 :d8 21
tll e5 tllxd6 22 tllc4 and White wins a piece.
1 9 'iic 2! ltieB 20 ltig5! f5 21 'iixc4+
�hB 22 .i.xd6 ltixd6 23 'iid 5 l:tdB 24
ltie6 ! 'iic B+ 25 �b1 l:td7 26 'iix d6 ! 1 -0
85
Q u e e n 's In dia n D e fe n c e

Summary
The positions after 5 'W'b3 are less well known than other lines in the QID, and that is already
a good reason for playing it. Furthermore it can give White some good pressure. After 5 . . aS
.

and 5 . 11i'e7 Black is probably always going to be very slightly worse, while after the more ac-
. .

tive 5 cS - with the idea of 6 a3 .i.aS! keeping the bishop - Black should be able to find a
...

path to equality. This latter line is particularly worthy of experimentation.

1 d4 lbf6 2 c4 e6 3 lbf3 b6 4 lbc3 i.b4 5 'iFb3 (D)

S ... aS Game 28
-

5 11i'e7 Game 29
... -

5 c5 (D) 6 a3 J.a5 7 i. g 5 (D) Game 30

. . . -

5 'ifb3 5 . . . c5 7 J.g5

86
I CHAPTER f/VE I
4 e3

1 d4 lLlf6 2 c4 e6 3 li:lf3 b6 4 e3
The 4 e3 line is often favoured by players
who do not like to study a lot of theory. Not
because knowledge of the line is limited but
because it has a tendency not to change.
Consequently in the books (as in this one)
Black is completely equal and the lines are
rather harmless, but some older GMs (who
might prefer playing with their kids) choose
this line over others that they feel will de­
mand more study. A large number of draws
should be anticipated.
In fact the e3-line is rather harmless for
Black and I simply cannot recommend it for Black should play this for safety. After
White, despite strong players such as Yusu­ 5 . i.e7!? 6 dS! White is playing against the
..

pov and Gelfand having found it useful for
their purposes occasionally. The system sim­
ply does not put Black under any pressure
and therefore he should be able to equalise
without too great an effort. Of course, that is
if he is prepared. I just think that playing a
more aggressive system will benefit your
game generally, no matter who you are.
Game 3 1
Gelfand-Karpov
Sanghi Nagar (7) 1995
1 c4 lLlf6 2 d4 e6 3 liJf3 b6 4 e3 .tb 7 5
bishop on b7. Nevertheless this is still play­
able for Black; only the main line is safer.
6 cxd5 exd5 7 .tb5+ !
Played in order to limit Black's influence
on e4.
7 . . . c6 8 .td3 .te7! ?
8. . .i.d6 appears equally but we are follow­
ing the path of the great masters here.
9 0-0 0-0 1 0 b3
Wlule is taking it slowly. 10 e4 dxe4 1 1
lllxe4 lllbd7 1 2 :et l:.e8 1 3 lllegS h6 1 4
lll e4 lllxe4 1 5 i.xe4 i.b4 was fine for Black
in Kurajica-Dizdar, Zagreb 1 993 .
. 1 0 . liJbd7
..

lLlc3 d5 This is where the knight belongs. After

87
Q u e e n 's In dia n D e fe n c e
10. . .lDa6 1 1 i.b2 c 5 1 2 l:.cl lDc7 1 3 dxc5
bxc5 14 lDa4 lDe4 (14 ... lDe6 1 5 i.xf6! would
be uncomfortable for Black), 1 5 i.xe4 was
played in Langeweg-Schneider, Plovdiv 1 983,
but 15 lDe5! is stronger, securing White an
advantage. The main threat is f2-£3, but
i.xe4 followed by liJd7 is also in the air.
1 1 i.b2
1 1 . . . i.d6
This move order is actually not so exact.
As we shall see lD£3-£5 is a relevant manoeu­
vre, and therefore Black should start with
1 1 ...l:le8!, securing equality and ensuring that
dropping back to f8 with the bishop no
longer wastes time.
1 2 l:c 1
Others:
a) 12 'ii'c2 :e8 13 l:.fel l:.c8 14 e4 (1 4
l:.acl c5 1 5 'ii'e2 lDe4 1 6 i.a6 i.xa6 1 7
'ii'xa6 liJdf6 1 8 dxc5 l:.xc5 1 9 h3 'ii'd7 20
'ii'd3 l:.ec8 was equal in Portisch-Hiibner,
Manila 1 990) 14 ... dxe4 15 lDxe4 lDxe4 1 6
i.xe4 liJf6 1 7 i. £5 l:.c7 1 8 :Xes+ 'ii'xe8 1 9
lDe5 c 5 20 'ii'd3 l:.e7 with equality, Po­
lugaevsky-Karpov, Roqueburne 1 992.
b) 12 lDh4! l:.e8 13 lD£5 i.£8 14 l:.cl was
a little better for White in Petrosian­
Taimanov, Zurich 1 953.
1 2 . . .:ea 1 3 li)e2 'iie 7 14 li)g3 g6 1 5
'iie 2 i.a3!
Apparently exchanging the 'good' bishop,
but in reality preventing any problems on the
long diagonal after ... c6-c5, and creating a
88
possible weak square on c3.
1 6 l:fe1 'ifd6
Black has equalised.
1 7 'iic 2 i.xb2
17 ... a5!? is also possible.
1 8 'iix b2 c5 1 9 l:ed 1 l:ac8 20 li)e2
l:ed8
20 ... a6 21 l:lc2 cxd4 22 lDexd4 gave White
a little something in Gelfand-Karpov, Sanghi
Nagar 1 995.
2 1 l:c2 cxd4 22 li)exd4 l:xc2 23 i.xc2
l:c8 24 l:c 1 li)e4 25 i.d3 l:xc 1 + 26
'iix c 1 'iic 5 27 'iid 1 b5 28 li)e2 b4 29
'ifa 1 'ifd6 30 'ikd4 'iic 5 31 'ika1 'iid 6 32
h3 li)dc5 33 i.c2 'iif 6 34 'if e5 'iix e5 35
li)xe5 f6 36 li)f3 li)e6 37 li)e 1 cj;>f7 38 f3
li)4c5 39 g4 a5 40 cj;>f2 .ta6 4 1 h4 h6
42 h5 gxh5 43 gxh 5 cj;>e7 44 li)g3 li)g7
45 i.g6 cj;>d6 46 cj;>g2 .tb5 4 7 li)c2 li)ce6
48 f4
White has played well and Black is still de­
fending. Now Karpov's superior technique
pays dividends.
48 . . . i.d7 49 cj;>f2 li)c7 50 li)d4 li)b5 5 1
li)xb5 + ? ! i.xb5 5 2 c,f;>e 1 i.d7 5 3 i.d3
i.g4 54 i.e2 li)f5!
Suddenly White has a difficult bishop
endgame on his hands.
55 li)xf5+
55 i.xg4 lDxg3 56 i.£3 lD£5 and White
has problems.
55 . . . i.xf5 56 cj;>d2 cj;>c5 57 i.d3 i.d7 58
i.c2 .tea 59 i.g6 i.c6 60 i.c2 i.d7 61

i.d 1 i.h3 62 i.f3 i.f5 63 �c 1 i.d7 64
�d2 a 4 65 bxa4
6 5 �d3 a3! and tricks like . . .i.a4 will al-
ways be in the air.
65 . . . i.xa4 66 i.e2 i.c6 67 i.f3
67 i.d3 d4 68 e4 i.e8 69 i.e2 i.f7 and
Black wins.
67 . . . i.e8 68 i.d 1 d4 69 �d3 i.b5+ 70
�d2 dxe3+ 71 �xe3 i.e8 72 �e4 �d6
73 �f5
73 �d4 i.f7 74 i.b3 i.xh5 75 �c4 i.f7+
and the pawn ending i s winning fo r Black.
73 . . . �e7 74 i.e2 i.f7 75 i.d 1 i.xa2 76
�g6 b3 77 �xh6 b2 78 i.c2 �f7!
Prevents all counterplay. After 78 ... bl'ii'
79 i.xb1 i.xb1 80 �g7 the position is
drawn.
79 i.g6 + �f81 80 i.d3 b 1 'ii' 8 1 i.xb 1
i.xb 1 82 f5 �f7 0-1
Game 32
Danner-Yu S haoteng
Gyu/a 2000
,_______________...
1 d4 �f6 2 c4 e6 3 �f3 b6 4 e3 i.b 7 5
i.d3 d5 6 0-0 i.e7 7 b3
As this game shows Black has no prob­
lems in this line.
7 . . . 0-0 8 �c3 c5 9 i.b2 cxd4 1 0 exd4
�c6
1 1 :tc 1
The main line. The alternatives are:
a) 1 1 'ii'e 2?! is not so good because the
4 e3
rook now gets stuck on a1 - 1 1 ... �b4! 1 2
i.b1 dxc4 13 bxc4 i.x£3 14 gxf3 'ii'xd4 1 5
�e4 'ii'd8 1 6 l:r.dl 'fic7 1 7 �xf6+ i.xf6 1 8
i.xf6 gxf6 19 i.xh7+ �g7! (1 9 ... �7 20
'i'e4+! is not something Black allows) 20 l:r.d4
l:r.h8 21 l:r.g4+ �f8 22 'ii'b2 :xh7 23 'ii'xb4+
'ii'c5 and Black had a positional advantage in
Grigorian-Karpov, Moscow 1 976.
b) 11 :el l? dxc4 (this does not look quite
right; after 1 1 ...l:lc8! White has nothing better
than 12 l:lc1 , transposing to the main line) 1 2
bxc4 :tc8 13 :c1 l:le8 14 d5! (suddenly this
is possible) 14... �b4 1 5 i.b1 exd5 16 a3 d4
1 7 �xd4 �c6 1 8 �fS 'i'xd1 1 9 l:lcxd1 i.f8
20 �d5 �d7 21 h3 l:lxe1+ 22 :xe1 l:r.d8 and
Black kept his stuff together in Portisch-
Hjartarson, Szirak 1 987, although the line
looks shaky.
1 1 . . . :tc8
Tbis is the most natural move. Black is
improving his position before taking action.
1 1 ...�b4 12 i.b1 :c8 1 3 �e5 �c6 14
�xc6 :Xc6 1 5 'ii'e2 l:le8 1 6 �d1 i.f8 1 7
�e3 was agreed drawn in Gelfand-Kramnik,
Novgorod, 1 997. lt seems to me that White
has more freedom to manoeuvre than is
usual in this line.
1 1 . ..'ii'd 6?1, as in Portisch-Adorjan, Szirak
1 987, is probably best met with 1 2 cxd5!
exd5 13 :e1 .l:.fe8 14 �b5 'ii' f4 1 5 g3 'ii'b 8
16 �e5 and White has the advantage.
1 2 lle 1 !
The rook belongs here. After 1 2 'ii'e2 :e8
13 l:lfd1 i.f8 14 h3 g6 15 i.b1 i.h61 1 6 �d2
dxc4 17 bxc4 �xd4 1 8 'ii'e 1 e5 Black was
much better in Polgar-Hansen, Aabenraa
1 989.
1 2 . . . �b4!
With this move Black takes control of the
centre� Others seem to be inferior.
a) Now 1 2...:e8!? comes out okay here,
but I do not trust it completely: 1 3 cxd5
�xd5 14 �xd5 'ii'xd5 (14 ... exd5 1 5 i.fS :c7
16 'ii'e2 gave White an edge in Yusupov­
Renet, Kaufbeuren 1 993) 1 5 i.e4 'ii'h 5
(1 5 ...'ii'd7? 1 6 l:r.xc61 i.xc6 1 7 �e5 'ii'b7 1 8
89
Q u e e n 's In dia n D e fe n c e
.txh7+! was winning i n Portisch-de Finnian,
Reggio Emilia 1 989; while 1 5 ...'iid6 1 6 d5!
led to White asswning the initiative in Yusu­
pov-Rodriguez, Novi Sad 1 990). Then 1 6
lbe5 (certainly not forced. . .) 1 6. . .'iixdl 1 7
l:.exdl lbxe5 1 8 i.xb7 l:.xcl 1 9 .txcl l:.d8
20 'it>fl l:.d7 21 i.c8 l:.d8 22 i.b7 l:.d7 23
i.c8 l:.d8 24 .tb7 was agreed drawn in Pet­
rosian-Peters, Lone Pine 1 978.
b) 12 ... .tb4 1 3 a3 i..x c3 14 i.. xc3 lbe7 1 5
lbe5 looked better for White in Ambarcum­
jan-Shabalov, Seattle 2002, but I am sure that
Shabalov only wanted to depart from the
normal paths and somehow unbalance the
situation.
1 3 i.f 1 lbe4
Clearly the most active move. Also possi­
ble is 13 ...l:.e8 14 a3 lbc6 1 5 cxd5 exd5 with
equality in Nogueiras-Farago, Aosta 1 990.
14 a3 lbxc3 1 5 :xc3 lbc6 1 6 'ifc2? !
White wants to prepare d4-d5, but this is
too slow and Black is excellently placed to
meet it. Better was 1 6 cxd5 'iixd5 1 7 i.c4
(1 7 i.d3? lbxd4 18 i.xh7+ 'it>xh7 19 'iixd4
'iixd4 20 lbxd4 i.f6 21 l:.d3 l:.fd8 was better
for Black in Van der Werf-Van den Doel,
Holland 1 994) and now 1 7 ...'i'd6?! 1 8 d5
lba5 (1 8 ... exd5 19 l:.d3! gives White the ini­
tiative) 1 9 l:.d3 lbxc4 20 bxc4 l:.fd8 21 lbd2?
(21 i.e5! 'iid7 [21 ...'iic 5! 22 d6! with an ad­
vantage to White] 22 i..xg7!! 'it>xg7 23 lbe5
and White has a winning attack on the way)
2 1 . ..exd5 (now Black holds the lead) 22 cxd5
90
.txd5 23 lbfl i.f8 24 lbe3 l:.c5 25 'it'd2
'i'h6 26 l:.dl i.d6 27 g3 i.e4 28 lbg4 'i'h3
29 £3 .txd3 30 'i'xd3 l:.d5 and White re­
signed in Lukacs-Atalik, Budapest 1 99 1 .
1 7 . . .'i'h5 heads fo r th e safe spot. 1 8 d5
exd5 1 9 'iixd5 'iixd5 20 i.xd5 i.. f6 21 l:.c2
l:.fd8 22 .txf6 l:.xd5 23 i.b2 l:.cd8 was equal
in Shariyazdanov-Chemyshov, Djakovo
1 994.
1 6 . . . i.f6 1 7 :d 1 'ifd6 1 8 cxd5
1 8 l:.e3 dxc4 1 9 bxc4 lba5! and Black's
pieces are much the better placed.
1 8 . . . 'ifxd5 1 9 i.c4 'ifh5 20 d5 lba5!
Black sacrifices a pawn for the two bish­
ops and to open up the position in order to
exploit White's uncoordinated pieces.
21 dxe6 lbxc4
21 .. .i.xc3? 22 'i'xc3 'i'g6 23 exf7+ 'it>h8
24 l:.d6 is given by Gershon as winning. Af­
ter 24...l:.xc4 25 bxc4 'it'xf7 it is, presumably,
not bad at all, but still - why should Black
want this position in the first place if he does
not win the exchange?
22 exf7 + 'ilxf7 23 :xc4 Ibc4 24 bxc4
.bb2 25 'ifxb2 i.xf3 26 gxf3 'ifxf3
The number of weaknesses in White's
camp spells the end.
27 :d4 h6
Creating an escape square for the king so
the rook can be activated.
28 a4 :t6 29 'ilc2 :e6 30 'ii' d 2 l:tg6+
3 1 �f1 'i'g2+ 32 'ifi>e2 :l.e6 + 33 �d 1
'i'f1 + 34 'ifi>c2 :l.e2 35 :d8 + 'ifi>h7 0-1
Game 33
Yusupov-Beliavsky
Austria 1998
This game is an illustration of one line that
Black should try to avoid, and into which
White players have occasionally had luck in
tricking their opponents.
1 d4 lbf6 2 lbf3 e6 3 e3 b6 4 i.d3 i.b 7
5 0-0 c5
Already the problems begin. After 5 ... d5!
Black can play 6 lbbd2 c5! and return to
 4 e3

these solid positions against the Colle or, in After this it seems that Black is in dire
the event of 6 c4 ii.e7, employ a transposi­ straits. Imperative is 13. .. l:.e8 14 ii.b5! a6 1 5
tion to the previous game. ll:ic6 .ixc6 1 6 i.xc6 l:.a7 1 7 ii. £3, where
6 c4 i.e7 7 lLic3 cxd4? ! White is simply better (Yudasin).
7 ... d5! is still good enough for Black. 1 4 'i'h3 lLidf6
7 ... h6!? has also been tried (to avoid ii.xh7+
and thereby d4-d5, as in the following exam­
ple) and now 7 ... 0-0?! 8 d5! exd5 9 cxd5 d6
10 e4 ll:ia6 1 1 l:.el ll:ic7 1 2 l:.bl l:.e8 13 a3
ll:id7 14 i.f4 l:.c8 1 5 •d2 gave White an
enormous advantage in Bukic-Ljubojevic,
Bugojno 1 978. The bishop is not well on b 7.
8 exd4 d5
Seems forced. After 8 ...d6 9 d5! White has
a better version of: 8 ... 0-0 9 d5! ll:ia6 10 .if4!
d6 1 1 dxe6 fxe6 1 2 i.g3, which gave White
an edge in Malaniuk-Tiviakov, Moscow 1 992.
9 cxd5 lLixd5 1 0 lLie5 0-0 1 1 'i'g4
Also possible is 1 1 �5 g6!? (1 1 .. .fS?! 1 2 1 5 d5! !
i.c4! secures White a strong position) 1 2 Opening the cl-file and thereby undermin­
� 3 ll:ixc3! 1 3 bxc3 ll:ic6 1 4 ll:ig4! with ing e4 as a strongpoint.
menacing play. 1 5 . . . lLixc3 1 6 bxc3 i.xd5 1 7 .1.g 5 h6
1 1 lLif6
. . . 1 7 ... g6 1 8 �4! h6 demonstrates how bad
Perhaps it was better to play 1 1 ...ll:ixc3 1 2 things are. In Polgar-Christiansen, San Fran­
bxc3 ll:id7 1 3 ii.h6 .if6 14 l:.adl , although cisco 1 991, White· was in the driving seat
White retains the initiative. after 19 i.xg6! fxg6 20 •xh6 ll:ih7 21
1 2 'i'h4? •xg6+ 'iith 8 22 .ixe7 •xe7 23 c4 �4 24
This move gives Black a chance to fight. cxd5 •xf2+ 25 'iith l l:.f6 26 •e4 exd5 27
After 1 2 •g3! 'iith 8 (1 2 ...•xd4 1 3 .ie3 �4 •xd5 l:.g8 28 h3 etc.
1 4 a3! is clearly favourable to White accord­ 1 8 i.xh6! gxh6 1 9 'i'xh6 'i'c7 20 'i'g5+
ing to L.Hansen) 13 .ig5 h6 1 4 •h3 ll:ig8 1 5 �h 8 2 1 'i'h4+ 1 -0
�5! Black was in serious trouble in Dizdar­
Hansen, West Berlin 1 988.
1 2 lLibd7?!
. . .

Necessary is 1 2... ll:ie4! 1 3 �5 ll:if6 14

•h4 ll:ie4 1 5 �5 ll:if6 1 6 •h4 ll:ie4 17
�5 ll:if6 1 8 �4 ll:ie4 1 9 �3?! •xd4 20
il.f4 ll:if6 21 ll:ie2 •a4 22 l:.fcl i.a6 23 i.c2
•e8 and Black held out in Kishnev­
Schlosser, Budapest 1 99 1 .
1 2. . .ll:ic6? 1 3 il.g5 g6 ( 1 3...h 6 1 4 .ixf6
ii.xf6 1 5 •e4 g6 1 6 ll:ixc6 •c7 1 7 .£3 il.g7
1 8 il.e4 f5 1 9 ll:ie7+ and wins) 1 4 i.a6! and
White was winning in Plaskett-Arkell, Lon­
don 1 99 1 . The moves •g3+ and ll:ig6+ will soon win
1 3 lld 1 lLie4? the queen.

91
Q u e e n 's In dia n D e fe n c e

Summary
The variation with e2-e3 is not really theoretical, but it can still be a minefield for Black. The
right way to meet it has been presented in this chapter. If you play like Black in Danner -Yu
Shaoteng, or the sidelines, you should not end up in any trouble. The most important lesson is ·

not to play like Beliavsky against Yusupov. Many grandmasters have suffered in this line.

1 d4 lL'lf6 2 c4 e6 3 lL'lf3 b6 4 e3 i.. b 7 (D) 5 i.. d 3

S liJc3 dS 6 cxdS exdS 7 .i.bS+ (D) - Game 3 1
5 d5
. • .

5 c S 6 0-0 .i.e7 7 ltJc3 cxd4 8 exd4 d S - Game 33

. . .

6 0-0 i.. e 7 7 b3 0-0 8 lL'lc3 c5 (D) - Game 32

4 . . . i.. b 7 7 i.. b 5 + 8 . . . c5

92
 CHAPTER SIX I
4 g 3 i.. b 4 + and
4 . . . i.. a 6 without 5 b3
1 d 4 l0f6 2 c 4 e6 3 l0f3 b6 4 g3 .i.b4 +
I n this chapter we are considering the
lines arising after 4 ... i.. b 4+ and also those
that occur after 4 ... i.a6 when White tries
any move other than 5 b3. 4 . . . i.. a6 is the
modem way of playing these Queen's In­
dian positions, and usually generates more
lively play than 4 ... i.. b 7, which will be con­
sidered in Chapter 8. After 4 ... i.. a 6, White
has several replies besides 5 b3. They in­
clude 5 'ii'c2 and 5 'ii'b 3, which are not
really dangerous for Black. Korchnoi has
tried to come up with some ideas after 5
'ii'c 2, but I find it hard to believe in them. 5
'it'a4 has been played for a long time now. It
was originally made popular by Michael
Rohde in the 1 9 80's, and has ever since
been a serious alternative. 5 liJbd2 is also a
popular alternative, but as this chapter illu s­
trates, Black can always find a safe path for
equality in these lines.
Game 34
Magerramov-Makarichev
Moscow 199 1
1 d4 l0f6 2 c4 e 6 3 l0t3 b 6 4 g3
.i.b4+ ! ?
This i s rather off-beat but does offer a
chance for transposition to standard lines.
5 i.d2 a5
The alternative is 5 ... i..xd2+ 6 'ii'xd2 i.. a6
7 liJa3!? (the knight is not so impressive out
here; 7 b3 transposes to 5 b3 i.b4+ 6 i.d2
i.xd2 7 'ii'xd2 and Game 47) 7 ... 0-0 8 i.g2
i.b7 9 0-0 d6 1 0 l:tfd1 liJbd7 1 1 l:tac1 'ii'e7
12 liJb5 a6 13 liJc3 ltJe4 14 'ii'd 3 £5 1 5 liJe1
liJdf6 16 d5 l:tae8 17 e3 b5 1 8 b3 ltJc5 1 9
'ii'c2 b 4 2 0 liJe2 e S with an unclear game in
Beliavsky-Miles, Tilburg 1 986. Garcia Pal­
ermo-Groszpeter, Zenica 1 987 went 10 liJh4
i.xg2 1 1 ltJxg2 c5 1 2 d5 exd5 1 3 cxd5 l:te8
1 4 £3 a6 1 5 liJc4 b5 1 6 liJce3 l:te5 with equal­
ity, while 1 0 liJb5 liJbd7 1 1 l:tfd1 a6 1 2 liJc3
ltJe4 13 ltJxe4 i.xe4 14 l:tac1 liJf6 1 5 'iie3
°iVb8 16 liJel i.xg2 17 liJxg2 'ii'b7 was equal
in Yusupov-Kuzmin, Minsk 1 982. White's
space advantage is not so important because
so many pieces have been exchanged.
6 .i.g2 0-0
Here White has a very slight advantage af­
ter 6 ... i.a6 7 i.xb4 axb4 8 ltJe5 l:ta7! 9 a3
i.c8 1 0 liJd3 bxa3 1 1 l:txa3 l:txa3 12 ltJxa3
0-0 13 0-0, as in Browne-Spassky, Manila
1 976. ,;
7 0-0 .i.a6
7 ... i.b7 8 i.g5 (8 'iic2 h6 9 i.. f4 i.. e4 1 0
'it'dl liJc6 1 1 a 3 i..d6 12 i..xd6 cxd6 1 3 liJc3
d5 14 cxd5 exd5 15 e3 ltJe7 16 ltJe5 i.. xg2 17
�xg2 d6 1 8 liJd3 was also better for White
93
Q u e e n 's In dia n D e fe n c e
in Ivanchuk -Speelman, Roquebrune 1 992)
8 ... i.e7 9 ll'lc3 ll'le4 10 ll'lxe4 i.xe4 1 1 i.xe7
'ii'xe7 12 'ii'd2 d6 13 'iie3 i.xf3 14 i.xf3
lta7 15 l:acl with advantage to White in
Karpov-Van der Wiel, Brussels 1 987.
8 b3
8 . . . d5
'lb.is is the main move here, but there are
some important alternatives:
a) 8 ...l:a7 9 i.f4 a4 (9 ... d5!? 10 c5 ll'le4 1 1
a3 i.c3 1 2 cxb6 cxb6 1 3 l:a2 b5 1 4 ll'lg5!
ll'lxg5 15 ll'lxc3 b4 16 axb4 axb4 17 ll'lb5
l:d7 1 8 'ifcl with an edge for White) 10 'i'c2
d5 1 1 l:dl 'ifc8 12 ll'le5 c5 13 dxc5 i.xc5 1 4
ll'lc3 and now, instead of 1 4... b5?, which was
played in Avrukh-Mikhalevski, Beersheba
1 997, Black should play 14 ... axb3 1 5 axb3
dxc4, although White is better after 1 6 ll'lxc4.
b) 8 ... c6 9 'i'c2 d5 1 0 i.xb4 axb4 1 1 a3
i.b7 12 ll'lbd2 bxa3 13 b4 ll'la6 14 'i'b3 dxc4
1 5 ll'lxc4 ll'ld5 1 6 :Xa3 ll'ldxb4 1 7 ll'lfe5 °flc7
1 8 l:.bl gave White excellent compensation
for the pawn in Zimmennan-Letreguilly,
Budapest 1 994.
c) 8 ... i.b7 transposes to Games 48-50
with 4 ... i.a6 5 b3 i.b7 6 i.g2 i.b4+ 7 i.d2
a5 8 0-0 0-0.
9 cxd5 ltlxd5
9 ... exd5 10 ll'le5 is a structural advantage
for White.
1 0 Ae 1 c5 1 1 e4 ltlf6 1 2 a3! �xd2 1 3
ltlbxd2 ltlbd7 1 4 e5 ltld5 1 5 ltle4
White has won the opening battle. The
94
d6-square secures him a slight pull.
1 5 . . .'i'bS? !
1 5 ... 'ii'e 7! (Magarramov) 1 6 ll'ld6 l:ad8! 17
l:cl cxd4 1 8 ll'lxd4 ll'lxe5 1 9 ll'lc4!? ll'lxc4 20 ·
bxc4 ll'lf6 21 ll'lc6 'ii'c7 22 ll'lxd8 :Xd8 with
just about enough compensation, although it
is still White who is trying for a win.
1 6 ltld6 cxd4? !
Now White gets the chance to start an at­
tack. Necessary was 16 ... h6.
1 7 'i'xd4 ltlc5
17 ... l:d8? 18 ll'lxf7! �fl 19 ll'lgs+ <j;e7
20 ll'lxe6! would eliminate Black completely.
1 8 ltlg 5! 'i'd8
1 8 ... h6? 1 9 ll'lgxfll l:.xfl 20 i.xd5 exd5 21
ll'lxfl <j;xfl 22 e6+1 ll'lxe6 23 'ifxd5 i.cB 24
l:acl and White wins.
1 9 'i'g4 l:a7
Black is trying very hard to defend.
1 9 ... h6? 20 ll'lgxfl l:.xfl 21 b41 and White
wins.
20 b4 ltld3 21 .i.e4! g6 22 ltlxe6 ltlxe5
Black is running out of defensive re­
sources. After 22 ...'ii'd 7? White wins with 23
i.xd5 ll'lxel 24 l:xel i.c8 25 h3! fxe6 26
ll'lxc8 exd5 27 ll'lxa7.
23 ltlxd8 ltlxg4
24 ltl8xf7!
A nice combination that in practice wins
the game.
24 . . . ltldf6 25 �c2 :Z.axf7 26 �b3! ltlh6
27 ltlxf7 ltlxf7 28 l:e6 �g7 29 l:xb6
�c8 30 bxa5 ltlg5 31 f4 ltlh3+ 32 �g2
 4 g 3 i. b 4 + a n d 4
.:es 33 1:.c 1 l:l.e2+ 34 �f 1 J:U2 + ? ! 35
cJi>e 1 llle4 36 .:c2 1 -0
Game 35
Anastasian-Brodsky
Moscow 1992
1 d4 lllf 6 2 c4 e6 3 lllf 3 b6 4 g3 .i.a6 5
'i'c2!?
1bis idea is not particularly popular. The
main reason is probably 5 ... d5, which appears
to be a very easy route to equality.
5 . . . c5
ECO suggests this move but does not in­
clude 6 e4 in its world view.
a) After 5 ... d5! we have:
al) 6 b3 .i.b4+ 7 J.. d2 i.. xd2+ 8 lt:ibxd2
0-0 9 i.g2 c5 promises Black equality, and 1 0
e 3 lt:ic6 1 1 dxc5 dxc4 12 bxc4 lt:ib4 1 3 'itbl
lt:id3+ much more than that in Vacek­
Hausner, Bmo 1 969.
a2) 6 lt:ibd2 .i.e7 7 J..g2 0-0 8 0-0 c5 9
dxc5 (9 .l:tdl lt:ic6 1 0 li'a4 li'c8 1 1 dxc5 dxc4
12 lt:ie5 lt:ixe5 1 3 i.xa8 b5 14 'iia5 .l:td8 1 5
.i.g2 i.xc5 was crushing in Udovcic-Rakic,
Sombor 1957) 9 ... J..xc5 10 a3 lt:ibd7 1 1 li'a4
i.b7 1 2 cxd5 exd5 1 3 b4 i.d6 14 J.. b 2 a5 1 5
b 5 .l:te8 1 6 e 3 lt:ie4 and Black had sufficient
counterplay in Blom-Johannessen, Marianske
Lazne 1 96 1 .
a2) 6 cxd5 i s the most logical move:
6 ... exd5 7 J.. g2 J.. b4+ 8 lt:ic3 0-0 9 0-0 .l:te8!
and Black is ready to give up the dark-
. . . i. a 6 with o u t 5 b 3
squared bishop to concentrate on the light
squares. To be successful in this strategy he
needs to keep the bishop outside the pawn
chain, meaning on the b1 -h7 diagonal. Lil­
loni-Gereben, San Benedetto 1 957 continued
1 0 .l:te1 c5 1 1 a3 J..xc3 12 bxc3 lt:ibd7 1 3 e3
'i'c7 14 'iia4?! (doing nothing about solving
the most important problem -where does the
queen's bishop belong?) 14 ... J..d3 1 5 .i.b2
(hardly a dream spot for the bishop) 15 ...b5
16 li'dl i.e4 17 J.. fl c4 18 lt:id2 .i.g6 1 9
.i.g2 lt:ib6 2 0 f3 lt:ia4 and Black was better.
b) 5 ...lt:ic6!? tries to exploit White's lack of
development. I am not sure Black equalises
with this line, but it does result in an interest­
ing position, e.g. 6 a3?! (6 J..g2! must be the
right move - after 6 ...lt:ib4 7 'itb3 d5 8 lt:ie5
[8 cxd5 li'xd5 is already equal] 8 ... dxc4 9
li'dl ! [9 lt:ixc4 li'xd4 1 0 .i.xa8 J..xc4 1 1 'i'dl
'ii'xdl+ 12 'it>xdl J..c 5 is terrible for White]
9 ... b5 10 a4 c6 1 1 axb5 .i.xb5 1 2 lt:ic3 lt:ibd5
13 e4 lt:ixc3 14 bxc3 and White has a strong
centre to compensate for his pawn deficit).
Returning to 6 a3, play might continue
6 ... i.b7 7 'ii'd3 eS! 8 dxe5 lt:ixe5 9 lt:ixe5
i.xhl 10 f3 .i.d6 1 1 'ii'c3 'i'e7 1 2 lt:ig4 lt:ixg4
13 'ii'xg7 0-0-0 14 'ii'xg4 h5 15 'ii f5 .l:tdg8 1 6
i.f4 i.xf4 1 7 'i'xf4 h4 1 8 g4 f5 1 9 gx f5 .i.g2
20 f6 'ii'd6 21 'i'e4 c6 22 f7 .l:tf8 23 .i.xg2
°i'xh2 24 'ii'g4 h3 25 .tfl :Xf7 26 lt:id2 'ii'e5
27 .i.xh3 .l:tg7 28 'ii'e4 'ii'g3+ 29 'it>d1 'ii'xh3
30 'it>c2 .l:tg2 31 'ii'd 3, and Black had great
winning chances in Korchnoi-Kachiani
Gersinska, Willingen 1 999.
6 .i.g2
After this move Black can equalise. I am
less sure about 6 e4!, although it has been
played only once - 6 ... cxd4 7 lt:ixd4 i.b7 8
i.g2 a6 9 0-0 'ii'c7 10 .l:tel lt:ic6 1 1 lt:ixc6
'ii'xc6 1 2 b3 .i.e7 13 lt:id2 d6 1 4 .i.b2 0-0 1 5
'i'c3 appears to be very slightly better for
White. After 15 ....l:tfcS 16 g4 lt:ie8 1 7 'ii'g3
i.f6 1 8 J..xf6 lt:ixf6 1 9 g5 lt:id7 20 .l:tadl 'i'c5
21 lt:ifl White is better, and in Korchnoi­
Lau, Hessen 1 999, Korchnoi more or less
outplayed his opponent by advancing the h-
95
Q u e e n 's In dia n D e fe n c e
pawn.
6 . . . �e6 7 dxe5 .be5 a 0-0 0-0 9 i.g5 ! ?
:ea 1 O 'ii'a4 i.b7 1 1 �e3 h6 1 2 i.f4!
'ii'e 7 1 3 a3 a5?
13 ... .i.d6 would have kept the position
more or less equal. Now Black has problems.
14 llad 1 llfda 1 5 �e5
The b-file is a terrible source of weakness
for Black.
1 5 . . . i.aa 1 6 �d3!
White has a clear advantage.
1 6 . . . �h5 1 7 i.e1 i.d4
17 ... lCid4 1 8 .i.xa8 l:txa8 19 .i.e3!? lCif5 20
.i.xc5 bxc5 21 lCie4, with the idea g3-g4 is
given by Dautov. White is close to winning.
1 a �b5 �f6 1 9 i.f4 �ea 20 ild2 i.f6
21 llfd 1 d5 22 exd5 exd5 23 lle2 �a7
24 �xa7 'ii'x a7 25 llde 1 llxe2 26 llxe2
"ile7 27 'ii' b 5 'ii'e 6 2a h4 g 5 ! ? 29 hxg5
hxg5 30 i.e 1 .:tba 3 1 �e 1 ! �d6 32 'ii'd 3
.lbe4 33 �f3 �e5 34 'ii'd 1 °ilf5 35 i.e3
i.e6 36 lld2
Entering some complicated tactics. 36
.i.xc5? .i.a4 37 e4 .i.xc2 38 'ii'xc2 'i'xe4
would turn the tables, but 36 b31?, as sug­
gested by Dautov, might improve.
36 . . . i.a4 37 llxd5 "ilxf3 3a i.xf3 i.xd 1
39 llxd 1 i.xb2 40 i.xg5 i.xa3 4 1 i.f6!
Now White also has an attack on the king­
side to make him happy.
4 1 . . . �e6? ! 42 lld7! �fa 43 lla7 �h7 44
i.e5 Ilda 45 i.h5 f6 46 i.f7+ �ha 47
i.e3 :ea 4a i.d2 i.e 1 49 i.e6 Ilda 50
96
i.e3 i.h6 51 i.f5 i.g7 52 i.xh7 �xh7
53 i.xf6 :l.ga 54 �f1 1 -0
Game 36
Barlov-Beliavsky
Yugoslavia 1992
1 d4 �f6 2 e4 e6 3 �f3 b6 4 g3 i.a6 5
'ii' b 3
This line can at best be described as com­
pletely harmless. After...
5 . . . �e6!
... White is forced to protect e4 an addi­
tional time.
6 .lbbd2 d51
Black's forces are quickly finding a way
into the game.
7 "ila4
Perhaps it is better to simply play 7 cxd5
exd5 (7 . .'ii' xd5!? must also equalise) 8 'i'a4
.
.i.b7 9 .i.g2 .i.d6 10 0-0 0-0, e.g. 1 1 lCibl
lCie4 12 .i.f4 l:le8 13 l:tc1 .i.xf4 14 gxf4 lCie7
15 lCic3 lCif5 and Black had equalised in Gel­
fand-Leko, Monaco 2002.
7 . . . i.b7 a i.g2 i.d6 9 0-0
9 lCie5 looks good, but after 9 ... 0-0! 1 0
lCixc6 1i'd7 Black retains his piece wi th no
minuses whatsoever, and complete equality.
Chekhov-Bareev, Germany 1 992 continued
1 1 0-0 :fes 12 1i'c2 .i.xc6 13 b3 (1 3 :d1
:ad8 with equality, as in Grivas-Delchev,
Vama 1 994, is probably better) 1 3 ... e5! 1 4
.i.b2 dxc4 1 5 .i.xc6 'ii'xc6 1 6 1"xc4 1"b 7 1 7
 4 g 3 i. b 4 + a n d 4 . . . i. a 6 with o u t 5 b 3
:acl ?! (1 7 dxeS .txeS 1 8 ilxeS :xeS is ac­
tive and quite pleasant for Black) 1 7 ... bS! 1 8
'ir'c2 exd4 1 9 ltJ £3 cS 20 b4 cxb4 2 1 .txd4
ltJe4 22 :fdl aS 23 ltJh4 Ji.f8 24 £3 ltJc3 2S
.txc3 bxc3 26 e4 b4 27 :as :ed8 28 :cdl
:xdS 29 exdS :as 30 'ir'd3 :xdS and White
resigned.
9 . . . 0-0 1 0 a3 a5! ? 1 1 'ii c 2 a4 1 2 l:e 1
i.e7?
1bis move makes very little sense at all.
Correct is 12 ... eS 13 dxeS ltJxeS 14 ltJxeS
Ji.xeS 1 S ltJ£3 .td6 16 ltJd4 .tcS 17 ltJf5
:es! with an unclear game. White is not very
well co-ordinated to exploit the well placed
knight.
1 3 e3 'iid 7 1 4 cxd 5 exd5 1 5 liJb 1 !
The knight belongs on c3.
1 5 . . . ltJa5 ! ?
An interesting pawn sacrifice.
1 6 ltJe5 'ilfe8 1 7 fixc7 i.d8 1 8 fic2 liJb3
1 9 l:a2 :tc8 20 ltJc3 ltJe4 21 liJd3 i.f6
More accurate is 21...£5!, with the idea of
22 :a1 gS followed by ... f5-f4 etc.
22 J:.d 1 h5 23 i.d2!
White has the advantage.
23 . . . i.g5 24 i.e 1 h4 25 .Uaa 1 ! ?
As the a4-pawn also falls, this makes good
sense.
25 . . . ltJxa 1 26 l:xa 1 fie6 27 'iiid 1 .Ufe8
28 ltJxa4 'iih 6 29 ltJc3 J:.cd8 30 f4!
Black is strategically lost. His rooks have
no play.
30 . . . i.f6 3 1 ltJxe4 dxe4 32 liJf2 'i:ig6 33
'iig 4?!
33 g4! is simpler.
33 . . . 'ii'h 7 34 %:tc1 i.c8 35 ft'e2 hxg3 36
hxg3 'i:ig6 37 g4?
Allowing a strong combination. After 37
:c6, planning g3-g4, Black is busted.
37 . . . i.xd4! 38 exd4?
38 :xc8 :xc8 39 exd4 retains the better
chances. Now Black takes over.
38 . . . e3 39 liJd3 i.xg4 40 'iic 2 i.f5 41
fic6 i.xd3 42 'i'xg6 i.xg6 43 i.c3 l:.d6
44 'it>h2 i.f5 45 .Ue 1 l:.h6+ 46 'it>g3
l:.g6 + 47 'it>h2 %:tg4 48 d5 l:.xf4- + 49
'ii.og 3 l:.g4+ 50 'ii.of3 l:.g6 51 i.h 1 i.e4+
52 'it>e2 i.xh 1 53 J:.xh 1 f5 54 i.b4 l:.g2+
55 'it>d3 J:.xb2 56 d6 e2 57 l:re 1 f4 58
'it>c3 J:rxb4 0-1
Game 37
Barbero-Cebalo
Caorle 1987
1 d4 ltJf6 2 c4 e6 3 liJf3 b6 4 g3 i.a6 5
liJbd2
1bis is one of the main systems and
should offer White a modest chance for an
advantage.
5 . . . i.b4!? 6 'ii b 3
6 'ii'c2 Ji.b7 7 .lig2 aS 8 0-0 dS 9 ltJeS 0-0
10 cxdS exdS 1 1 ltJdf3 ltJe4 1 2 ii.e3 :es led
to equality in Porcisch-Bronstein, Las Palmas
1 972, as did 6 'ii'a4 cS 7 a3 ii.xd2+ 8 .txd2
cxd4 9 ii.g2 .tb7 10 0-0 ii.c6! 1 1 'ii'd l .txf3
1 2 .txf3 ltJc6 13 .tf4 0-0 14 Ji.d6 :es 1 S
ii.xc6 dxc6 1 6 'ii'xd4 c S 17 'ii'f4 ltJhS 1 8 'ii'e S
f6 19 'ii'xhS 'ii'xd6, Dizdar-Palac, Medulin
1 997.
6 . . . c5
Best. 6 ... ltJc6 7 dS .txd2+ 8 .txd2 ltJe7
(8 ... ltJa,? 9 'ii'a4! ltJe4 10 .txaS bxaS 1 1 dxe6
fxe6 12 .tg2 0-0 13 0-0 gave White a clear
advantage in Vaganian-Nogueiras, Montpel­
lier 198S) 9 .tc3 ltJf5 10 ltJd2 favours White.
Kasparov-Speelman, La Valetta 1 980 contin­
ued 10 ...ltJd6?! 1 1 £3! 0-0 1 2 e4 and White
was now clearly better. After 1 2.:.exdS 1 3
97
Q u e e n 's In dia n D e fe n c e
cxdS i.xfl 1 4 .l:.xfl aS 1 5 eS a4 1 6 'i'c2 'i'e8
1 7 �£2 lllxdS 1 8 'i'd3 'i'e6 1 9 exd6 the lead
had grown to being technically winning. An
improvement is 1 0... 'i'e7 1 1 0-0-0 exdS 1 2
cxdS llle4 1 3 lllxe4 'i'xe4 1 4 £3 'i'e3+ 1 5
i.d2 'ii'xb3 1 6 axb3 llld4 1 7 b4 with a plus
for White (Kasparov).
6 ...'i'e7 7 i.g2 i.b7 8 0-0 i.xd2 9 i.xd2
0-0 10 .l:.ad1 gave White a slight edge in Al­
exandria-Chiburdanidze, Thilisi 1 981 .
·1 i.g21
7 a3 i.xd2+ 8 i.xd2 lll c 6 9 dS (9 i.g2
cxd4 10 'i'a4 'i'c8 1 1 0-0 i.b7 12 b4 llle4
was a pawn in Nesis-Sanakoev, corr. 1 986,
while 9 dxcS bxcS 10 'i'c2 looks best, but
White has no chance of an advantage)
9 ... lllaS 1 0 Las bxaS 1 1 dxe6 fxe6 1 2 'i'c2
.l:.b8 13 i.h3 'i'b6 and Black has the better
prospects, Fedorowicz-Seirawan, South Bend
1 981 .
7 . . . t'tlc6
7 ... i.b7 8 dxcS i.xcS 9 0-0 0-0 10 'i'c2
i.e7 1 1 b3. d6 12 i.b2 lllbd7 was equal in
OU-Andersson, Tallinn 1 998. But the simple
8 0-0 looks enough for an edge.
8 dxc5
8 a3 transposes to the note to White's 7th
move.
8 bxc 5 ! ?
...
8 ...LcS 9 0-0 0-0 seems to equalise, e.g.
10 a3 i.e7 1 1 'i'a4 i.b7 12 e4 'i'c7 13 'i'c2
lll e S! 14 b3 lllxf3+ 1 5 i.xf3 i.d6 16 i.g2
.l:.ac8 1 7 i.b2 i.eS with equality, Tibensky-
98
Hracek, Slovakia 2000, or 1 0 'i'a4 i.b7 1 1
lllb3 i.e7 12 .l:.d1 'i'c8 1 3 i.f4 lllb4 14 llleS
i.xg2 1 5 @xg2 'i'b7+ 16 �g1 dS, Redze­
pagic-Popovic, Novi Sad 1 985.
9 0-0
9 . .1:1.bS?
..
The rook is misplaced here due to a future
i.f4. Both alternatives are better.
a) 9 ....l:.c8 1 0 'i'c2 0-0 is interesting. We
don't know much about it because 1 1 lllb3
lll d4 12 ltibxd4 cxd4 13 b3 dS 14 ltixd4 eS
1 5 a3 i.cS 1 6 ltJEi dxc4 17 bxc4 i.xc4 was a
display of poor opening play that left Black
in front in Dlugy-Ehlvest, Mazatlan 1 988.
But this looks like the route that Black
should pursue.
b) 9 ... 0-0 1 0 a3 i.xd2 1 1 i.xd2 dS 12 cxdS
exdS 1 3 l:fc1 was, potentially, slightly better
for White in Kempinski-Hansen, Hamburg
1 999.
1 0 'i'c2 0-0 1 1 t'tlb3! 'i'e7 ? !
1 1 ...lLlaS 1 2 .tf4! d6 1 3 lLlfd2 gives White
the advantage but is nonetheless preferable.
1 2 i.f4
1 2 a3! .taS 1 3 lllxaS ltixaS 14 'i'a4 ltib3
15 'i'xa6 ltixa1 16 i.f4 .l:.b6 17 'i'xa7 :Xb2
1 8 l:xal has been suggested as giving White
an advantage.
1 2 . . . e5
Forced. After 1 2 ...d6? 1 3 .l:.fd1 l:lfd8 1 4 a3
i.aS 1 5 lLlgS White wins a piece.
1 3 i.g5 .1:1.fcS? !
Imperative is 13 ... ltid4 14 lllbxd4 cxd4 1 5
 4 g 3 i. b 4 + a n d 4 .
'i'a4 .l:tb6 1 6 a3 i.d6 1 7 ti:Jd2! i.b8 1 8 b4 d6
1 9 bS i.c8 20 i.c6!, when White has some
positional pluses. Now he is winning.
1 4 a3 i.a5 1 5 ti:\xa5 ti:\xa5 1 6 ..W a4!
i.xc4 1 7 ..Wxa5 i.xe2 1 8 lUe1 i.xf3 1 9
i.xf3 h6 20 i.xf6 ..Wxf6 2 1 ..Wc3 d 6 22
i.d5 c4 23 .l:.e3 .l:.b3 24 ..Wd2 J:.xe3 25
fxe3 ..Wg5 26 .l:.c1 ..Wg4 27 ..Wf2 �h8 28
..Wxf7 ..We2 29 ..Wf2 ..Wd3 30 e4 a5 3 1
.l:.xc4 .l:.xc4 32 ..Wf8+ �h7 3 3 i.g8 + 1 -0
------.
Game 38
Salov-Karpov
Wy:k aan Zee 1998
1 d4 ti:\f6 2 c4 e6 3 ti:\f3 b6 4 g3 i.a6 5
tll b d2 i.b7
5 ... cS 6 i.g2 .tb7 is just a transposition,
but 6 ... ti:Jc6! is much stronger, as can be seen
in the next game.
6 .tg2 c5
6 ... i.. e7 7 0-0 0-0 8 b3 normally leads to
transposition after 8 ... dS.
7 e4!
Time has given this move the stamp of
approval. 7 dS!? exdS 8 ti:Jh4 g6! 9 cxdS i.. g7
1 0 tt:Jc4 0-0 1 1 ti:Jd6 (1 1 0-0 d6 1 2 i.f4 i.a6!
13 l:tcl i.xc4 14 .l:txc4 bS is fine for Black)
1 1 ...i.a6 1 2 a4 ti:Je8 1 3 tt:Je4 d6 1 4 0-0 ti:Jd7
15 .l:tb1 ti:Jdf6 led to equality in van Wely­
Psakhis, Leeuwarden 1 993.
7 . . . cxd4
Black cannot exploit the e-pawn's mo-
. . i. a 6 wi th o u t 5 b 3
mentarily weakness: 7 ...tt:Jxe4?! 8 ti:JeS! dS
(8 ... ti:Jc3?? 9 libs g6 10 'iih3) 9 'ii'a4+ ti:Jd7
1 0 cxdS ti:Jxd2 1 1 i.xd2 i.xdS 1 2 i.xdS
exdS 13 0-0 f6 14 l:tfe1 fxeS 1 5 .:r.xeS+ i.e7
1 6 i.gS?! bS! with a quick draw in Goldin­
Smirin, Moscow (GMA) 1 989, but Avrukh
recommends 1 6 .l:tae1 ! (with the threat of
i.gS), when 1 6 ...'it>f7 1 7 l:txdS! appears to be
decisive .
After 7 ... d6 8 dS exdS 9 cxdS i.a6 1 0 i.£1
'i'c8 1 1 'i'a4+ ti:Jfd7 12 i.xa6 'i'xa6 1 3 'i'xa6
tt:Jxa6 14 0-0 ti:Jb4 15 tt:Jc4 f5 16 exf5 ti:Jf6 1 7
i.gS ti:JbxdS 1 8 .l:tadl White was clearly bet­
ter in Sosonko-Miles, Wijk aan Zee 1 981 .
8 0-0!
The modem treatment - Black cannot
cling on to the pawn. Other options promise
nothing:
a) 8 eS tt:Je4! 9 0-0 ti:Jxd2 1 0 i.xd2 i.xf3
1 1 'i'xf3 ti:Jc6 12 'i'xc6 dxc6 1 3 i.xc6+ 'i'd7
14 i.xd7+ 'it>xd7 was completely equal in
Salov-Karpov, Wijk aan Zee 1 993.
b) 8 ti:Jxd4 i.cS! and now:
bl) 9 ti:J4b3 i.e7 10 0-0 'ii'c 7 1 1 .l:te1 d6
12 ti:Jd4 0-0 13 b3 a6 14 i.b2 ti:Jbd7 was
equal in Piket-Salov, Wijk An Zee 1 992.
b2) 9 ti:Jc2 'i'c7 10 0-0 i.e7 and now in
Agdestein-Hjartarson, Belgrade 1 989, White
should have maintained the balance with 1 1
b3. Instead 1 1 eS?! was too optimistic. After
1 1 ...i.xg2 12 exf6 i.xf1 13 fxe7 i.d3 14 'ii'f3
i.xc2 1 5 'i'xa8 'it>xe7 1 6 b3 l::td 8! 1 7 i.a3+
'it>e8 1 8 'i'f3 'i'eS! 1 9 l:tcl i.g6 Black had a
clear lead.
b3) 9 ti:JbS a6 10 ti:Jc3 'i'c7 1 1 'ii'e2 i.e7
with complete equality in Welin-Chemin,
Lugano 1 989.
8 . . . d6
The alternatives favour White:
a) § . . . tt:Jxe4?! has never been played but
has been analysed by Cifuentes, among oth­
ers. After 9 ti:JeS dS 1 0 cxdS exdS 1 1 'ii'a4+
ti:Jd7 1 2 tt:Jxe4! dxe4 1 3 i.h3 bS (1 3 ... i.c8 1 4
'ii'c 6 .l:tb8 1 5 ti:Jxf7! 'i'f6 1 6 i.e6! .l:tg8 1 7
i.gS and White wins) 14 'ii'xbS i.c8 1 5 'i'c6
l:tb8 1 6 ti:Jxf7! White wins. Black's best op-
99
Q u e e n 's In dia n D e fe n c e
tion seems to be 9 ... .td6 10 &Dxf7 �xf7 1 1
&Dxe4 .txe4 1 2 he4 &Dc6 1 3 1W£3+ <ii>g8 14
.txc6 dxc6 1 S 'i'xc6 l:tc8 1 6 'iWe4, although
White still has a pleasant structural advan­
tage, while 9 ... &Dd6 10 i..xb7 &Dxb7 1 1 'iWf3
does not work out well...
b) 8 ... &Dc6?! 9 eS &Dg4 1 0 h3 &Dh6 11 &De4
gives White a clear advantage.
c) 8 ... .tcS?I 9 eS &De4 1 0 &DgS! lLic3 1 1
bxc3 .txg2 1 2 coftxg2 'iWxgS 1 3 tLl f3 is given
as clearly better for White by Ftacnik. In fact
13 cxd4! looks even stronger...
9 lL!xd4 a6! ?
Here there are a number o f alternatives
but it would seem to be an enormous waste
of energy trying to memorise them all. White
has a simple plan in a2-a4-a5 followed by a
knight transfer to aS. That is really all you
need to know, and that ... &Dbd7 often runs
into e4-eS after l:te1 , as we shall see.
9 ... &Dbd7 10 :et (1 0 a4 &Des is premature)
1 0 ... 'i'c8 1 1 a4! a6 12 aS eS 1 3 lLif5 g6 1 4
&De3 .te7 1 S axb6 0-0 1 6 b 4 was poor for
Black in Oll-Kengis, Riga 1 99S. 12 ... .:.b8 has
been suggested as an improvement, but to
me it still looks as if White is a whole lot
better. 10 ... .te7?! has also been investigated,
1 1 eS i..xg2 12 exf6 .txf6 1 3 lLixe6 fxe6 14
�xg2 &Des 1 S &De4 giving White a clear ad­
vantage according to Avrukh. Browne­
Bradford, Dallas 1 996 continued 1 0 ... 'iWc7?!
11 &DbS 'iWb8 12 cS! (freeing the c4-square
for the knight) 1 2 ... dxcS 1 3 lLic4 eS 14 f4
1 00
.tc6 1 S lLic3! bS 1 6 fxeS and White had a
powerful initiative. Finally 1 0 ... eS!? 1 1 &DbS
a6 12 lLic3 i.. e7 13 lLif1 l:tc8 14 lLie3 0-0 1 S
a4 was slightly worse fo r Black in Cramling­
Chiburdanidze, Groningen 1 997, although
Black's position remains solid.
1 0 l:.e 1 •c7
10 ... lLibd7 11 eS! i..xg2 12 exf6 i..b7
(1 2 ...'iWxf6 13 lLixe6 fxe6 14 coftxg2 &Des 1 S
&De4 lLixe4 1 6 l:txe4 wa s clearly better for
White in Browne-Burger, Philadelphia 1 990)
1 3 fxg7 i..xg7 14 &De4 0-0? (1 4 ... i..xe4 1 S
.:.Xe4 0-0 1 6 l:tg4 gives White an edge) 1 S
&Dxd6 lLieS 1 6 lLixb7 'iWxd4 1 7 'iWxd4 lLif3+
18 �f1 lLixd4 1 9 l:td1 l:tab8 20 &Dd6 l:tfd8 21
&De4 f5 22 lLic3 and White was winning in
Tregubov-Grooten, Amsterdam 2001 .
1 1 a4!
The reason why this is possible here and
not in other hedgehog positions is the knight
on d2 and the excellent support it adds to the
c-pawn. White is also more flexibly placed to
push with b2-b4.
1 1 lL!c6
. . .
11 ... lLibd7 12 aS i.. e 7 13 &D2b3 l:td8 14
axb6 tLlxb6 15 tLlaS was better for White in
Bonsch-Chuchelov, Berlin 1 996.
1 2 lL!xc6 .bc6 1 3 a5 l:.bS?!
This is probably because Karpov over­
looked White's 1 Sth move. Now Black is in
trouble. 13 ... i.. e 7, developing, is necessary.
1 4 axb6 •xb6
1 4 ... l:txb6 leaves the rook awkwardly
 4 g3 i. b 4 + a n d 4 . . . i. a 6 wi th o u t 5 b 3

placed without solving the problem of the 22 c5 ilc7

a6-pawn. After 1 5 b3, with the idea J.b2-d4, 22 ... fib5 23 °it'xb5 axb5 24 J.xe5 and
Black has problems and is still not devel­ White wins.
oped. 23 i.xe5 dxe5 24 i.xb7 fl.xb7 25 c6
1 5 fl.e3! ltld7 ltlb6 26 ilxa6 ilxc6 27 ltlc4!
Now Black is in trouble as developing is
not possible. Note that 15 ... J.e?? 16 l:.b3
fie? 17 .l:.xb8+ "ilxb8 1 8 e5 .i.xg2 1 9 exf6
wins for White.
1 6 fl.b3 ila7
Or 16 ...fic? 17 :xa6 l:.xb3 18 itJxb3 'ifb7
1 9 l:.al ..ixe4 20 ltJa5 "ila8 21 ..ixe4 �xe4
22 J.f4 and the lead in development is again
a key factor.
1 7 fl.ba3
White exerts maximum pressure on the
primary weakness.
1 7 . . . i.b7
17 ... ltJc5 18 itJb3! ltJxb3 19 l:.xb3 :Xb3 Winning a piece and the game. The threat
20 "it'xb3 J.e7 21 "it'a2 and the a-pawn is is ltJa5.
difficult to protect without allowing 'it'a4+. 27 . . . fl.b8 28 ltlxb6 0-0
1 8 b4 "ilb6? ! 28 ... :Xb6 29 Wa8+ and White wins a
Black is wasting time. 1 8 ... J.e? has to be rook.
tried, although White has many strong op­ 29 ltlc4 ile4 30 fl.e3 °ild 5 3 1 ltlxe5 fl.fc8
portunities. 32 °ild3 1 -0
1 9 ila4! i.e7 20 i.b2 i.f6?
The lesser evil 20 ... e5 is tantamount to po­ Game 39
sitional resignation. Tregubov-Shaposhnikov
Samara 2000
1 d4 ltlf6 2 c4 e6 3 lLlf3 b6 4 g3 i.a6 5
ltlbd2 c5

21 e5! !
Fully exploiting the lead in development.
21 . . . i.xe5
21...dxe5 22 :d3! and Black will never es­
cape as c4-c5 is coming. In my opinion this line is unnecessarily

101
Q u e e n 's In dia n D e fe n c e
risky for Black.
6 e4
The only serious option. 6 i.g2 lDc6 7
lDe5!? (White can also play safe, but he will
gain no advantage) 7 ... lDxd4 8 e3 lDfS 9 'il'a4
'il'c8 10 i.xa8 'il'xa8 1 1 .l:r.g1 i.c8 1 2 b3 lDe4
1 3 i.b2 lDxd2 1 4 �xd2 i.e7 1 5 .l:f.adl f6 1 6
lDd3 i.b7 gave Black the better chances in
Timman-Dautov, Forchheim 2000.
6 cxd4 7 e5!
•••
White needs to act now to gain an advan­
tage. 7 i.g2?! is a waste of time. Remember
that White has sacrificed the d4-pawn! von
Hennan-Hiibner, Altenkirchen 1 999 went
7 ... lDc6 8 e5 lDg4 9 0-0 :b8 1 0 .l:f.e1 i.c5 1 1
h3 lDe3, and White resigned.
7 lbg4 8 h3 lbh6
. . •
Now the knight is worse off and White
can take time to develop.
9 -*.g2 lbc6 1 0 0-0 lbf5
10 ... i.e7 1 1 a3 0-0 1 2 b4 i.b7 13 lDe4
lDxb4? (1 3 ... lDfS!? is better, but White ap­
pears to be well placed) 14 lDf6+I i.xf6 1 5
exf6 lDc6 1 6 i.g5 'il'c7 1 7 i.xh6 gxh6 1 8
'il'd2 �h8 1 9 'il'xh6 .l:r.g8 20 lLJg5! .l:f.xg5 21
i.e4 .l:f.xg3+ 22 fxg3 'il'xg3+ 23 �h1 and
Black resigned in Hellsten-Astrom, Ronneby
1 998.
1 1 a3!
.
Menacing. 11 'il'a4 has also been played
(Hellsten-Shipov, Gistrup 1 997 being one
example) but never with much success.
White has to realised that he is well placed
1 02
and that the pawn on d4 is very difficult for
Black to hang on to, so he should just play
around it.
1 1 -*.e7
. . •
This move has been given a ?! by one
commentator but no improvement has been
suggested. The following lines demonstrate
the problems faced by Black:
a) 1 1 ... l:c8?! 1 2 l:e1 i.e7 1 3 b4 i.b7 1 4
lDe4! and Black i s in trouble a s h e cannot
develop freely. After 1 4... 0-0 comes 1 5 g4!
lDh4 1 6 lDxh4 i.xh4 1 7 lDd6 .l:f.c7 1 8 lDxb7
.l:f.xb7 1 9 b5 lDa5 20 i.xb7 lDxb7 21 'il'xd4
f6 22 l:a2 and the compensation for the ex­
change is not apparent.
b) 1 1 .. .'il'c7 1 2 b4 i.b7 1 3 lDe4 h5 1 4 i.f4
a5 1 5 b5 lDd8 1 6 lDxd4 'il'xc4 1 7 lDxfS exfS
1 8 :et 'il'e6 1 9 lDg5 gave White a clear ad­
vantage in Sherbakov-Galliamova, Cheliab­
insk 1 989.
c) 1 1 ...b5 12 g4 lDh4 13 lLJxh4 'il'xh4 14
tDn 'il'd8 15 lDxd4 lDxd4 16 'il'xd4 looks
good for White (Gershon).
1 2 b4 -*.b7 1 3 lbe4!
This is the most logical move. Before
White takes any action he improves his
pieces to the maximum. 1 3 g4 has proved to
be less dangerous.
1 3 d5
• • .
Here Black has a senous alternative in
1 3 ... 0-0!? 14 g4 d5 (only move) 1 5 cxd5! exd5
1 6 lDed2 (1 6 lDeg5 lDh4! is only in Black's
interest; the knight is misplaced on g5)
1 6 ... lDh4 1 7 b5 lDa5 1 8 lDxh4 i.xh4 1 9 tDn
i.e7 20 lDxd4 and White has an appealing
position, but Black has some good things
going for him, too. The c4-square is a
strongpoint for the knight and the c-file will
also come in handy. However, I believe that
the potential storm on the kingside is very
serious, and White should be optimistic.
1 4 cxd5 'ii'x d5
After this White gets time to irritate Black.
Preferable is 14 ... exd5 1 5 lDed2 a6I (directed
against b4-b5) 1 6 lDb3 .l:f.c8!, and now 1 7 g4?!
lLJh4 1 8 lDxh4 i.xh4 1 9 lDxd4 lDxd4 20
 4 g 3 j,, b 4 + a n d 4 .
°flxd4 l:tc4 gives Black a strong llllttative.
Instead after 17 .ib2 1 7...0-0 1 8 g4 tLlh4 1 9
lLixh4 .ixh4 White has only a slight edge.
1 5 l:.e 1 d3?!
Black is in trouble and now self-destructs.
After 1 5 ... 0-0 16 g4! tLle3 (1 6 ... tiJh6 17 .ixh6
gxh6 18 °fld2 looks very dangerous for
Black) 17 fxe3 "flxe4 18 exd4 "fldS 19 i.e3
White is slightly better.
1 6 j,, b 2 0-0-0?
This seems completely ridiculous.
1 7 g4 tt'ih6 1 8 tt'itd2 'iWb5??
Losing. After the necessary 18 ... °fld7 19
tLlc4 'ifi>b8 20 lLied6 Black is still in trouble.
1 9 tt'ic3! 1 -0
White wins material.
-------
Game 40
Lautier-Karpov
Monte Carlo (rapid) 2000
1 d4 tt'if6 2 c4 e6 3 tt'if3 b6 4 g3 j,, a 6 5
tt'ibd2 d5
1bis is the main reason why 5 lLibd2 is
not such a popular choice.
6 j,,g 2 j,, e 7 7 0-0 0-0 8 lt'le5
Here White could transpose with 8 cxd5!
exd5 9 tLle5. Other options causes no prob­
lems for Black. With the move order in the
game Black has some additional options.
In Aagaard-Gavrikov, Gothenburg 1 998 I
discovered that the queen is not well placed
after 8 "flc2: 8 ... .ib7 9 b3 a5 1 0 i.b2 tLla6 1 1
. . j,, a 6 with o u t 5 b 3
:acl a4 and Black was already doing fine.
8 b3 c5 9 dxc5 i.xc5 10 i.b2 tiJbd7 1 1
l:tcl i.b7 1 2 lLid4 l:tc8 1 3 tLlb5 °fle7 led to
equality in Piket-Karpov, Wijk aan Zee 1 998.
8 . . . j,, b 7!
1bis move, followed by ... a 7-a5 and
...tLla6, is the main option for a good reason.
With the knight on d2 White has to play b2-
b3 and i.b2 to develop his dark-squared
bishop, but cannot meet ... a5-a4 very well.
However, there is another way to go, namely
8 ... c5!?, e.g. 9 cxd5 exd5 1 0 b3 l:te8
(10 ... .ib7! followed by ...tLla6 is the right
path; Black does not play the opening alto­
gether well here) 1 1 .ib2 cxd4?! 1 2 tiJdf3!
(standard in these lines) 12 ... .ic5 13 a3 lLibd7
14 lLid3! and White has a very clear advan­
tage, Polugaevsky-Rodriguez, Sochi 1 988. 1 0
dxc5!? i s interesting. Black should probably
not play 10 ... bxcS, when White was better in
Romanishin-Psakhis, Moscow 1 983 after 1 1
b3 .ib7 1 2 .ib2 aS 1 3 e4! d4 1 4 tLlec4 lLifd7
1 5 l:.el tiJb6 1 6 lLixb6 °flxb6 1 7 tLlc4. In­
stead 10 ... .ixcS! 1 1 lLid3 (heading for f4)
1 1 ...l:.e8 1 2 lLif3 ll'ic6 secures equality. Earlier
9 tLlb3 is possible, e.g. 9 ... .ib7 1 0 dxc5 bxc5
1 1 cxdS i.xdS 12 .if4 "flb6 13 "flc2 i.xg2
1 4 �xg2 tLlc6 1 5 lLixc6 "flxc6+ 1 6 'iti>gl tiJd5,
which was perhaps slightly preferable for
White in Jusupov-Polugaevsky, Toluca 1 982.
9 cxd5!
Clearly the most dangerous option. Alter­
natives:
1 03
Q u e e n 's In dia n D e fe n c e
a) 9 'ii'a4. What the queen i s supposed to
do out here is not altogether clear. Kempin­
ski-Nisipeanu, Krynica 1 998 continued
9 ... .id6 10 cxd5 .ixd5 1 1 e4 .ib7 1 2 tDec4
.ie7 with a good opening for Black, although
it was about to get even better: 1 3 e5?! .ixg2
14 exf6 .ixf6 1 5 'it>xg2 'ii'ds+ 1 6 tlJ£3 b5 and
Black had a clear advantage.
b) 9 b3 c5! 10 dxc5 bxc5 1 1 i.b2 t2Jbd7
12 t2Jxd7 'ii'xd7 with equality, 011-Khalifman,
Pamu 1 996. After 1 0 .ib2 cxd4! a difference
from positions where cxd5 exd5 has been
played is that White cannot delay the recap­
ture - 1 1 .ixd4 t2Jc6 was equal in Uhlmann­
Stahlberg, Prague 1954. Note that 10 cxd5
does not transpose to the main lines because
after 1 0 ... .ixd51 Black has already equalised.
c) 9 'ii'c2 c5 10 cxd5 exd5 1 1 dxc5 .ixc5
12 t2Jdf3 t2Jbd7 13 .if4 :e8 was also level in
Sutter-Kindermann, Bem 1 995.
9 . . . exd 5
The interesting 9 ... .txd5!? 1 0 e4 .ib7 1 1
'ii'c2 (1 1 'ii'a4 'ii'e8 1 2 'i'c2 tDc61, with an
even game, is better) 1 1 ...t2Ja6 1 2 l:.dl 'i'e8?
was seen in Dizdar-Granda Zuniga, New
York 1 997. After 12 ...'i'xd4! 1 3 tDdc4 'i'c5
14 .ie3 'i'b5 it is clear that White has genu­
ine compensation for the pawn. A GM once
told me that the most important thing he had
learned from Fritz was that you should take
pawns when it seems dangerous but you
cannot actually see any real danger. If you
look through Tal's old games you will find
that he bluffed half his way to the World
Championship title.
1 0 b3 a5
Here Black can also try 10 ... c5 t t .ib2
t2Jc6!? t2 tDxc6 .ixc6 t3 :et .ib7 with
counterplay according to Dolmatov. Instead
t t ...t2Jbd7 t2 t2Jxd7 'i'xd7 t 3 dxc5 bxc5 t 4
e4 seemed slightly better for White in Topa­
lov-Gelfand, Monte Carlo t 999, while
t 1 ...tDa6 t 2 :et t:fJc7 1 3 dxc5 .ixc5 t4 l:iJd3
.id6 t 5 t:fJ£3 tDe4 t 6 l:iJd4 was a good deal
better for White in Goldin-Ulibin, Kazan
t 995.
1 04
1 1 -.c2
Again I do not like this move at all. White
should concentrate on developing his minor
soldiers first before sending the queen into
battle. t t .ib2 l:iJa6 t 2 l:.ct c5 1 3 l:iJd3! (an­
other standard manoeuvre in such positions)
t 3 ...:e8 t4 l:iJf4 was played in Khalifman­
Leko, New Delhi 2000. Such situations are
what the main lines of the Queen's Indian
Defence are all about. White might have a
small plus and can manoeuvre around, trying
to create something more, while Black will
try to equalise or use any chance that might
be given him to get some kind of an initia­
tive.
1 1 . . .ltia6 1 2 i.b2 c5 1 3 dxc5 bxc5 1 4
.:tad 1 ? !
This move is rather careless. The only
black piece that is not really well placed is the
queen's knight, and now Black is allowed to
bring it back into the game. 14 a3 'i'b6 1 5
:adt a4 1 6 bxa4 l:.fc8 t 7 l:[bt 'i'a7 results in
a kind of dynamic equilibrium.
1 4 . . . a4!
Black is activating all his pieces this way.
1 5 bxa4 ltib4 1 6 -.b 1 -.c7 1 7 e4
Black is about to advance his c-pawn and
White uses the moment to claim squares for
his knights. t 7 a3 t:fJc6 t 8 tDxc6 .ixc6 also
gives Black the advantage .
1 7 . . . d4 1 8 a3 ltic6 1 9 ltiec4 .Z:.xa4 20
.Z:.fe 1 ? !
Inviting Black to exploit hi s more active
 4 g 3 j_ b 4
pieces. It was better to try to disturb Black
with 20 'it'c2 l:.fa8 21 f4!?, when White's
forces are not easily kept out.
20 . . . ltia5!
Now this is possible. After this the de­
fence of a3 is more difficult for White.
21 ltixa5
21 'it'c2 �xc4 22 �xc4 l:.fa8, followed by
...�d7-b6 with advantage.
21 . . . :1.xa5
White has no counterplay and Black has
the c-pawn as a strong force and the a-pawn
to put pressure on.
22 "ii' c 2 "ii'c 6 23 ltic4 :l.a7 24 f4 ltid7 25
"ii'd 2 ltib6 26 :l.c 1 ltixc4 27 .l:.xc4 :I.dB 28
J:.ec 1 "ii' b 6 29 a4 j.a6 30 a5 "ii' b 3 3 1
J:txc5 i.. x c5 32 J:.xc5 :l.b7 3 3 i.. xd4 "ii'a4
34 �f2 :l.b4 0-1
Game 4 1
Sakaev-Berzinsh
Duisburg 1992
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 g3 i.. a 6 5
"ii'a4
1bis line can easily be compared with 5
�bd2. It is by no means dangerous for Black
from a theoretical point of view but there is
still plenty of play in the position, allowing
for innovations and improvements.
5 . . . c6
1bis line is somewhat dubious. The supe­
rior 5 ... c5 is considered in the following
+ and 4 . . . j,, a 6 with o u t 5 b 3
games.
6 ltic3 b5!?
The point. Passive play with 6 ... .i.e7 7
i.g2 0-0 8 0-0 d5 9 �e5 1i'c8 1 0 cxdS cxd5
1 1 i.f4 favoured White in Hubner-Bauer,
Leon 2001 .
7 cxb5 cxb5 8 ltixb5 "ii'b 6 9 tbc3 i.. b4
Black also fails to equalise after 9 ...�c6 1 0
i.g2 .i.b4, e.g. 1 1 0-0 .i.xc3 1 2 bxc3 .i.xe2
13 l:.e1 "ii'a6 14 'it'xa6 i.xa6 15 �e5 l:.c8 1 6
.i.xc6! dxc6 17 .i.a3, when White had a last­
ing advantage in Nikolic-Ljubojevic, Monte
Carlo 1995.
1 0 j_g2 0-0 1 1 "ii'd 1 !
1bis is the dangerous move for Black. Af­
ter 1 1 0-0 i.xc3 12 bxc3 i.xe2 1 3 l:.e1
.i.bS!? 14 "ii'b 3 dS (Black must prevent c3-c4:
White stood better in Skembris-Timman,
Corfu 1 993 after 1 4... .i.c6?! 1 5 c4! �a6 1 6
i.f4!) 1 5 �e5 l:.c8 16 i.e3 �bd7 1 7 �4
�£8 1 8 a4 .i.a6 19 'it'xb6 axb6 with equality
in Vukic-Bronstein, Sarajevo 1 971 . Mean­
while in Tukrnakov-Gulko, Leningrad 1 977,
11 "ii'c2 �c6 12 0-0 :fc8 1 3 a3 .i.xc3 14
bxc3 �dS 15 :d1 �ce7 16 .i.b2 l:.ab8 1 7
l:.abl "ii'b3 gave Black enormous compensa­
tion for,,the pawn.
1 1 . . .:ca
1 1 ....i.xc3+ 12 bxc3 :cs 13 .i.d2 dS 1 4
0-0 �e4 1 5 l:.c1 leaves Black with compen­
sation for the pawn, but perhaps not enough.
1 2 i..d 2 d5 1 3 0-0 i.. xc3 1 4 i.. x c3 ltie4
1 5 :l.c 1 ltid7 1 6 :l.e 1 !
1 05
Q u e e n 's In dia n D e fe n c e
White is taking his time to untangle. After
16 ttld2 ttlxc3 1 7 bxc3 l:.c7 Black will exert
pressure on c3, and 1 8 c4 'i'xd4 19 cxdS
l:txc1 20 'ii'x c1 :lc8! 21 'i'd1 exdS is level.
1 e . . . :tab8
16 ... l:.c7 17 'i'a4! and Black will find it
hard to improve his situation due to the
threat of �aS.
1 7 'i'a4 .tb5 1 8 'i'a3 'i'b7 1 9 .b5! :txe 1
20 :txe 1
White has more or less untangled.
20 . . .:te8
After 20 ... �xe2 21 l:tc7! °WWbS 22 'i'b3!!
'i'xb3 23 axb3 ttlf8 24 b4 a6 25 ttleS White
loses his pawn but, suddenly, his pieces to­
tally dominate Black and he can hope for a
successful result.
21 :txe8+ 'i'xe8 22 .tf1 'i'e 1 23 <;Pg2
'i'd 1 24 'i'b3 .txe2 25 'i'xd 1 .txd 1
Again Black has won back his pawn, but
at a price. White has the two bishops and an
extra pawn on the queenside, so the ending
is, in fact, dreadful for Black.
2e .tb5 ©be 27 b3 .te2 28 ©e5 . ©de 29
.tae .tb1 30 ©ee .txa2 3 1 ©xa7 ©d7
32 b4 e5 33 .te7 ©e4 34 ©ee exd4 35
©xd4 .tb1 3e .tb7 ©dbe 37 .tee 'iW8
38 f3 .td3 39 �f2 �e7 40 .tb5 ©d7 4 1
.tee ©ebe 42 �e3 .t b 1 4 3 © b 5 .ta2 44
�d4 �ee 45 ©e3 .tb3 4e ©e2 g 5 47
�e3 .te4 48 ©d4+ �e7 49 b5 ©a4+ 50
�b4 ©ae5 51 ©t5+ <;Pee 52 ©d4+ �e7
53 �a5 te 54 ©t5 + �ee 55 ©d4+ �e7
1 06
5e be ©ae 57 ©f5 + �ee 58 ©d4+ '3iie 7
59 ©b5 ©db8 eo .tb7 ©e5 e1 .tde+
1 -0
Game 42
Karpov-Pol gar
Buenos Aires 200 1
1 d4 ©te 2 e4 ee 3 ©f3 be 4 g3 .tae 5
'i'a4 .tb7 e .tg2 e5 7 dxe5
This is practically the only line played
these days. After 7 0-0 cxd4 8 ttlxd4 ifc8 9
ttlc3 .txg2 10 �xg2 'Wib7+ Black has an
improved version of the hedgehog system, as
with little space each exchange is normally an
advantage. Nikolic-Karpov, Monte Carlo
2000 continued 1 1 f3 �e7 1 2 l:.d1 a6 13
.tgS h6 14 �xf6 �xf6 1 5 ttle4 .te7 1 6 'i'b3
ifc7 17 l:tac1 0-0 1 8 f4 ttlc6 1 9 ttlf3 l:tfd8 20
'i'e3 l:tac8 21 g4 dS and Black was okay.
7 . . . bxe5!?
More usual is 7 ... �xcS, as can be seen in
the two following games. After the text Black
is rock solid but has few chances of an ad­
vantage.
8 0-0 .te7 9 ©e3 0-0 1 0 .tf4 'i'be 1 1
:tfd 1 :td8
After 1 1 ...d6 12 l:.d2 Black has an extra
option in 1 2 ... ttlc6!? 1 3 .txd6 �xd6 14 l:.xd6
'i'xb2 1 5 l:tb1 'ii'xc3 1 6 l:txb7 ttld4 1 7 'i'd1 ,
as in Karpov-Polugaevsky, Biel 1 990. Now
17 ... l:.ad8! would have kept the balance. No­
tice it is important to use the appropriate
 4 g 3 j,, b 4 + a n d 4 . . . j,, a 6 with o u t 5 b 3
rook, as 1 7 ... l:f.fd8 gives White the additional
option of 1 8 l:f.xa7!?, with the main point
being 1 8 ... 'i*'xc4?? 1 9 ll:ie5! etc. What worries
me most here is 1 2 'iWb3! l:f.d8 with transposi­
tion to the main game.
1 2 'i'b3!
12 l:f.d2 d5 13 cxd5 exd5 14 'i*'c2 ttJ a6 1 5
a3 h 6 1 6 i.e5 d4 1 7 ll:ia4 'it'b5 1 8 b4? cxb4
1 9 ll:ixd4 l:f.xd4! and Black was winning in
Olafsson-Bareev, Hastings 1 990. Instead 14
ll:ie5 ll:ia6 1 5 l:f.ad1 'it'e6 1 6 ll:ic4 ll:ib4 1 7 a3
i.c6 1 8 'it'a5 d4 1 9 axb4 i.xg2 20 �xg2
cxb4 was equal in Tukmakov-Aseev, USSR
1 989, as was 14 e3 ll:ia6 15 l:f.ad1 h6 16 h4
l:f.ac8 in Sturua-Arkhipov, Moscow 1 992.
1 2 . . . d6
12 ...'it'xb3 is to premature in view of 1 3
axb3 d 6 1 4 ll:ib5!, forcing concessions.
1 3 l:d2
White is improving his position. 13 'i*'xb6
axb6 14 ll:ib5 tLle8 15 tLlg5 i.xg2 1 6 �xg2
tLlc6 gave White nothing in Mirzoev-Dautov,
Istanbul 2000, but 14 l:f.d2! improves, with a
chance to gradually develop an initiative.
1 3 . . . tl\a6
13 ... 'it'xb3 14 axb3 ll:ic6 15 l:f.ad1 tLle8, as
in Olafsson-Lautier, Antwerp 1 998, is possi­
bly enough for equality. But one tends to ask
what the rook is really doing on d1 besides
attacking the d6-pawn.
1 4 'i'xb6 axb6 1 5 l:ad 1 tl\e8 1 6 h4 h6
1 7 tiJh2 j,, x g2 1 8 'it>xg2 tl\ac7 1 9 e4 'it>f8
20 b3 .:l.dbS
Seems too optimistic; perhaps Black failed
to appreciate the implications of his oppo­
nent's reply.
21 e5! d5 22 cxd5 exd5 23 tl\xd5 tl\xd5
24 l:xd5 .:1.xa2 25 l:d7 !
Now th e white pieces are active, which
cannot be said of Black's. In fact Black is in
trouble.
25 . . ..:l.aa8 26 tl\g4 .:1.dS 27 tl\e3 l:l.ac8 28
tl\f5 .:l.xd7 29 l:xd7 j,,d 8 30 l:l.b7 j,, c 7 3 1
h 5 l:l.d8 3 2 tl\e3 l:l.d3 3 3 e 6 fxe6 34
j,, x c7 l:l.d7 35 1:Xb6 l:l.xc7 36 l:xe6 l:l.b7
37 I:.e5 l:l.xb3 38 l:.xc5 tl\f6 39 g4 .:l.a3
40 l:l.cS + 'it>f7 41 tl\c4 I:.a7 42 f3 'it>e6 43
J:tc5 tl\d5 44 .:!.c6+ 'it>e7 45 tl\e5 tl\f4+
46 'it>g3 I:.a4 47 lfo2 'it>f6 48 tiJd7 + <3ie7
49 tl\c5 l:l.d4 50 l:l.a2 'it>f7 51 l:a 7 + 'it>f8
52 tl\e4 tl\e6 53 I:.a6 <3ie7 54 <3if2 I:.c4
55 'it>e3 l:l.c 1 56 l:l.a7 + l:l.c7 57 l:l.a4 l:l.b7
58 l:l.a3 l:l.c7 59 tl\g3 <3if6 60 tl\e4+ 'it>e7
61 tl\c3 l:.d7 62 f4 tl\c5 63 tl\e2 <3if7 64
tl\d4 tl\e6 65 tl\f3 l:.b7 66 tl\e5 + <3if6 67
tiJd3 <3ie7 68 'it>e4 I:.d7 69 <3if5 .l:.d6 70
l:a7 + 'it>f8 71 tl\e5 tiJd4+ 7 2 'it>e4 tl\e6
73 tl\g6 + 'it>gS 74 <3ie5 I:.b6 75 .:l.e7 1 -0
Game 43
Borges Mateos-Spraggett
Cienfuegos 1997
1 d4 tl\f6 2 c4 e6 3 tl\f3 b6 4 g3 j,, a 6 5
'i'a4 j,, b 7 6 j,, g 2 c5 7 dxc5 j,, x c5 8 0-0
0-0 9 tl\c3 j,,e 7!
The bishop is not well placed on c5, a
square that is more suited for the knight.
1 0 l:.d 1 a6! ?
This has become fashionable over the last
couple of years. Black is trying to establish a
hedgehog position. 10 ...tLla6 - the main line ­
is considered in the following game.
1 1 j_f4 d6 1 2 l:.d2
With this move White is trying to put d6
under pressure but, as we shall see, Black can
easily protect it. The alternative is 1 2 l:f.ac1
'it'c7:
a) 1 3 �c2 l:f.d8! (1 3 ... ll:ibd7 1 4 'it'd2 tLle8
107
Q u e e n 's In dia n D e fe n c e
is a standard Hedgehog but Black wants the
rook on the other side of the knight) 14 'i'd2
ll:ie8 1 5 i.gS i.xgS 16 'i'xgS ll:id7 1 7 b3
:ac8 18 'i'e3 ll:ief6 with equality, Jakobsen­
Hansen, Greve 2002.
b) 13 'i'b4 (more dangerous than 13 'i'c2)
1 3 ... ll:ie8 14 i.gS (14 ll:ia4 ll:id7 1 5 'i'd2 ll:ic5
16 ll:ic3 ll:ie4 1 7 ll:ixe4 i.xe4 was fine for
Black in Hansen-Onischuk, Bundesliga 1 995)
14 ... ll:if6 15 i.xf6 gxf6 16 li:idS!? (forcing a
draw, but it seems that White has no advan­
tage anyway) 1 6 ... exdS 1 7 cxdS 'i'd8 1 8 ll:id4
..ti>h8 1 9 ll:ic6 'i'c7 20 li:id4 'i'd8 21 ll:ic6
'i'c7 22 ll:id4 and a draw was ag-reed in lvan­
chuk-Lautier, Moscow 200 1 .
1 2 . . . 'ii'c 7 1 3 :ad 1 :ds 1 4 ltig5
Others:
a) 14 ll:ie 1 i.xg2 15 ll:ixg2 ll:ic6 16 i.gS
ll:iaS 17 ll:ie3 h6 1 8 i.xf6 i.xf6 19 ll:ie4 i.e7
20 :c2 'i'b7 with equality in Bonsch­
Stempin, Polanica Zdroj 1 987.
b) 14 'i'c2 h6. Black addresses the possi­
bility of ll:ig5-e4 and i.gS. This is not strictly
necessary, but quite a common move when
White has no chances of a kingside pawn
storm. Grivas-Ionescu, Dubai 1 986 contin­
ued 1 5 e4 ll:ihS 1 6 i.e3 ll:id7 1 7 b3 :ac8 1 8
a 4 ll:ihf6 1 9 i.£4 ll:ieS 2 0 ll:ixeS dxeS 21
:xd8+ i.xd8 22 i.e3 i.e7 23 f3 ll:id7 and
Black was okay. Normally Black should be
careful when allowing White a pawn majority
on the queenside, but because the pawn is on
a4 there is no problem here.
c) 14 'i'b3 ll:ie8 15 e4 ll:id7 16 eS ll:ixeS 17
ll:ixeS dxeS 1 8 .:txd8 :xd8 1 9 i.xeS 'i'xeS 20
i.xb7 aS with equality, Meister-Wegner,
Germany 1 987.
1 4 . . . .1xg2 1 5 <it>xg2 ltic6 1 6 ltige4 ltie8!
Black should be careful and remember
this lesson. White has a knight on c3 and
cannot find any good use for it, so an ex­
change on e4 would be silly. Indeed
1 6 ... ll:ixe4?! 1 7 ll:ixe4 eS leaves the dS-square
susceptible to future attention.
1 7 .i.g5
More sensible than 17 g4!? which, in
1 08
Zaltsman-Amason, New York 1 986, met
with 17 ... ll:iaS!, a brilliant manoeuvre, albeit
quite simple. There followed 1 8 b3 ll:ib7 1 9
f3 ..ti>fB! (guarding the minor pieces and ad­
dressing the king's own security), and now
White should have played 20 'i'a3 with a
sensible position. Instead after 20 i.g3?1
'i'c8! Black assumed the initiative: 21 i.£2 bS
22 cxbS axbS 23 'i'xbS dS 24 ll:ia4 ltaS 25
'i'b6 dxe4 26 :xd8 exf3+ 27 exf3 i.xd8 28
:xd8 'i'xd8 29 'i'xb7 :ds and Black had a
clear advantage.
1 7 . . . f6!
This is not really necessary. Also possible
is 17 ... i.xgS!? 1 8 ll:ixgS h6 1 9 ll:if3 ll:iaS 20
:d4 (20 b3? bS! is in Black's interest)
20 ... :dc8 with sufficient counterplay.
1 8 .1e3!
Forced. The bishop is misplaced after 1 8
i.f4? ll:iaS! 1 9 b 3 f5 2 0 ll:igS bS! 2 1 'i'a3
i.xgS 22 i.xgS l:f.db8, when Black has a
powerful initiative.
1 8 . . . ltie5 1 9 :d4 Ades 20 ltid2 f5 21 f3
.1f6
Black is more than comfortable here.
22 .1f2 ltic6 23 :d3 ltia5 24 e4 g6 25
.1d4 .i.xd4 26 :xd4 ltic6 27 :d3 ltie5
28 :d4 ltif6 29 'ii' a 3 :ds 30 exf5 gxf5
31 ltie2 ltic6 32 l:.h4 ltie5 33 'ii' e 3 :ea
34 l:th6?
What the rook is supposed to be doing up
here is not clear. After 34 'i'gS+ 'i'g7 35
'i'xg7+ @xg7 36 .:td4 the endgame is level.
 4 g 3 i. b 4 + a n d 4
34 . . .-.g7 ! 35 l:h4 b5 36 b3 l:ad8 37
tllt4? tllg 6! 38 tll x g6 hxg6 39 l:d4
39 'ii'd3 g5 40 l:.d4 d5! is also clearly bet­
ter for Black because 41 cxd5? e5! is decisive.
39 . . . e5 40 l:d3 'ii' b 7?
The traditional 40th move blunder... Black
can win with 40 ... d5! 41 cxd5 e4! 42 fxe4 fxe4
and White will suffer material losses.
41 'iitg 1 �f7 42 l:e 1 d5 43 cxd5 tll x d5
44 'ii'c 5 tllf 6 45 l:xd8 J::.x d8 46 •ea -.d5
4 7 l:e2 'ifd4 48 •xd4 l:xd4 49 lllf 1 'iite 6
50 l:c2 �d7 51 Wf2 tll d 5 52 l:d2 l:xd 2+
52 ...tt:'lc3 53 lk2 l:td3 54 tt:'le3 �d6 main­
tains the pressure, but now we have a draw.
53 tllx d2 Yz -Yz
Game 44
Piket-Chuchelov
Netherlands 2000
1 d4 tllf 6 2 c4 e6 3 lllf3 b6 4 g3 .ta6 5
•a4 .tb7 6 .tg2 c5 7 dxc5 i.xc5 8 0-0
0-0 9 tllc 3 .te7 1 0 l::. d 1 tll a 6 1 1 .tf4
The most normal move, but dangerous is
1 1 i.e3!?, with the idea of putting the bishop
on d4. This has led to problems for Black but
seems to have been solved by 1 1 ...°iib BI, with
the idea of overprotecting the d6-square.
Georgiev-lvanchuk, Tilburg 1 993 continued
12 Aac1 llic5 1 3 'ji'c2 llice4 1 4 .if4, trans­
posing to the main game, but 14 ...'ii'c B 1 5
llixe4 i.xe4 1 6 'ii'a4 'ifb7! (the obvious
move) 17 i.ft Afe8 1 8 llie1 a6 was level.
Here 14 i.d4 .ic5 looks fine for Black.
1 1 . . . -.c8 1 2 l:ac 1 tll c 5 1 3 •c2 tll ce4 1 4
tlld4
14 llixe4 i.xe4 1 5 'ii'a4 (both 15 'ii'd2
'fib7 1 6 i.d6 i.xd6 17 'ii'xd6 Aac8, Dautov­
Smirin, Daugavpils 1 989, and 1 5 �3 'iib 7
16 i.d6 i.xd6 17 Axd6 Aac8, Dlugy­
Adorjan, New York 1 984, give Black a good
game) 1 5 ... �7! 1 6 .id6? (this exchange
favours Black; 1 6 i.f1 is seen in the note
above) 1 6 ... .ixd6 1 7 l:.xd6 Afc8! and Black
was better in Blees-Gershon, Tel Aviv 1 999.
Already the threat is ... b6-b5.
. . . .t a 6 with o u t 5 b 3
In Bareev-Karpov, Tilburg 1 99 1 , 1 4 h3?!
Ad8 15 g4? d6! left Black in the lead as White
had achieved nothing but weakening his
kingside. Karpov offers 1 5 tt:'ld4 with equal­
ity.
1 4 . . .tll x c3 1 5 •xc3 a6 1 6 'ii'b 3 .txg2 1 7
�xg2 'ii'b 7+ 1 8 'ii'f3 I:ta7 !
Stronger than 1 8 ...'ii'xf3+ 1 9 �f.3 l:tac8
20 e4, when White has more freedom to
manoeuvre thanks to his space advantage.
1 9 ii'xb7 l:xb7 20 f3 l:c8
20 ... tt:'lh51? 21 .i.e3 £5!? was an interesting
alternative.
21 e4 d6
21...tt:'leB!? 22 h4 h5! 23 a4 i.f6 24 i.e3
was slightly better for White in Piket-Karpov,
Monaco 1 999. However, as in the main
game, White failed to make anything of the
modest lead, so perhaps Black should not
fear it so much. Nevertheless I feel that
1 0 ... d6 is the safest option for Black.
22 tlle 2 tlle8 23 b3 h5!? 24 h4 g6 25
l::.d 2 I:tbc7 26 l:d3 l:c5 27 a4 l:5c6 28
l:cd 1 I:tc5 29 �f1 .tf8 30 tll c 3 l:b8! 3 1
I:t 1 d2 l:b7 ! 32 �e2 .te7 33 �d 1 �f8 34
l:c2 l:c8 35 tlle 2 l:bc7 36 l:a2 l:b7 37
.th6+ .*08 38 .te3 .tt6 39 .tf4 .te7 40
ltic3 .tf6! 41 tlle 2 .te7 42 l:c2 l:bc7 43
tll c3 .tf6 44 tll b 1 b5?! 45 axb5 axb5 46
tll a3 bxc4 47 l:xc4 .te7 48 .1:1.xc7 .l:l.xc7
49 tll c4 l:c6 50 �c2 f5! 5 1 l:c3 fxe4 52
fxe4 lllf 6 53 tlld 2 l:xc3+ 54 �xc3 Wf7
55 �d3 Yz-Yz
1 09
Q u e e n 's In dia n D e fe n c e

Su mmary
The 4 ... i.b4+ line invariably leads to a small advantage for White. 4 ... i.a6 is more modem and
more dynamic. After moves like 5 'ii'c2, 5 c!Dbd2, 5 'ii'b3 and 5 'ir'a4 Black can easily achieve
equality. The real test (if that is what you want to call it) is found in the following chapter. It is
a notable feature of these lines that, after the bishop returns to b7 at a later date, it seems that
White has used his extra move to his own disadvantage. The only real dangerous move con­
sidered here is 5 c!Dbd2, but the idea of ... d7-d5 and ... i.b7 followed by ... a7-a5 and ... c!Da6
should give Black a good game.

1 d4 l'Llf6 2 c4 e6 3 l'Llf3 b6 4 g3 (D) .ta6

4. . . i.b4+ 5 i.d2 a5 6 i.g2 0-0 7 0-0 i.a6 8 b3
8 ... i.b7 - Chapter 7; 8 ... d5 - Game 34
5 'ifa4
5 b3 - Chapter 7
5 'ir'c2 - Game 35
5 'ir'b3 - Game 36
5 c!Dbd2 (D)
5 . . . i.b4 - Game 37
5 ...i.b7 6 i.g2 c5 - Game 38
5 . . . c5 6 e4 - Game 39
5 ... d5 6 i.g2 i.e7 7 0-0 0-0 - Game 40
5 . . . .tb7
5 ... c6 - Game 4 1
6 .tg2 c 5 7 dxc5 .txc5
7 ... bxc5 - Game 42
8 0-0 0-0 9 l'Llc3 .te7 1 0 l::l. d 1 (D)
1 0 . . a6 - Game 43; 1 0 . c!Da6 - Game 44
. . .

4 g3 5 liJbd2 1 0 .l::l.d 1

1 10
 CHAPTER SEVEN
4 i.a6 :
. . .
Main Line with 5 b3
1 d 4 tllt 6 2 c 4 e 6 3 tllf 3 b 6 4 g3 .ta6 5
b3
This is where the main theoretical battle­
ground for the Queen's Indian and maybe
even for 1 d4 is taking place. Black has a
multitude of variations to choose from.
Some of them · are more experimental and
lead to an unbalancing of the position, while
others aim for long forced lines hoping to
equalise. The balance of play currently fa­
vours Black in the sense that White appar­
ently cannot engineer an advantage in the
main lines towards the end of the chapter.
But if you, like me, not only desire to play
perfect chess, but also want to win some
games now and again, then you might find
more fun in some of the sidelines presented
in this chapter, as all top players have done
from time to time.
Game 45
Kasparov-Gelfand
Novgorod 1997
1 d4 tllf 6 2 c4 e6 3 tllf 3 b6 4 g3 .ta6 5
b3 d 5 ! ?
This move has been played more than it
deserves, or at least it has been played with
an idea neither I nor practice can fully ap­
prove 0£ Another dubious alternative that
I
some might consider worth a try is 5 ... cS, e.g.
6 i.g2 and now:
a) 6 ... dS seems to be the only serious op­
tion. Bach-Schilow, 1 997 continued 7 0-0
�c6 8 cxdS exdS 9 .tb2 .l:c8 10 .l:el with
(probably) an edge for White in one of those
'better player wins' situations. 7 ... .te7 8 cxdS
�xdS 9 .tb2 0-0 1 0 �bd2 �d7 1 1 �c4 .l:c8
12 e4 �Sf6 13 dS· exdS 14 exdS was better
for White in Neverov-Sazonov, Yalta 1 996.
b) 6 ... �c6 7 dS! exdS 8 cxdS �e7?!
(8 ... �b4 9 �c3 .tb7 10 e4 .ta6 1 1 .tfl
.txfl 12 <ifiixfl , with some advantage to
White, is a lesser evil) 9 �c3 g6? (9 ... d6 1 0
0-0 g6 1 1 i.b2 .tg7 merely points to the
weird knight on e7) 10 i.gS .tg7 1 1 d6 �c6
12 �dS with a clear advantage in Popov­
Zvjaginsev, St. Petersburg 1 998.
c) 6 ... i.b7 7 dS exdS 8 �h4 g6 9 .tb2
.tg7 10 cxdS d6 1 1 0-0 bS 12 a4 b4 13 �d2
was better for White in Tratar-Slak, Slovenia
1 994.
6 .tg2 dxc4?1
After this move White has a large advan­
tage. :Sut 6 ... .tb4+! 7 .td2 .txd2+! is
interesting (rather than 7 ... .td6 8 �c3 �bd7
9 e4! cS 10 exdS exdS 1 1 'it'e2+!, as in Nes­
torovic-Rajkovic, Belgrade 1988). Then 8
�bxd2 0-0 (8 ... cS 9 dxcS bxcS 1 0 0-0 0-0 1 1
l:.el �c6 1 2 cxdS exdS 1 3 l:.cl �6 1 4 'i'c2
111
Q u e e n 's In dia n D e fe n c e
lllb4 1 5 1i'xc5 lllxa2 1 6 'ii'xb6 axb6 17 l:.c6
was better for White in Nikolic-Yudasin,
Tilburg 1 993) 9 0-0 lllbd7 1 0 l:.c1 c5 1 1 cxd5
(1 1 l:.e1 i.b7 1 2 cxd5 exd5 13 'i'c2 l:.e8 1 4
'ii'b2 'ii'e7 1 5 e 3 a 5 with equality in Tischer­
Eismont, Biel 1 994) 1 1 ...exd5 1 2 l:.e1 l:.e8 1 3
e 3 l:.c8 1 4 .i..h3 ltc7 1 5 i.f1 ended peace­
fully in Altennan-Yudasin, Haifa 1 995, al­
though I have a feeling that White might be
slightly better here. Instead 8 'ii'xd2 0-0 9 0-0
'ii'e7 10 cxd5 exd5 1 1 lllc3 .i.. b7 12 ltac1
lll a6 13 l:.fdl l:.fd8 14 'ii'h2 c5 1 5 e3 was a
prototypical position in Chernin-Razuvaev,
Tilburg 1 994.
7 'll e 5 i.. b 4+ 8 c.t>f 1 i..d 6
Here 8 ... lllbd7?? 9 lllc 6! is embarrassing.
The other alternatives are:
a) 8 ... lll fd7 9 lllxc4! c6 1 0 i.b2 0-0 (after
1 0 ... b5 1 1 llle 3! Black appears to have noth­
ing better than 1 1 ... 0-0, transposing) 1 1
lllbd2 b5 1 2 lll e3 .i..b7 1 3 'ii'c2 'ii'b 6 1 4 ll:\£3
tll f6 15 lll e5 .i.. e7 16 lll3g4?! was Karpov­
Timman, Kuala Lumpur (8) 1 990. Karpov
himself indicates that 1 6 l:.c1 ! followed by
.i.. £3 and �g2 was better, securing an advan­
tage.
b) 8 ... c6 9 bxc4! 0-0 10 .i..b 2 .i.. d6 1 1 tlld2
'ii'c 7 1 2 'ii'c2 secured White a small plus in
Bykhovsky-Gofshtein, Tel Aviv 1 995.
9 'llx c4 'lld 5 10 e4 'lle 7
1 1 i.. b 2
11 .i.. £3 ?! lll bc6 1 2 'at>g2 .i..xc4 1 3 bxc4 e5
14 d5 lll d4 15 .i.. e3 c5 is fine for Black, Jas-
1 12
nikowski-Lau, Polanica Zdroj 1 986, while 1 1
e5? .i..xe5 1 2 .i..xa8 'i'xd4 1 3 'ii'xd4 i.xd4 1 4
lll c3 .i..xc3 1 5 l:.bl c6 is just stupid. Black is
better.
1 1 . . . 'll b c6 1 2 .'ll b d2 0-0
12 ... e5 13 d5 lll a5 14 �g1 lllxc4 1 5 lllxc4
i.xc4 16 bxc4 0-0 1 7 .i..h3 £5 1 8 'at>g2 gave
White an overwhelming advantage in Kas­
parov-Atlas, Catonsville (Simultaneous Dis­
play) 1 996.
1 3 �g 1 b5 14 'llxd6!
An improvement at the time.
1 4 . . . cxd6 1 5 h4 'ii' b 6
15 ... d5 is suggested as an improvement by
Dokhoian, Kasparov's second, but after 1 6
h51 Gelfand does not like Black's position.
1 6 h5!?
Here White could play even better with 1 6
d5!, with an advantage.
1 6 . . . h6?!
16 ... lllxd41 17 h6 g6 1 8 lll c4 lll e2+! 19
'ii'xe2 bxc4 lead to complications that are not
necessarily in White's favour.
1 7 d5 'lle 5 1 8 l't)f1 b4 1 9 i..d 4 'ifa5
19 ... .i..e2 20 i.xb6 .i..xd1 21 l:.xdl axb6 22
f4 lllg4 23 dxe6 l:.xa2 24 l:.xd6 gives White a
considerable advantage.
20 'lle 3 .Z:.ac8 21 .Z:.h4 .Z:.c7 22 'ii'd 2
White i s firmly in the driving seat.
22 . . . .Z:.c3?
22...lll d 3.
23 i..x c3 bxc3 24 'ifd4
Now it is over.
 4. . . i. a 6 : Ma in L in e with 5 b 3

24 . . . exd5 25 exd5 •c7 26 •d 1 :ea 27 Korinthos 1 998.

i.e4 •bs 2a :t4 i.b7 29 :c 1 •as 30 9 . . . 0-0 1 0 a3 i.e7
:c2 ;.,ha 3 1 i.g2 i.a6 32 :a4 •b6 33 If Black exchanges the bishops on d2 he
ltic4! i.xc4 34 bxc4 ltif5 35 .l:xc3 ltid4? will experience permanent problems on the
36 c5 :xc5 37 •xd4 1 -0 cS-square: Korchnoi's 10 ... .txd2 1 1 'ii'xd2
--
---. .tc6 12 lLJc3 dS 13 lLJeS .tb7 14 .l:.fcl , with
Game 46 a pleasant position for White, illustrates this
Tkachiev-Nisipeanu well.
Naujac-sur-Mer 2000 1 1 ltic3 i.c6
1 1 ....i.a6 12 'i'c2! dS 13 l:fcl lLJbd7 14 e4
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 g3 i.a6 5 gave White the advantage in Karpov-
b3 b5!? 6 cxb5 i.xb5 7 i.g2 i.b4+ Christiansen, Wijk aan Zee 1 993.
1 2 •c2 .d 5
12 ... .tb7!?, with the idea 13 e4 dS 1 4 eS
lLJe4 with counterplay has been suggested by
Gershon, but I think that White has the ad­
vantage after the simple 1 5 l:.fcl.
1 3 ltie5 i.b 7 1 4 ltia4 ltifd7 1 5 ltid3

7 ... cS 8 0-0 .tc6 9 .ta3 lLJa6 1 0 lLJbd2

.te7 1 1 .l:tc1 cxd4 1 2 .txe7 'ii'xe7 1 3 lLJc4
0-0 14 'ii'xd4 dS 1 5 lLJfeS l:.fc8 1 6 lLJxc6
:.Xc6 17 lLJeS l:.c7 18 :.Xc7 'Wxc7 19 lLJd3
gave White a slight advantage, due to the
superiority of bishop against knight, in van
der Sterren-Timman, Holland 1 987. White has the advantage.
7 ... dS 8 0-0 lLJbd7 9 .ta3 .txa3 10 lLJxa3 1 5 . . . i.c6 1 6 ltiac5 ltixc5 1 7 lLixc5 :a7
.ta6 1 1 lkl ! 'We7 12 lLJbl l:.c8 1 3 .l:tc2 was 1 a :tc 1 i.b5 1 9 e3 lLid7?
Wells-Dautov, Bad Worishofen 1 997, but Losing.
White could have played better with 1 3 l:.c6! 20 a4! i.e2 21 i.xa5! ltixc5 22 dxc5
.tb7 14 l:c2 with an edge according to :xa5 23 b4 i.f6 24 :a3 i.c4 25 bxa5
Dautov. •bs 26 l:l.b 1 •a7 27 i.f1 ! •xc5 28
8 i.d2 a5 9 0-0 l:l.ab3 1 -0
Here White has a refutation according to
all the books, but after 9 .txb41? axb4 10 Game 47
'ii'd2 lLJc6! (the new move) 11 'WgS l:.aS 1 2 Bogdanovski-Ca brilo
'ii'xg7 ri;e7 1 3 'ii'h6 'ii'a 8 1 4 'Wd2 lLJe4 1 5 Bije/jina Dvorovi 2002
'Wb2 d S Black has enough compensation for
the pawn and had a good game, until a draw 1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 g3 i.a6 5
was settled in Belakovskaia-Mastrovasilis, b3 i.b4+ 6 i.d2 i.xd2+ ! ?

1 13
Q u e e n 's In dia n D e fe n c e
1bis sideline is not as popular as one
might imagine when we look at the evalua­
tion after the opening. I think this is because
the position is rather dull and Black does
have a slight structural disadvantage in hav­
ing a c-pawn compared to the e-pawn (both
pawns are in half-open files). One difference
is that £2 is a better protector for the e-pawn
than b6 for the c-pawn. It is also significant
that White's king can assist the e-pawn better
than the Black can assist the c-pawn. These
are, of course, relatively minor issues but,
structurally, White does possess a slight ad­
vantage. It should not matter theoretically,
but in practice such things often do.
7 'ifxd2
The only serious move. 7 lbbxd2 i..b7 8
i..g2 cS 9 0-0 0-0 1 0 'ii'c2 lbc6 1 1 'ii'b2
lbxd4 1 2 lbxd4 i..xg2 1 3 �xg2 cxd4 14
'ii'xd4 dS was completely equal in Anders­
son-Short, Na:stved 1 985.
7 . . . d 5 8 cxd5
Also interesting is 8 i..g2 and now:
a) 8 ... 0-0 9 0-0 cS 10 lbc3 lbc6 1 1 .:!fdl
.:!c8 1 2 cxdS exdS 13 dxcS bxcS 14 lbxdS
ltJxdS 1 5 'ii'xdS i..xe2 1 6 'ii'xd8 .:!fxd8 17
:Xd8+ .:!xd8 1 8 .:!el .:!dl 1 9 :Xdl i..x dl 20
lbd2 lbb4 21 a4?! was Gligoric-Szabo, Mos­
cow 1 956. Chekov suggests 21 a3! lbc2 22 a4
ltJd4 23 .i.. d S �f8 24 f4 f6 25 �£2 �e7 26
�e3, with a slight advantage, as being more
prudent. After 9 ... c6 10 'ii'c2 lbbd7 1 1 .:!dl
.:!c8 12 lbeS?! (1 2 lbbd2 cS 13 dxcS is a
shade better for White) 1 2... cS 1 3 lbxd7
'ii'xd7 14 cxdS cxd4 1 5 'ii'b 2 eS Black
emerged ahead in Zhu Chen-Christiansen,
Seattle 2001 .
b) 8 ... cS 9 cxdS exdS 1 0 'ii'e3+ 'ii'e7 1 1
'ii'xe7+ �xe7 1 2 lbc3 lbc6 1 3 dxcS bxcS 1 4
lbh4 lbd4 1 5 0-0-0 .:!hd8 1 6 .:!hel i..b7 1 7
�b2 g6 1 8 e4, Ruzele-Balashov, Boeblingen
1 998, and White looked slightly better, al­
though it is very very slight.
8 . . . exd5 9 'ife3 +
9 lbc3 0-0 1 0 i..g2 .:!e8 1 1 lbeS! (White
must be careful here: 1 1 0-0?! lbe4! 12 lbxe4
1 14
dxe4 1 3 lbgS lbc6! 14 .:!fdl lbxd4 1 5 lbxf7
lbxe2+ 1 6 'ii'xe2 .i..xe2 1 7 .:!xd8 .:!axd8 1 8
lbxd8 .:!xd8 1 9 .i..xe4 cS left White struggling
for equality in Eingom-Gelfand, Moscow
1 990) 1 1 ...i..b 7 12 0-0 cS 1 3 e3 'ii'e7 14 'ii'c2
lba6! 1 5 .:!acl lbc7 1 6 .:!fel lbe6- with equal
play in Hansen-Epishin, Tilburg 1 993. In
Georgiev-Christiansen, Las Palmas 1 993
Black chose the less accurate 10 ... lbe4?!, the
game continuing 1 1 lbxe4 dxe4 1 2 lDg5 lDc6
13 i.xe4 lbxd4 14 i.xh7+ �h8 15 i.d3
.i..xd3 1 6 'ii'xd3 'ii'xgS 17 'ii'xd4 .:!ae8 1 8 e3
'iib S 1 9 ..Wc4 (Dautov's 1 9 0-0-0!? cS 20 'iVdS
'ii'b4 21 �b2 looks even stronger) 1 9 ... 'ii'xc4
20 bxc4 .:!eS!, and now Dautov gives 21
�e2! .:!cS 22 .:!acl .:!d8 23 h4 .:!d6 24 .:!hdl
.:!dc6 25 .:!d4 .:!aS 26 l:k2, retaining a clear
advantage.
9 . . . 'ife7 1 0 'ifxe7+ <Jilxe7 1 1 lLic3 l:le8
1 1 . ..lbbd7 12 i.h3 .:!he8 13 0-0 �f8 14
.:!fcl c6 was equal in Hjartarson-Agdestein,
Gausdal 1 987.
1 2 J:lc 1 c6 1 3 i..g 2 tLlbd7 1 4 0-0 �8 1 5
J:fe 1 i.. b 7 1 6 i.. h 3 J:lad8 1 7 e3 h6
Black has an equal position. However, this
endgame is far from decided and White just
plays on and on, his longer arms deciding the
game ...
1 8 b4 g 5 19 i..f 5 lLie4 20 h4 gxh4 21
tLixh4 lLid6 22 i..d 3 tLlf6 23 a4 J:le7 24
a5 b5 25 tLlf5 tLixf5 26 i..x f5 lLie8 27
lLie2 i.. c 8 28 i.. d 3 .i.d7 29 �g2 lLic7 30
l:th 1 J:le6 3 1 i..f 5 J:ld6 32 .i.xd7 .1:8xd7

33 ltif4 ltia6 34 ltid3 .l:!.e7 35 .l:!.h5 .l:!.ee6
36 '1t>f3 <iit9 7 37 We2 liJbS 3S ltif4 .l:!.e7
39 l:e5 �S 40 J:xe7 <iitx e7 41 .l:.h 1 ltid7
42 94 ltif6 43 f3 liJ9S 44 <iitd 3 l:.f6 45
l:h3 �S 46 <iite 2 l:d6 47 ltih5 ltie7 4S
lti93 l:96 49 <oitf2 l:f6 50 l:h 1 l:.d6 5 1
l:c 1 <oi?97 52 @e2 <iit9 6 5 3 .l:!.9 1 W97 54
@d3 �S 55 e4 dxe4+ 56 fxe4 l:.96 57
ltih5 liJgS 5S ltif4 .l:!.95 59 ltih3 .l:.g6 60
@e3 <>ties 6 1 ltif4 .l:!.95 62 <iitt 3 ri;e7 63
.l:!.c 1 ri;d7 64 ltid3 t6 65 l:h 1 Wd6 66
ltit4 .l:!.g7 67 ltih5 l:g6 6S lti93 '1t>e6 69
ltif5 Wd7 70 ri;f4 We6 7 1 J:c1 ri;d7 72
.l:.c5 a6 73 d5 cxd5 74 l:xd 5 + Wc7 75
l:d6 '1tib7 76 .l:td7+ @bS 77 ltid6 l:95 7S
.l:tb7 + was 79 .l:!.f7 @bS so l:.b7 + <iita S
S 1 .:r.h 7 <oi?bS S2 ltif5 .l:!.96 S3 ltid4 l:g5
S4 ltic6 + <oi?cS S5 ltid4 <iitb S S6 ltie6 l:e5
S7 ltic5 'itcS SS ltixa6 ltie7 S9 ltic5 ri;c7
90 l:.xh6 ltic6 91 l:h7 + .l:!.e7 92 l:.xe7+
ltixe7 93 a6 <iitb 6 94 ltid7 + <iitx a6 95
ltixf6 <l.>b7 96 ltid5 lti96 + 97 'itf5 ltih4+
9S 'itf6 <oi?c6 99 95 @d6 1 00 96 ltixg6
1 -0
In TWIC this game is quoted as ending in
a draw here. I would say that this is very
unlikely, but could have happened as a result
of time trouble. Whether the result was 1/2-1/2
or 1 -0 is of no great importance except, of
course, to the players. White outplayed Black
from an equal position, which is all we need
to know.
Game 48
Petursson-Polugaevsky
Reylefavik 1987
1 d4 lbt6 2 c4 e6 3 lbt3 b6 4 g3 i.a6 5
b3 i.b7 6 i.g2 i.b4+ 7 i.d2 a5 ! ?
This i s the choice of the ambitious player.
The typical positions in this line might be
marginally better for White, as all positions in
the QID tend to be, but results are almost
always decided by who is the better player on
the day.
s 0-0 0-0 9 i.g5
4 . . . i. a 6 : Ma in L in e with 5 b 3
The main line, 9 ll'ic3, is considered in the
next two games.
Another option is 9 'it'c2, when Black can
equalise with 9 ... c5!, e.g. 10 l:tdl cxd4 1 1
ll'ixd4 .ixg2 1 2 'it>xg2 'it'c7 1 3 .ig5 'it'e5 1 4
.ixf6 'it'xf6 1 5 ll'ic3 ll'ic6 1 6 ll'idb5 l:fd8 1 7
.l:.acl 'it'e7 1 8 'it'e4 f5 1 9 'it'f3 .ic5 20 ll'ia4
ll'ie5 21 'it'c3 ll'ig4 22 ll'ixc5 'it'xc5 23 'it'd4
ll'if6 with a draw in Brenninkmeijer�Hiibner,
Wijk aan Zee 1992. Black should avoid 9 ... h6
10 a3, when 10 ... .ie7 1 1 ll'ic3 d5 1 2 .l:.fd l !
'it'c8?! (1 2 ...ll'ia6 improves) 1 3 cxd5 exd5 14
.if4 l:td8?! (1 4... ll'ih5) 1 5 ll'ih4! left White in
charge in Karpov-Lobron, Baden-Baden
1 992, while 10 ... .ixd2 1 1 ll'ibxd2 d5 12 cxd5
exd5 13 .l:.acl ll'ia6 14 e3 c5 1 5 .l:.fd 1 fol­
lowed by ll'ib1 -c3 gives White a slight advan­
tage according to Dautov.
9 . . . a4! 1 0 bxa4
10 a3?! axb3! does not benefit White.
1 0 . . . h6 1 1 a3 .i.a5! 1 2 i.xf6 'ii'xf6
Black has at least equalised.
1 3 e3
13 'it'd3 is probably better.
1 3 . . . c5!
1 4 .:r.a2
Anoths:r Polugaevsky game goes 14 ll'ibd2
.ixd2 1 5 ll'ixd2 .ixg2 16 �xg2 cxd4 17 exd4
'it'xd4 1 8 .l:.bl .l:.a6 19 .l:.el l:c8 20 .l:.e3 ll'ic6
21 llbb3 ll'ie5 22 l:e4 'it'd6 23 'it'e2 'ii'c 6 24
�gl d6 25 'it'e3 'it'c5 26 l:tb4 'i'xe3 27 llxe3
and Black won the ending in Epishin­
Polugaevsky, New York 1989.
1 15
Q u e e n 's In dia n D e fe n c e
1 4 . . . .:l.a7 1 5 llle 5?!
1 5 dxc5 bxc5 1 6 �bd2 with approximate
equality.
1 5 . . . cxd4! 1 6 'i'xd4 d6! 1 7 lllf3
17 ltlg4!? 'ii'xd4 18 exd4 i.c8! 19 ltle3
i.d7 also gives Black something. The bishop
on a5 is potentially very strong.
1 7 . . . e5 1 8 'i'd 1
1 8 . . . llld 7
Preferable is 18 ... 'i'e61 1 9 �fd2 i.xg2 20
�xg2 d5 21 'i'b3 d4! with a definite advan­
tage for Black.
1 9 lll fd2 .txg2 20 ltxg2 'i'e6 21 .:l.b2!
lll c 5 22 .:l.b5 .:l.fa8 23 lll b 3 lllx b3 24
'i'xb3 .:l.c8! 25 .:l.d5?!
25 J:r.d1 :Xc4 26 .l:xd6 'ii'xd6 27 'ii'xc4
.l:c7 keeps White's disadvantage to a mini­
mum.
25 .:l.ac7 26 .:l.fd 1 .:1.xc4 27 .:l.xd6 'i'f5
. • .
28 :I.dB + lth7 29 .:l.xc8 .:1.xc8 30 llld 2
e4! 3 1 ltg 1 ?
31 �f1 l:r.c3 32 'ii'b 1 .l:d3 might be terri­
ble but it is White's only chance of survival.
3 1 . . . .txd2! 32 .:1.xd2 .:l.c1 + 33 ltg2
33 l:td1 'ife6! picks up a rook.
33 . . . 'i'f3+ 34 lth3 .:l.g 1 0-1
Game 49
Karpov-Salov
Unares 1992
1 d4 lllf 6 2 c4 e6 3 lllf3 b6 4 g3 .ta6 5
b3 .tb7 6 .tg2 .tb4+ 7 .td2 a5 8 0-0
1 16
0-0 9 lll c 3
9 . . . d5
The more flexible 9 ... d6 is considered in
the next game. The main reason for not play­
ing this move is an attempt to push ... d7-d5
in one go, investing the extra tempo some­
where else. After the alternative 9 ... l:te8 1 0
'ii'c2 h 6 1 1 l:tfe1 i.xc3 1 2 i.xc3 i.e4 1 3
'iVb2 d 5 14 i.f1 Black has tried:
a) 14 ... i.xf3!? 15 exf3 �bd7 16 i.d3 a4
17 l:ad1 was Polugaevsky-Anand, Roque­
brune 1 992, and 1 7 f41 would have secured
White an edge.
b) 14 ...�bd7 1 5 �d2 'iVe7 16 l:tac1 c6 1 7
e 3 was slightly better for White in Pinter­
Garcia Trobat, Spain 1 993.
c) 1 4... c6 15 ltld2 i.g6 16 a3 gave White a
tiny advantage in Alterman-Korchnoi, Beer­
Sheva 1 992.
1 0 "it'c2 llla 6 1 1 .:1.ad 1 ! h6
1 1 ....l:c8 12 cxd5 exd5 13 i.h3 favours
White.
1 2 llle 5 'i'c8 1 3 .th3! ? "it'e8 1 4 cxd 5
exd5
14 ...ltlxd5 15 �xd5 i.xd5 16 e4 i.b7 17
i.f4 is not to be recommended for Black.
1 5 .tf4 :I.dB 1 6 'i'c 1 ! .tc8
16 ... �e4 17 ltlxe4 dxe4 18 'ii'c4 is only a
little better for White.
1 7 .tg2 .tb 7 1 8 a3 .td6 1 9 lll c 4!
The knight seeks fresher pastures. Black is
in a bad situation.
1 9 . . . .txf4 20 'i'xf4 .tc8 2 1 llle 3 .te6 22

tt::\t 5 i.xt5 23 �xt5 •ea 24 •d3 tiJbB 25
b4 axb4 26 axb4 tt::\c 6 27 b5 tt::\a 5 28
:te 1 :tea 29 :c1 •d6 30 :c2 :e7 3 1
e 3 g6 32 J:ec 1 �g 7 3 3 tt::\e 2 :as 34 tt::\t4
:a7 35 :a2 ltd7 36 �a3! •xa3 37 .C.xa3
�ta 38 tt::\d 3 �ea 39 tt::\ b4 h5 40 :ac3
�dB 41 :c6! tt::\e4 42 i.h3 :d6 43
:xd6+ tt::\x d6 44 i.f1 ! tt::\e4 45 tt::\x d5
tt::\ b 3 46 :c2 g 5 47 i.e2! :a 1 + 48 11i>g2
tt::\c 1 49 t3 tt::\d 6 50 i.t1 g4 51 txg4 hxg4
52 tt::\t 6 ::ta2 53 :t2! :xt2+ 54 �xt2
tt::\a 2 55 e4 tt::\ c 3 56 �e3 tt::\ c xb5 57 d5!
c5 58 dxc6 tt::\ a 7 59 e5 tt::\d c8 60 c 7 + !
�xc7 6 1 i.c4 tt::\ c 6 62 tt::\xg4 1 -0
----
Game 50
T opalov-Adams
FZ!Jitsu Siemens Masters Frankfurt 2000
1 d4 tt::\t 6 2 c4 e6 3 tt::\t 3 b6 4 g3 i.a6 5
b3 i.b7 6 i.g2 i.b4+ 7 i.d2 a5 8 0-0
0-0 9 �c2 d6 1 0 tt::\c 3 tt::\b d7 1 1 :te 1
This is the most difficult fo r Black to
meet. Others:
a) 1 1 .l:!.adl i.xc3 12 i.xc3 i.e4 13 °ili'cl
a4 14 i.h3 b5 15 tiJd2 axb3 16 axb3 bxc4 1 7
bxc4 .l:!.a2 led to full equality in Morovic Fer­
nandez-Adams, Istanbul 2000. The position
is about equal, perhaps a bit better for Black.
b) 1 1 tDel "ili'c8 12 l:kl l:.e8 13 tiJb5
i.xg2 14 ttJxg2 'ib8 1 5 a3 i.xd2 1 6 "ili'xd2
c6 17 tiJc3 d5 with equality in Korchnoi­
Adams, Wijk aan Zee 2000.
4 . . . i. a 6 : M a in L in e with 5 b 3
1 1 . . . i.xc3 1 2 i.xc3 i.e4 1 3 �b2
The great expert of the QID-g3, GM
Dautov, plays the more flexible 13 °ili'cl !?,
when the queen is yet to commit to a flank.
Dautov-Romanishin, Essen 2001 continued
13 ... h6 14 i.h3 i.h7 15 tiJd2 tDe4 1 6 tDxe4
i.xe4 1 7 f3 i.h7 1 8 e4 a4 19 °iWh2 with a
pull, while 13. .. "ili'h8 14 i.h3 c6 1 5 tiJh4 h6
16 f3 i.h7 17 e4 b5 18 °ili'd2 gave White
similar prospects in Dautov-Zeller, Boe­
blingen 1 996.
1 3 . . . a4
Just one of several possibilities.
a) 13 ... c6!? 1 4 i.f1 b5 15 ttJg5 i.g6 1 6
i.g2 d5 17 cS?! h6 1 8 tiJf3 ttJe4! led to equal-
ity in Pinter-Almasi, Budapest 1 997. Avrukh
suggests 1 7 f3!? as an improvement. Sjoberg-
Almasi, Malmo 1994 went 1 5 d5?!, and now
the strongest seems to be 1 5 ... exdS! 1 6 cxb5
c5 1 7 i.xf6 tDxf6 1 8 tiJd2 i.g6 and Black
has a pleasant grip on the centre.
b) 13 ... d5 14 i.h3!? i.£5 (14... dxc4 1 5
bxc4 c5 1 6 d5! exd5 17 cxd5 i.xd5?! 1 8
.l:tadl gives White a dangerous initiative ac­
cording to Chemin) ' 1 5 i.xf5 exf5 16 tiJh4
tDe4!? 17 f3 tDxc3 1 8 "ili'xc3 f4 1 9 .l:tacl with
a slight edge for White in Chernin­
Romanishin, Taastrup 1 992.
c) 13 ... °ili'b8 14 i.fl (in Petursson-Lemer,
Moscow 1987 White tried 14 i.h3 bS 1 5 dS
e5 16 tiJd2 i.g6 17 cxb5 °ili'xb5 18 e4 ttJc5 1 9
i. f1 °iWb6 and was already struggling) 1 4. . .c5
1 5 .:t.adl .l:!.d8 1 6 i.h3! a4! 17 tiJd2 axb3 1 8
axb3 i.b7 1 9 d5 with compensation for the
pawn, should Black decide to take it, in van
Wely-Karpov, Tilburg 1 996. 14 ... i.xf3!? 1 5
exf3 d5 1 6 .l:!.acl c6 was Hulak-Polugaevsky,
Zagreb 1 987. Now 1 7 f4 should give White
some advantage.
1 4 .tn c5
1 4 ...°iWb 8!? 15 tiJd2 b5 1 6 f3 Jf.. c 6 17 b4
bxc4 1 8 tDxc4 d5 1 9 tiJa3 .l:!.e8 gave Black a
good game in van der Sterren-Timman,
Gouda 1 997. But we should still question
whether White could have played better.
14 ... i.b7 15 tiJd2 ttJe4 16 i.g2 tiJdf6 1 7
1 17
Q u e e n 's In dia n D e fe n c e
tt::lxe4 tt::lxe4 1 8 d 5 tt::lxc3 19 'Wxc3 e5 2 0 b4
f5 offered Black sufficient counterplay in
Pinter-Romanishin, Balatonbereny 1 995, but
it seems that White was not out to test his
opponent in this game.
1 5 ll.ld2 i.b7 1 6 b4
1 6 i-h3 'ir'c7 17 :adl :a6 1 8 d5 e5 19 f4
l:tfa8 was the course of Kragelj-Romanishin,
Pula 1 994, with a good game for Black.
1 6 . . . a3 1 7 fib3 cxd4 1 a i.xd4 e5 1 9
i.e3 fic7 20 l:l.ac 1 'iWc6 2 1 t3 fia4 22
fib 1
22 . . . d5!
It is obvious that Black has outplayed his
opponent in the opening and now has a bet­
ter game.
23 i.t2 e4 24 i.d4 :I.tea 25 cxd5 i.xd 5
26 t4 e3 27 ll.lc4 i.e4 2a 'iWa 1 'iWxb4 29
ll.lxe3 b5 30 i.h3 l:.ada 31 :!.ed 1 i.aa 32
l:.b 1 fia4 33 'iWc3 ll.lta 34 i.e5 l:.xd 1 +
35 l:.xd 1 ll.ld5 36 l:l.xd5 i.xd 5 37 i.xg7
hl.xe3 3a 'iWxe3 'iWd 1 + 39 �t2 �xg7 40
'iWe5 + t6 41 'iWe7+ i.t7 42 'iWxa3 'iWd4+
43 'iWe3 'iWxe3+ 44 �xe3 i.xa2
As you can see it is not only club players
who play on in embarrassing positions. 1bis
probably has something to do with a fighter's
attitude...
45 �d4 i.c4 46 e3 ll.le6 + 47 �c3 ll.lc5
4a i.t5 h6 49 h4 i.e6 50 i.c2 ll.la6 5 1
i.d3 b4+ 52 �d4 ll.lc7 5 3 e 4 b3 5 4 �c3
�ta 55 i.e2 ri;e7 56 g4 �d6 57 g 5 hxg5
5a txg5 txg 5 59 hxg5 �c5 60 �b2 �b4
1 18
6 1 g6 ll.lb5 62 g7 ll.lc3 63 i.g4 ll.la4+ 64
�b1 i.ga 65 e5 i.h7 + 0- 1
Game 5 1
Ehlvest-Chandler
Lucerne 1989
1 d4 ll.lt6 2 c4 e6 3 ll.lt3 b6 4 g3 i.a6 5
b3 i.b7 6 i.g2 i.b4 + 7 i.d2 c5!?
Initially this seems quite strange but, when
you think about it, Black is trying to gain
some control over the dark squares on the
queenside, as well as opening the c-file. If
you compare it with 7 ... a5, then 7 ... c5 better
meets 9 tt::lc 3.
a 0-0
Harmless is 8 dxc5 i.xc5 9 4.Jc3 tt::le4 10
tt::lxe4 i.xe4 1 1 i.c3 0-0 12 0-0 tt::lc 6 13 'ir'd2
d5 14 cxd5 'ir'xd5 with complete equality,
Hjartarson-Korchnoi, Barcelona 1 989.
a . . . 0-0 9 i.c3
9 dxc5 transposes to 8 dxc5. 9 tt::l c3 cxd4
1 0 4.Jxd4 i.xg2 1 1 'it>xg2 d5 results in
immediate equality, a draw being agreed in
Olafsson-Petursson, Reykjavik 1 989, for
example.
9 . . . d5 1 0 cxd 5
Or 10 tt::l e 5 4.Jc6! 1 1 tt::lxc6 (1 1 i-xb4
tt::l xb4! does not trouble Black) 1 1 ...i.xc6 1 2
i.xb4 (1 2 e3, a s in Goldin-Douven, Palma
de Mallorca 1 989, leads to equaliity after
1 2... i.xc3 13 tt::lxc3 :c8 14 dxc5 dxc4!)
1 2 ... cxb4 13 'ir'd3!? dxc4 1 4 bxc4 i.xg2 1 5

rJolxg2 b5! 1 6 cxb5 'iid5+ 1 7 f3 l::t fd8 with a
complex position which should be okay for
Black, Nikolic-Salov, Brussels 1 988.
10 i-xb4 cxb4 11 ltJbd2 ltJc6 1 2 e3 l::tc 8
13 'iib 1 l::t c7 14 l:tc1 ike7 1 5 'tib2 l::t fc8 1 6
lDe5 was equal in Tukmakov-Amason, Reyk­
javik 1 990. I suspect that Black might have
the better practical chances here.
1 0 . . . exd5
Only move. 1 0 ...ltJxd5 1 1 i-b2 cxd4 1 2
ikxd4 'iif6 1 3 a3! 'iixd4 14 lDxd4 i-c5 1 5
ltJb5! lDc6 1 6 b4 i- e7 1 7 e4 lDf6 1 8 e 5 lDd5
19 lD1c3! was a solid plus for White in Kar­
pov-Korchnoi, Biel 1 992, and 10 ... i-xd5 1 1
a3 i-xc3 1 2 lDxc3 i-b7 1 3 dxc5 bxc5 1 4
ikxd8 l::txd8 1 5 lDe5 i-xg2 1 6 rJolxg2 was
uncomfortable for Black, if not a great deal
worse, in Kishnev-Romanishin, Munich
1 992.
1 1 i.b2
11 dxc5 i.xc5 12 i.b2 ltJa6 13 lDc3 ltJe4
14 l::tc 1 ltJc7 1 5 ltJd4 l::te 8 favoured Black in
Novikov-Hjartarson, Tilburg 1 992.
1 1 . . .lLle4 1 2 a3
Brenninkmeijer-Amason, Groningen 1 990
resulted in a level game after 12 dxc5 bxc5 1 3
a3 i-a5 1 4 b 4 itb6!? 1 5 lDc3 lDxc3 1 6 i-xc3
cxb4 17 axb4 i-xb4 1 8 i.xb4 'iixb4 19 l::tb 1
'iie7 20 ltJd4!? 'iid 7! etc.
1 2 . . . i.a5 1 3 lLlfd2 f5!?
Black has had a successful opening. The
situation is unclear.
1 4 lLlxe4 fxe4 1 5 lLlc3 lLla6!
4 . . . i. a 6 : M a in L in e with 5 b 3
1 5 ... ltJc6? 1 6 dxc5 i.xc3 1 7 i-xc3 bxc5 1 8
b4 is very good for White.
1 6 J:!.c1 'it>h8 1 7 e3 i.xc3 1 8 i.xc3 'ii'g 5
19 i.b2 J:!.ad8 20 f4! ? exf3 2 1 J:!.xf3 'iie 7
22 'iid 2 J:!.xf3 23 .i.xf3 J:!.e8 24 J:!.e1 lLlc7
25 i.g2 lLle6 26 .:lf 1 lLlg5 27 J:!.f5 h6 28
J:!.e5 'ii'f7 29 .=.xe8+ 'ii'xe8 30 'ii'f 2?
Allowing the creation of a passed pawn.
Instead 30 dxc5! bxc5 31 b4 is unclear.
30 . . . c4! 31 bxc4 dxc4 32 d5 .i.xd5 33
h4 lLle4 34 'ii'f 5?
34 i-xe4 i-xe4 35 'if'f4 is an unattractive
lesser evil.
34 . . . c3!
Now this is possible.
35 'ii'xd5
This loses on the spot. There is still some
resistance left in 35 i.a1 c2 36 i-b2 lDf6!? 37
i.xd5 ikxe3+ 38 rJiih2 c1\i' 39 i.xcl 'it'xc1,
with an extra pawn for Black.
35 . . . lLlf6! 36 'ii'd 4 cxb2 37 'ii' x b2 'ii'x e3 +
38 'ii'f 2 'ii'x a3 39 'ii'f4 'ii'c 5+ 40 'it>h 1 b5
4 1 'ii'd 2 'iic 7 0-1
Game 52
Dokhoian-Romanishin
Yerevan 1989
1 d4 lLlf6 2 c4 e6 3 lLlf3 b6 4 g3 i.a6 5
b3 i.b7 6 .i.g2 .i.b4+ 7 �d2 c5 8 0-0
0-0 9 .i.xb4 cxb4
1 0 'iid 3!
The most serious move.
1 19
Q u e e n 's In dia n D e fe n c e
a) After 10 tllbd2 d6 1 1 'i'c2 it is obvious
that the queen is less well placed on c2.
Shariyazdanov-Baklan, Neum 2000 contin­
ued 1 1 ...tll c6 1 2 e4 e5 13 d5 tllb 8 1 4 a3 bxa3
(14 ... llla 6!, a suggested of Rabinovich, seems
to be more natural) 15 b4 a5 1 6 c5!? (this
double pawn sacrifice was, of course,
planned, and is rather complex) 1 6 ... lll a6 17
c6 lllxb4 1 8 'i'c3 i.a6 and Black had good
counterplay in this very double edged posi­
tion.
b) 10 a3 and now Black has:
bl) 10 ...bxa3 1 1 lllxa3 a5 12 lll e l i.xg2
was a short draw in Bauer-Baklan, Germany
1 999. It seems that Black should be equal.
There is also 1 2 tllb 5! to consider, fighting
for e4. This seems more dangerous, although
I believe Black is fine.
b2) 10 ... llla6 1 1 e3 'iie7 12 a4!? llle4 13 a5
b5! 14 lll fd2 tll c 3! 1 5 ilc2 i.xg2 16 �xg2
lllx bl 1 7 l:.axbl l:.fc8 with sufficient coun­
terplay for Black in Epishin-Petursson, Bem
1 9 9 1 . 1 1 axb4 lllxb4 12 lll c3 llle4 13 'i'c1
lllxc3 14 'iixc3 a5 15 tlle 5 was a draw in
Yusupov-Benjamin, Saint John 1 988, while
1 1 lll e 5 i.xg2 1 2 <ifiixg2 'i'b8 13 tll d3 bxa3
14 :xa3 'i'b7+ 1 5 �gl d5 1 6 tlld 2 lllb 8 1 7
e 3 lll c 6 was yet another draw in Hracek­
Dorfman, Hamburg 2001 .
1 0 . . . .te4?
Serving only to create problems for Black.
Both alternatives are superior:
a) 1 0 ... d5 and now:
al) 1 1 lllbd2 tllbd7 12 l:.fcl 'i'e7 13 'i'e3
(1 3 lll e5 l:.fc8 14 f4 tll e8 1 5 e4 tll df6 was
more or less equal in Psakhis-Lutz, Baden­
Baden 1 992) 13 ...l:.ac8 14 tll e5 lllxe5 1 5
dxe5 llld7 1 6 cxd5 i.xd5 1 7 i.xd5 exd5 1 8
ll\ £3 :c5 1 9 iid4 l:.e8 with equality (or
something very close) in Nikolic-Amason,
Moscow 1 990.
a2) 1 1 llle 5 lllbd7 12 lllxd7 'i'xd7 13
lll d 2 .i::r. fd8 1 4 l:.fdl dxc4 1 5 lllxc4 with
equality in Gavrikov-Korchnoi, Swiss League
1 995.
b) 10 ... d6 1 1 tllbd2 a5 (1 1 ...lllbd7 12 a3
1 20
bxa3 13 l:.xa3 e5 14 lllh4 i.xg2 1 5 lllxg2
d5!? 1 6 llle3 dxc4 1 7 tlldxc4 gave White an
advantage in Ftacnik-Romanishin, Maribor
1 995) 1 2 l:.adl ilc7 1 3 e4 e5 1 4 lllh4 lllc6
was equal in . Gruenberg-Hjartarson, Novi
Sad 1990. 13 e3!? stays flexible in the centre
and seems to be stronger. After e3-e4 ... e6-e5
Black is always equal.
1 1 'i'e3 �c6 1 2 �bd2 d5 1 3 �xe4?
Later this variation was refuted with 13
cxd5! exd5 (1 3 ... i.xd5 14 'i'd3 gives White
the advantage as e2-e4 is coming) 1 4 llle5
i.xg2 15 lllxc6 'i'd7 16 tlle 5 'i'h3 17 l:.fcl
with a clear advantage for White in Dautov­
Romanishin, Kecskemet 1 989.
1 3 . . .�xe4 1 4 �d2
Preferable is 14 a3 bxa3 15 l:.xa3 lllb4 1 6
l:.cl a5, with equality in Lemer-Romanishin,
Nikolaev 1 995.
1 4 . . . �c3
Now the threat of ... lllxd4 dictates White's
moves, and Black already have an advantage.
1 5 �b1
1 5 ll\£3 seems to be more natural.
1 5 . . . �xd4! 1 6 �xc3 �c2 1 7 'i'c 1 �xa1
18 �b5 a6 19 �d4 �xb3 20 axb3 .:l.c8
21 .:l.d 1 'i'f6 22 e4 .:l.fd8 23 'i'b 1 dxc4?
23 ...dxe4! 24 lll c2 l:.xdl+ 25 'i'xdl :d8
26 'iie l 'i'b2! secures Black a clear advan­
tage.
24 �c6 .:1.xd 1 + 25 'i'xd 1 e5 26 �xb4
cxb3 27 �d5 'i'd6 28 'i'xb3 .:l.c 1 + 29
.tn b5 30 �g2 'i'c5 31 .td3 'i'd6 32

i.c2 'i'c5 33 i.d3 'i'd6 34 i.c2 'i'c5
%-%
Game 53
N ielsen-Anand
Moscow 200 1
1 d4 ltif6 2 c4 e6 3 ltif3
An important point here is that this can
lead to transposition to the Catalan with 3 g3
i.b4+ 4 i.d2 i.e7 5 i.g2 d5 6 �£3 0-0 7 0-0
c6 8 'i'c2 b6 9 b3 i.a6 1 0 l:.dl �bd7, and
you have the move after 10 ... 0-0. Of course
Black can also delay castling in this line but,
as we shall see, this has no positive sides.
3 . . . b6 4 g3 i.a6 5 b3 i.b4+ 6 i.d2 i.e7
1bis is the main line everywhere, al­
though, as we have seen, not Black's only
reasonable line.
7 i.g2
White also has 7 �c3 0-0 8 e4 as an inter­
esting possibility, but time has shown that
after 8 ... d5 9 cxd5 i.xfl 10 <!ixfl exd5 1 1 e5
�e4 1 2 <!ig2 'i'd7 1 3 'i'e2 �xc3 1 4 i.xc3
�c6 1 5 .z:thel �dB 16 �g1 �e6 1 7 f4 g6 1 8
�£3 �g7, as in Tukmakov-Vaganian, Odessa
1 989, Black is fine, although, of course, there
is a lot of play in the position. The main line
is marked by there being very little place for
individual performance or any other kind of
competition between the two players, so
such lines do attract a lot of players who see
chess as 'going to the office'. There are many
4 . . . i. a 6 : Main L in e with 5 b 3
GM between 2450 and 2550 who view chess
as a random profession, playing a sort of
accountant's chess. These people often play
such technical lines, which are equal but not
dead, and then they make their 75% score
without any real risks. Here the position is
just equal.
7 . . . c6
Black has two major alternatives:
a) 7 ... i.b7 8 �c3 0-0 9 0-0 d5 (best -
9 ... �a6 1 0 e3 d5 1 1 'i'e2 c5 1 2 l:lfd1 �c7 13
�e5 i.d6 14 i.el 'ike7 1 5 f4 l:lfd8 1 6 g4 was
better for White in Van der Sterren-Sokolov,
Amsterdam 1 994, while 9 ... c5 10 d5! exd5 1 1
�el �a6 1 2 cxd5 d6 1 3 �d3 �c7 1 4 e4
�d7 15 f4 was quite pleasant for White in
Timman-Franco Ocampos, Mar del Plata
1 982) 10 cxd5 exd5 1 1 i.cl l �a6 1 2 i.b2 c6
13 �e5 i.d6 14 �d3 l:.e8 1 5 l:lc1 �c7 1 6
l:lel �e6 1 7 e 4 with a slight advantage for
White in a typical position, Karpov-Kurajica,
Tilburg 1 994.
b) 7 ... d5 has a reputation for achieving
equality, but a poor .record when it comes to
practical results. The position arising after 8
cxd5 �d5 9 �c3 �d7 1 0 �xd5 exd5 1 1
0-0 0-0 1 2 l:lc1 l:le8 1 3 l:.el c5 1 4 i.e3 i.b7
15 i.h3 cxd4 1 6 i.xd4 �f6, as in Karpov­
Sokolov, Linares 1 987, might be equal, but I
would not like to have such a position with
Black, where White has a simple blockade of
the isolated pawn. All in all it might be equal
from an objective point of view, but Black
will find it hard in practice.
8 0-0 d5 9 'i'c2 ! ?
1bis variation was previously thought to
be harmless, but after 8 i.c3 d5 9 �e5 had
been drained of new ideas, White tried to go
elsewhere to find an advantage. I do not
thing that this is the way, but nor can I sug­
gest another path. Other moves have been
tried here, but none so serious that it de­
serves a mention in a book like this.
9 . ltibd7 1 0 l:d 1
. .
The most precise approach. The alterna­
tives were:
121
Q u e e n 's India n D e fe n c e
a) 1 0 a4 0-0 1 1 aS ltie4 12 axb6 axb6 13
.if4 bS 14 ltibd2 ltixd2 1 5 ltixd2 bxc4 1 6
bxc4 .i.f6 with equality Gurieli-Sokolov,
Basle 2001 .
b) 1 0 .if4 l:tc8 1 1 ltibd2 (1 1 l:tdl is, of
course, still the main line) 1 1 ...cS 1 2 dxcS
ltixcS 1 3 l:tacl 0-0 14 'i'b2 bS 1 5 cxdS ltixdS
1 6 .ieS f6 1 7 .id4 b4 with a better position
for Black in Babu-Prasad, Kuala Lumpur
1 992. But White did not play the best possi­
ble lines. Meanwhile 12 ... i.xcS 13 'i'b2 dxc4
1 4 ltixc4 .ixc4 1 5 bxc4 0-0 1 6 l:tadl 'fie7
was equal in Tjiam-Van der Wiel, Holland
1 993.
1 0 . . . :cs
Unless some improvement is found over
the main game here, Black will have to go
back to castling and slow play: 1 0 ... 0-0 1 1
.if4 l:tc8 1 2 ltic3 and now:
a) 1 2 ... ltihS 1 3 .icl ltihf6 1 4 e4 dxc4 1 5
ltid2 b S 1 6 bxc4 bxc4 1 7 ltia4 (1 7 'fia4 i.bS
1 8 'fic2 'i'aS 19 lti£3 h6 20 .id2 .ib4 21
l:tabl .ixc3 22 .ixc3 i.a4 23 'fid2 'i'hs and
Black has a good position, Van Wely-Adams,
Wijk aan Zee 2001) 17 ... cS 1 8 dS exdS 1 9
exdS ltieS 20 .ib2 .id6 resulted i n a com­
plex struggle in Georgiev-Mitkov, Skopje
2002, where the best player won.
b) 12 ...bS 13 cS b4 14 ltia4 .ibS 1 5 ltib2
ltie4 1 6 ltid3 'i'aS is extremely complex,
Hansen-Emms, Esbjerg 2000. Perhaps Black
is equal, but it is a considerably difficult posi­
tion to handle.
c) 12 ... dxc4!? seems to be the strongest, al­
though both the alternatives are fine. Kram­
nik-Leko, Hungary 2001 continued 13 ltid2
cS 14 dxcS .i.xcS 15 ltide4 ltixe4 1 6 ltixe4
cxb3 17 'i'b2 bxa2 18 l:txa2 .ic4 19 l:txa7
.idS 20 ltixcS bxcS 21 e4 .ic6 22 l:td6 l:ta8
23 'i'a3 l:txa7 24 'i'xa7 'fic8 25 .ifl ltif6 26
.ia6 'i'e8 27 £3 .ibS 28 'i'xcS i.xa6 with a
draw. White obviously had compensation,
but hardly an advantage.
Earlier White has also tried 1 1 a4 cS 1 2
ltic3 dxc4 1 3 bxc4 .ixc4 14 ltieS ltixeS 1 5
dxeS ltidS with a complex but even position
1 22
in Gelfand-Lautier, Biel 2001 . White will
expand with f2-f4 and e2-e4 and send his
knight to d6, while Black maintains his extra
pawn.
1 1 ..tf4! c5?!
11 ... 0-0 still transposes to the 1 Oth move.
1 2 dxc5 .bc5 1 3 lbc3!
This remarkable piece of home prepara­
tion should have refuted Black's set-up and
will, in the future, make Black return to
1 0 ... 0-0. The latest try at the top level looked
like this: 1 3 'i'b2 0-0 14 ltic3 .ib4! 1 5 ltieS
bS 1 6 a3 .ixc3 1 7 'i'xc3 bxc4 and Black had
the advantage in Van Wely-Polgar, Hoogev­
een 2001 .
1 3 . 0-0 1 4 e4!
. .
The notes that follow are based on Peter
Heine Nielsen's annotations in the splendid
Swedish magazine Schacknytt.
1 4 . . . lbg4
14 ... dxc4 15 eS ltig4 16 ltigS g6 17 ltige4
would also give White the advantage.
1 5 exd5
15 l:td2!? is a more careful alternative ac­
cording to Nielsen, but after 1 5 ... eS! it seems
to be just as complicated - only now I prefer
Black.
1 5 . . . lbxf2 1 6 :n e5
1 6 ... ltih3+ 1 7 �hl ltixf4 1 8 gxf4 exdS 1 9
ltig5 ltif6 20 ltixdS g6 2 1 l:tadl gives White a
winning attack. The plan is 'i'c3-h3, with
numerous threats.
1 7 i.. c 1 e4 1 8 lbh4 e3!

In their preparation Peter's team had fo­
cused on 1 8 ...'i'f6 19 i.b2 �d1+ 20 <it>h1
�e3 21 'ii'e2 �fl 22 Axfl 1i'd8 23 �e4,
and White has overwhelming compensation.
1 9 �a4 .i.d4
1 9 ...�3+ 20 i.xh3 e2+ 21 �xc5 exf11i'+
22 i.xfl �xc5 23 i.b2 is not as promising
as earlier exchange sacrifices but still leads to
a clear advantage for White. One line is
23 ... b5 24 �£5 bxc4 25 i.xg7, and White's
attack decides.
20 .i.b2
20 l:.b1 ?! �h3+ 21 i.xh3 e2+ 22 �h1
exfl'ii' + 23 i.xfl i.f6!, followed by ... b6-b5,
gives Black reasonable chances.
20 . . . .i.xb2 21 'ifxb2?
A blunder. 21 �xb2! g6 is necessary ac­
cording to Anand. 22 1i'c31 l:.e8 23 Aae1
leads to a clear advantage to White according
to Curt Hansen.
21 . . . b5 22 �f5 'ifg5
The move White had missed.
23 �d6 bxa4 24 �xca .i.xc8 25 'ifd4
�f6 26 J:ae 1 :ea 27 d6 �h3 + 2a .i.xh3
.i.xh3 29 l:f4 'ifa5 30 J:xe3 :Xe3 3 1
'ifxe3 axb3 3 2 d7 .i.xd7 3 3 axb3 .i.h3
0-1
-----
Game 54
Kamsky-Karpov
Elista 1996
,________________.
1 d 4 �f6 2 c 4 e 6 3 �f3 b6 4 g3 .i.a6 5
4 . . . .i. a 6 : M a in l in e with 5 b 3
b3 .i.b4+ 6 .i.d2 .i.e7 7 .i.g2 c 6 a .i.c3
d5 9 �e5 �fd7 !
9 ... 0-0 10 �d2 �fd7 1 1 �d3 has been
known to be an unnecessary extra possibility
for White since Lputian-Portisch, Manila
1 990. White is slightly better.
1 O �xd 7 �xd7 1 1 �d2 0-0 1 2 0-0 :ea
This has been the main line almost for­
ever, but the Portisch line with 1 2.:.�f6 13
e4 b5 has brought Black good results. Kar­
pov-Timman, Hoogeveen 1999 continued 1 4
Ae1 dxe4 1 5 1i'c2 Ac8 1 6 Aad1 �d5 1 7
i.xe4 (1 7 cxd5 cxd5 1 8 a3 i.xa3 gives Black
good compensation for the pieces, while
Black also had excellent compensation after
1 7 i.b2 f5 1 8 cxd5 cxd5 1 9 1i'b1 b4 20 �fl
1i'b6 21 f3 i.f6 in Karpov-Polgar, Buenos
Aires 2000) 1 7 ... �xc3 1 8 'ii'xc3 i.f6 1 9 1i'c2
g6 20 �£3 bxc4 21 bxc4 'ii'a 5 22 c5 i.b5
with a level game.
12 ... b5?! 13 Ae1 bxc4 14 bxc4 �b6 1 5 c5
�c4 16 e4 led to a plus for White in No­
vikov-Hellers, Copenhagen 1 991 .
1 3 e4 c5
The main move. 1 3 ...b5 14 Ae1 clxc4 1 5
bxc4 bxc4 1 6 1i'a4 i.b5 1 7 1i'c2 l:.e8 1 8 a4
i.a6 1 9 :ad1 i.£8 20 l:.e3 1i'c7?! (20 ... c5 21
d5 exd5 22 exd5 with a slight advantage is
probably better) 21 i.fl c5 22 d5 exd5 23
exd5 Axe3 24 fxe3 �e5 25 1i'e4 and White
was well in control in Grabliauskas-
Shariyazdanov, Swidnica 1 997. 14 ... dxe4 1 5
i.xe4 c5 1 6 cxb5 i.xb5 1 7 d5 exd5 1 8 i.xd5
1 23
Q u e e n 's In dia n D e fe n c e
tiJb6 1 9 �g2 was also favourable for White
in Shipov-Gagarin, Moscow 1 994. After
1 9 ... �f6?! 20 �xf6 ir'xf6 21 l2Je4 'i'g6 22
'i'd2! Black was in trouble, with 'i'a5 threat­
ened.
1 4 exd5 exd5 1 5 dxc5 dxc4 1 6 c6 cxb3
1 7 l:l.e 1 i.. b 5?!
1 7 ...b2! was once thought to be refuted
and · thus replaced with the text move, but
current trends go in the opposite direction.
17 ... b2 is considered in the next game.
1 8 axb3 i..x c6 1 9 i..x c6 l:l.xc6 20 :xa 7
i..f 6
20 .. Jhc3? 21 tiJbl ! and White has a clear
advantage, e.g. 21 ....l:k7 22 !hc7 'i'xc7 23
.l:r.xe7 .l:r.d8 24 'i'd5 'i'c1+ 25 'iii>g2 .l:r.f8 26
:Xd7 'i'xb1 27 .l:r.d6 and Black has a horrible
ending to face which, in practice, is drawn
very rarely.
21 lll c4! i..x c3
21....l:r.xc4? 22 bxc4 �xc3 23 .l:r.e3 and
White wins material. The best chance lies in
21...tiJc51 22 'i'xd8 .l:r.xd8 23 �xf6 gxf6
(23 ... .l:r.xf6 24 b4 t2Je6 25 t2Jxb6 gave White a
clear advantage in Karpov-Beliavsky, Linares
1 993) 24 t2Je3!, an improvement over a pre­
vious game. Van Wely-Karpov, Cap d'Agde
1 996, went 24...ttJxb3 25 tiJf5 .l:r.c1 26 :Xc1
liJxc1 27 .l:r.b7 b5 28 'iii>g2 h5 29 :Xb5, with a
promising ending for White.
22 l:l.xd7 'i'f6
22 ...'i'c8 23 .l:r.e3 �f6 24 'i'd5 b5 25 t2Ja3
b4 26 l2Jc4 h6 27 'iii>g2 .l:r.c7? 28 tiJb6! and
White was clearly better in Khalifman-Van
der Wiel, Ter Apel 1 993. Instead 27 ... .l:r.c5 28
'i'b7 'i'xb7+ 29 .l:r.xb7 �c3, with an endgame
with no other prospects than a fight for a
draw, was the only chance.
23 l:l.e4 'i'f5?
Karpov and his seconds prepared this po­
sition before the game and none of them saw
the coming combination. The most amazing
thing about this oversight is that a similar
combination had already been played in the
following game: 23 ... .l:r.e6 24 .l:r.f4 'i'g6?
(24...'i'xf4 25 gxf4 .l:r.e1+ 26 'i'xe1 �xe1 27
1 24
tiJxb6 is the only chance, but of course Black
cannot hope for anything other than a draw)
25 .l:r.dxf7! 'i'xf7 26 .l:r.xf7 .l:r.xf7 27 'i'd8+ l:r.f8
28 'i'd5 and White won, Chernin-Veingold,
Seville 1 993.
24 l:l.f4 'ii'e 6
25 lldxf7!
Picking up a pawn.
25 . . . :es
The sacrifice could never be accepted:
25 ... l:r.xfl 26 'i'd8+ and 25 ...'i'xfl 26 l:hf7
l:r.xf7 27 'i'd8+ l:r.f8 28 'i'd5+ are both
straightforward winning lines.
26 'i'f3 i..f 6 27 l:l.b7 h6 28 c,i;lg2 �hS 29
h4 c,i;>gS 30 �h2 �hS 31 'i'h5 l:l.dS 32
l:l.f7 i.. d 4 33 llfS+ .::l.xfS 34 l:l.xfS + c,i;lh7
35 'i'f3 i.. c 5 36 l:l.f5 J::.c S 37 h5 .:I.dB 38
l:l.e5 'i'd7 39 'i'e4+ �hS 40 �g2 l:l.fS 41
f4 l:l.dS 42 'i'f3 i..d 4 43 l:l.e2 b5 44 llld 2
i.. b 6 45 llle4 'ii'd 1 ? ! 46 lllf 2 'i'b 1 47 lllg 4
'i'f5 48 llle 5 �g8? 49 lllc 6 1 -0
Game 55
Gheorghiu -Cserna
Berlin 1986
1 d4 lllf 6 2 lllf 3 e6 3 c4 b6 4 g3 i.. a 6 5
b3 i.. b4 + 6 i.. d 2 i.. e 7 7 i.. g 2 c6 8 i.. c 3
d5 9 llle 5 lllfd7 1 0 lllx d7 lllx d7 1 1 llld 2
0-0 1 2 0-0 l1c8 1 3 e4 c5 14 exd5 exd5
1 5 dxc5 dxc4 1 6 c6 cxb3 1 7 l:l.e 1 b2!
Once refuted, now the drawing line, kill­
ing all life in the position.

1 8 .i.xb2 lDc5
18 ... �f6 19 �b3 .tb4 20 l:.e3 'i'�d1+ 21
.l:xd1 l:.fd8 22 .l:d4! gives White a clear ad­
vantage due to the pawn on c6.
1 9 .i.a3
1bis was the refution of the ... b2 line, but
improvements have been found. Others:
a) 19 �b3!? �d3 20 l:e2 .tb4! 21 .td4
.tc5! was a draw in Gelfand-Anand, China
2000. The reason is 22 �xc5 (22 .tc3? �f4!
23 .l:d2 �e2+ 24 'itih1 'i'e7! 25 .tb2 .txf2 is
enormous for Black) 22... bxc5 23 .tc3 .tb5!
24 .l:e3 c4 25 'i'g4 g6 26 a4 i.xc6 27 'i'xc4
.txg2 28 'i'xd3 with equality. Here 21...�c5?
22 .l:c2! .td3 23 �xc5! .txc2 24 1i'g4 gave
White a strong position in Dautov-Sax,
Germany 1 998.
b) 19 1i'g4?! .tf6 20 .txf6 'i'xf6 gave
White nothing but problems in Timo­
schenko-Bagaturov, Enakievo 1 997.
c) 1 9 �c4 .txc4 (White is better after
both 1 9 ... �d3 20 .l:xe71 .txc4 21 .l:d7 �xb2
22 .l:xd8 �xd1 23 l:.dxd1!, and 1 9 ... .tf6? 20
'i'xd8 .l:fxd8 21 .txf6 gxf6 22 �e3) 20 'ii'g4
.tg5! 21 'ii'xc4 �d3 with a further branch:
cl) 22 .tc3 �xe1 23 l:.xe1 .l:e8 24 .l:xe8+
(24 .l:b1 .l:e6 25 h4 .tf6 26 .tb4 'i'd4 Barus­
Shneider, Jakarta 1 997 and Black has a clear
advantage according to Khalifman)
24 ...'i'xe8 25 'ii'g4 .th6, planning to return
the exchange, gives Black a better game
4 . . . .i. a 6 : Main L in e with 5 b 3
according to Khalifman.
c2) 22 l:.e2? �xb2 23 l:.xb2 .tf6 wins for
Black.
c3) 22 i.e5! �xe1 23 :Xe1 .tf6! 24 .txf6
(Komarov gives 24 .th3? 'ii'd2 25 'iif fl l:.ce8)
24...'i'xf6 25 c7 and a draw was soon agreed
in Georgiev-Komarov, Yugoslavia 1 997.
1 9 . . . .i.d6?
19 ... .tb5?! 20 �b3 .tf6 21 �xc5 bxc5
was Gheorghiu's recommendation, but after
22 l:.c1 ! 'i'xd1 23 .l:exd1 .txc6 24 .txc5
White seems to have some chances. The a­
pawn is probably falling.
1 9 ... .tf6! 20 l:.ct .tg5 21 l:.c2 i.b5 22
�b3 'ii'xd1 23 .l:xd1 .txc6 24 �xc5 .txg2
25 rJ:;xg2 .te7 26 .l:dc1 .l:tfe8 27 .l:tc3 bxc5 28
.txc5 .txc5 29 .l:xc5 l:.xc5 30 .J:.xc5 g6 31
l:r.c7 l:r.e2 32 l:xa7 h5 33 'iif f3 l:r.c2 34 h4 r/;g7
35 l:r.a6 was the course of Dautov-Alterman,
Germany 1 998. 1bis endgame is a theoretical
draw. Black can even lose a tempo once the
a-pawn starts rolling and still make the draw.
However, many players find it difficult to
draw such endings in practice. One example
is Georgiev, who lost it to Akopian in Las
Vegas 1 999. That game can be found in my
book Excelling at Chess, page 1 04. Alterman
also lost this endgame, but those intending to
play this line should look this ending up and
then they would find that there is nothing to
fear. 20 .txc5 bxc5 21 l:r.c1 .tb5 22 l:.xc5
was agreed drawn in Shneider-Anand, Bastia
2000. Taking the difference of strength of
the two players into account, one might sus­
pect that Black was not altogether confident
about his position, but I think that if Anand
really wanted to win against Shneider, he
would have chosen a less forcing system.
20 lDb3 lDd3
20 ...�xb3 21 .txd6 �xa1 22 .txf8 and
White can hope for a considerable advantage.
2 1 .i.xd6 'i'xd6 22 lDc 1 ! J:fd8 23 lDxd3
.ixd3 24 1i'a4 a5 25 J:ad 1 'i'c7 26 .i.h3
b5 27 1i'h4 1 -0
1 25
Q u e e n 's In dia n D e fe n c e

Summary
After 5 b3 dS 6 i.g2, game 45 shows that 6 ... dxc4?! is past its sell-by date. White has found the
right path to a small but very clear advantage. After S ... bS!? White will always be a little better,
as his pawn structure is slightly superior. Still this line has brought many players good practical
results, and is good for winning attempts, as the position easily becomes unbalanced. Also
interesting is 5 ... i.b4!? with the idea to exchange on d2. As far as I can see this line is theoreti­
cally okay, even though the practically results indicate that it is easier to play the White side.
Both 7 ... aS!? and 7 ... cS!? are probably very slightly worse for Black, but again they give posi­
tions where there remains much play, and where the better player most often succeed. The
long lines with 5 ...i.b4+ and 6 ... i.e7 seems to be the main reason for Black to play 4...i.a6.
There are good reasons for this: If White plays the tough lines, Black can liquidate the game to
a draw, and if not, Black is probably fine all the same.

1 d4 liJf6 2 c4 e6 3 liJf3 b6 4 g3 ..ta6 5 b3 (DJ i.. b4 +

5 ... dS 6 i.g2 dxc4 - Game 45
5 ... bS - Game 46
5 ... i.b7 6 i.g2 i.b4+ 7 i.d2 (DJ
7 ... aS 8 0-0 0-0
- 9 i..gS Game 48
9 ltJc3
. - 9 . . dS Game 49; 9 . d6 - Game 50
. .

7 ... cS 8 0-0 0-0

- 9 i.. c 3 Game 51; 9 i.xb4 cxb4 10 'it'd3 Game 52
-

6 ..td2 ..te7
6 ... i.xd2+ - Game 47
7 i.. g 2 c6 8 ..tc3
8 0-0 dS 9 'ii' c2 - Game 53
8 . . . d 5 9 ltJe5 ltJfd7 1 0 ltJxd7 liJxd7 1 1 ltJd2 0-0 1 2 0-0 .l:c8 1 3 e4 c5 1 4 exd5 exd5
1 5 dxc5 dxc4 1 6 c6 cxb3 1 7 :e 1 (DJ
1 7 ... i.. b S - Game 54; 1 7 ... b2 - Game 55

5 b3 7 ..td2 1 7 .l:.e T

1 26
 CHAPTER EIGHT
4 g 3 i.b7
1 d 4 lbf6 2 c 4 e 6 3 lt:if3 b6 4 g 3 .i.b7
4 ... J..b 7 is the old classical move, which
was much played in the days when few
people believed in such modem moves as
4... J..a 6. In this chapter we will consider all
the classical lines as well as some of the
sidelines. As you will find out, if you choose
to play anything other than 5 J.. g2 J.. e 7 6
tLlc3 then there is really no way to play for
an advantage, or even create an interesting
game. All these d5, b3 and l:e1 lines are not
dangerous for Black, once they have been
looked at under the microscope.
So the real battleground is in the main
lines and here it is a matter of taste. All lines
should lead to an even game. But what kind
of even game suits you best? Take a look
and then decide.
Game 56
Karpov-Gavrikov
USSR 1988
1 lt:if3 lt:if6 2 c4 b6 3 d4 e6 4 g3 .i.b7 5
J.g2 c5
This move is more or less refuted (this
game). Other alternatives are:
a) 5 ... J.. b4+. The difference with the
Bogo-Indian is hard to find after this. After 6
i.d2 Black has four possibilities:
I
al) 6 ... c5 7 J..xb4 cxb4 8 0-0 0-0 9 tLlbd2
d6 10 l:te1 'ilc7 1 1 e4 e5 is very similar to
Game 52, except that White has not played
b2-b3. This is an advantage that secures
White the better prospects in this position.
12 'ilb3 a5 13 'ile3 tLlc6 and now, instead of
14 d5, which was seen in Gavrilov-Seferjan,
Rjazan 1 993, Gavrilov recommends 14 h3
ltfe8 1 5 d5 tLlbB 1 6 tLlh4 J..c 8 1 7 @h2 tLla6
1 8 f4 'ilc5 1 9 'ile2 with a slight pull.
a2) 6 ... a5 7 0-0 0-0 8 i.f4 J.. e7 9 tLlc3 tLle4
· 10 'ilc2 tLlxc3 1 1 bxc3 l:a7 12 e4 d6 13 l:tfe1
tLld7 14 l:tad1 'ile8 with a modest advantage
to White in Piket-Korchnoi, Nijmegen 1 993,
after 15 a4.
a3) 6 ... J..e7 7 tLlc3 0-0 (7 . . c6 8 e4 d5 9
.
cxd5 cxd5 10 e5 tLle4 1 1 0-0 0-0 12 lte1
tLlxd2 13 'ilxd2 tLlc6 14 J.. f1 favoured White
in Savchenko-Hulak, Portoroz 1 996) 8 0-0
tLla6 9 l:tc1 lte8 10 l:e1 d5 1 1 cxd5 exd5 12
J.. f4 c5 1 3 'ila4 tLle4 1 4 lted1 with another
edge in Filippov-Kalinin, Moscow 2002.
a4) 6 ...'ile7 7 0-0 i.xd2 8 'ilxd2 0-0 9
tLlc3 d5 10 tLle5 tLla6 1 1 l:tfd1 , Ilincic­
DamljanO'vic, Niksic 1 997, and we have yet
another of these typical positions where
White is only very slightly better.
b) 5 ... g6?1 6 0-0 J..g7 7 'ilc2 0-0 8 tLlc3 d6
9 e4 tLlbd7 1 0 b3 e5 1 1 i.b2 exd4 1 2 tLlxd4
l:e8 13 l:tfe1 simply led to a clear plus for
127
Q u e e n 's In dia n D e fe n c e
White in Milev-Simagin, Russia 1 959. The
weakening of the light squares on the queen­
side does not work well within such a King's
Indian structure.
6 d5!
White asswnes control of the centre. Ei­
ther Black will have to live with an undesir­
able structure or fight for d5. As this game
shows, White is 'fitter' for this fight.
6 . . . exd5 7 ltlh4 b5 8 0-0!
8 cxd5 merely justifies Black's play this far.
8 . . . bxc4 9 ltlc3 i.e7 1 0 ltlf5!
1 0 i.. g5?! is inferior. Now White clearly
has the advantage.
1 0 . . . 0-0 1 1 ltlxe7 + ! 'it'xe7 1 2 i.g5 h6
1 2...ll:la6?! 13 lllxd5 i..xd5 1 4 i..xd5 .l:lab8
1 5 Lc4 secures White a very clear advan­
tage. The knight is badly placed on a6.
1 3 i.xf6 'it'xf6 14 ltlxd5 i.xd5 1 5 'it'xd5
ltlc6 1 6 'it'xc4 'ii'x b2 1 7 e3
1 28
White will soon regain the pawn, after
which the superiority of both his bishop and
his pawn structure will guarantee an advan­
tage.
1 7 . . .fl.ab8 1 8 Wxc5 .l:l.b6?!
1 8 ... .l:lb5 is better. This slow play presents
White with an opportunity to complete de­
velopment without any concerns.
1 9 fl.ad 1 ltlb8 20 i.d5!
Now the game is more or less decided.
The rest, as they say, is a matter of tech­
nique ...
20 ... 'it'b5 2 1 'flc7 'fla6 22 fl.c1 'it'a5 23
fl.fd 1 fl.b5 24 'ifd6 'flb6 25 'fle7 'it'g6 26
i.e4 'ii'e 6 27 i.h7 + �xh7 28 'flxf8 'flxa2
29 'it'd6 a6 30 'ii'd 3+ f5 31 fl.b1 'ii'e 6 32
fl.xb5 axb5 33 'flxb5 ltlc6 34 fl.d5 �g6
35 'it'c5 'fle4 36 fl.d6+ �h7 37 'it'd5
'ii' b 1 + 38 'ii'd 1 'ii'e4 39 'ifd3 'it'g4 40 fl.d5
1 -0
Game 57
Polugaevsky-Wojtkiewicz
Haninge 1990
1 d4 ltlf6 2 ltlf3 e6 3 c4 b6 4 g3 i.b 7 5
i.g2 i.e7 6 0-0 0-0 7 b3 d5
8 cxd5 exd5
Black has an easier way to equalise in
8 ... ll:lxd5! 9 i..b 2 ll:ld7 (9 ... c5 is a good
alternative: 10 lll c3 cxd4 1 1 lllxd5 i..xd5 1 2
lllxd4 i..xg2 13 �xg2 i.. f6 1 4 e3 i..xo4 1 5
i..xd4 gave White some advantage in Aa-

gaard-Christensen, Copenhagen 1 999, but
simply 14 ...�ds+! would have kept the game
drawish) 1 0 lDbd2!? c5 1 1 .l:r.cl .l:r.c8 1 2 lDc4
ltJ5f6 with equality in Dautov-van der Ster­
ren, Ter Apel 1 994.
9 .i.b2
9 lDc3 lDa6 10 ii.b2 �c8 11 .l:r.cl .l:r.d8 12
a3 c5 13 dxc5 bxc5 14 b4 �d7 1 5 bxc5
lDxc5 1 6 lDe5 �e6 17 lDa4 ltJfe4 1 8 lDxc5
ii.xc5 19 �a4 ii.b6 20 4:)£3 .l:r.ac8 led to
equality in Nikolic-Khalifman, Manila 1 990.
9 . . .:ea 1 O lt:lc3 lt:la6
10 ... �c8, with equality, is mentioned
above.
1 1 lt:le5 c5 1 2 .l:lc 1 �f8
12 ... cxd4 1 3 lDb5! favours White.
1 3 e3 lt:lc7 1 4 lt:la4!?
White has emerged from the opening with
a small plus.
1 4 . . . lt:le6 1 5 dxc5 bxc5 1 6 lt:lg4! lt:le4? !
1 6 ...lDxg4! 1 7 �xg4 �e7 1 8 .l:r.fdl .l:r.ed8
was better according to Polugaevsky.
1 7 f3 lt:ld6 1 8 f4! lt:le4
1 8 ... d4!? was apparently necessary.
1 9 lt:lf2! lt:lt6 20 f5 lt:lg5 21 lt:lxc5 .i.xc5
22 .:1.xc5 lt:lge4 23 lt:lxe4 lt:lxe4 24 'ifd4
f6 25 .:l.c2 'iVd6 26 .l:lfc 1 .l:le7 27 'iVa4
'iVd8 28 �a3 .l:lf7 29 .l:ld 1 'ifb6 30 'ifd4
'ifa5 31 .i.b4 'iVb5 32 �xe4 ! dxe4 33
.l:lc7! ! .l:lxc7 34 'iVd8 + 'ife8 35 'ifxc7 .i.c6
36 .:ld6 .i.b5 37 a4 :ea 38 'iVa5 .i.d3 39
'iVd5 + 'it>h8 40 .:l.d7 h6 41 g4! 'ife5 42
'ifxe5 fxe5 43 .i.d6 .l:lc 1 + 44 'itg2! .l:ld 1
4 g3 �b 7
45 l:!.xa7 .i.f1 + 46 'it>f2 .l:lxd6 47 'itxf1
.:l.b6 48 a5 .l:txb3 49 a6 1 -0
Game 58
Freitag-O ral
Graz 1998
1 d4 lt:lf6 2 c4 e6 3 lt:lt3 b6 4 g3 .i.b 7 5
.i.g2 .i.e7 6 0-0 0-0 7 d5!?
This pawn sacrifice had its heyday in the
1 980's. These days it is not considered to be
,
very dangerous but it must be remembered
that it is more messy than the main lines.
7 . . . exd5
Also possible is 7 ...lDa6!?, as tried in Haus­
rath-Tiviakov, Zwolle 2000. I suppose this
was employed as a means to steer the lower
rated opponent out of his 'book' ... 8 lDc3
ii.b4 9 lDe5 ii.xc3 10 bxc3 lDc5 1 1 ii.a3 �e7
12 lDd3 d6 13 lDxc5 dxc5 14 e4 e5 1 5 .l:r.e 1
lDe8 1 6 ii.fl ltJd6 and Black was doing well.
8 lt:lh4!
This was Polugaevsky's discovery after the
following game: 8 lDd4?! ii.c6! 9 cxd5 ii.xd5
10 ii.xd5 lDxd5 1 1 e4 ltJb4 1 2 lDc3 ii.f6 1 3
lD 5 .l:r.e8 14 f4 d 6 1 5 �g4 lD8c6 1 6 e 5 dxeS
17 ltJe4 exf4 1 8 ltJh6+ �fB 1 9 lDxf6 �xf6
20 .l:r.xf4 .l:r.el+ 21 �g2 lDe5 22 .l:r.xf6 lDxg4
23 :Xf7+ �e8 24 :Xg7 lDxh6 25 a3 .l:r.d8 26
axb4 .l:r.d7 27 .l:r.xd7 �xd7 and the endgame is
a disaster for White, as was seen in-the game
Polugaevsky-Korchnoi, Buenos Aires 1 980.
1 29
Q u e e n 's In dia n D e fe n c e
8 . . . c6!
The only good move. After 8 ... .!lle4 9
cxd5 .i.xh4 10 .i.xe4 .i.f6 1 1 'i'c2 White was
slightly better in Gheorghiu-Unzicker, Wijk
aan Zee 1 981 . 8 ... .!lle 8 9 lll f5 llld6 1 0 .!llxe7+
'iii'xe7 1 1 cxd5 c5 12 .!llc3 .!lla6 13 .i.f4 fa­
voured White in Barlov-lvanovic, Yugoslavia
1 984.
9 cxd5 �xd5
Also good is 9 ... cxd5!? 10 .!ll c3 .!lla6 1 1
.!ll f5 .!llc7 12 .tf4 :e8 13 .i.d6 i.£8 1 4 'iii'b 3
.i.c6 1 5 i.xf8 :xf8 1 6 l:lad 1 :e8 17 l:lfe 1
l:le5, with equality in Romanishin-Razuvaev,
Jurmala 1 987. In fact I think this is probably
the safest line for Black. On the other hand,
it does not offer much in the way of winning
chances.
1 0 �f5 �c7
Or 10 ...i.f6!? 1 1 e4 .!lle7 1 2 .!lld 6 .i. a6 1 3
l:.e1 i.e5 14 f4 .i.xd6 1 5 'iii'xd6 c 5 1 6 b4
cxb4 1 7 'iii'xb4 .!llbc6 1 8 'i'a4 'i'c8 with un­
clear play in Milos-Sunye Neto, Sao Paulo
2000. Actually I would prefer to be Black in
this line, and in the game, too, but I know
that a creative player will also be able to get
something out of White's position.
1 1 �c3
One of the two main lines. 1 1 e4 d5 1 2
lte t l i s Barlov's variation, which appears to
lead to an unclear but equal game: 1 2... dxe4
13 lllc 31? i.c8 14 .!llxe7+ 'i'xe7 1 5 .i.xe4
.i.e6 1 6 .tf4 :d8 1 7 'i'h5 h6 1 8 :e3 was
Piket-Van der Wiel, Rotterdam 1 997, with
1 30
compensation for the pawn, but Black has
his chances too and did okay.
White had previously played 1 3 'iWg4 i.f6 14
.!llc3 i.c8 1 5 .i.xe4, but after 1 5 ... .i.xf5
(1 5 ....!lld 5? 1 6 .th6 g6 17 .i.xf8 <ittx f8 1 8
.i.xd5 .i.xf5 1 9 'i'b4+ .i.e7 20 'iii'b3 cxd5 21
.!llxd5 .i.e6 and Black resigned in Barlov­
Abramovic, Yugoslavia 1 994 - simply 22
:Xe6 fxe6 23 'ii' f.3+ followed by .!llxe7 wins)
1 6 'i'xf5 g6 17 'i'f.3 .!ll e6 1 8 i.h6 .!ll d4 19
'ii'g2 l:le8 20 .i.e3 a6 21 .!lle2 l:.a7 22 lllxd4
.i.xd4 23 .i.xd4 'ii'xd4 24 .i.xc6 :Xel+ 25
l:lxel .!llxc6 26 'iii'xc6 <Jl;g7 27 b3 there was
little life left in the position in Pigusov­
Tiviakov, Beijing 1 997. Of course a man like
Pigusov is well known for his love for ex­
changes. Like my first trainer, who said that
the sacrifice of a pawn would only make you
try to win it back, so therefore it was a waste
of time ...
1 1 . . .d 5
1 1 ...lll e 8 12 .i.f4 .!ll a6 1 3 'i'd2 d5 14 e4
.!llac7 1 5 l:r.ad1 i.f6 1 6 exd5 lllxd5 1 7 lllxd5
cxd5 1 8 llle3 .!llc7 19 i.xc7 'ii'xc7 was al­
ready drawn in Timman-Karpov, Tilburg
1 983.
1 2 e4 lle8! ?
The new move. Previously known was
1 2 ... .i.f6 13 :et i.c8 14 'ii'f.3 .i.e6 1 5 g4
.!lld7 16 'i'g3 dxe4 1 7 .i.xe4 with compensa­
tion but no more, Romanishin-Epishin, Bel­
grade 1 988.
13 .i.f4 �e6 14 'i'g4 �d 7

1 5 e5
This looks like a positional error, and now
Black does not have to worry about his cen­
tre.
1 5 . . . g6 1 6 lll h 6+ 'ot?g7 1 7 tllt 5+ �g8 1 8
tll h 6+ 'ot?f8 1 9 :te 1 j,,g 5 20 h4 j,,xf4 2 1
gxf4 f5 22 exf6 tllx f6 2 3 'iff 3 j,, c 8 24
:e5 �g7 25 :ae1 �xh6 26 f5 gxf5 27
:xt5 �g7 28 :e3 a5 29 'ifg3+ �h8 30
'ife5 tllg 7 0-1
--------..
Game 59
Van Wely-Korchnoi
Hoogeveen 200 1
1 d 4 tllf 6 2 c4 e 6 3 tllf 3 b 6 4 g 3 j,, b 7 5
j,,g 2 j,,9 7 6 0-0 0-0 7 :e 1
This line was very popular in the mid-
1 990's, when Kramnik championed it. It is
now known to be harmless in view of...
7 . . . tlla 6!
It is this move that has more or less done
away with any modem interest in 7 :et .
Instead 7 ... dS 8 cxdS exdS 9 ttlc3 might give
White a little plus.
8 tlle 5
Other attempts at an advantage are:
a) 8 ttlc3 ttle4 9 ttlxe4 (9 i.d2 dS 10 cxdS
exdS was equal in Barsov-Llgterink, Hoogev­
een 1 997) 9 ... .ixe4 1 0 'ii'a4 'Wc8 1 1 ttle5
.ixg2 12 �xg2 cS! 13 dS?! exdS 14 cxdS
ttlb4! 1 5 .l:.dl ! d6 1 6 ttlf3 'itb7!? 1 7 a3 bS 1 8
'itb3 ttla6 with a good game for Black, Polul-
4 g 3 j,, b 7 .
jahov-Tiviakov, St. Petersburg 1 998. White
has not been able to establish the powerful
centre that secures him play in Benoni struc­
tures. Necessary here was 1 3 .ie3 'itb7+!? 1 4
ttlf3!? cxd4 1 5 .ixd4 with complete equality.
10 ttle5 has also been played, Bator­
Stefansson, Mariehamn 1 997, continuing 10
ttle5 hg2 11 c;l.>xg2 'Wc8!? 1 2 e4 1i'b 7 (al­
ready equal) 1 3 ttlxd7 :fd8 1 4 ttleS .if6 1 5
.if4 .ixe5 1 6 .ixe5 f6 1 7 .if4 e 5 1 8 .id2
.l:.xd4 and Black had improved his prospects.
10 d5!? is another try. Then the safe
1 0 ... exd51? 11 cxd5 c5 1 2 dxc6 dxc6 1 3 ttle5
.ixg2 14 �xg2 'Wc8 leads to approximate
equality, while the ambitious 1 o ... .if6 1 1
!LJd4 hg2 1 2 c;l.>xg2 leaves Black with a
modest lead in development, meaning he
must act quickly in order to avoid e2-e4-e5.
Delchev-Marin, Istanbul 2000 went 12 ... c5
1 3 ttlb5 exd5 1 4 'Wxd5 ttlb4 15 'Wdl a6 16
ttld6 b5 1 7 e4 bxc4 1 8 e5 ttld3! 19 exf6 'ii'xf6
20 .if4 g5 21 'Wg4 ttlxf4+ 22 gxf4 'Wxd6 23
'ii'xgs+ 'Wg6 with a draw.
b) 8 a3!? is an interesting try. White is aim­
ing for the standard position and invests a
tempo to do so. 8 ... d5 (8 ... c5 9 d5 exd5 10
ttlh4 ttlc7 11 ttlc3 d4 1 2 .ixb7 .l:.b8 1 3 .ig2
dxc3 1 4 bxc3 leaves all the white pieces well
placed) 9 cxd5 exd5 10 ttlc3 c5 1 1 .if4 ttlc7
12 l:tct ttle6 13 .ie5 ttlg4 was seen in
Wojtkiewicz-Zagrebelny, Dhaka 1 999. In
many ways this fits in with White's plan to
increase pressure on the d5-pawn. White was
willing to surrender the bishop for the knight
and now it is Black who expends energy to
arrange this. 13 ...'ii'd 71? is a sensible alterna­
tive, preparing to put more weight on the d­
file. After (1 3 ...ttlg4) 14 dxc5 bxc5 1 5 h3
ttlxeS 1 6 ttlxeS i.f6?! 1 7 ttlg4 .ixc3 1 8 bxc3!
White had a slight but distinct advantage, but
the improvement 1 6 ... ttlc7!?, keeping the two
bishops, would have given Black a fine game.
1 2 .ixc7 'Wxc7 1 3 :et 'ii'd7 1 4 lileS 'We6
1 5 ttld3 c4 1 6 ttleS .l:.ab8 17 e4 dxe4 1 8
ttlxe4 .ixe4• 1 9 .ixe4 ttlxe4 20 l:txe4 'ii'dS 2 1
'ii'e2 f5 22 .l:.e3 b S wa s the course o f Ibragi-
131
Q u e e n 's In dia n D e fe n c e
mov-Pushkov, Elista 2001 , with a level game.
Tb.is shows that the idea is just to get a posi­
tion and then play, rather than attempt to
refute Black's plans.
8 . . . d5!
The improvement over 8 ... .ixg2. The dif­
ference with 7 ... dS is that White's knight is
not very well placed on eS so early, so Black
has no problems in equalising.
9 cxd5 exd5 1 0 lll c 3 c5 1 1 i.f4 llle4 1 2
lllxe4 dxe4 1 3 dxc5 ll:\xc5 1 4 1i'c2 1i'c8
Black has equalised. The ES-square is a
brilliant outpost for the queen, as is e6.
1 5 :ed 1 g5 1 6 i.e3 1i'e6 1 7 ll:\c4 :ac8
1 8 b3 h6 1 9 i.d4 i.d5 20 i.b2 :td8 2 1
:ac 1 b5 2 2 1i'c3 f 6 2 3 llle 3
23 . . . lll x b3 24 Wxc8 :xc8 25 :xc8+
Wxc8 26 :xd5 lll c 5 27 lllf 5 i.f8 28 i.h3
c,t>h7 29 c,t>g2 h5 30 llle 3 Wc6 3 1 i.f5 +
c,t>g7 32 i.d4 c,t>f7 3 3 i.xc5 i.xc5 34
1 32
:d1 + c,t>t8 35 llld s i.h6 % - %
Game 60
Fom inyh-Chernyshov
Smolensk 2000
1 d4 li:lf6 2 c4 e6 3 lllf3 b6 4 g3 i.b 7 5
i.g2 i.e7 6 0-0 0-0 7 ll:\c3 ll:\a6 ! ?
An interesting sideline. Tb.is illustrates
how good this is against 7 l:tel . When it is
good here, it must be brilliant there! 7 ... dS 8
llle S lll a6 9 .if4 transposes to 8 .if4 in the
note to White's 8th move.
8 a3
Others:
a) 8 b3 l:te8 9 i.b2 dS 10 lll eS cS 1 1 e3
'iic 7 12 cxdS exdS 13 lZ:lbS 'iic 8 14 a3 lllc7
led to a balanced game in Zaremba­
lbragimov, Chicago 2002.
b) 8 �f4 d6! (8 ... dS 9 lZ:leS cS 10 cxdS
exdS 1 1 l:tcl lll e4 12 lllxe4 dxe4 13 dxcS
lZ:lxcS 14 lll c4 lll e6 1 5 �eS f6 16 llld6 �xd6
17 .ixd6 gave White a little something in
Paunovic-Tiviakov, Cacak 1 996) 9 a3 cS 10
dS lllc 7 11 e4 a6 1 2 b4 exdS 1 3 exdS cxb4 14
axb4 bS with a complicated game with
chances for both sides in Ribli-Markowski,
Moscow 1 994.
c) 8 'iia4 'iic 8! (protecting the knight on
a6 just once more) 9 J:tdl lll e4 1 0 lllxe4
.ixe4 1 1 .if4 d6 1 2 l:tacl 'iib 7 1 3 llle l
�xg2 1 4 lllxg2 cS with equality in Tatai­
Tiviakov, Arco 1 998.

d) 8 .i.g5 c5 9 d5 exd5 10 cxd5 d6 1 1 liJh4
1i'd7 12 1i'd2 b5 13 e4 b4 14 liJd1 .i.d8,
Shneider-Tiviakov, Moscow 1 992, was very
messy.
8 . . . ltie4 9 �d2 f5 1 0 b4 �f6 1 1 :b 1
1 1 l:r.c1 c5 1 2 dxc5 bxc5 1 3 b5 liJc7 does
not look dangerous for Black.
1 1 . . .ltixc3
1 1 ...c5 12 .bxc5 liJaxc51 13 dxc5 liJxc3 1 4
.i.xc3 .i.xc3 1 5 1i'd 6 .i.c6 i s another route to
equality.
1 2 �xc3 �e4 1 3 :b3
1 3 . . . c6!
Part of a magnificent but not too complex
plan. Black wants to establish a giant knight
on d5.
1 4 �b2
14 b5?! cxb5 15 cxb5 liJc7 favours Black.
1 4 . . . ltic7 1 5 �h3
15 a4 b5 16 axb5 cxb5 17 c5 liJd5 18 e3
a5! and Black has a positional plus.
1 5 . . . b5! 1 6 ltid2 bxc4 1 7 ltixc4 �d5 1 8
'i'd3 'i'b8 1 9 �g2
19 £3!? Wb5 20 l:lc1 l:r.fb8 21 e4 .i.xc4 22
1i'xc4 fxe4 23 fxe4 a5! leads to equality.
1 9 . . .'i'b5 20 :c 1 :tb8 21 f3 a5!
Black has the initiative. White was not able
to counter quickly enough.
22 e4 �xc4 23 :xc4 d 5 ! 24 exd5 exd5
25 :c5 'i'xd3 26 :xd3 axb4 27 :xc6
bxa3 28 �xa3 :b 1 + 29 �c 1 ltib5
Black's pieces are literally springing to life
with powerful effect.
4 g3 �b 7
30 f4 �xd4+ 3 1 �f1 :aa 1 32 �xd5+
�8 33 :d 1 �c3 34 �e2 :xc 1 35 :xc 1
:xc 1 36 �d3 �b2 37 :b6 :d 1 + 38
�c4 ltia3+ 0-1
Game 61
Andersson -Karpov
Ti/burg 1983
1 ltif3 ltif6 2 c4 b6 3 g3 �b7 4 �g2 e6
5 0-0 �e7 6 d4 0-0 7 ltic3 ltie4 8 'i'c2
This line is so classical that many players
still play it with some odd kind of respect for
the old masters. The newer and more dan­
gerous 8 .i.d2 is treated in the following
games.
8 liJxe4 .i.xe4 9 liJh4 is the ultimate draw­
ing line. After 9 ... .i.xg2 1 0 liJxg2 d5 1 1 'ii'a4
dxc4 1 2 1i'xc4 c5 13 .i.e3 cxd4 . 1 4 J.xd4
1i'c8 the position is completely equal. Apart
from a blunder once deciding a game be­
tween Polugaevsky and Andersson (White), I
cannot recall this line being dangerous for
anything other than one's interest in chess.
Yet this characteristic is precisely the reason
why some people play it. By somehow reach­
ing a technical position where they cannot
create any winning chances they leave the
floor free, as it were, for their opponents to
create losing chances. Andersson-Karpov,
Reggio Emilia 1 991 continued 1 5 l:.fc1 Wxc4
1 6 l:lxc4 liJa6 1 7 liJf4 :fd8 1 8 .i.e3 l:.d7 1 9
liJd3 f6 2 0 l:lac1 l:lad8 21 <iltg2 <J;; fl with
equality.
8 . . . ltixc3 9 'i'xc3 c5
This is the most drawish line here. Equally
good is 9 ... £5, e.g. 10 d5 .i.f6 1 1 'ii'c2 liJa6 1 2
l:ldl 1i'e7 1 3 liJd4 liJc5 1 4 .i.e3 .i.xd4 1 5
.i.xd4 d6 1 6 b4 liJd7 1 7 a4 a5 1 8 bxa5 l:lxa5
with equality in Krogius-Kholmov, Kiev
1 964. Meanwhile Andersson-Akesson,
Skelleftea 1 999 went 10 b3 .i.f6 1 1 .i.b2 d6
1 2 l:ladl a5 13 liJe1 .i.xg2 14 liJxg2 liJc6 1 5
'ii'd2 'ii'd7 1 6 d 5 liJd8 1 7 .i.xf6 l:lxf6 1 8 dxe6
liJxe6 1-9 liJf4 liJxf4 20 'ii'xf4 l:le8, again with
complete equality. It should be said such
1 33
Q u e e n 's In dia n D e fe n c e
positions can be seen a million o r s o times in
GM games, and some players, like Anders­
son, continue to play them because they
know them so well.
1 0 l:ld 1
Black also has nothing to fear after the
line 1 0 b3 i.f6 1 1 i.b2 cxd4 1 2 liJxd4 i.xg2
1 3 @xg2 liJc6 14 'i'e3 'i'c8 1 5 l:tfdl l:d8, as
in the encounter Beliavsky-Adams, Belgrade
1 995.
1 o . . . d6 1 1 b3 .i.t6 1 2 .i.b2 tlld 7
Ibis is the most flexible choice, but an­
other route to equality is well known, namely
1 2 ... 'i'c7 1 3 'i'd2 l:d8 14 dxc5 dxc5 1 5 'i'f4
liJa6 1 6 'i'xc7 ltJxc7 17 ltJe5 (1 7 i.xf6 gxf6
1 8 liJel i.xg2 1 9 @xg2 @£8 20 l:td3 l:txd3
21 liJxd3 @e7 22 l:tdl liJe8 23 g4 was equal
in Hort-Karpov, Buenos Aires 1 980)
1 7 ... i.xg2 1 8 �xg2 i.xe5 1 9 i.xe5 liJe8 20
f4 f6 21 i.c3 @f7 22 g4 l:txdl 23 l:txdl �e7
24 <t>£3 l:td8 with a draw to follow in Tim­
man-Karpov, Amsterdam 1 980.
1 3 'ifd2 'ile7 14 dxc5 .i.xb2 1 5 'ifxb2
dxc5 1 6 l:ld3 l:lad8 1 7 l:lad 1
1 7 . . . ill b 8 1 8 h4 l:lxd3 1 9 l:lxd3 l:ld8 20
'ild2 l:lxd3 21 'ilxd3 Ya - Ya
.------
Game 62
Pinter-Tompa
Hungary 1993
______________...
1 d4 ltif6 2 c4 e6 3 illf3 b6 4 g3 .i.b7 5
.i.g2 .i.e7 6 ltic3 � 7 .i.d2
1 34
.., . . . .i.f6!
I believe this is the best way to deal with
these lines. 7 ... 0-0 8 d5!? is considered in the
next game. Note that if 7 ... 0-0 8 0-0 then it is
interesting to play 8 ... d5!?, for if you want to
play such positions this is certainly the right
place to push to d5. Actually these lines are
considered to give White a slight plus in
ECO, but I guess that is primarily because
there was originally a guy who tried a line and
secured a modest advantage and then he has
been quoted for it ever since. Perhaps White
is a little better, but then he always is in these
situations. Korchnoi-Polugaevsky, Moscow
1 972 continued 9 cxd5 exd5 10 l:tcl liJd7 1 1
.i.f4 c5 1 2 ltJxe4 dxe4 1 3 liJd2 £5 1 4 liJb3
i.d5 1 5 dxc5 with a draw.
8 ... d6 9 d5! liJxd2 1 0 liJxd2 e5 1 1 f4! exf4
12 gxf4 liJd7 1 3 liJ£3 i.f6 14 liJd4 l:te8 1 5 e4
gave White a substantial advantage in Va­
ganian-Furman, USSR 1 971 .
8 . . . £5 is considered in the last game.
8 0-0
8 'i'c2 liJxd2 9 'i'xd2 looks like a loss of
time, and I suspect that that is what it is.
After 9 ... 0-0 10 0-0 d6 1 1 e4 (1 1 d5 i.xc3 1 2
'i'xc3 exd5 13 liJh4 d4! 1 4 'i'xd4 i.xg2 is
equal, as pointed out by Makarichiev)
1 t .. .liJd7 1 2 .:r.ad1 g6 1 3 h4 i.g7 14 h5 'i'e7
15 .:r.fe1 l:tfd8 16 hxg6 hxg6 17 i.h3 c5 1 8
d5 ltJe5 1 9 ltJxe5 i.xe5 Black had no prob-
lems whatsoever in Tukmakov-Nielsen, Co-
penhagen 1 996.

8 .l:r.c1 has not been tested since Black
came up with 8 ... ..ixd4! 9 il)xd4 il)xc3 1 0
..ixb7 il)xd1 1 1 .l:r.xd1 c6! 1 2 ..if4 0-0 13
i.d6 .l:r.e8 1 4 ..ixa8 'ii'c 8 1 5 b4 il)a6 1 6 b5
'ii'xa8 17 bxa6 c5 1 8 ll)£3 'ii'e4 with sufficient
counterplay in Karpov-Salov, Rotterdam
1 989, a game Black went on to win after a
fierce struggle.
8 . . . 0-0 9 l'lc1 c5
9 ... d5 is a tempo too late. After 10 cxd5
exd5 1 1 ..if4 il)a6 1 2 ..ie5 .l:r.e8 1 3 ..ixf6
'ii'xf6 14 e3 c5 1 5 il)e5 'ii'e 7 1 6 .l:r.e1 il)c7
White should not play 1 7 il)d3?! (Korchnoi­
Salov, Belgrade 1 987) but 1 7 'ii'a4!? il)xc3 1 8
.l:r.xc3, with the idea o f 1 8. . .c4 1 9 l:xc4 b5 20
:Xc7 bxa4 21 :Xe7 .l:r.xe7 22 .l:r.c1 with a
small advantage to White according to
Korchnoi.
1 0 d5 exd5 1 1 cxd5 .!Lixd2
Also possible is 1 1 ...il)xc3!? 12 ..ixc3 d6
13 il)d2 .l:r.e8 14 il)e4 ..ie7!? 15 g4 b5 with a
messy game in Rogers-Speelman, Oropesa
del Mar 1 996.
1 2 .!Llxd2
1 2 'ii'xd2 .l:r.e8 1 3 h4 d6 14 il)g5 ..ia6 1 5
il)ge4 ..ie5 1 6 l:.fe1 g6 1 7 a4 'ii'e7 1 8 b 3 il)d7
resulted in stone-cold equality in Epishin­
Oms Pallise, Sevilla 2000.
1 2 . . . d6 1 3 .!Lide4!
1bis is the only dangerous move. White
should remember how strong the bishop is
on f6. After 1 3 il)c4 ..ia6 1 4 'ii'b3 ..ixc4 1 5
'ii'xc4 il)d7 1 6 e 3 a6 1 7 a4 'ii'e7 1 8 'ii'e2 .l:r.fb8
4 g 3 i.. b l
1 9 .l:r.a1 .l:r.e8 only Black had realistic pros­
pects of pushing for a win in Vyzmanavin­
Khalifman, USSR 1 990.
1 3 . . . i.. e 5!
1 3 ... ..ie7?! had been played quite often be­
fore 13 ... 1'.e5 took over. Actually, I wonder
why? The bishop does not look better than
the knight once it gets to e 7. (I would suspect
that even 13 ... il)a6 is better than this) . White
earned himself a clear advantage in Karpov­
Salov, Linares 1 993: 14 f4 il)d7 1 5 g4! a6 1 6
a4 :e8 1 7 g5 ..i f8 1 8 ..tiih 1 ! etc. Shneider­
Beliavsky, USSR 1 990, went 1 6 g5 b5 1 7
il)g3 .l:r.e8 1 8 h4 ..i f8 1 9 h 5 with the better
game for White. Let us compare the bishop
with the knight now -why run away from the
exchange like that?
1 4 'ilfd2
14 f4 ..id4+ 15 ..tiih l ..ia6 16 .l:r.£3 .l:r.e8
(1 6 ... 'ii'e7 is even better according to Almasi)
17 g4 'ii'e 7 1 8 g5 il)d7 1 9 e3 ..ixc3 20 il)xc3
b5 21 l:g3! £5! and Black was slightly better
in the game Yusupov-Almasi, Germany
1 994. It should be noted that the line itself
does not give Black the advantage - White
should be able to improve on this, but the
positive trend seems to be with Black, none­
theless.
1 4 . . . i.. a 6 1 5 �h 1
1 5 .l:r.fe1 g6 1 6 f4 ..id4+ 1 7 e3 ..ig7 1 8 a3!?
'ii'e7 19 b4 il)d7 20 g4! h6 21 g5!? was un­
clear in Brenninkmeijer-van der Weil, Hol­
land 1991.
15 ... 'ilfe7 16 f4 i..d4 17 l'lfe 1
The most natural. After 17 .l:.£3 il)d7 1 8 e3
..ixc3 1 9 il)xc3 b5 20 ..ifl l:ab8 Black had
good counterplay in Horvath-Zagrebelny,
Budapest 1 993, while 1 7 b4 g6 1 8 b5 ..ic8 1 9
e 3 introduced an early draw in Yermolinsky­
De Firmian, Seattle 2000.
1 7 . . . g6 1 8 e3 i..g 7 1 9 g4
1 9 a3 il)d7 20 b4 cxb4 21 axb4 ..ic4 22 il)e2
.l:r.fc8 23 il)g5 ..ixe2 24 .l:r.xe2 a5 gave Black a
better game in Van der Sterren-Yemelin,
Hamburg 1 997.
1 9 . . . h6 20 g5 hxg5 2 1 .!Llxg 5
1 35
Q u e e n 's In dia n D e fe n c e
The game i s rather unclear.
2 1 . . . llld 7 22 'ii'f 2 ..i.h6!
22 ... tiJf6 23 �4 followed by e3-e4-e5
would be very dangerous for Black.
23 h4 ..i.xg5 24 hxg5 f6! 25 gxf6 'ii'h 7+
26 �g 1 lllxf6 27 'ii'g 3 :tae8 2 8 ..i.h 3 �g7
29 :tcd 1 :th8 30 ..i.g2 'ii'h 4 3 1 'ii'x h4
:txh4 32 a4? ! g5! 33 fxg5 c!lld 7 34 :ta1
..i.d3 35 ltlb5 :tg4 36 ltlxd6 .:ta 37 e4
:txg5 38 ltlf5 + �6 39 :te3 ltle5 40
:tg3! :tfg8 41 :txg5? :txg5 42 ltle3 ..i.xe4
43 .:n + �e7 44 :tf4 ..i.xg2 45 ltlxg2
llld 3 46 :te4+ �d6 47 �f 1 lllx b2 48 a5
:tf5+ 49 �g 1 :txd 5 50 a6 :td 1 + 51 �f2
.:a1 52 lllf4 :ta4 53 :te6 + �d7 0-1
· slight advantage to White in Karpov­
Game 63
Romanishin-Korchnoi
Lvov 2000
1 d4 lllf 6 2 lllf3 e6 3 g3 b6 4 ..i.g2 ..i.b7
5 c4 ..i.e 7 6 ltlc3 0-0
This move makes a lot of sense. Black
wants to see if White's move order holds any
real threats to his wellbeing.
7 d5!
7 'ili'c2 is, as always, answered with 7 ... c5!
and Black equalises with 8 d5 exd5 9 tlJg5
tLia6 1 0 cxd5 tiJb4 1 1 'ili'd2 tiJfxd5 1 2 tlJxd5
�xg5 1 3 'ili'xg5 1i'xg5 14 �xg5 �xd5 1 5
�xd5 tlJxd5 1 6 0-0-0 tiJb4 1 7 a3 tlJc6 1 8
l:.xd7 .l:tfe8 1 9 �dl f6 20 �d2 l:lad8, a s in
Bogdanovski-Marin, Yerevan 1 996.
1 36
7 . . ..i.b4
.
7 ... tiJa6 8 0-0 tlJc5 9 tlJe5 a5 10 �e3 exd5
1 1 cxd5 �a6 1 2 'i'c2 �d6 13 f4 gave White
the advantage in Atalik-Stoica, Herculane
1 996.
8 0-0 ! ?
This line allows the exchange on c3, after
which Black is okay. The critical lines arise
after 8 �d2! and now:
a) 8 ...Lc3 9 �xc3 exd5 1 0 tiJh4 tlJe4
(10 ... c6 1 1 tiJf'S dxc4 12 'i'd6 �h8 13 tlJxg7
c.fiixg7 14 'ii' f4 h6 1 5 g4 gave White a promis­
ing attack in Halasz-Lendwai, Miskolc 1990)
1 1 cxd5 tLixc3 12 bxc3 d6 13 0-0 tiJd7 14 e4
tLic5 1 5 1i'd4 �a6 1 6 l:lfdl 1i'd7 1 7 f4 with a
Ivanchuk, Linares 1 993.
b) 8 ... c6 9 0-0 �xc3 10 �xc3 (1 0 dxc6!?
dxc6 1 1 �xc3 tiJbd7 12 'i'c2 also looks bet­
ter for White and the two bishops) 10 ... cxd5
1 1 tiJh4 tlJe4 1 2 cxd5 tiJxc3 13 bxc3 'ii'c7 14
c4 tLia6 1 5 l:lct with an advantage according
to Ribli.
c) 8 ...tiJa6!? 9 0-0 tLic5 and now 10 t:De1
l:le8 11 t:Dc2 �xc3 12 �xc3 exd5 13 cxd5
t:Dce4 14 tbe3 t:Dxc3 1 5 bxc3 :Xe3! 1 6 fxe3
d6 1 7 'ii'd4 tbd7 1 8 l:lf4 'ili'e7 19 .l:taf1 t:De5
gave Black brilliant play for the exchange in
Razuvaev-Tiviakov, Rostov 1 993. It is hard
to find any use for the rooks as all the files
are closed or semi-closed. An improvement
for White is 1 0 tbh4!, e.g. 1 0 ... l:le8 1 1 l:lct a5
1 2 a31 (1 2 .l:r.el ?! d6! was good for Black in

Van Wely-Tiviakov, Gausdal 1 992)
1 2 ... .i.xc3 1 3 .i.xc3 exd5 14 cxd5 .i.a6! 1 5
l::te 1 with a modest plus fo r White thanks to
the strong and uncontested dark-squared
bishop. 14 ... lDce4 has been analysed by Bar­
lov. After 1 5 llJf5 lDxc3 1 6 llxc3 .i.a6 1 7 d6!
White has the initiative. Note 17 ... .i.xe2 1 8
dxc71 1i'c8 1 9 liJd6, and White wins in a mil­
lion ways. White is on top after 1 5 ... lDxd5?
16 .i.xg7 1i'g5 1 7 g4 etc.
8 . . . i.xc3
There are altematives, but this is the
move.
9 bxc3 l0a6 1 0 l0h4 :tb8!
Time has shown this to be the strongest
option. 10 ... e5 1 1 e4 d6 1 2 f4 exf4 1 3 lD£5
.i.c8 14 lDxg7?! <l;xg7 1 5 1i'd4 �g8 1 6 gxf4
lDc5 1 7 e5 lDg4 was good for Black in Filip­
pov-Tiviakov, Elista 1 995, but simply 1 4
.i.xf4 would have kept White slightly better.
10 ...lLJc5 1 1 .ta3 l:tb8 1 2 .i.xc5 bxc5 1 3
l::tb 1 exd5 1 4 cxd5 .i.a6 1 5 l:lel d6 1 6 llxb8
1i'xb8 17 1i'd2 l:le8 1 8 e4 gave White some
pressure in Shneider-Brynell, Aars 1 995.
1 1 e4 d6 1 2 i.e3
12 f4 lDd7 13 lDS e5 1 4 fxe5 (1 4 .i.e3
lDac5 15 lDd2 .i.a6 is equal) 14 ...lDxe5 1 5
lDxe5 dxe5 1 6 .i.a3 lDc5 1 7 .i.xc5 bxc5 1 8
1i'a4 1i'd6! with a good game for Black in
Filippov-Savon, St. Petersburg 1 996.
1 2 . . .:tea 1 3 :te 1 l0d7 14 f4
After 14 lDS lDac5 1 5 1i'c2 1i'f6 1 6 .i.d4
1i'h6 1 7 h3 .i.a6 1 8 .tf1 exd5 1 9 exd5 'ii'g6
4 g 3 i. b 7
Black had nothing to fear in Romanishin­
Markovsky, Biel 1 995.
14 . . . l0ac5 1 5 i.f2 i.a6 1 6 l0f3
White decides to forget about the c4-pawn
and generate his own chances on the king­
side.
1 6 . . . e5!
1 6 ... .i.xc4 1 7 llJd4 liJd3 18 .i.ft lDxf2 1 9
�xf2 .ixft 20 lDc6 gives White something
to bite on. Now the d4-squares is protected.
1 7 f5! ?
White refuses to compromise.
1 7 . . . .txc4 1 8 l:te3! b5 1 9 g4 f6 20 h4
:tea 21 l0e 1 c6 22 dxc6 :txc6 23 l0c2
:tc7 24 l:tg3 l0b6 25 g5 fxg5?
I qualify this move as a suicide attempt!
Actually, Black built up his position in order
to be able to push ... d6-d5, and he had to do
it now! Then White would be at a crossroads:
25 ... d5 26 .txc5! :Xc5 27 1i'h5 <it>h8 28 lDe3
1i'e7 29 l:ldl d4 30 cxd4 exd4 31 :Xd4 lDd7
with equality is a suggested improvement
from Korchnoi.
26 hxg5 d5 27 i.xc5! :txc5 28 'ii'h 5 l0d7
29 f6?
29 lDe31.
29 . . . d� 30 fxg7 i.f7 31 'ii'h 3 :txc3 32
l:txc3 d xc3 33 'ii'x c3 'ii'x g5 34 l0e3 l0f6
35 :tf 1 l0xe4 36 'ii'a 3 l0g3 37 :tf2 'ii'x g7
38 'ii'x a7 :tf8 39 'ii'b 6 'ii'g 5 40 'ii'a 7 .te6
41 l0d5! l:txf2 42 �xf2 l0h 5 43 'ii' b 8 +
�g7 44 'ii'x b5 l0f6 45 'ii' b 7+ �h6 46
l0xf6? 'ifxf6 + 47 'ii'f 3 'ifxf3 + 48 �xf3
137
Q u e e n 's In dia n D e fe n c e
..i.d5+ 49 �g3 ..txa2 5 0 ..te4 ..tf7 5 1
�g4 ..tg6 52 .ib7 �g7 5 3 �g5 h6+ 54
�g4 �f6 55 ..tea ..tt5 + 0 - 1
Game 64
Yermolinsky-Gulko
Scattle 2000
1 d4 tLlf6 2 c4 e6 3 tLlf3 b6 4 g3 ..tb7 5
..tg2 ..te7 6 tLlc3 tL!e4 7 ..td2 0-0 8 d 5 !?
f5 9 'i'c2
9 0-0 transposes to the main lines.
9 tLld6
• • .
It is also possible for Black to play for
equality with 9 ...exd5 10 lllxd5 lll c6 1 1 0-0
a5 1 2 l:tadl i.. £6 13 i.. e3 �hS 14 lllxf6 l:Lcf6
1 5 lllg5 lllb4 1 6 °i'hl 'ties 1 7 a3 llla 6, as in
Horvath-Atalik, 1991, which ended in a draw
after 26 moves.
1 0 ..tf4
A possibly better shot for an opening ad­
vantage is 1 0 b3 i.. £6 1 1 l:.dl 'ii'e 7 12 0-0
lll a6 13 llle l c6 14 dxc6 dxc6 1 5 e4 lllb4 1 6
°i'b 1 c 5 1 7 i.. £4 lll fl 1 S lllb5 e 5 1 9 i.. c l
l:.adS with a complex struggle in Van der
Sterren-Salov, Wijk aan Zee 1 99S, where
White might get to occupy d5 at the right
moment if he plays his cards right.
1 0 . . . tLia6
10 ... lllxc4 1 1 lllb 5 i.b4+ 1 2 �fl i.xd5
13 lllxc7 lllc 6 14 lllxd5 exd5 1 5 lllg5 gives
White some initiative, while 10 ... i.. £6 1 1 0-0
'ii'e 7 12 i.xd6 cxd6 13 l:tacl (13 l:.fdl g6 1 4
1 38
llld4 e5 1 5 tlldb5 l:.cS 1 6 e4 was a little bet­
ter for White in Ivanchuk-Salov, Linares
1 993, although Black can improve) 1 3 ... llla6
14 dxe6 dxe6 15 l:.fdl l:.acS 1 6 'ii'a4 l:.fdS
was completely equal in Epishin-Korobov,
Ohrid 2001 .
1 1 0-0 tL!xc4 1 2 dxe6 tLlb4
12 ...dxe6 13 'ii'b3 i.. d5 14 lllxd5 exd5 1 5
llld4 'ii'd7 1 6 'ii'h5 'ii'xb5 1 7 lllxb5 l:.adS l S
b3 gives White a small plus.
1 3 'i'c 1
1 3 'ii'b 3? llla 5! 14 'ii'a4 dxe6 is better for
Black.
1 3 . . . d5
Black cannot enter 1 3 ... dxe6 1 4 l:.dl ! llld5
15 lll d4 'ties 16 lllxe6 lllx f4 17 'ii'xf4 as he
clearly has severe problems.
1 4 b3 tL!a5
1 4 ... llld6 15 a3 llla6 16 lll e5 looks good
for White.
1 5 tL!e5 .:t6?!
15 ... c5!? 1 6 l:.dl i.. £6 and Black should be
able to keep the balance. Now he has to play
some precise moves in order to equalise.
1 6 'i'b 1 !
1 6 . . . d4
16 ... c5 17 a3 lll a6 lS ndl and d5 is weak.
1 7 ..txb7 tLlxb7 1 8 l:td 1 c5? !
1 S ... l:.xe6 1 9 'ii'x f5 l:.£6 20 'ii'g4 i.. d 6! 21
llle4 i.. xe5 22 i..xe5 l:.g6 23 'ii' B lllc 6 gives
Black an acceptable position, although it is
still White who is dictating events.
1 9 e3 .:xe6 20 'i'xf5 .:ta 21 'i'g4 tLld6?

This is just bad. Necessary was 21...�d6!
22 exd4 cxd4 23 �bS when White's pros­
pects are slightly preferable.
22 exd4 cxd4 23 llxd4 ltic2 24 .!tic&
24 . . . 'ii'f 8 25 ltixe7 + 'ii'x e7 26 ltid 5 'ii'f 7
27 ltixf6 + 'ii'xf6 28 llxd6 'ii'x a 1 + 29 lld 1
'i'f6 30 W'd7 llf8 3 1 W'd3 llc8? 1 32
'i'd5+ Wh8 33 .i.e5 'i'g6 34 'i'd8 + W'e8
35 'ii'd 7 1 -0
Game 65
Khalifman-Korchnoi
St Petersburg 1997
._______________..
1 d4 ltif6 2 c4 e6 3 ltif3 b6 4 g3 .i.b7 5
.i.g2 .i.e7 6 ltic3 ltie4 7 .i.d2 f5 8 d 5 .i.f6
I am not too fond of this line. 8 ... 0-0
transposes to Game 64.
9 'ii'c 2 'ii'e 7
9 ...�d6 gave White a better game after 1 0
�f4 'ile7 1 1 �xd6 cxd6 1 2 0-0 0-0 1 3 l:lfdl
g6 14 �d4 eS 1 5 �dbS l:lc8 1 6 e4 in lvan­
chuk-Salov, Linares 1 993.
1 0 lld 1
White is concentrating his play on the dS­
pawn. Both alternatives are worse. 10 0-0
J.xc3 1 1 �xc3 exdS 12 cxdS �xdS looks
acceptable for Black, Hulak-Naumkin, Palma
1 989, while Tyomkin gives 10 �xe4? fxe4 1 1
'W'xe4 exdS 1 2 'ilf4 �a61 1 3 �d4 0-0 1 4 �£5
'W'cS as being better for Black. Clever guy.
1 0 . . . exd5
Khalifman writes in his notes: ' ... the text
4 g 3 .i. b 7
move was played before and brought Black
good practical results. I think it's strategically
dubious ... '
The alternative is 1 0 ... �a6 1 1 0-0 �d6
(1 1 ...�acS 12 �bS 0-0 13 �xc7 l:lac8 14
�bS exdS 1 5 cxdS �xdS 1 6 'i'b1 gave
White a plus in Naumkin-Marinelli, Formia
1 995) 1 2 �bS (1 2 dxe6!? dxe6 13 'i'a4+ rJitfl
1 4 �e3 favoured White in Petursson­
Rozentalis, Malmo 1 993) 12 ...�xbS 1 3 cxbS
�cS 14 dxe6 dxe6 1 5 �b4 �e4 1 6 'ilc1 and
White was slightly better, Olafsson­
Rozentalis, Lyon 1 994. Rozentalis generally
plays a lot of · these typical QID positions
without too much hesitation, as he knows
them very well.
1 1 cxd5
1 1 �xdS?! �xdS 12 cxdS 'ilcS! 1 3 'ilxcS
bxcS 14 �c1 �a6 and Black is better
(Olafsson-Arnason, Kopavogur 1 994).
1 1 ... cS?!
This move finds a refutation in this game.
Better is 1 1 ...0-0 12 0-0 �a6 (probably the
best -White has only a slight advantage now)
1 3 �f4 �c3 14 bxc3 l:lae8 1 5 �d4 'i'cS 1 6
'i'b3 �h8 17 �e3 with an for White accord-
ing to Khalifman.
1 2 dxc6! dxc6
·
1 3 ltixe4!
This was a new move. The e-pawn is now
terribly weak.
1 3 . . . fxe4 1 4 ltih4! .i.xh4 1 5 gxh4 0-0!
1 6 'ii'c 4+ 'ii'f 7 1 7 'ii'xf7+ l:txf7 1 8 .i.xe4
1 39
Q u e e n 's In dia n D e fe n c e
White is close to winning the ending.
1 8 . . . c5!? 1 9 .bb7 l:xb7 20 ..tc3 lt:lc6 2 1
l:g 1 :ea 2 2 l:g5 l:e4 2 3 h5 l:be7 24 e3
li:ld4 25 'lfi>f1 h6 26 l:d5 lt:lb5 27 ..te 1
l:f7? ! 28 l:d8 + 'lfi>h7 29 l: 1 d 7! l:f5 30
J:d5 .:as 31 J:bS! l:d6 32 :t5 :ts 33
l:xf6 gxf6 34 l:d8 c4? 35 a4 c3 36 axb5
cxb2 37 l:d 1 l:c4 38 ..td2! l:c5 39 l:.b 1
l:xb5 40 ..tc3 l:c5 41 ..txb2 llb5 42 �e2
a5 43 'lti>d3 a4 44 'lfi>c2 J:c5+ 45 ..tc3 b5
46 �d3 l:xh5 4 7 ..txf6 1 -0
Game 66
lvanchuk-Timman
Wljk aan Zee 200 1
1 li:lf3 li:lf6 2 c4 b6 3 g3 ..tb 7 4 d4 e6 5
..tg2 ..te7 6 lt:lc3 0-0 7 0-0 lt:le4 8 ..td2
f5
Because of the trouble I feel Black faces in
this line I highly recommend 8 ... i.f6.
9 'i'c2!
1bis quieter variation is an attempt to in­
crease the tension. After 9 d5 i.f6 1 0 l:tcl
l2Ja6 Black normally equalises completely.
The theory goes as follows: 1 1 a3 c6! 12 dxe6
dxe6 1 3 'i'c2 c5 1 4 l:tfdl 'i'e7 1 5 lDel (1 5
l2Jxe4 i.xe4 1 6 'i'a4 liJb8 1 7 b4 l2Jc6 1 8 b5
liJd8 1 9 i.c3 i.xc3 20 l:txc3 liJfl was equal
in Beliavsky-Brodsky, Koszalin 1 998)
15 ... l2Jxd2 1 6 i.xb7 'i'xb7 17 'i'xd2 l:tad8 1 8
'i'e3 'i'e7 with equality in Hansen-Korchnoi,
Malmo 1 996. Or 1 1 i.el c5! (the new way to
play) 1 2 dxc6 dxc6 1 3 'i'b3 l2Jac5 1 4 'it'c2
'i'e7 1 5 b4 l2Jxc3 1 6 i.xc3 i.xc3 1 7 'i'xc3
l2Je4 1 8 °Wb2 c5 with complete equality, Be­
liavsky-Stefansson, Istanbul 2000.
9 . . . ..tf6
9 ... l2Jxc3 10 i.xc3 i.e4 1 1 'i'a4 d6 1 2
l:tacl 'i'd7 1 3 'i'dl ! c 5 14 'i'd2 liJc6 1 5 l:tfdl
and White had a little something in Skem­
bris-Huzman, Beersheba 1 993.
9 ... l2Jxd2 10 'i'xd2 i.b4 1 1 l:tacl i.e4 1 2
a3 i.xc3 1 3 l:txc3 'ii'f6 14 ltJ e1 i.xg2 1 5
lDxg2 lDc6 was very slightly better fo r White
in Stahlberg-Castaldi, Hamburg 1 955.
140
1 0 l:ad 1
1 o . lt:la6
. .
1bis is certainly not the main move, and in
this game Black is nowhere near equality .
The alternatives are:
a) 10 ...l2Jxc3 is the critical line given in
Chess Informant, and also the most natural.
But I think Black needs to come up with a
new concept somewhere to play like this. 1 1
i.xc3 i.e4 (1 1 ...'ii'e7 allows 1 2 d5! i.xc3 13
'ii'xc3 l2Ja6 14 liJd4 l2Jc5 1 5 b4 l2Je4 1 6
i.xe4 fxe4 1 7 dxe6 dxe6 1 8 b 5 , when White
is just a whole lot better, Ovsejevitsch-Rehm,
Berlin 1 997) 1 2 'iicl!, Azmaiparashvili­
Gulko, Elenite 1 995. The queen is much
better placed here than on d2 because it
leaves the way clear for the traditional plan
with i.h3 followed by liJd2, driving the
bishop away from e4 and seizing control of
the centre. Now 1 2...'i'e8, to prevent i.h3, is
answered with 13 l:tfel with the idea of i.fl .
However, it could b e argued that the inclu­
sion of these points benefits Black as well as
White. Instead the game continued (1 2 'i'cl !)
1 2 ... d6 1 3 i.h3! 'it'e8 (Gulko does not like
this move and improves it with 1 3 ... l2Jd7!? 1 4
liJd2 i.b7 1 5 e 4 c5! wi th counterplay - he is
probably correct, although it is far from clear
that Black has a potentially equal position,
and 16 d5 seems good for White to me) 14
l2Jg5 'ii'h 5 1 5 lDxe6! and now White found
the strong combinatorial continuation
1 5 ...'i'xh3 1 6 f3 i.b7 17 lDxc7 and Black
 4 g 3 j,, b 7

was in trouble, although he went on to win. Interesting was 15 ... cSI? 16 dxc5 lDxc5 1 7 b4
Earlier 1 2 'ii'd2 leads to a more or less d6, which Fritz came up with, although it
equal position after 12 ... d6 13 lDe1 .ixg2 1 4 looks like White is better after 1 8 :Xd6 (1 8
lLJxg2 1i'e7!, when th e idea behind develop­ bxc5 dxe51 gives Black a good game)
ing the queen first is seen after 1 5 1i'c2, when 1 8 ...'ii'e4 1 9 'iid2 .ixe5 20 .ixe5 1i'xe5 21
Black has the deep reply 1 5 ...1i'f7! intending f4! (21 bxc5 1i'xc5 and Black is fine) 21 ...'ii'e4
1 6 e4 fxe4 1 7 1i'xa4 d5! followed by 1 8 ... lDc6 22 bxc5 bxc5 23 'i'd3! etc.
and Black has good and natural counterplay. 1 6 dxe5 l:f7
Karpov -Polugaevsky, Amsterdam 1 981 went
1 5 lDf4 lDd7 1 6 d5 .ig5 1 7 dxe6 .ixf4 1 8
1i'xf4 1i'xe6 1 9 :fe1 lDc5 and Black fine.
b) 10 ... 1i'c8 1 1 d5! lDd6 12 b3 lDa6 1 3
.if4 favoured White in Lundin-Skold, Swe­
den 1 967.
c) 10 ... d6 11 d5 .ixc3 12 .ixc3 lLJxc3 1 3
'ii'xc3 e 5 1 4 b 4 saw White assume an advan­
tage on the queenside in Larsen-Levitt, Lon­
don 1 990.
d) 10 ... c5 1 1 d5 .ixc3 12 .ixc3 exd5 1 3
cxd5 d 6 1 4 lDh4 lDxc3 1 5 bxc3 1i'f6 1 6 f4
lDa6 1 7 .ih3 g6 1 8 e4 was also better for
White in Navrotescu-Marin, Bucharest 2001 . 1 7 b4 c6 1 a l:d6 b5 1 9 l:ed 1 bxc4 20
1 1 .!Lle5! a3 :ea 21 j.e 1 .!Llba 22 a4 c5 23 b5
11 a3 c5 1 2 d5 .ixc3 13 .ixc3 exd5 1 4 We4 24 Wxe4 fxe4 25 j.c3 �fa 26 �g2
cxd5 d 6 1 5 lDh4 lDxc3 1 6 bxc3 achieved l:c7 27 j.d2 :cs 2a g4 c3 29 j.e3 a6
nothing in Gritsak-Brodsky, Ordzhonikidze 30 l:c 1 axb5 31 axb5 �ea 32 l:xc3 c4
2000. 33 h3 :ta 34 j.a7 �e7 35 l:d4 e3 36
1 1 . . . .!Llxc3 1 2 j.xc3 j.xg2 1 3 �xg2 'ii'c a fxe3 l:c7 37 i.b6 l:b7 3a i.c5 + �ea 39
14 l:fe 1 Wb7 + 1 5 �g 1 j.xe5 b6 l:ga 40 l:a3 d5 41 exd6 e5 42 l:a7
This is a tough decision, after which the 1 -0
pressure down the d-file and the potentially A tough struggle, where the knight never
very strong bishop gives White an advantage. got into the game.

14 1
Q u e e n 's In dia n D e fe n c e

Summary
4 ... .tb7 still seems to be playable according to theory. The older attempts to get an advantage
with 7 dS and 7 b3 do not trouble Black and neither does 7 l:.et lll a 6!. The most dangerous
tries are the logical developing moves with 6 lll c3 and 7 .id2 in reply to 6 ... llle4. Black should
have a good grasp of the theory to play these lines, but if he does, then he can also count on
equality. My investigations have clearly indicated that ... .tf6 is a better defence than . fS, but
. .

who knows what time will show.

1 d4 .!bf6 2 c4 e6 3 .!bf3 b6 4 g3 .i.b7 5 .i.g2 {D) .i.e7

5 . cS - Game 56
..

6 lLJc3
6 0-0 0-0 (D)
7 b3 - Game 57
7 dS exdS 8 lllh4 - Game 58
7 l:.et - Game 59
7 lll c3
7 ... lll a6 - Game 60
7 ... lll e4
8 .id2 - 6 lllc3
8 'fi'c2 - Game 6 1
6 .!be4
• • .

6 . 0-0 7 dS i.b4 - Game 63

..

7 .i.d2 (D) 0-0

7 ... .tf6 - Game 62; 7 . fS 8 dS .if6 - Game 65
..

8 0-0
8 dS f5 9 'ii' c2 - Game 64
8 f5
. • .

8 ... .if6 - 7 ... .tf6

9 'if c2 Game 66
-

5 .i.g 2 6 . . . 0-0 7 .i.d2

142
 INDEX OF COMPLETE GAMES
I

Alterman-Liss, Israel 1999 ............................................................................. 76

Anastasian-Brodsky, Moscow 1992.............................................................. 95
Andersson-Karpov, Ti/burg, 1983 ............................................................... 133
Bacrot-Adams, Cannes (rapid) 200 1 .............................................................. 39
Barbero-Cebalo, Caorle 1987 ........................................................................ 97
Bareev-Eingom, "Kiev 1986 ........................................................................... 40
Bareev-Karpov, Moscow (blitz) 1993 ............................................................. 18
Barlov-Beliavsky, Yugoslavia 1992 ............................................................... 96
Barlov-Sakaev, Budva 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .�
................... . 36
....

Belozerov-Nikolenko, Moscow 1999 ........................................................... 67

Bogdanovski-Cabrilo, Bijefina Dvorovi 2002 ........................................... 1 13
Borges Mateos-Spraggett, Cienfuegos 1997 ............................................. 107
Cramling-Almasi, Horg,en 1995 .................................................................... 51
Crouch-Harikrishna, London 200 1 ............................................................. 73
Danner-Yu Shaoteng, Gyula 2000 .............................................................. 89
Dokhoian-Romanishin, Yerevan 1989 ..................................................... 1 19
Dreev-Seirawan, Wyk aan Zee 1995 ............................................................ 29
Dzhandzhava-Chernin, Lvov, 1987............................................................ 37
Ehlvest-Chandler, Lucerne 1989................................................................. 1 18
Fominyh-Chemyshov, Smolensk 2000 ..................................................... 132
Freitag-Oral, Graz 1998 .............................................................................. 129
Gelfand-Karpov, Sanghi Nagar (7) 1995 ...................................................... 87
Gelfand-Kasparov, Novgorod 1997 .................... , .......................................... 30
Gershon-Anastasian, Saint Vincent 2000 ........: .......................................... 45
Gheorghiu-Csema, Berlin 1986 ................................................................. 124
lvanchuk-Timman, Wyk aan Zee 200 1 .................................................... 140
Kamsky-Karpov, Elista 1996 ...................................................................... 123
Kamsky-Yudasin, Biel 1993 ......................................................................... 63

1 43
Q u e e n 's In dia n D e fe n c e

Karpov-Gavrikov, USSR 1988 ................................................................... 127

Karpov-Polgar, Buenos Ains 200 1 .............................................................. 106
Karpov-Salov, Unans 1992 ......................................................................... 1 16
Kasparov-Gelfand, Novgorod 1997 ............................................................. 1 1 1
Kasparov-van der Wiel, A.msterdam 1988 . . . . . . . . . . . . . . . . . . . . . � ............................ 55
Khalifman-Korchnoi, St Petersburg 1997 .................................................. 139
Khalifman-Short, Paernu 1996...................................................................... 59
Lautier-Karpov, Monte Carlo (rapid) 2000 ................................................. 103
Lautier-Timman, Amsterdam 1996.............................................................. 42
Magerramov-Makarichev, Moscow 1991 ................................................... 93
Miles-Timman, Ti/burg 1986 ........................................................................ 71
Miles-Unzicker, Johannesburg 1979 .............................................................. 16
Neverov-Stefanova, Reykjavik 2002 ............................................................ 47
Nikolic-Lautier, Moscow (Fide knockout) 200 1 ............................................ 19
Petursson-Polugaevsky, Rey�javik 1987 ................................................. 1 15
Piket-Anand, Monte Carlo (blindfold) 1997 .................................................. 49
Piket-Chuchelov, Netherlands 2000 ........................................................... 109
Piket-Tiviakov, Wijk aan Zee 1994 .............................................................. 69
Pinter-Tompa, Hungary 1993 ..................................................................... 134
Polugaevsky-Wojtkiewicz, Haninge 1990 ............................................... 128
Portisch-Nikolic, I..:fub!Jana 1985 ................... .-. ............................................ 24
Portisch-Popovic, l./ubb'ana 1985 ................................................................ 65
Romanishin-Korchnoi, L.vov 2000 ........................................................... 136
Sakaev-Berzinsh, Duisburg 1992 . .� . . .
. .......... ...
. .... .. ......... . . 105
. .........................

Salov-Karpov, Wijk aan Zee 1998 ................................................................. 99

Seirawan-Timman, Hilversum Match (5thgame) 1990 ............ :.................. 8 1
Silman-Gheorghiu, Palo Alto 198 1 . � . .
... .... ........ .
...... . . 35
.............. .............. . ....

Teske-Paehtz, East Ge1'111a'!J 1988 ............................................................... 21

Timman-Karpov, Hoogeveen 1999 ................................................................ 79
Timman-Polgar, Bali 2000 53
...........................................................................

Tkachiev-Nisipeanu, Naujac-sur-Mer 2000 ............................................. 1 13

Tolnai-Adotjan, Hungary 1992 . . .
........... ........... . . . .· . . . . 62
............................. ... . ...

Topalov-Adams, Fujitsu Siemens Masters Frankfurt 2000 ........................ 1 1 7

Tregubov-Shaposhnikov, Samara 2000 .................................................. 101
Van Wely-Korchnoi, Hoogeveen 200 1 ......................................................... 13 1
Yermolinsky-Gulko, Seattle 2000 .............................................................. 138
Yusupov-Beliavsky, Austria 1998 ............................................................... 90

1 44