Escolar Documentos
Profissional Documentos
Cultura Documentos
defence
by A. Raetsky
& M. Chetverik
EVERYMAN CHESS
Gloucester Publishers pic www.everymanchess.com
First published in 2005 by Gloucester Publishers pic (formerly Everyman Publishers
plc), Northburgh House. 10 Northburgh Street. London EC1 V OAT
The right of Alexander Raetsky and Maxim Chetverik to be identified as the authors
of this work has been asserted in accordance with the Copyrights, Designs and Pat
ents Act 1988.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, electronic, electrostatic, magnetic
tape, photocopying, recording or otherwise, without prior permission of the publisher.
Distributed in North America by The Globe Pequot Press, P.O Box 480,
246 Goose Lane, Guilford, CT 06437-0480.
Everyman is the registered trade mark of Random House Inc. and is used in this
work under license from Random House Inc.
Bibliography 4
lntnxluction 5
1 e4 e5 2 lLif3 lLif6
BIBLIOGRAPHY I
Books
Ji'a.rhionable Varia/ion in tbe PetrrJ.ffDye11ce, IUctsky and Chetvcrik (Voronczh 1992)
Modi!Ji 1•aritmt msskoi parlii, IUetsky and Chctvcrik (Voronezh 1990)
Rltssisci.Je Ptlflie, Schwarz (Hamburg 19H6)
JI.Jtssischc f.'erfridigNI'!,. Konikowski and Sicbenhaar (Germany 1992)
The Cochrtmc G01nbil, Matsukevich (Moscow 1994)
TIJc PeiiT!ff,Janjgava (Gambit 2001)
The J>etrq!J Defence, l'orintos and Haag (Batsford 1991)
17:Je PetroffDefence, Yusupov (Oirns 1999)
Periodicals
Cl.�ess 11ifo1711ttnl 1-90
Nen' i11 Chess YeariJOok. 1-72
Software
Chess Assistant 7.1
4
INTRODUCTION I
'1'/Je e111irr lheol)• �/chm opmi11gs is diiJided i11/o lbe Pellf!lf D�Je11ce tmd the �jerted J>rllf!UDtjellce (1 d4, 1 t4
tllld other li11ts). Alexander Ractsky
-
TI1e Petroff Defence has been one of the most fashionable 'open games' in the last decade,
but it also has a long history. After 1 c4 e5 2 ltlf3, instead of defending his e5-pawn, Black
prepares a counterattack on White's e4-pawn with 2...ltlf6. J .ucena mentioned this idea as far
back as 1479, while in 1512 Damiano analysed 3 ltlxe5 ll'lxe4 and, of course, after 4 tfe2
reached conclusions that were discourah>i.ng for Black. No wonder the symmetrical opening
was forgotten for many centuries after that! However, in I H24 Russian maestro, Alexander
Petroff, found out that 3 ltlxcS could be met by 3 ...d6!, and only after that should Black cap
ture the e4-pawn. In 1842 another Russian expert, Jaenisch, published valuable analysis in
l'ttltlfllede. The opening was given a name of the Petroff or the Russian Defence Qn Russia,
Germany and Scandinavia).
It is quite narural that two Russian masters write about the Petroff Defence for the pub
lisher Everyman Chess. Just like ballet, the Russian Defence could be labelled as property of
the Russians. Grandmaster Arrur Yusupov is an outstanding thl:orctical expert of this opening,
while Russian champions Smyslov, Karpov and Kramnik often usc it in practice. Raetsky and
( :hctverik have not made a substantial contribution to the theory of the Petroff Defence so
far. However, it's worth pointing out that our first published articles and booklets were de
vntL'l.l to this opening in the early 1990s. Alexander Raetsky has considered the Petroff De
fence to be his f.1vourite opening for more than twenty years and can be proud, at least, of the
l(uantity of his b>amt."S played in the Petroff Defence (about 200, including 60 correspondence
1-�amcs), if not the quality.
Emotional players arc scared off by the S)•mmetrical trend in the Petroff Defence; they arc
afraid of the drawish aspirations of weaker opponents who play White. But what can be more
symmetrical than the initial position in chess, which has not yet been ruined by the notorious
'draw death' despite Capablanca's indications? A more skilful strategic player triumphs in the
Petroff Defence regardless of the colour of his pieces and the position's symmetry. The
healthy strategical foundation of this opening allows Black to defend a lot of systems in the
Petroff Defence even while playing against stronger opponents. Even the lines declared
'doubtful' by theory are normally better than their reputation.
5
The Petroff Defence
Apart from minor alternatives (see Chapter 10), White has to choose between 3 d4 and 3
'Oxe5. The authors believe that after 3 d4, the move 3.../0xe4 is stronger than 3...exd4. Here
the sharp variations like 4 dxe5 .i.c5!? and 4 .i.d3 d5 5 'Oxe5ltld7 6'0xd7 ..i.xd7 7 0-0 �4!?
can replace the popular, solid, but uninteresting alternatives (4 dxe5 d5 and 4 .i.d3 d5 5/0xe5
/Od7 6 'Oxd7 .i.xd7 7 0-0 .i.d6 respectively). After 3 'Oxe5 it is more difficult for Black to
initiate an open battle and the positional niceties of the fight come to the fore.
One final advantage for the fans of the Petroff Defence: you don't have to study numerous
complicated openings like the Ruy Lopez, Two Knights Defence, Italian Game and Scotch
Gamel After 2'0f3'0f6 you are home and dry!
Finally, some acknowledgements. Special thanks go to the founder of the Petroff Defence,
Mr Petroff, and many thanks to our Danish/Scottish friend Jacob Aagaard for his technical
help. Also thanks go to Zoya Nayshtut for her excellent translation into English and also to Mr
Yusupov for his great book on the Petroff Defence- the real bible for people from our 'cast'
(Petroff Defence players).
Play the Petroff Defence and be happy!
6
CHAPTER ONE I
3ltJxe5: The Main Line
with 8 c4
7
The Petroff Defence
White can play 1 1 !.DeS (Games 3-4) and 1 1 Pava.'IOvic, Istanbul Oly mpia d 2000 - and
.l.e3 (Games S-6). here Pavasovic claims an edge for Black after
The most fashionable line at the moment
16...exd3 17 Wb3 'ifd7) 1 S...d3 1 6 �3 dxc2
is 10....i.f5, which was utilised by Kramnik in
1 7 1i'xe2 .l.e6 1 8 llcl1 1Wc7, which Pavasovic
his 2004 World Championship match with
regards as unclear.
Leko. lbis move is discussed in Games 1 -2.
b) 1 1 �3 leads fairly dirccdy to a draw:
1 1 ...dxc4 12 .l.xc4/.Df6 1 3/.Des 'ifxd4 (Bolo-
Game 1 gan pointed out the error 13 ...1.Dc2? 14/.Dxf7
Leko-Kramnik llxf7 IS .i.xf7+ �f8 1 6 .l.e3 when White is
World Ch. (Game 1), Brissago 2004 much better) 14 .i.xf7+ Wh8 1 S I.Lif3 •d7 16
'------• /.Des ..d4 17 I.Lif3 'Wd7 was agreed drawn in
1 e4 e5 2 �f3 �f6 3 �xe5 d6 4 �f3 Bologan-Zamicki, Buenos Aires 2000.
Q)xe4 5 d4 d5 6 .i.d3 &6 7 0-0 J.e7 8 1 1 ...�c3 1 2 bxc3 &6
c4�9 J.e2
Of course saving the bishop is the most
popular option. Alternatives arc studied in
Games 9-1 0.
9 0-0 10 �c3 J.t5
.•.
1 3 lZ.e1
After 13 cxdS .xdS White has a few
choices:
a) 14 lle 1 llfe8 transposes to the main
game.
1 1 a3 b) 14 .i.e3 l'lfd8 1S :e1 &s 1 6/.Des f6!
This is the natural move but we should (16.....b3? is just a blunder: 1 7 .l.c41 .xd 1
also consider two others: 1 8 .i.xf7+ � fB 19l:taxdt /.Dc6 20 .l.ct i.LixeS
a) 1 1 �5 cS and now: 21 AxeS .i.g4 22 llde1 .i.f6 23 ll5e4 and
a 1) 1 2 dxcS is not a dear error but it does White has won a clear pawn, Movsesian-1
allow pressure on f2: 12..ixcS 13 I.DxciS P.Nielsen, Blmdesliga 1 999) 1 71.Dd3 i..d6 1 81
(White must be careful: if 1 3 a.1? then I.Db4 •n is unclear.
1 3...1.L!xf2! 1 4 llxf2/.Dc2 1 5 lla2 llc8 1 6/.Dd3 c) 14 c4 1i'd6 IS dSI.DeS t6/.Dd4 .i.d7 1 7
.l.xf2+ 17 I.Lixf2 dxc4 and Black has a clear a 4 c6!? 1 8 l:tb t (perhaps White shoula prefer
advantage) 13...lDxdS 14 ..xd5 'irxdS 1 5 1 8 'ifb3 cxdS 1 9 cxdS 1Wb4 20 'ifxb4 .l.xb4,
cxdSlLd8 16 i.c4 llfc8 t7/.Df3 h6 1 8 i..d2 when his d-pawn is marginal!)' more of a
I.Lixcl2 1 9 I.Lixd2 i.cl4 with an unclear posi strength than a weakness) 18 ...cxdS 19 cxd5
tion. ..xdS 20 I.Lif5 .c6 21 I.Lixe7+ ..xc7 22
a2) 12 a3 cxd4 13 I.Lixc4 dxe4 14 axb4 f6 llxb7 1We6 was level in Van Den Doei
1 5 I.Dg4 (15 �t3?1 d3 1 6 .i.xd3 - J.Polgar- Schandorff, Esbjerg 200 1.
8
3 liJxe5: The Main Line with 8 c4
tl) 14 .if4lfr.ts 15 .i.xc7 b6 16 .i.f4 Lc8 22 g41 1Wf4 (even worse is 22. ..1!rd7?! 23
17 .i.t.12 .i.d6 18 llc1 h6 19 a4 :C7 20 h3 .i.xf7+! Wh8 - the point is 23...Wxf7 24
Wh3 21 .i.b5 (Krakops-Ulescas, Leon 2001). llxa51 llxa5? 25 1i'b3+ Wg6 26 llc6+ when
Now Black cook! have played 2t ...'exdl 22 Black will soon have to give up his queen -
l:lcxd1 �b3 23 L2 �xd2 24 �xd2 llxc3 24 ..ixd6 1i'xf7 25 ..tc5 and White was close
with equality. to victory in Karpov-Portisch, Torino 1982)
13 .. .1Z.e8 23 .ixd6 'irxd6 24 'ifo •d7 25 :C2lDc4 26
I\ simple, sensible developing move but llacl and White has a strong initiative, espe
131ack has other options. cially since 26...1le8?? loses to 27 lle7!.
13....i.f6 is probably not qui te good b2) 17...i.e6 18 .i.xdi fxe6 19 �0 :aes
enough to cyualise: 14 .i.f4 �aS 1 5 cxdS 20 .ih4 1t'c6 21 1tc2 h6 and Black is just
WxdS 16 Wa4 (if 16 .i.xc7 :lacS 17 .i.xa5 active enough to hold et:juality: 22 .i.g3 1i'd5
Wxa5 18 c4 llfd8 Black has pleasant com 23 a4 �4 24 lle4 a6 25 1i'e2 1i'c6 26 lle1
pensation) 16...b6 t7lDd2 .i.d7 18 'irt.11 and .i.xg3 27 hxg3 tLki6 28 llxe6 1i'xc3 29 lieS
\Vhitc had an cdbJC in Short-Bologan, Skan 1i'b3 was now agreed drawn in 1\.dams
tlcnborg 2003. This advantage grew after Anand, FlOE World Championship, New
IK.J:ac8 19 i.d3 .i.f5 20 �4 .i.e7 21 'irh5 Dell1i 2000.
i.xc4 22 1txd5 .i.xdS 2.l llxe7 llfe8 24 14 cxd5
Jl:ac I �f8 25 llxe8+ llxe8 26 llxe8+ Wxc8 14 .i.f4 is discussed in Game 2
27 .i.xc7 . 14 ....xd5 1 6 ..tf4
13...dxc4 14 .i.xc4 has been well tested: White has a minor alternative in 1S �e3
.i.f6 (:tlso fine is IS...�aS!? 16 �d2 'ird7 17
.to .i.d6 18 c4 c6 19 'lfa4 .i.e? with an un
clear position) 16lDd2 .l:lad8 17 .i.c4 11d7 1 8
�b3 .i.e7 19 a4 .i.d6 20 'lfd2 (Bologan
Degrnevc, Belfort 2002) ands here Bologa.n
gives 20...�7! 21 .i.f4 b6 as equal.
1 5.. .1Z.ac8
9
The Petro ff Defence
the more comfortable game) 23 l:lxe7 lbxc7 .to •d7 20 1i'a4 lbe5 21 '1Vxd7 lbxf3+ 22
24 lbes .i.xe5 25 .i.xe5 l:le8 26 .i.g3 lbf5 27 lbxf3 ltxd7 23 lbe5 l:ld5 as in Leko
.i.xc7! ..xc7 28 .xf5 'lrxc4 29 d6 and Bologan, Dortmund (rapid) 2004. Igor Zait
White's strong passed pawn gave him the sev now suggests 24 c4 l:ldd8 25 lbxg6 fxg6
edge in Leko-Anand, Unares 2003. 26 d5 �fl 27 �ft l:ld7 28 l:lab 1 b6 29 ltb3
A few days after this main game Lcko as a way for White to gain an edge.
tried 16 c4 'lre4 1 7 .i.e3 and now:
a) 1 7...'ifc2!? (Kramnik's novelty) 1 8 d5
lbas 19 lbd4 •xdt 20 l:lcxdt (if 20 l:laxdt
Parkin analysed 20....i.d7 21 lbb5 b6!? 22
lbxa7 l:la8 2' lbb5 .i.xb5 24 cxb5 .i.xa3 as
equal) 20....i.d7 21 .i.d2 .i.f6 22 .i.xa5 .i.xd4
23 l:lxd4 l:lxe2 Vz-112 Leko-Kramnik, World
Championship (Game 3), Brissago 2004.
b) Also reasonable is 17 ....i.f6 18 l:la2 b6
1 9 h3lba5 with a further split
b1) 20 .i.d2 lbb7!? (20...�1?1 leads to a
very unpleasant ending: 21 'ifxb1 Lb1 22
l:lxb1 l:lxe2 23 'itft l:lce8 24lbg1 ll2e4 25
.i.xa5 bxa5 26 lle2, Kotronias-Marjanovic, 1 8 c4?!
Kallithea 2003) 21 'ifa4 .i.e6 22 d5 lbc5 23 This allows Black the chance to give up
1i'xa7 .i.d7 when White's awkwardly placed his queen for a definite equality. A better try
queen provides Black with compensation for for the advantage is 1 8 lbdz!? .i.f5
the pawn. (1 R.ixg2? may look worrying for White,
b2) 20 g4 .i.g6 21 g5 .i.c7 22 lbe5 .i.d6 23 but I. Zaitsev provided the clever refutation:
lbxg6 hxg6 24 c5 .i.ffi (Kasparov analysed 19 c4 .c6 20 d5 .g6 21 .i.hS .i.f3+ 22
24...bxc5 25 dxc5 .i.f4 26 .i.g4 llcd8 27 l:ld2 Lg6 .i.xdt 23 .i.f5) 1 9 .t.n 'lid7 20 lbb3
.i.xg5 28 .i.d7 lieS 29 l:ld4 as slighdy better lbxb3 21 1i'xb3 c6 when White has the mer�
for White) 25 .i.g4 l:lcd8 26 l:lae2 1i'c6! 27 est of edge£.
1i'c2 (27 cxb6 was immediately abandoned as 1 8...li»lc41 1 9 .i.xc4 1i'xc4 20 �d2 'ttd5
a draw in Anand-Adams, Unares 2002) 21 li»le4 ••e4 22 i.g5 1i'xe1 + 23 1i'xe1
27...bxc5 28 dxc5 •d5 29 1lrc3 lbc6 is equal .i.xg5 24 ••5 i.f& 25 •••7 c5 26 •xb7
according to Dokhoian. i.xd4 27 :S2
Brissago (3) 2004. Belov gives 27 l:tdt llb8 28 .d7 l:le2 29
1 6....i.e41? ltxd4 cxd4 30 11Vxc.J4 as equal.
Previously 1 6..if6 had been played: 1 7 27 ...c4
lbh2!? 'ifa5 1 8 .i.d2 lbc7 (or 1 8. ..l:lcd8 1 9 Black must avoid the back rank tric�1
.i. f3 h6 20 lbg4 .i.xg4 21 hxg4 .i.g5 22 27.. 1J..c7? 281le2!.
.i.xg5 hxgS - Kramnik-Anand, Wijk aan Zee 28 Jle2 lled8 29 e4?!
2003 - and here Hu?.man gives 23 .cl! White should have played 29 l:ld2 .i.e5 30
l:txet+ 24 'IVxel 'itffi 25 l:tbt lle8 26 .cl as £4! .i.f6 (not 30 ...i.xf4?! 31 llxd8+ llxd8 32
.
clearly better for White) 1 9 lbg4 .i.xg4 20 'ifc6 hS 33 'ifxc4 .i.g3 34 �ft :Cs 35 1rc3
Lg4 ltcd8 when Anand bclit.-ves White is h4 36 a4 when White has some winning
slighdy better. chances) 31 ltxd8+ l:lxd8 32 .c7 c3 33 a4 g6
1 7 .i.e3 c!lla 5 34 aS l:ld2 35 a6 lta2 36 �h2 c2 37 a7 after
Another try is 17 ...l:lcd8 18 lbd2 ..tg6 19 which the game would have been drawn.
10
3 li)xe5: The Main Line with 8 c4
Game 2
Grischuk-Adams
Halkidiki 2002
11
The Petroff Defence
is 2t...1i'f6 22 l:tet �5 23 1Wf3 «i>ffl24 �f4 kingside pawns decide the game.
(Anand-Karpov, Prngue 2002). Anand now 30 ll:!f4 .i.h7 31 .ta2 �7 32 ll:!e6 llxf2
gives 24...1Wxh4!? 25 .i.xf7 1i'g4 26 'ifxg4 33 �xf2 .igS 34 �f8+ �es 35 .bg8
.i.xg4 27 .i.d.5 �e7 28 .i.xb7 l:b8 as offering �xf8 36 .idS!
Black enough compensation. White has kept the extra pawn and now
2211'g3 dominates the kniJ.,rht: Black's position is
Black can also defend against 22 �f4 �eS hopeless.
23 dxeS 1i'xc4 24 exd6 1i'xc3 25 l:tf1 .i.e4! 36...�e7 37 �e3 ¢>f6 38 ¢>f4ll:lc6 39
(not 25...D.d8? 26 �xg6 fxg6 27 d7 1i'c6 28 .i.xc6 bxc6 40 c4 �e6 41 g4 ¢>f6 42
l:tdl bS 29 hS! gxhS 30 .xhS when White g5+ hxg5+ 43 Wg4 1 -0
has a huge advantage, Mortensen
M.Andersson, Sweden 2003) 26 �6 i.d3 27 Game 3
'ilfOI fxe6 28l:td1 1i'xa3 29l:txd3 Wet+ 30 Anand-Shirov
'it>h2 1i'c6 31 1i'xc6 bxc6 32 l:l.c3- the ruok Moscow2001
ending is drawn.
22 ll:!a5 23 .ia2 ...d3 24 llf1 •xg3 25
••• 1 e4 e5 2 ll:!f3 ll:!f6 3 ll:!xe5 d6 4 ll:!f3
fxg3 lieS 26 hS! .id3 ll:!xe4 5 d4 d5 6 .id3 ll:!c6 7 0-0 .ie7 8
Black avoids the obvious double attack: c4ll:lb4 9 .te2 0-0 1 0 �c3 .i.e&
26...i.xh5? 27 :5 b6 28 l:.xhS :Xd 29 l:lf5
when White should win.
27 .txf7 + ¢>f8
27...'it>h7?! is too passive: 28 l:.f3 ..ie4 29
l:.e3 d5 30 �f4 �4 31 l:.e2 �xa3 32 .ixd5
i.xdS 33 �d5 and White has good winning
chances.
28 llf2
1 1ll:le5
Apart from 11 .i.e3, which is studied in
Games 5-6, there arc two alternatives to note:
a) 11 .if4 cS 12 l:lc1 .i.f6 13 �bS? (this
expedition fails to some brilliant resources;
White should prefer 13 11'a4 �xc3 14 bxc3
�c6 15l:tab1 'ilfd7 with an unclear position)
13 ...dxc4 14 �7 �5! 15 �xc6 fxc6 16
28 . . .oh7? .i.g3 (Adams-Shirov, Dortmund 1998). Shi
A fatal slip. After 28...l:.xc3 29 ..ig(rr �g8 rov now analysed 16...�xg3 17 hxg3 bS! HI
White has no more than a draw. a4 cxd4 19 �xd4 'tWb6! 20 �f3 (the lx:autiful
29 .td5 llf8 point is 20 �xbS? ..xf2+! 21 �xf2 .i.d4
The problem with 29...:Xd is 30 l:.f31 mate!) 20... .i.xb2 21 l:.bt �d! 22 ..c2 �xa4
J:ct+ 31 Wh2 i.c4 32 2:f7+ Wd8 33 .i.xc4 23 l:.xb2 �xb2 24 '1Vxb21lac8 when Black's
%lxc4 34 :Xg7 :Xd4 35 g4, when White's excellent pawns promise a clear advantage.
12
3 fDxe5: Th e Main Line with 8 c4
b) 11 a3 is safe but unthreatening. For ex plan. 15 f4 was tried in Anand-Lcko, Leon
nrnplc, 1 t...lL'lxc3 12 bxc3 tL'lc6 13 cxd5 2001, the game being level after 15...f6 16
.lxd5 14 c!Od2 tL'la5 15 .i.d3 b6 16 •c2 h6 tL'lg4 .i.xg4 17 .i.xg4 fS 18 .i.e2 .i.f6 19 l:tbt
17 .i.b2 .i.b7 18 lbe4 b5! 19 a4 (tf 19 .i.xb5 i.d4+ 20 Wh1 tL'lc1 21 b4 b6 22 l:tb3 lbes 23
t'5 20 "Wa4 c6 21 .i.c4+ lL'lxc4 22 •xc4+ •d5 i.e3 i.xe3 24 .l:txe3 llkt6 25 l:tb3 .f6 26
2.3 llk12 .i.g5 24 'lfd3 c5 Black's active bish bxc5 bxc5 27 11fa1 ..xa1 28 llxa1 Wfl.
ops provide ample compensation) 19...a6 20 15 f3!? is more enticing. Motylev analysed
uxb5 axb5 21 .i.xb5 fS 22 tL'lc5 .i.xcS 23 15...•c7 16 i.f4! i.d6 17 fxe4 l:tc8 (of
(lxcS •d5 24 c6 lL'lxc6 25 •d3 ll:le5 26 course the idea is 17..ixe5 18 d6) 18 .a4
Vxd5+ .i.xd5 with an equal position, Lcko l:tc7 19 tL'ld3 l:txe4 20 .i.xd6 .xd6 21 ..if3
Krnmnik, Dortmund 1999. with a slight edge for White.
1 1 ... c5
This early break gives White the chance to
ntlvance in the centre. The alternative 11...f6
is the subject of Game 4.
1 5...f6
Black threatens the knight at a moment
when it has no choice but to rctteat. Instead
15....i.d6 16 f4 f6 gives White the additional
1 2 l0xa4 option of 17 .xe4!? l:te8 (if Black takes the
If 12 .i.e3 cxd4 13 .i.xd4 lL'lxc3 14 bxc3 piece with 17... fxc5 18 fxe5 .l:te8 White has
li:k6 15 lL'lxc6 bxc6 16 .a4 cS 17 ..ie3 rl4 strong passed pawns and an attack as com
Ulack equali.<;es comfortably acconling to pensation after 19 .i.d3 rfi 20 c6) 18 .i.h5
Yusupov. Ae7 19 i.fl+ (Morgado assessed 19 lL'lf7
12 ...dxa4 1 3 d5 .i.e& '1Vc7 20 lL'lxd6 l:txe4 21 lL'lxe4 as unclear)
Black must retreat since after 13.....id6?! 19...Wh8 20 tL'lg6+ hxrfi 21 'IVxrfi i.g4 22 fS!
14 a3 i.xeS 15 axb4 ..t£5 16 bxc5 White's l:txf7 23 Vxf7 .c7 24 ..xc7 ..ixc7 25 .tf4
pawns are too strong. Baklan-Timman, and Black's minor pieces cannot find any
Neum 2000 continued 16, ..•c7 17 g3 11fxc5 activity so White has an edge.
IK i.e3 •d6 19 c5 •f6 20 �3 ..Lb2 21 1 6�
J:a41 (Baklan pointed out that 21 lLb1 ..id4 The play is certainly complicated, but
22 'ifxb7 ..ih3 23 .l:tfd1 ..ixe3 24 fxe.l Wc3 is White seems to have the better of it.
just unclear) 21..ie5 22 11fxb7 ..th3 23 Act 1 6.....d6
aS 24 d6 and the pawns promise White an In Topalov-Shirnv, FIDE World Cham
obvious advantage. pionship, Moscow 2001 Black tried 16..id6
14 a3 �6 1 5 •c2 17 f4 cxf3 (or 17....txg4 18 .txg4 fS 19 .i.e2
Anacking the pawn Js White's natural .te7 20 ..ie3 ..i.f6 21 l:tadt and White's ad-
13
The Petroff Defence
vantage is small but defmite) 18 i.xf3 'ifc7 tage with the calm 28 'ifa3.
19 g3 i.d7 20 liJt2 bS 21 b3 :ae8 22 'iPg2 26 .i.f4
i.eS 23 :b1 i.d4 24 i.d2 and White re Now this simple move is enough to con
tained the usual nagging edge. finn a solid advantage.
1 7 f3 26 ...h6 27 .i.d6!
White tries to destroy his opponent's cen
tre. Anand also analysed 17 g3 fS 18 i.f4
'ifb6 19liJe3 i.f6 (not 19...g5?1 20 i.eS f4 21
gxf4 gxf4 22 'iPh I! and it is White who at
tacks along the g-file) 20 h4 ...xb2 21 'ifxb2
i.xb2 22 :a2 when White ha.c; some com
pensation but no definite advantage.
17 f5
... 1 8 lbf2 .i.f6
Similarly, White would rather return the If 40...i.f6 then 41 :xf6! gxf6 42 'ifg6+
pawn than allow the fB-rook into the game. forces mate. For example, 42...Wh8 43
21 .....xe5 22 Wh 1 .i.d7 'ifxf6+ Wh7 44 ...f5+ and the bishop will
Black's queen bishop settles for a modest soon join the attack with check.
square. 22 .f4 gains the £5-squarc but only at
.. 41 11Ve6+ Wh7 42 'iff5+ Wg8 4a
the expense of surrendering g4: 23 liJd3 i.£5 11Vc8+ �h7 44 .:tf8 1 -o
24 i.g4! (not 24liJxe5?! i.xc2 25 i.d3 i.xd3
26liJxd3 gS 27 h4 h6 28 hxgS hxgS 29 g3 bS, Game 4
when White's centre collapses) 24... g5 25 Leko-Grischuk
.id2 and White retains the initiative. Wijk aan Zee 2002
23 ltld3 .i.a4 24 •xa4 •xe2 25 .:l.f3
.:tae8 1 e4 e5 2 ltlf3 ltlf6 3 ltlxe5 d6 4 l£lf3
Again this is simple, sensible development. lbxe4 5 d4 d5 6 .i.d3 lbc6 7 0-0 .i.e7 8
Black had a tricky try in 25...b5 26 cxbS c4 lbb4 9 .i.e2 0-0 1 0 lbc3 .i.e& 1 1 ltle5
liJb4!? 27 axb4 c4. The idea is to kick the f6 1 2 ltlf3
knight and follow with ...1ict+ and ...i.eS+, This simple retreat causes Black the most
but White keeps control and a clear advan- problems.
14
3 et:Jxe5: The Main Line with 8 c4
balanced play) 18 a3 llk6 1 9ltixc6 1fxc6 20 ltic6 1S cxdS (1 5 ltixdS?I was ttied in lvan
ll\xc4 1i'xc4 (Minakov-Moq,rado, correspon chuk-Ponomariov, FIDE World Champion
dence 1 999) and here Morgado gives the ship, Moscow 2002, but after I5 ...Ld5 1 6
l'lJUalising line 21 i.d2 1fxd4 22 'iVxb7 'iVxd2 cxdS 'IVxdS 17 'iVa4 i.f6 1 8 lld1 llad8 1 9
23 1i'xe7 c4. i.e3 f4! 20 i.xf4ltixd4 21 ltixd4 .i.xd4 22
12 . �h8
. . i.e3 cS Black was already a bit better)
Stepping away from any future trouble on 1S ....txd5 16 .i.f4ltixc3 17 bxc3 .i.d6 Black
the a2-g8 diabJOnal. has comfortable equality.
12...c5?1 is premature: 1 3 i.e3 :C8 14 c) 13 lle l, however, is a reasonable alter
dx.cS i.xcS 1 S .i.xcS D.xcS 16 1Wb3 aS 1 7 a3!? native to 13 a3. 13 ... c5 1 4 .i.e3 f5 1 5 a3 f4
(17 lladt is less clear: 1 7...'iVe7 1 8lDa4 i.f7 (Grischuk assesses 1 S...ltixc3 16 bxc3 llk6
19 11'e3 llc7 20 a3 ltia6 21 cxdS ltid6 22 17 cxdS i.xdS 1 8 dxcS f4 1 9 i.d4 as slighdy
1Wxe7 llxe7 23 .i.xa6 bxa6 is very messy but better for White) 16 i.d2ltic6 17ltixe4 dxe4
lllack seems to have compensation, Khalif 18 dS i.xd51 1 9 cxdS cxf3 20 dxc6 fxc2 21
man-Karpov, Denpasar 2000) 17 ...ltixc3 11'xe2 bxc6 22 i.xf4 and White's better
(17.. dxc4 allows White to play a very con
. structure gives him a small advantage, Gris
vincing queen 'sacrifice': 18 ltixe4 cxb3 1 9 chuk-Motylcv, FIDE World Championship,
�xeS .i.f7 20 axb4 axb4 21 i.d 1 and White Moscow 2001 .
has a healthy matc..'t'ial advantage) 1 8 1fxc3 13 ...�c3 14 bxc3 �
75
Th e Petroff Defence
quite goo d enough to equalise: 16 cxd5 (or 26....i.d8 27 Wxh6+ i..h7 28 .i.bS) 27
.i.xd5 (or 16...1ixd5 17 .i.b2 'irg5 18 c4 .i.d6 'ifxc7 'ifxc7 28llxc7 Q..eko) .
19 .i.c3 and White has the easier game) 17 c4 26 h47
.i.f718 .i.b2 and now: This is a clever idea but there was an in
a) If 18....i.d6 White must be a bit careful. stant win with 26 .i.bS! gxf4 27 D.xa7 as
19 ..c2 (supporting c4 is essential; not 19 Black cannot escape the pin.
.i.d3?1 .i.f4! 20 dS c6 and White's centre col 26 ...gxf4 27 •xf4 ..tg7 28 •xf5 'ird8?
lapses disasttously) 19....i.g6 20 .i.d3 .L:d3
21 'irxd3 and White has an edge (Belov).
b) 18...5 19 .i.c3 cS 20 dS .i.f6 21 1ic2 b6
22 .i.d3 .i.xc3 23 ..xc3 �b7 24 �f3 llXI6
25 ILleS and the protected passed pawn
makes White's poslbon preferable,
Kasimdzhanov-Adams, FIDE World Cham
pionship, Tripoli 2004.
1 6 lle1 ..tf6 1 7 a4
Trying to gain space on the quecnside. 17
�b3 has also been tried: 17...b6 (an intrigu
ing alternative is 17... dxc4 18 ILleS .i.c8 19
.L:c4 1Llxd4!? 20 cxd4 ..xd4 21 'irxd4 .i.xd4
22 .i.e3 .i.xa1 23 D.xa l with equality) 18 .i.f3 Now White is wmrung again. After
.i.gS 19 cxdS .i.xdS 20 .i.f4 D.c8 (or 20...1Lle7 28.....e8 29 1if4 lie7 30 .i.b5 We6 31 .i.xc6
21 i..es c6 22 .L:f6 :Xf6 23 IL!ct �g6 24 ...xc6 32lLa7 White has only a small advan
00 h6 - Leko-Kramnik, Dortmund 2000 - tage.
and now C.alkin suggests that White gains an 29 l£ig511 hxg5 30 hxg5 Wf8
edge with 25 IL!b4 .i.xf3 26 1Wxf3) 21 .i.eS This is the only move to save the king.
16
3 liJxe5: The Msin Line with 8 c4
3 1 1th3+ -*.h6 32 .i.f5 b) 14 111'a4 has also been tried but Black
The simple 32 gxh6 l:r.e8 33 llxc7 :Ct+ 34 has no real problems: 14...f415 .i.d2 �h8 16
.tn etlc7 35 llxa7 is also winning. :ab1 l:tb8 17 l:tfc I dxc4 18 .i.xc4 .i.g4 19
32 ...J:le8 33 J:lxc7 J:le1 + 34 Wh2 liJe7 35 i..e2 a6 20 h3 .ihS 21lbg5?1 (now the tactics
g41 .i.e6 work out in Black's favour; White should
l.cko supplies the neat winning variation have settled for 21 etlh2 .i.xe2 22 l:txe2 'ifdS
'5. .l:te6 36 �g2 :a6 37 c6!? llxc6 38 l:r.xc6
. with equality) 2 l ....i.xe2 (2t....i.xg5?! is not
Q'\xc6 39 'irxh6+ 1Wxh6 40 hrxh6. the answer: 22 .i.xhS f3?1 23 .i.xgS 1i'xg5 24
36 Wxh6+ 1i'xh6+ 37 gxh6 liJxf5 38 .i.xf3! l:r.xf3 25 1i'xc61 l:txf2 26 l:r.e8+ l:lfB 27
gxf5 .i.xf5 39 J:lf7 l:r.xb8 9e3+ 28 �h 1 l:r.xb8 29 1i'xc7 and
White's pawns win easily. White should win this ending) 22 etle6 'irdS
39....i.g6 40 J:lxa7 J:lc1 41 Wg3 23 l:r.xe2 (if 2.1 etlxfB f3 Black is already much
:Xc3+ 42 �4 .i.h5 43 �5 llf3 44 c6 better because 24 gxf3 .i.xf3 25 etle6 1i'f5
.tg4 45 a5 J:lxf2 46 a6 llc2 47 llc7 1 -0 gives a winning attack) 23... f3 24l:r.ce1 fxg2!
P""'"-----. and Black was clearly better in Rohde-
Callie 5 Scirawan, Estes Park 1986.
Shirov-Gelfand 12 1Vb3
Leon 2001 Increasing the central pressure. 12 l:r.cl is
the subject of Game 6, while a worthwhile
1 e4 e5 2 liJf3 liJf6 3 lbxe5 d6 4 liJf3 option is 12 etlt:S. flor example, 12....i.f6
/li)xe4 5 d4 d5 6 .i.d3 .i.e7 7 0-0 liJc6 8 (Short assesses 12...etlxc3 13 bxc3 etlc2 14
c4 liJb4 9 .i.e2 0-0 1 0 liJc3 .i.e& 1 1 .i.e3 l:r.c I etlxc3 15 fxc3 .i.e4 as unck-ar) 13 g4
.if5 .i.e6 14f4etlxc3 15 bxc3 etlc6 16 .i.f3 .i.xeS
11 ... f5 is a h'Dod alternative. flor example, 17 dxeS d4 (White has the initiative after
12 a3 etlxc3 13 bxc3 etlc6 17...dxc418 fS .ic8 19 .i.cS 1i'xd1 20 l:r.axd1
l:r.e8 21 l:r.fcl) 18 cxd4.ixc4 19 dS etle7 20
.i.cS .ixfl 21 �ft 'ird7 22 'lfb3 and
White's bishops and impressive centre pro
vide ade<Juatc compensation, Short-Anand,
Amsterdam 1993.
1 2...c6
Choosing to bolster the centre but
12...dxc4 is also acceptable: 13 .i.xc4 aS t 4
etlxe4(14a3 can lead to a neat draw by repe
tition after 14...etld2! 15 .ixd2 .i.c2 16
.ixf7+ l:r.xf7 17 1fe6 .ifS 18 '1Vb3 .ic2 19
'ire6) 14....ixe4 15 a3 a4 16 'ifdl etlc2 17
.J:r.cl etlxc3 18 fxe3 .id6 19 ..tbS 'iff6 20 l:r.c3
and now: 'irh6 is unclear, Barua-Manesh, Raipur 2002.
a) 14 cxdS .ixdS 15 c4 .ixf3 16 .ixf3 f4 1 3 cxd5
17 .idS+ �h8 18 .ict etlxd4! 19 l:tb1 (19 White has various alternatives:
i.b2 does not alter the assessment: l9...c5 20 a) 13 c5 etlxc3 14 bxc3 etlc2 15 1fxb7
.ixd4 cxd4 21 .i.xb7 llb8 22 .idS .icS is etlxat 16 l:r.xa1 1fe8 17 'ira6 f6 18 etle1 'ireS
l"lJUal) 19....ic5 20 l:txb7 f3! 21 .ixf3 etlxf3+ (Kramnik states that White has compensa
22 gxf3 112-1/z Hubncr-Yusupov, Rotterdam tion after 18 ...1Wd7 19 etld3 .ixd3 20 .ixd3)
191'18. 19 'ffxc8 .ixc8 20 .i.f4.idS 21 etlc2 l:r.eB 22
17
The Petroff Defence
.i.bS.I:le7.
14 bxc3 .!bxd5 Game 6
Kotronias-Motylev
Moscow2004
1 5 1ixb7
White can play safe with 1S c4 tllxe3 16
fxe3 'ii' c7 17 .i.d3 .i.xd3 18 'lf'xd3 .llfe8 with
an equal game.
1 5 lL'lxc3 1 6 ..i.c4 llb8
••.
18
3 lC.xe5: The Mein Lin e with 8 c4
a) 10 'i'a# is tempting but Black can de lle7 and ilie position is finely balanced)
fend: 10...'i'd7 11 'i'xd7+ �d7 12 ltk3 15.. ll.b8 16 .xa7 (defending ilie b-pawn
J:lhd8! (this is the way to equalise: after wiili 16 llab1 has other drawbacks: 16. ..if5
.
12...ltlxc3 13 lLk5+ �e8 14 bxc3 ltk6 15 17 llbc1 gS!? 18 .ig3 g4 19 tDes llxb2 20
cxd5 .ixd5 16 .ic4 Shirov claims an edge lDxe4 .i.xe4 21 lDxg4 .i.g7 with an unclear
for White) 13 a3 lDxc3 14 lDe5+ �e8 15 game) 16 ...1lxb2 17 lDa4 (or 17 ltlxe4 dxe4
bxc3 lDc6 16 lbxc6 bxc6 17 cxd5 .ixd5 18 18 Wa3 llc2 19 lDe5 .ixe5 20 .ixe5 f6 21
J:ld �f8 19 .if4 .id6 20 .i.xd6+ cxd6 21 .if4 .ig4 wiili a messy position) 17...1lb4 18
.id3 g6 22 f3ltdb8 231leb1 �e7 and in this a3 (White must be careful wiili his a4-knight;
dead level position a draw was agreed, Rau instead 18 llab1 ? 'i'bsl 19 Wxb8 llfxb8 and
sis-Bacrot, France 2003. White was losing in Al-Modiahki
b) After 10 cS!? Black must defend 3b'llins t lordachescu, Dubai 2004) 18.. ll.c4 19 ltlb2
ideas with 'i'b3. llxd4 20 .i.e3 llc4 21 lDxc4 .ixa1 22 llxa1
b1) The natural 10 ...0-0?! runs straight into dxc4 with an 'unclear' verdict from Ior
!rouble after 11 'ifb3! aS 12 aJ lDa6 13 dachescu.
1i'xh7. The attempt to trap the queen with 10 0-0 1 1 .te3 .tf5 1 2 llc1 dxc4 1 3
.•.
19
Th e Petroff Defence
b) 17 ....i.f6 18 .i.d2 .d6 19 llfet lladS Exploiting the loose bishop on c4. Black
20 c4!? is Nijboer's suggestion (20 h4 c5 21 could also play the simple 19... b5 2U .i.d3
h5 cxd4 22 cxd4 &i:k7 21 h6 lDg6 24 hxg7 llfcS 21 life1 .i.d6 22 lL!e3 ltlxe3 23 .i.xe3
.i.xg7 led to a messy position in Nijboer .i.d5 24 .e2 'itf7 with equality .
foressinet. Leon 2001). After 20. .�7 21
. 20 .i.b3
.i.a5 lieS 22 .i.c3Q)g6 White's space advan White ambitiously keeps the tension. The
rnge makes his position preferable. quiet 20 .i.xe6+ ltlxe6 21 lL!e3llfeS 22 ltlf5
b) 17...ltlxe3 1S fxe3 .i.d619lL!c4 .i.c7 20 .i.fB is level.
e4 f6 is unclear - Black's bishop pair com
pensates for White's centre.
c) 17... .i.d6 tS .i.d2 'lh14 19llfet ltlf6 20
a4ltlg4 21 .i.f4 .i.dS 22 .i.g3 lLixeS 23 clxe5
e'h6 (Hubner pointed out the blunder
23...'itxa4? 24 .i.xh7+1 *xh7 25 'lh15+ WgS
26 exd6) 24 .i.e4 .i.xe4 25 "ifxe4 .i.c7 is
equal, Klimov-Smikovski, Toljatti 2003.
23lQe31lae8 24lQf5lQe7
A sharp try is 24....i.f4!? 25 'W'g4 .i.xd2 26
lLlh6+ gxh6 27 11'xd7 .i.xe1 28 'trxb7 lle2
with an interesting and unbalanced position.
25lQxd6 ••d6 26 c4lQg6
The greedy 26...1Vxd4?! allows a deadly
pin: 27 .i.b4llf7 28lle6.
1 7 16
••• 27 .,5 ••d4 28 .i.b4 l0114
Less accurate is 17 ... .i.gS 1S .i.xg5 'itxgS Black has to play creatively to survive. If
19 llfel llaeS 20 g3 "iff5 21 "ifdt ltlb6 2S..Jlxcl+? 29llxe1 lidS 30 ..e6+ Wh8 31
(White has a promising initiative after .eS+ .!DEB 32 "fif7 Black's weak back rank
2t..lle7
. 22 .i.d3 1i'f6 23 c4 ltlb6 24lle4- costs him the game.
Topalov) 22 .i.d3 "iff6 23 1i'c21!f> 24 llbt 29 'lrh5 g& 30 1ra5
.i.cS 25 lle2 lie7 26 llbe1 and White had Motylev points out that 30 .c5 'trxc5 31
some prcsswe in Topalov-Akopian. Linares .i.xc5 llxet+ 32lLet lids 33 .i.xa7 Wf7 is
1995. equal.
1 8� 30 Jle51
••
Or tSlLld3 1i'd7 19llfe1 .i.d6 20 h3 .i.f7 Once again 30....1Let+ leads to back rank
21 .i.b3llae8 22 "ifg4llxet+ 23llxc1 lids problems: 31 llxe1 llf7 32 .i.c5 'lrh2 33
24 Wxd7 llxd7 25 lL!es Le5 26 dxe5 fxe5 .d8+ Wg7 34 Wh2 (one clever preparatory'
20
3 �xe5: Th e Main L ine with 8 c4
move is required; instead 34 l:le8?1 allows 41 b4 llf3 42 lld7 + �f7 43 .i.e3 Jlxh3
Black to escape with a draw with 34...'iVct+ 44 :Xa7 ¢>16 45 hb6
JS 'it>h2 �f3+! 36 gxf3 'iVf4+- the king must After the alternative 4S l:la6 �S 46 l:lxb6
!(II to hl and allow a perpetual because mov �e6 47 bS llh4 48 gS �£5 49 bxc6 �f3+
in� to the g-file allows ...'iVgS+ and 1fxc5) Black's countcrplay arrives just in time to
_\4...�£5 35 lleH and White has a dangerous save the draw.
111tack (Motylev). 45...�5 46 .i.dB+ ¢>e6 47 ¢>g2 llb3
31 llcd 1 ! 48 lle7+ ¢>d6 49 :Xh7 I.Oxc4 50 .i.g5
Now White has t o be careful. I f 31 l:lxeS?! 1.0&5 51 llg7 ¢>d5 52 .i.f4 lbxg4 53 :Xg6
rxeS 32 �xfB 'iVe4 33 �ft 'iVxg2+ 34 �e2 lbxf2 54 lld6+ ¢>c4 55 ¢>xf2 llf3+ 56
'i'c4+ 35 Wd2 �f3+ 36 �c3 'iVd4+ 37 'iPb4 ¢>e2 llxf4 57 We3 llh4 58 llxc6+ Wxb4
'.f.>xffi Black's position is preferable: two 59 lle6 Wc4 60 :Xe4+ llxe4+ 61 Wxe4
pawns and White's exposed king provide Y.z-%
more than enough compensation for the
l'XChangc. Gamel
31 ...1rxd1 32 •xeS fxe5 Kasparov-Anand
After 32 ...1Vxb3 Black's exposed king al Unares2000
lows White to draw with 33 'iVd6 l:lf7 34
l:tc!t+ <l;g7 35 l:lc7 gS 36 :Xf7+ �xf7 37 1 e4 e5 2 I.Of3 �f6 3 I.Oxe5 d6 4 lbf3
'i'e7+ '.ti>g6 3M '1Fe4+. lbxe4 5 d4 d5 6 .i.d3 �c6 7 0-0 .i.e7 8
c4 � 9 .i.e2 o-o 1 o &3 b&
21
The Petroff Defence
2000). Here Hracek suggests that 1 6 .LIJ is 23 We2 'ifh4 24 g4 hS 25 i.e3 hxg4 26 i.d5
worth an edge, the point being 1 6...ltlxc3?! (Kasparov) when White escapes the attack
1 7 ltlxf7! 'ifdS 1 8 l:r.xe7 'ifxf3 19 �h6+ gxh6 and has a good advantage.
20 .i.xh7+ Wh8 21 gxf3 .i.a6 22 .i.xh6 when 1 7 d5 lL!b8
White is dearly better.
b) 12 ... c51? (the aggressive approach is
best) 1 3 .i.f3 cxd4 1 4 �xdS f5 1 5 �xe7+
1Wxe7 16 a3 �!? 1 7 �!? hxg6 1 8 1Wxc2
'ifh4 and after some very sharp play the posi
tion remains unclear.
1 1 ...lL!xc3 1 2 bxc3 l0c6 1 3 cxd5 1txd5
14 �1
The simple 14 .i.f4 is only level after
14....i.d6 1 5 .i.xd6 cxd6 1 6 c4 •as.
1 8 .c2
Possibly this commits the quL>en too early
Instead 1 8 ltleSI? may be the best choice
Kasparov-H.Olafsson, Kopavogur (rapid)
2000 continuing 1 8...i.f6 1 9 .i.b2 g6 (or
1 9...c6 20 d6 g6 21 cS and White's pawns are
very strong) 20 'ifd2 ti)d7 21 �xd7 i.xb2 22
�xffi .i.xal 23 ltlxg6 l:r.xel+ 24 'ifxe1 hxgt)
25 .xa1 c6 26 dxc6 i.xc6 27 1i'c3 and Kas
parov was dearly better. After 1 8 ... i.d6 1 9
14....i.b7 1 5 .i.d3 Aae8 1 6 c4 i.b2 9£6 20 l:r.e3 White is building a strong
h is wisest to hit the queen now. If 1 6 attack, the tactical justification being
'lrc2 Black has two reasonable replies: 20...lLe5?! 21 i.xeS i.xeS 22 'ifhs :CS 23
a) 1 6...h617 i.h7+ Wh8 1 8 i.e4 (Kas l:r.aet gS 24 h4 when White is already dose to
parov-Kramnik, I.inares 2000) and here Kas victory.
parov analysed the equalising line 1 8 ... '1rh5 1 8...g6 1 9 �5 .i.f& 20 .i.b2 �7 21 f4
1 9 �5 i.d6 20 .i.f3 'ifh4 21 �xf7+ Wg8 22 If 21 lDg4 then 2t..ixb2 22 Wxb2 hS 23
ltlxh6+ Wh8. �3 �cS 24 i.c2 c6 is just unclear.
b) The dangerous 1 6...'tth5!? is a good rea 21 ...i.g7 22 .,2 lL!c5
son to avoid 1 6 'ifc2 here: 1 7 J:r.bt �xd4! This is fine but 22... c6!? 23 d6 ltXs 24
(Black targets the h2 square) 1 8 cxd4 {White llad 1 lle6 with a hard fight ahead wa.o; also
must avoid the back rank trick 1 8 �xd4? reasonable.
i.d6) 1 8... i.xf3 1 9 gxf3 i.d6 and Black has a 23 .i.c2 •d& 24 1le3 f6 25 � llxe3 26
threatening attack (Rogers). 1txe3 ..i.c8 27 f5 ..i.xf5 28 hf5 gxf5 29
1 6...ifd8 l0f2
If Black again tries 1 6 ...'1rh5 then White is The alternative 29 �h6+ also leads tO
better placed to meet the threats to h2: 1 7 dS equality after 29 .ixh6 30 11fxh6 ltle4.
.
i.e21 i.xf3 21 .i.xf3 Wxh2+ 22 Wft 'ifh1+ Here the players agreed a draw. Kasparov
22
3 liJxe5: The Main L ine with 8 c4
wves the variation 30...fxe4 31 1Wc4 f5 31 is also fine: after 1 6 dS ltleS 17 �xeS �xeS
Wc2 1Wc5+ 33 Wh 1 .i.xb2 34 ...xb2 1fxc4 3S Black has strong, active piece play) 1 6 llb1
We5. cS 17 dxcS ltle4 (the tricky 1 7...�xb2?! back
�------., ftres after 18 1lxb2 �xc3 1 9 cxd6 �xb2 20
Game S .i.xf7+! <ifhf7 21 ...dS+ Wf8 22 ...xf5+ ...f6
Grischuk-Pavasovic 23 1i'xh7 when White is clearly better) 1 8
Istanbu/2003 �c2! �xb2 (after 1 8...lt)g3?! 1 9 fxg3 l:lxe3
-------• Black appears to be very active but a few
1 e4 e5 2 .!Of3 .!Of& 3 .!Oxe5 d& 4 .!Of3 accurate moves from White defuse the dan-
lf)xe4 5 d4 d5 6 ..id3 .!Oc& 7 0-0 i.e7 8 ger: 20 1i'd2 .i.d4 21 �xd4 1Wxd4 22 Wh2
c4 l0b4 9 ..te2 dxc4 �g6 23 ltldS! :Cs 24 �f4 1We3 25 •xe3
Immediately clarifying the siruation in the l:lxe3 26 Ld3 llxd3 27 llbe1 and White is
c.:cntre to give White an IQP. simply a pawn up) 1 9 .xd8 l:laxd8 20 llxb2
10 .bc4 0-0 1 1 l0c3 .i.xc3 21 l:lxb7 �xeS 22 .i.xcS .i.xc2 and
Also possible is the straightforward I I White still holds a slight advantage, Karpov
li)cS. For example, 1 1 ...c6 12 �3 lbxc3 1 3 Kasparov, World Championship (Game 41),
hKc3 �5 1 4 1i'b3 (14 1id3 .td6 1 S �b3 Moscow 1 984.
WVf6 16 ...g3 .i.�.-6 is also murky) 1 4......c7 I S b2) Direct action with 1 3 ltle5 is also
.i.xd5 cxd5 1 6 .i.£4 (Sax-Pavasovic, Baden possible: 1 3...�6 (not 13...lbf5? 1 4 lbxf7!
I 'JIJ9). Now 1 6...'iraS 1 7 llfe1 .1£6 leads to llxf7 1 5 .i.xf7+ Wxf7 16 �3+, a fairly
unck-ar play. common tactical theme in this line) 1 4 .1£4
li:lf5 (Black should prefer 14...liJaS!? I S .i.e2
.i.f5 with unclear play) I S lbxc6 bxc6 16 dS
cS 17 lba4 .i.a6 18 lle1 (A.Sokolov
Ag7.amov, Riga 198S). Now Black should try
l!L.c4 1 9 .i.c2 ltld6 20 lik5 .i.c8 21 .i.e5,
but White stiU has the initiative.
1 2 bxc3 li)d5
1 1 . . ..!0xc3
Exchanging is narural but retreating is also
pc 1ssible:
a) 1 t ...lbf6 is rather passive. After t2 ltleS
r6 1 3 .i.gS lbfdS 1 4 .i.xe7 Wxe7 1 5 lle1
.iw 1 6 .i.xdS lbxd5 1 7 lbxdS cxdS 1 8 '1Vb3
White's superior minor piece is enough for
n n edge. 1 3 'ttc2
b) More active is 1 1 ...�6 12 .i.b3 �f6 1 3 'lrb3 only wastes time. After l3. ..lbb6
and now: 1 4 .1d3 cS I S 'irc2 h6 Black has equalised
bl) 1 3 h3 (simply preventing ... .1g4) comfortably.
1 3. . ..i.f5 1 4 .tc3 l:le8 I S a3 lbd3 (1 5...�6!? 1 3 .....ta&
23
The Petroff Defence
27 ...1WhS 28 h3 'tWaS
Threatening mate on e I as well as the
knight, but there is a solution.
29 1Wd21 1Wxa7 30 llxf6
This sacrifice is enough for a draw.
30 ...gxf6 31 1Wxh6+ Wg8 32 1Wh4
Black steps out of the pin rather than 'tWaS % -%
blocking it with 17 ... c6. After 1 8 ltle5 Wd6 Black cannot escape the checks.
19 i..d2 lLcS 20 c4 ti)f6 21 .i.£5 White
keeps a nagging initiative. Game 9
1 8 c4 lnf6 Anand-I .Sokolov
Black could have exploited the c-fLic line Dortmund 1999
up with 1 8...c6!?. After 19 :C1 .Li6 the posi
tion is complex. 1 e4 eS 2 lnf3 li)f6 3 lDxeS d6 4 lDf3
1 9 i.b7 Ab8 20 i.e& i.d6 21 lld1 lnxe4 S d4 dS 6 i.d3 lDc6 7 0-0 i.e7 8
Advancing with 21 cS does not achieve c4 lnb4 9 cxdS lnxd3 10 1txd3 1txd5
any ck"U after 21 ....te7 22 :Ct bxcS 23 This is much more reliable than 10...ti)f6
l:xb8 Wxb8 24 dxcS Wb4 25 i..e3. which gives White the useful option of 1 1
21 .'W'c8 22 Ab3 .i.fS 23 'irb2 i.d7 24
•. 'ifb5+ and now:
lOes i.xc6 2S lnxc& 1Wg41 a) 1 1 ... i..d7?l I 2 'ifb3 and Black cannot
This excellent move is forced. If 2S .. ..z:r.aS?! regain the pawn. Timman-Kovacevic, Zagreb
then 26 .i.xh6! gxh6 27 dS opens the long 1985 continued 12...0-0 1 3 tlX3 bS 1 4 i..g5
diagonal with decisive effect. b4 (now Black's kingside is shattered but
26 llf3 even if 14..llb8 15 :lfet b4 1 6 i..x f6 i..x f6
Playing to win material with 26 f3 leads 1 7 ti)e4 i.e7 1 8 tiXS his position is very
only to a messy position after 26 ...Wh5 27 g4 unpleasant) 1 5 i..xf6 gxf6 16 ti)e4 �h8 1 7
ti)xg4 28 ti)xbS ltxb8 29 cS i..xh2+ 30 'itg2 l:fet a S 18 l:lact .i.bS?! (it wa s already bad
Wh4 31 fxg4 'Wxg4+ 32 'itxh2 Wxdl. but now Black is lost) 1 9 ti)h4 l:lgR 20 d6�
26 ..llbe8 27 lnxa7!?
.. i..xd6 21 .xf7 .tf4 22 l:lcS l:lg7 23 Wxf6
This pawn may seem poisoned, but White Wxm 24 ti)xf6 and Black could resign.
has seen deeply into the position. Also inter b) Black should play 1 t ...Wd7 but White
esting is 27 .i.xh6l? We4 28 �5 i..xeS 29 still has an edge after 12 'Wxd7+ .i.xd7 13
dxeS tl)g4 30 i..f4 with an unclear position. ti)c3 0-0-0 14 l:t:l :Ides 15 ltleS i.. b4 16
24
3 lt:Jxe5: The Main L ine with 8 c4
The solid option is 1 2 llk3 t.Oxc3 1 3 Since ....i.e6 is no longer an option, Black
1fxc3 and now: must fmd a different way to escape the pin
a) The natural 1 3 ... c6?1 allows a beautiful on the e-file, and here he prepares to unpin
tncric: 14 .i.h6! .i.e4 (the point is 14...gxh6 I S with ...�fl. An interesting alternative is
:cs 1i'd7 1 6 llaet .i.e6 1 7 dSI cxdS 1 8 :Xe6 14 ...Wd6!?, for example 1S .i.gS f6 1 6 .i.h4
fxc6 19 11fxh8+ .i.fB 20 1i'f6 and White wins) c6 I 7 .i.g3 1i'd7 with an unclear position.
1 5 .i.xg7 llg8 16 llxc4 1i'xe4 17 lle1 .xe1+
I K t.Oxe l llxg7 1 9 dS and White is much
better.
b) After 1 3... .i.e6 Black will be able to cas
ric without any great difficulty:
b 1) 14 lieS looks aggressive but Black has
1wo acceptable replies.
b 11) 1 4...1i'c4?? is an ugly blunder: lS
W'xc4 .i.xc4 16 b3 .i.c6 1 7 dS and White will
win a piece.
bl2) 14 .....c6 1 S ..e1!? 0-0-0 1 6 .i.gS
.txgS 1 7 t.OxgS llhe8 1 8 llcl ..d7 19 •as
'it>bK 20 t.Oxe6 llxc6 21 llxe6 fxe6 22 •cs g6
2:l g3 .d6 was equal in Milos-Zarnicki. Villa 1 5 11fxc7
Martelli 1998. Taking the pawn docs not force an advan
b 1 3) 1 4.....d7 1 5 .i.gS f6 1 6 lle2 fxgS 17 tage, but neither do the alternatives:
:acl 0-0 18 llxe6 .i.f6 is unclear. a) 1 S .i.f4 lld8 1 6 l:te3 �f7 17 ..xc7 lld7
b2) 14 ..xc7 (White wins a pawn, but 18 ..c3 l:lhd8 19 a3 .ite4 20 I.Od2 (the solid
Ulnck will have strong compensation based way out is 20 l:tdt .i.xf3 21 l:txf3 ..e4 22
on his bishop pair and light square control) l:te1 .xd4 with dead equality - Varavin)
1 4. .i.d6 1 S ..c3 (1 S .c2 leads to similar
. 20....i.h l l 21 �3 .xb3 22 I.Oxb3 .i.c6 23
25
The Petroff D e fen ce
:C3 h5 and Black has plenty of compensa 'ilig2 1i'c4+ with a perpetual check.
tion, Varnvin-Egin, Kaluga 1998.
b) 1 5 b3 �fl 1 6 1i'xc7 llhe8 17 "lfc4
..xc4 18 bxc4 b6! (fixing then c-pawn as a
target; Motylev points out that 18 ...:Cc8?! 1 9
c5 b6 20 .i.a3 bxc5 2 1 llact gives White a
dear advantage) 19 .i.b2 llac8 20 l:lacl .i.d6
21 l:lxe8 l:lxe8 22 c5 .i.f4 (Zhang Pengxiang
Motylev, Shanghai 2001). Motylev now sug
gested that White should play 23 l:let llxel+
24 tfue1 bxc5 25 dxc5 �e6 26 lDg2 .i.d2 27
lDe3 �d7, assessing the position as equal.
1 5 0-0 1 6 1lxe7
•••
Game 10
Nyysti-M.Rychagov
Gothenbutg 2003
•xg4+
White's mating threat is demonstrated by
18....i.e4?? t9 l:lxh7+.
1 9 •g3 'tfxd4 20 .lbg6!
White very wisely heads for a draw. In
stead 20 l:lxb7? allows Black a deadly attack
on the g-file with 20...l:lg8 21 l:le1 ..d5!
(2t....i.ffi! fails to make the most of Black's
chances: 22 llxfl l:lxg3+ 23 hxg3 llg8 is only
unclear) 22 llee7 11fdt+ 23 <i>g2 .i.e4+ 24
l:lxe4 l:lxg3+ 25 hxg3 11fd5 26 l:tbe7 f5 27 'This is cet"tainly a threatening-looking
l:t7e5 "lfxe4+ 28 llxe4 fxe4 and Black should move, but Black has convincing replies.
win this ending. 1 2 o-o-O
..•
26
3 l0xe5: The Main L ine with 8 c4
retreating: 13 ....te7 14 Wf3 g6 1 5 g4! and 1 4 lbc3 is a solid option. However, after
White wins a piece) 14 1Vf3! lbe4 15 lbc3 14...lbxc3 15 bxc3 Wx£3 16 gxf3 Black has
�xc3 16 bxc3 Wxf3 1 7 lbxf3+ i..e7 1 8 i..a3 no problemS. For example, 1 6 ...llde8!? 1 7
Ac6 1 9 .i.xe7 Wxe7 20 d5 and White wins lbxf7 llhfB 1 8 l£le5 .i.h3 with compensation.
cas�y. Zuidema-Barendregt, Amsterdam 1 4...-*.h4
1 966.
b) 12 ... f6?! is more complicated but still
bnd: 1 3 lbc3 (13 Wf3 is only enough for
Cl(uality after 1 3...g6 L4 lbc3 lbxc3 1 5 bxc3
1i'xf3 1 6 lbx£3 c;ild7) 1 3 ...lbxc3 1 4 1fxf5
�b5 1 5 ..g4 l£lxd4 (of course Black would
like to take the piece with 1 5... fxe5 but after
1 6 llxe5 Wxd4 1 7 llxc7+ Wxe7 1 8 .i.g5+
�d6 1 9 lld1 White is dearly better) 1 6 l£ld3
and the pin causes Black immense difficulty:
bl) 16 ...Wf7 1 7 l£lf41 1ff5 18 llxe7+!
�xe7 1 9 1fxg7+ �d6 20 .i.e3 llhg8 21 1ff7
cS Qf 21 ... b6 then 22 Ld4 1Vxf4 23 .i.xb6!
clears the d-file and continues the attack) 22 1 5 �3
1i'xb7 llgb8 23 Wa6+ llb6 24 Wa3 and After this White seems to be in trouble.
White's attack was obviously progressing weU FoUowing 15 lbc3 there arc many paths bu t
In Jonsson-Schandorff, Panormo 2001 . White cannot force more than equality:
b2) If 1 6...lbc2 White must find 1 7 lbb4! 1 5...lbxc3 1 6 bxc3 .i.e6 17 g5 (or 17 .xd5
�xb4 1 8 .xb4 c5 1 9 Wg4 �f7 20 i..h6! .i.xcl5 1 8 g5 f6 1 9 gxf6 .i.xf6 with a balanced
gxh6 2 1 llad1 with a massive attack, De Fir game) 1 7.. .llhe8 1 8 c4 .xd4 1 9 llb1 c6 20
mian-Plaskett, Copenhagen 1 985. 17 llxe 7+ .i.f4 (if 20 lbxc6 Black bails out to a draw
leads to a draw after 1 7...Wxe7 1 8 1Ve2+ Wf7 with 20 ...Wxt2+1 21 1Vxf2 .i.xf2+ 22 Wxf2
19 'irxc2 llad8 20 lbe 1 1i'd1 21 1Vxc7+ Wg6 bxc6) 20...1Wc5 21 lbxc6 lld31 22 llxe6 (after
22 1Vg3+, as the pins force White to setrle for 22 Wxd3 Black can force a draw with
n perpetual. 22..Wxf2+ 23 Wh t .i.dS+ 24 Wxd5 l'lxet+
c) A reasonable alternative is 1 2...g6 1 3 25 llxe1 .xet+ 26 Wg2 '1Ve2+ 27 Wh3 .ft+
1'13 (not 1 3 g4?1 lbxt21 1 4 W n 'eht+ 1 5 28 Wxh4 1i'xf4+; instead 22 lle3 continues
�xf2 i..h4+ when White is i n trouble) the fight: 22...llxc3 23 lbxa7+ Wxa7 24 .i.xe3
1 l .. Wxd4 1 4 lbc3 1i'xe5 1 5 .i.f4 (or 1 5 Wb8 25 We4 .i.xgS 26 i..xgS and with oppo
lDxe4 0-0-0 1 6 .i.f4 1i'a5 1 7 llact lld7, when site bishops and both kings vulnerable, the
the game is unclear) 1 5 •a5 1 6 b4!? 1i'a3!
... position is totally unclear) 22... fxe6 23 ..xd3
(lhis is the best way to contain White's initia Wxt2+ 24 Wht .xf4 25 lbxa7+ �b8 26
tive; 1 6 ...1Wxb4?1 is obviously bad after 1 7 lbc6+ Wc7 27 lba5 lidS 28 1Ve2 lld2 29
ltld5, while 1 6. .ixb4 is not a clear mistake llxb7+ Wd8 30 llb8+ Wc7 lf2-1/z Wahls
hu1 after 1 7 llxe4+ .i.e6 1 8 .te5 White's Yusupov, Germany 1 992.
initiative is very dangerous) 17 lDd5 .x£3 1 8 1 5 10xf2
.•.
!(X f3 i..dS 1 9 fxe4 .te6 20 lbxc7+ i..xc7 21 This neat trick turns the game in Black's
j_xc7 (the game has fizzled out to equality) favour. Another attractive choice is
2 l ...f6 22 a3 llc8 2l llact <t>f7 1/z- 1/z Be 1 5...llhg8!? 1 6 .ic3 hS 17 h3 hxg4 18 hxg4
liavsky-Smyslov, Re�o Emilia 1 986. .i.e6 when Black has a dangerous initiative.
1 3 .,3 g6 1 4 g4 1 6 'lfxd5 l0h3+ 1 7 �2 llxd5 1 8 gxf5
27
The Petroff Defence
.i.xe1 1 9 �xe1 lle8 20 �f3 llxf5 21 11tis little trick enables Black to save both
Wg3?1 rooks, so White resigned.
White should prefer 21 tik3, but after
21 ...�f4+ 22 .i.xf4 lLf4 Black is still obvi Game 1 1
ously on top. Shirov-Gelfand
Astana 2001
21 �1 1 22 lbbd2
•..
Anothc:r way to win is 22...:C3!? 23 'iPf2 8....tc6 possibly commits the bishop too
.l.ex£3+ 24 Wxgt l:tg5+ 2s Wht Af2. early:
23 Wg2 g5 24 �b3 g4 25 �1 a) 9 cxdS is interesting: 9 ....i.xdS 1 0 W
There is also little hope aftc:r 2S �eS lilxc3 11 bxc3 0-0 1 2 i.f4 .td6 13 .txd6
�xd4 26 �xd4 AfxeS 27 �g3 £5. Wxd6 14 lL\g5 h6 (14. . £5 1 S c4 .i.xg2! 1 6
.
28
3 lDxe5: Th e Main Line with 8 rc 4
Also possible is 8 ..ig4 9 ltic3 (9 cxd5 is R.xd7 23 t'fJxf7 'ttfB 24 t'iJxh8 Wxh8
loo simple an approach: 9...Wxd5 10 ltic3 (Sutovsky-Huzman, Tel Aviv 2000) . Now
lt:\xc3 1 1 bxc3 0-0 12 IZ.c l ..i.xf3 13 Wxf3 White should have continued 25 WcS :<12
Wxf3 14 gxf3 ..i.d6 1 5 :b1 lLb8 1 6 ..i.e3 b6 26 'ttc3 :d5 27 1Z.c7 with a dear advantage.
17 f4 t'iJe7 1 8 c4 c6 1 9 Wg2 :rds and Black 1 0 c5
had equalised in Kotronias-Langrock, Ham
burg 2001) 9...t'fJxc3 (9...t'iJf6 transposes to
H ...t'fJf6 9 t'fJc3 ..i.g4) 1 0 bxc3 0-0 1 1 IZ.c1
tlxc4 1 2 ..i.xc4 Wd6 13 :b1 ltab8 14 :b5
nnd White has hi� usuaJ edge.
9 .!003
Black is okay in the variations after 9 cxd5
i;)xd5. For example, 10 t'iJc3 0-0 11 :e1 t'iJf6
1 2 a3 ..i.g4 1 3 d5 t'iJd4 14 ..i.xh7+ t'iJxh7 1 5
Wxd4 R.xf3 1 6 gxf3 R.f6 1 7 1i'd3 Wd7 18
.if4 t'iJg5 �t is important to improve the
badly placed knight immediately; instead in
Svidler-Ovetchkin, Tomsk 2001 Black played
IH..1lfe8 when 19 t'iJc4 stranded the knight Ensuring a space advantage at the expense
on h7 - White was a touch better after of taking the pressure off Black's centre.
19...b6 2.0 :e2 c5 21 llact R.d4 22 d6) 19 With the bishop already on e6, 10 cxdS
.lxg5 ..i.xg5 20 IZ.c4 WfS and White's weak seems less logical but it is b)• no means bad.
ened structure means that Black has no prob For example, lO.. .t'fJxdS 1 1 IZ.c1 0-0 12 .i.e4
lems. .i.f6 1 3 a3 h6 14 1r'd3 t'fJce7 1 5 .i.d2 c6 1 6
ltad1 llc8 17 h3 t'iJxc3 1 8 bxc3 b S 1 9 .i.f4
..i.dS 20 t'iJd21 llc8 21 1i'g3 Wh8 22 R.c21?
(dearing the way to centraJise the knight;
instead 22 Wg4 R.c6 23 'i'hs t'iJg81 24 .i..b I
rfJ 25 'irf3 .idS 26 1r'g3 .ig7 was equaJ in
Movscsian-Yusupov, Batumi 1999) 22....d7
23 t'iJe4 and White's position is mildly prefer
able.
1 0.....tg4 1 1 ..tb5 0-0 1 2 ..txc& bxc& 1 3
h3 ..txf3 1 4 11'xf3 Ae8 1 5 ..te3 1ib8
Black can provoke many exchanges with
1S ...t'iJe4 but after 16 t'fJxe4 dxc4 17 1r'xe4
..i.xcS 1 8 •xc6 .ixd4 19 ltad1 1tf6 20 'irxf6
9 .i.e6
..• .ixf6 21 b3 White's better structure is a sig
9...R.g4 exerts less influence on d5, so it's nificant factor.
luwcaJ for White to continue 10 cxd5. For 1 6 .:ab1 .i.f8 1 7 l:lfe1 eb4
example, 1 0...t'iJxd5 1 I ..i.e4 R.e6 1 2 Wd3 1 7 ...t'iJe4 still does not help. Gelfand
�cb4 1 3 Will c6 14 a3 t'fJa6 1 5 t'fJxdS ..i.xdS anaJyses 18 .if4 t'iJxc3 19 bxc3 llxct+ 20
J (, l'.lc1 t'iJc7 17 R.gS R.e6 1 8 ..i.xc7 1r'xe7 1 9 llxe 1 Wb2 21 lle2 as dearly berter for White.
li]c5 0-0-0 20 Wc2 l:hd4?1 (20. . 1r'd6 21 . 18 a3
Jlnd1 is better, after which White has only a 1 8 Wf4!? l:tac8 1 9 lle2 also keeps an edge.
Hmall advantage) 21 .i..xc6! 'iVd6 22 ..i.d7+! 18 11'c4 1 9 l:lbc1
..•
29
The Petroff Defence
30
3 l0xe5: The Main Line with 8 c4
31
The Petroff Defence
35 11'b4?1
The d-pawn had to be taken, even though
Black is a bit better afu:r 35 ltxd2 llxd2 36 52 lDb1
/t)xd2 ..xh4 37 Le6+ �h8. The queenside pawns are also decisive af
35 .Aed8 36 l0g5
•• ter 52 /t)f3 lZ.e8+ 53 Wft ltxdl+ 54 .i.xdl aS.
Now it is too late for 36 llxd2?!. After 52. . .Ae8+ 53 �1 llxd1 + 54 .*.xd1
36 ...�1+ 37 Wh2 ltxd2 38 /t)xd2 Wet the lDe4 55 .*.c2 a5 56 lDB3 lDd2+ 57 �2
pin is deadly. llb8 58 g4 b4 59 �b1 l0xb1 60 .i.xb1
36 •c2 37 '*.b3
••• AcB 81 .*.f5 llc3 0-1
32
3 /i)xe5: The MBin Line with 8 c4
Summary
The line 6... .i.e7 7 0-0 liX6 8 c4 has been enduringly sruuied, but it is still very popular. Strate
lo{ic understanding of the variation is the key to succL"SS - there is no real need to memorist: the
det:uls.
Although 8...lbf6 looks quite reasonable, at the moment aD the attention L� on 8 ...lbb4.
White should really retreat his bishop to e2 - Games 9-1 0 show that 9 cxdS does not lead to
�uccess against accurate black defence. After 9 .i.e2 the bishop fianchetto to b7 (Game 7) may
well not equalise (see the note to White's 1 8th move). In the case of 10 ... .i.e6 t i lDeS neither a
counterattack in the centre with 1 1 ...c5, nor ousting the centralised knight with 1 1 ...f6 shakes
off aU Black's operung problems. Mobilisation on the qut:enside with 1 1 .i.t:J seems to give
lc:ss chances for an advantage after both t t ...fS and 1 t....i.f5.
With the knight on b4 it is logical to develop the bishop to f5 as soon as possible. After 9
.le2 0-0 10 lbc3 .i.f5 1 1 lbeS, t l...cS is quite effective, while in the event of 1 1 a3 lbxc3 1 2
bxc3 llX6 the complicate<.!, strategic play has not revealed any appreciable advantage for White
�o far.
1 e4 e5 2 /i)f3 /i)f6 3 /i)xe5 d& 4 /i)f3 /i)xe4 5 d4 d5 6 ..td3 ..te7 7 0-0 Nc& 8 c4 /i)b4
8...lbf6
9 llX3 - Ga111e 1 1; 9 h3 - Gtllllt: 12
8 ..te2
9 cxdS lbxU3 10 1Wxd3 1i'xd5 t l llet .i.fS (D)
1 2 g4 Ga111e 9; 1 2 lDc5 - Ga111t 10
-
8...0-0
9...dxc4 - Ga111e 8
10 lllc3 ..tf5
10.. b6 G1111�e 7
. -
1 1 . ..tf5
. . 10 ..te6
. . . t3. . . :es
33
CHAPTER TWO I
3ltJxe5: The Main Line
with 8 lle1
1 e4 e5 2 l0f3 lOt& 3 l0xe5 d& 4 .!Df3 to equality (the latter in view of 1 3 li)xc7
l0xe4 5 d4 d5 6 .i.d3 .i.e7 7 0-0 .!De& 8 �d7 1 4 .i.t4Q. Black can take on dS with his
l:le1 queen without harm if he precedes this with
After 1 e4 eS 2 li)f3 li)f6 3 lilieS d6 4 li)f3 l O... il..xO 1 1 .xf3 (Game 1 4), while in
li)xe4 5 d4 dS 6 .i.c.J3 il..e7 7 0-0 liJc6, it's true Game 1 5 Black takes the strategic decision to
that the immediate attack on the dS-pawn with occupy the square in front of the isolated
8 c4, as covered in the previous chapter, is in pawn with 1 0. .li)xd5.
.
1 0...li)xd4 or 10 ... .i.xf3 11 .xf3 li)xd4. This •d6! and .0-0-0 (Mikhalchishin). to li)bd2
..
is the reason White often releases the tension .d6!? 1 1 c4!? (Game 1 6) makes some sense
in the centre with 1 0 cxdS (Games 14-1 5). because ... d5xc4 does not work, while the
Pillsbury introduced I O...•xdS, but after 1 1 manoeuvre li)d2-ft deserves consideration
llJc3 neither the retreat 1 t...1i'h5 nor the after 1 0 li)lxl2 0-0. However, this plan is less
queen swap I t....i.xO 1 2 li)xd5 il..xdt leads appeaJjng after 1 0 li)lxl2 •d6 in view of
34
3 liJxe5: The Main L ine with 8 lie 1
n2-g8 with 1 1 ...';;,h8 (Game 1 7) offers a sacri- even in this symmetrical position.
nee of the b-pawn, while Botvinnik proposed However, Black should seriously consider
prot1.:cting b7 with 1 1 ...�5 (Game 1 8). 10 ...�xd4 here: 1 1 cxd5 i.xf3 (1 1 ...c5?! is
inaccurate: 1 2 i.bS+ WfB 1 3 .ie2 and White
Game 13 has a useful k·ad in development) 12 gxO
Shirov-Vusupov c5!? 13 dxc6 �xc6 14 i.bS 0-0 15 '5'xcl8
European Team Ch., Batumi 1999 i.xd8 16 .i.xc6 bxc6 17 i.c3 i.b6 1 8 .J:r.ad l
-------• llfd8 and the game was completely level in
1 e4 e5 2 li'Jf3 li'Jf6 3 li'Jxe5 d& 4 li'Jf3 J.Polgar-Karpov, 1-loogt:Veen 1999.
�xe4 5 d4 d5 6 .i.d3 liJc& 7 0-0 .i.e7 8 1 1 1fxf3 lt'lxd4 1 2 'W'd1
.l:le1 ..llg4 9 c4 li'Jf6 White can also try 1 2 'iVg3, when 1 2...�e6
Trying to win a pawn immediately is a is a little better for White after 1 3 .i.f5 0-0 14
mistake: 9...i.xf3?1 1 0 'if'xO �xd4 11 ..d cxd5 �xd5 1 5 �3 �f6 1 6 i.e3. However,
li:\15 12 1ih3! �fd6 1 3 cxcl5 �f6 1 4 i.g5 Black is okay after 1 2...dxc4 13 i.xc4 (Kar-
�.:'lxd5 1 5 �c3 �fB 1 6 llxe71 �xe7 1 7 �5 pov pointed out the losing blunder 13 'iVxg7?
�)xdS (this gives up material and admits that �0+1 14 �hl llg8 1 5 'ifxf6 �xc1) 1 3...0-0
It has all gone wrong; if Black tries 1 7 ... f6 1 4 .tg5 i.d6 1 5 �4 h6 1 6 i.xf6 1i'xf6 1 7
rhcn 1 8 �xf6! gxf6 1 9 i.xf6 W f7 20 •o is 1i'xf6 gxf6, Lobron-Karpov, Hanover 1 983.
fnr too strong) 1 8 i.xd8 llxd8 1 9 �4 �f6 1 2 li'Je& 1 3 .i.f5
.••
35
The Petroff Defence
tllf5 0-0 and Black keeps a useful extra 21 llxc7+ 11'xc7 22 .rlxe7+ Wxe7 23 �+
pawn) 17...¢'£8 18 lies •d6 19 %15 lids 20 �f6 24 11'xb7 %lfc8 and White has good
lOe3 ..xdl+ 21 %lxd1 :lxd l+ 22 c!tlxd1 etle6 winning chances.
23 %lxb5 b6 24 ttlc3 q;e7 and unswprisingly 1 8 -*.e3
a draw was agreed in Kasimdzhanov
Yusupov, Essen 2001 .
1 3 dxc4
•.•
36
3 liJxe5: The Main Line with 8 :e 1
'4 Wf2 �e6 35 «li'e3 Wd6 36 Wd4 (Yusu bxc6 16 .i.g5 l:tfe8 with an unclear position)
pov). 13 lllc3 'iFd7 14 'iFxd7+ �d7 1 5 .i.e3 llle6
29...liJd3 30 lte2 :dB 31 i.c7 ltd4 32 1 6 l:tad1 .i.d6 17 .i.f5 We7 1 8 lllb5 Zl.hd8 1 9
g3 % -% lllxd6 cxd6 20 h 3 (White has obvious com
pensation for the pawn) 20...b6 21 g4 h6 22
Ga1f/e 14 .i.d4 l:tac8 23 .i.c3 g6 24 .i.c2 hS 25 f3 1/z-1/z
Sax-Vusupov Kasparov-Karpov, World Championship
Thessaloniki Ofyntpiad 1984 (Game 28), Moscow 1984.
_______________. b) 12 'iFxdS should not be dangerous:
1 e4 e5 2 liJf3 lilf6 3 liJxe5 d6 4 liJf3 12 ...lllxd5 13 lllc3 0-0-0 (Black is well ad-
�xe4 6 d4 d5 6 i.d3 liJc6 7 0-0 i.e7 8 vised not to chase material with 1 3.. llldb4 14
.
J:le1 i.g4 9 c4 liJf6 1 0 cxd5 .i.e4 lllxd4 because after 15 .i.e3 lld8 1 6
..ixd4 l:r.xd4 17 a3 lllc6 18 lllb5 White will
win back the pawn and force an advantage)
14 .i.e4 i.b4! 1 5 i.d2 lC!f6 1 6 i..xc6 bxc6 17
..ie3 lCJds 18 a3 (Kamsky-Karpov, Linares
1 994 continued 1 8 %tact lllxd 1 9 fxe3 cS 20
l:tft f6 21 llf5 Zl.he8 and Black had an edge,
which grew considerably after 22 dxcS?
llxe3) 1 8...i.xc3 19 bxc3 lC!xc3 20 llact
lC!bS 21 llxc6 lC!xd4 22 Zl.c4. Despite Black's
extra pawn Karpov assessc..-s the position as
el(Ual.
1 2...11fxd4 1 3 lilc3
After 13 'iFxc7 one of BL1ck's options is to
10 i.xf3
••. force e']uality with 1 3...lllg4!? 1 4 'iff4 'ifxd3
This is much safer than 10 ...'iFxd5 1 1 lllc3 1 5 'ifxg4 0-0.
.ixf3 Qf 1 t ...'iFh5 1 2 i.b5 ..txf3 13 gxf3 �fB 1 3 0-0 14 liJb6 .g4
...
14 i.xc6 bxc6 15 i.f4 White's easier devel Black has a sound alternative in 14 ...1i'b41?
opment is worth a slight advantage) 1 2 lC!xd5 I S i.gS llfd8 1 6 'iFh3 '!fl 1 7 lllxc7 IL1cR 1 8
i.xd1 1 3 lllxc7+ Wd7 14 i.f41 ..tg4 (Kavalck i.xf6 .i.xf6 1 9 llld s 'iFd6 20 lllxfM 'ifxf6,
nnnlvsed 14...lC!h5 15 i.f5+ Wd8 1 6 i.c5 achieving easy equality.
li'lx�S 1 7 lllxa8 lllc6 1 8 .llaxdt i.d6 1 9 Zl.e5! 1 5 11fxg4
as winning for White) 1 5 d5 lCJd4 1 6 lllxa8 1 5 lllxc7 Zl.ad8 1 6 'ifxg4 lllxg4 1 7 i.e2
l:lxa8 1 7 i.c51 .i.f5 (or 1 7 ... ..tc5 1 8 llect lC!ge5 1 8 i.f4 ..if6 is also level.
.ib6 19 llc4 and White still has a clear ad 1 5...liJxg4 1 6 i.f5?1
vantage) 18 i.ft lllc2 1 9 i.bS+ Wd8 20 d6 Ths is inaccurate; 16 ..te2 lC!f6 17 lllxc7
ll'lxct 21 llxe l i.e6 22 dxe7+ Wxe7 23 .i.d4 llac8 l8 lllb5 ..ib4 is simply eCJual.
h6 24 a4 (Kavalck-Toth, Haifa 1976), when 1 6 �f6 1 7 liJxc7 lladB 1 8 i.e3 a61
•••
rhc tactics have burned out but White's A later game, Kamsky-Khalifman, FIDE
llishop pair still ensures a small advantage. World Championship, Las Vegas 1999,
1 0.. lllxd5 is studied in Game 1 5.
. continued 1 8.....ib4 1 9 lled1 g6 20 i.h3
1 1 •xf3 •xd5 12 •g3 a6 (or 20....i.d6 21 lllbS ..ie.S 22 lllxa7 lllxa7
White has two other moves: 23 .i.xa7 Zl.xdl+ 24 llxdl l:ta8 25 -'.c5
a) 1 2 .,3 lllxd4 (also playable is ..ixb2, again with equality) 21 i.h6 llxdl+ 22
12 ....xd4 13 lllc3 l:td8 14 i.b5 0-0 15 i.xc6 Zl.xd 1 l%d8 23 llxc.l8+ lllxd8 24 i.g5 i.e7
37
The Petroff Defence
Perhaps White should try 1 9 :ed1 .i.d6 llb3 bxa4 35 llb7 l:tb1 36 �3 a3 0-1
20 .i.b6 .i.e5 21 llab1 (after 21 l:txd8 :Xd8
White has some problems since 22 l0xa6? Gan1e 1 5
fails to 22...:d5) 21 ..J:lxd t + 22 :xd1 .i.xb2, Psakhis-Cooper
when White's activity is enough to hold the Pott Erin 2003
balance.
1 9....tb41 20 llf1 lbd4! 1 e4 e5 2 o!bf3 lDf6 3 lbxe5 d6 4 o!bf3
Ths keeps up the pressure. 20..id2?! al o!bxe4 5 d4 d5 6 .td3 lbc6 7 0-0 .te7 8
lows White to complicate matters with 21 lle1 .tg4 9 c4 lDf6 1 0 cxd5 fud5 1 1
l:lxc61 bxc6 22 .i.c5. o!bc3 0-0 1 2 h3
21 llc4?1
Now Black's advantage is serious. After 21
.i.xd4 l:lxd4 22 l:lfdt llfd8 23 llxd4 :Xd4 24
�fl .i.d6 Black has very little.
21 ...o!bxt5 22 llxb4 l:td7! 23 :C1
Black's advantage is confmned by the fact
that White cannot grab on b7: 23 llxb7?
loses to 23...l0xe3 24 fxc3 llc8 25 llcl l0e8,
while the line 23 .i.f4? lieS 24 llxb7 li)d4! is
also hopeless.
23 . .llc8 24 llbc4 l:tcd8 25 h3 lbxe3 26
.
fxe3 �8 27 e4 �7
38
3 l/Jxe5: Th e Main L in e with 8 Ae 1
�b6 1 6 dS .ixe4 17 .i.xc4 ll:le7 1 8 .i.e3 treat: 21 .i.f6! l:tb8 22 9h5 �g8 23 lte3 and
li.)cc8 (the greedy 1 8 ...lLlexdS docs not work: the attack is devastating.
I IJ .ixb6 lLlxb6 20 .i.xb7 l:b8 21 l:tad1 lL!c.I7 21 �5 h6 22 ..i.f61 �g8 23 Ae3 l/Je7
22 'il'bs l:te8 2.1 :Xd7 l:txel+ 24 lLlxet •e8 24 Ag3 l/Jg6 25 Axg6 fxg6 26 •xg6 1 -0
25 �f3 l:txb7 26 ...xb7 1fxd7 27 ..xa7 and
White has good winning chances) 1 9 l:tacl Game 16
�16 20 .i.bt 'i6'd7 (Shirov-I.Sokolov, FIDE Leko-Yusupov
World Championo;hip, Las Vegas 1 999). Dottmund 1998
Now Ivan Sokolov suggests that 21 .i.gS!?
i.xgS 22 lLlxgS g6 2.1 ...g3 gives an edge to 1 e4 e5 2 liJf3 li.:lf6 3 l/Jxe5 d6 4 lDf3
White. lDxe4 5 d4 d5 6 ..i.d3 ..i.e7 7 0-0 lDc6 8
14 bxc3 ..i.f6 Ae1 ..i.g4 9 c3 f5
This is more aggressive than 9. .ll:\f6 lO
.
39
The Petroff Defence
1 1 ...0-0?1
Now castling short is a mistake - Black
has two better moves:
1 0.....d6 a) 1 t ...ltlxd2 12 .i.xd2 dxc4 (not
Possibly preparing to castle long. Instead 12...ltlxd4? 13 lihd4! .i.xdt 1 4 ltlxf5 1fc5 1 5
1 0... 0-0 is also reasonable, with the foUowing: b4, when White wiU have a decisive material
a) t 1 '1Fb3 transposes to Game 17. advantage) 1 3 i.xc4 0-0-0 14 .i.c3 .i.f6 with
b) 1 1 lbft with a further split: an unclear position.
bt) t t ....i.d6 is tricky but probably inaccu b) 1 1 ...0-0-0!? leads ro sharp play: 12 cxd5
rate after 1 2 ll)g3 (not 12 lbe3? .i.xh2+1, ltlxd2! (once again 12 ...ltlxd4? is a mistake:
when 13 �? is imposs.ible because of 1 3 lihd41 .txd1 14 ltlxfS ..eS 1 5 ltlxe7+
13...l£lxf2 1 4 ..c2 ltlxd3 1 5 1Wxd3 .i.xt3 ..xc7 1 6 .i.xe4 and White's extra material
winning for Black - the problem with 1 6 ensures a clear advantage) 1 3 dxc6 ltlxf3+ 14
gx f3 is 16...1ih4+, winning the rook) 12.....f6 gxf3 i.h3 1 5 cxb7+ Wb8 16 .tn i.xfl 17
1 3 '1Fb3 and White is a bit better. D.xft i.f6 with a very mes.c;y position (Leko).
b2) t t .. .th4 12 g3 .tgs 13 .txg5 l£lxg5
. 1 2 cxd5 c!Dxd4
14 .i.e2 l£lh3+ 1 5 �g2 f4 16 ltl3d2 .i.xe2 17 White a bit better after 12...l£lxd2 1 3 dxc6
1fxe2 ltlg5 1 8 1Wg4 fxg3 19 hxg3 ..f6 with l£!xf3+ 1 4 gxf3 i.hS 1 5 cxb7 llab8 1 6 i.c4+
an unclear pos.ition, Ljubojevic-Hjartarsson, ll.f7 17 'irb3.
Belgrade 1989. 1 3 .*.xe4l fxe4 1 4 1llxe4
1 1 c41?
The other option is 1 1 ltlft. For example,
1 1 ...0-0-01? 1 2 ltk3 h5!? 1 3 h3 g6! 14 hxg4
(mstead 1 4 i.c2 i.xt3 1 5 i.xf3 .th4 1 6 1le2
lbe7 1 7 l£lf1 i.f6 1 8 1lc2 'itb8 19 a4 1lh7 20
aS g5 was unclear in Adams-Makarychev,
Oviedo 1992) 14...hxg4 15 .txc4 (White
must avoid 1 5 ltld2? because of 1 5...11fh2+
1 6 'itft ltlxf21 17 'iPxf2 .th4+ 18 Wc2 1ldc8
with an immense attack) 1 5...dxc4 1 6 ltlc4
..f6 and Black has good compensation - 1 7
.i.gS? loses to 1 7...1fg7 1 8 Le7 1ih6 19
l£lh4 l£lxc7 20 g3 g5.
40
3 '!Jxe5: The Main L ine with 8 .l:l e 1
14 .i.xf3
..• 49 'it>e3 .l:lb3+ 50 �4 .l:lb4+ 51 'it>g3
Leko assessed 14...�xf3+ 1 5 gxf3 .i.xf3 White prefers to avoid the pawn ending
H i �xd6 .i.xdl 17 �xb7 l:labB 18 l:he7 51 WgS ltbS+ 52 �h6 1lxh5+ 53 WxhS even
:xb7 1 9 d6 as clearly better for White. !hough he wins neatly after 53 ... b5 54 f4 b4
1 5 1i'xd4 1fb4 1 6 1t'xb4 ..txb4 1 7 .l:le3 ss 5 b3 56 f6 b2 57 fl b1W ss f89+ Wa2
.i.h5 1 8 .l:lb3 .i.a5 59 Wa8+.
White also has a clear plus after 1 8...a5 19 51 b5 52 h4 .l:lb1 53 .l:lxh7 b4 54
.••
a."\ .i.g6 20 f3 .i.xc4 21 fxe4 .i.d6 22 .i.d2. .l:la7 + �b2 55 �4 .l:lh1 56 'it>g5 b3 57
19 ..te3 ..tf7 20 .!Lig5 h5 'it>c3 58 .l:lc7 + �d4 59 .l:lb7 Wc3 60
Or 20 llxb7 .i.xdS 21 :lbS .i.xc4 22 l:lxaS h6 b2 61 g4 b11i' 62 .l:lxb1 .l:lxb1 63 h7
n6 23 :lc1 and, despite the opposite-coloured .l:lb8 64 �6 .l:lb6+ 65 �g7 .l:lb7+ 66
bishops, �'hitc has an obvious advanta�:,>c. ¢>h6 .l:lb6+ 67 Wh5 .l:lb8 68 g5 1 -0
20 ..tb6 21 .!Lixf7 .l:lxf7 22 ..txb6 axb6
.••
11 •.• Wh8
32 �3 Sacrificing a pawn. 1 1 ...�5 will be stud
Or 32 :lel+ �d6 33 l:le8 :lcS 34 l:lxcS ied in the next game.
l:.xcS 35 l:lb8 and White's advantage is t:vi 12 ..xb7
tlcnt. White can opt to include 12 h3 .i.hS be
32 .l:lc5 33 .l:lxc5 .l:lxc5 34 'it>d3 .l:lc4 35
..• fore playing 13 Wxb7. For example, 13 ...l:lf6
lle1 + Wd6 36 .l:le8 .l:lf4 37 f3 .l:lh4 38 h3 1 4 9b3 l:lg6 (or 14 ...g51? 1 5 .!?lfl l:lbB 1 6
b6 39 .l:lb8 'it>c5 40 .l:lc8+ 'it>d6 41 .l:lf8 g5 Wc2 with a murky position) 15 .i.e2 .i.d6
42 .l:lf6+ 'it>c5 43 b4+ 'it>b5 44 .l:lf5 �a4 (1 5 ... .i.h4?! is the wrong way to build up the
45 .l:lxg5 'it>xa3 46 .l:lxd5 attack: 1 6 l:ln .i.xf3 1 7 o!?lxf3 and White is a
46 bS �b4 47 l:lxdS is even more straight safe pawn up because 17....i.xf2+?! fails to 1 8
forward. llxf2 .!?lxf2 1 9 � f2 Wd6 20 o!?lgS! l:lf8 21
46 .Axb4 47 .l:lh5 llb3+ 48 �e4 .l:lb4+
•• 9a3 Wd8 22 .i.f4 h6 23 o!?lf3 lle8 24 .i.d3,
41
Th e Petroff Defence
1 7 Wf 1 17
Trying to escape the pressure along the g
file. In this complex position White has two
1 3 ...,5 other main options:
The other retreat 1 3 'tl'b3 also allows a) 1 7 I!Dxe4 leads to very unclear play after
Black counterplay: 1 3 ...ltg6 (1 3...ltb8 14 .a4 1 7 ... fxe4 1 8 li)gs :Xgs 1 9 .ixgs •xgs 20
simply transposes to the game) and now: .i.xhS lik71 21 .d 1 g6 22 .i.g4 h5 23 .i.d7
a) 1 4 .ib5 I!Dxd2 1 5 I!Dxd2 .i.d6 1 6 g3 (or lids 24 .i.b5 I!D5.
1 6 .ixc6 Lh.2+1 17 Wxh.2 1Wh4+ 1 8 �gl b) Curiously, 1 7 lbcs allows Black a
lth6 1 9 f3 •xel+ 20 .!Dn lthl+l 21 �ht choice of forced draws: 17 ...�e5 1 8 dxe5
.xft+ 22 Wh2 .i.xf3 and now White should .ixe2 19 exd6 ltxg2+! 20 Wxg2 .gS+ 21
accept that Black has a perpetual check be �h 1 I!Dxf2+ 22 Wh2 f4! 23 lOft � 24
cause if 23 'ii'c2? .ie4 24 .d2 :ciS Black has hxg4 1Wh4+ 25 Wg2 1i'xg4+ with a perpetual
a deadly attack) 1 6 ...ltk7 1 7 .i.d3 h5 1 8 .!Dfl check, or 1 8...•g5 1 9 .if1 I!Dxf2! 20 exd6
h4 and Black has a dangerous initiative, .!l)xh.}t 21 Wh1 I!Df2+ (lvanchuk).
Peshkov-Raetsky, correspondence 1985. 1 7 -*.14 1 8 �3 .i.xc1 1 9 llaxc1 •d6
.•.
much better for White, Gagunashvili- It is vital to support f3. If 22 �5?! then
42
3 liJxe5: The Main L ine with 8 lle 1
22... ..i.xf3 23 ..i.xf3 'Wh2 24 g3 .1Lg31 25 41 'it'e8+ �h7 42 'it'xc6 'ti'd2+ 43 'iVg2 e2 44
.lxe4 'ti'xh3+ 26 We2 .J:g6 and Black's attack liJgS+ 'it'xg5 45 'it'xg5 e11i' 46 'it'hS+, with
i� venomous. White having reasonable winning chances.
33 hxg4
After 33 fxg4? Black can draw with
33...Wfe4+ 34 tll f3 l:txg4+! 35 hxg4 'ifxg4+ 36
'ith l 1i'xf3+ 37 .llg2 c2 38 �gl .e3+ 39
'itht •n.
33 ...llh6 34 1txd5 /:i)e7 35 •e5
White had a faster win with 35 'ild7! l:tb8
36 'ifxe7 llxb2 37 l:lxb2 llh2+ 38 Wfl l:txb2
39 l:lxe3.
35 ....-xe5 36 dxe5 /:i)d5 37 �g3 lie&
38 /:i)h3 llxe5 39 /:i)f4 /:i)xf4 40 �xf4 lla5
41 a3 llb5 42 �xe3 llcb6 43 b4 liaS 44
Jla1 c5 45 llb2 cxb4 46 llxb4 llba5 47
22 1lge6 23 liJxe4 fxe4 24 g3! 1tf5 25
•.. a4 llc5 48 Jla3 1-0
t;)g1 .i.xd1 26 llxd1 e3?1
This is the wrong approach. 26...h5 leads Game 18
In an unclear position after 27 c4 e3 28 cxd5 Anand-Gelfand
1i'xd5 29 'it'c4 liJe7. However, the best line is Moscmv 2004
1hc simple 26...l:tf6 27 l:te2 l:tcf8 when Black
has a definite edge. 1 e4 e5 2 .!Llf3 .!tlf& 3 c!Dxe5 d6 4 .!tlf3
27 �2 1We4+?! o!Llxe4 5 d4 d5 6 .i.d3 .i.e7 7 0-0 lbc& 8
One slip follows another... After 27 ...l:tffi l:te 1 .i.g4 9 c3 f5 1 0 'lrb3 0-0
2flf3 fi:k7 29 l:te2 White would have b�o.-cn A major alternative here is lO...'ifd6 with
only slightly better. the idea of ca!ltling lJUeenside. White has
28 f3 1Wf5 29 h2 1lg6 30 llde1 1Wf4 31 three main answers:
g4 h5 32 'lrb51 hxg4
43
The Petroff Defence
allows a winning attack after 1 6....txf2+! 17 .!DeS i.xf2+ (White is also clearly better after
�£2 9xh2+) t 4...fxe4 1 s .tn .th4. the alternatives 17...f4 1 8 :e6 9d8 19 l0d3
b) 1 1 .!Dfd2!? is interesting but also poten f3 20 g3 and 17 ...lbd8 18 .tn b6 19 .!Ddl
tiaUy risky: 1 1 ...0-0-0 1 2 f3 i.h4 (the tricky .!De6 20 1tb4! - Anand) 1 8 Wxf2 ....2 19
12 ....!De5?! fails to 13 i.xe4 dxe4 1 4 fxg4 i.xc6 bxc6 20 11'xc6 f4 21 ..xdS+ Wh8 22
.!Dxg4 1 5 .!Df1 and White is clearly better - 1WxhS f3 23 11'xf3! :Xf3+ 24 Wxf3 and White.
Timman) I 3 :n and now: has cleverly consolidated his advantage, An·
bl) The creative 1 3...i.h3?! is refuted by and-Kramnik, Tilburg 1998.
accurate defence: 1 4 ..c2 11'g6 1 5 .!Db3 :h£8 b) 1 3 ...i.h4! with a further branch:
16 .!Da3 llde8 17 i.f4! {but not 17 Wht??
.!D£2+! t 8 llxf2 i.xg2+! 0-1 Ivanchuk-Anand,
Reggio Emilia 1 989) 17 . .:e6 18 .!DbS and
.
j
14 .!Dxd2 '1Fd6 I S h3 i.hS 16 .!Db3 i.h4 17 Koziak-Motylcv, Lvov 1999.
44
3 lbxe5: The Main Line with 8 lle 1
45
Th e Petroff Defence
Summary
Nowadays the active continuation 9 c4 docs not seem to offer White an advantage in the
opening. It can be recommended to fans of the isolated quecn's pawn, which arises after
9 ... l0f6 10 cxd5 l0xd5. However, we should not forget that 1 0... .A.xf3 1 1 1i'xf3 l0xd4 (Game
1 4) has proven to be a reliable way of equalising.
Despite initial appearances, 9 c3 is one of the sharpest system� in the Petroff Defence. Of
course, opposite side castling after 9. . . f5 1 0 'irb3 1Vd6 and ...0-0-0 leads to many complica
tions. After 1 0 ...0-0 11 l0bd2 l0a5 Black maintains material equality but loses control over the
important eS-square, and following 1 2 "it'c2 .i.d6 (or 1 2. ..c5) 1 3 lOeS White's chances are pref
erable. The line with 1 1 ...'�h8 (Game 17) is extremely intricate and every move must be care
fully analysed. After 1 2 1Wxb7 Black's queenside becomes (.."VCO more vulnerable but the offside
position of White's queen exposes his king to danger. Black has .. .llf8-f6-g6 at his disposal
increasing his initiative on the kingside, and despite many games it is still impossible to give an
exact assessment of this position.
1 e4 e5 2 lDf3 lDf6 3 lDxe5 d6 4 .IL!f3 lllxe4 5 d4 d6 6 i.d3 i.e7 7 0-0 .IL!c6 8 .lle 1
i.g4 (D) 9 c4
9 c3 f5
10 l0bd2 - Game 16
10 1Wb3 0-0 1 1 li.:lbd2 (D)
1 t ...lOas - G0111e 18
1 1...Wh8 - Game 1 7
9••. lllf6 (D) 10 cxd5
10 l0c3 - Go1ne 1 J
1 0...i.xf3 - G0111e 14
10...l0xd5 - Come 15
46
CHAPTER THREE I
3ltJxe5: Black Plays 6 . . . J..d 6
call it the Marshall Variation. At the lx.-ginnin� Despite the obvious transposition after 9
of d1c 20th century Marshall often playt:d the lDc3 lDxc3 10 bxc3 .i.g4 1 1 cxdS cxdS and 9
J..,'lUllbit 7 0-0 .i.g4?! 8 c4 0-0 9 cxdS f5, a line cxdS cxdS I0 lDc3 lDxc3 11 bxc3 .i.g4, this
that is nut quite correct, and eventually White docs not mean that there arc no signitlcant
players learned how to dampen Black's dan differences berween 9 lDc3 and 9 cxdS. After
�-,rcrous initiative (sec Game 1 9). 9 lDcJ (Games 24-25) Black can proceed
·n,e tabiya of the system arises after the with 9...lDxc3 10 bxc3 dxc4 1 1 .i.xc4 .i.J..,r4,
moves 7 0-0 0-0 8 c4 c6. Black has strcllbrth when the sharpest line is 12 'ii'd3 lDd7 13
ened the <.IS-pawn. which lx.>Came vulnerable lbgs lDf6 1 4 h3 i.hS 1 5 f4 h6 1 6 g4 as
;�ftcr ....i.d6. 9 'iVc2, attacking the kni1,rl1t on played by Capablanca (Game 25).
e4, is covered in Games 1 9-2 1 . Defending the The theory is very far advanced in the line
knight with 9...f5, as played in the historic of 9 cxdS cxdS 10 lDc3 lDxc3 11 bxc3 .i.g4,
J..,>amc Williams-Staunton, London 1851 is and here White manages to develop a strong
certainly possible (see the notes to Grune 19), initiative on the kingsidc. After 12 llb1 the
but these days most Black players prcfer game branches: 12 ...b6 is studied in Game
Krau7.e's su&�,rcstcd pawn sacrifice 9 ...lDa6!?. 26, while 12 ...c!'Dd7 is the subject of Games
'I hen to .i.xe4 dxe4 11 'iVxe4 (Game 19) gives 27-28.
Black active play as compcm;ation for the
pawn after either 1 1.. .lle8 or 11...lDb4!? (Raet Gamc 19
sky's b'llmes are important here). Nowadays, Burkov-Raetsky
j.,>randmasters, including the elite, prefer to Correspondence 1985
prcvent the possibility of ...lDa6-b4 with 10 a3,
;�ftcr which Black continues with 1O...lle8, 1 e4 eS 2 c!Llf3 I.Llf& 3 c!Llxe5 d6 4 c!Llf3
47
The Petroff Defence
l'Dxe4 5 d4 d5 6 .i.d3 .*.d6 7 0-0 li)xd5 17 'tth5 li)f6 18 1Wh4 White is much
White has a couple of minor alternatives: better - Cordel) 1 6 .tgS .te7 17 .i.xe7 li)xe7
a) 7 lbbd2 f5 8 �5 0-0 9 0-0 c5 to c4!? 1 8 .i.e4 .i.xf3 19 .1xf3 li)f5 20 11fxb7 l:lb8
(the passive 1 0 c3 allows Black to create 21 ..xa7 lbxd4 22 .lg4 l:lxb2 23 l:lad1 l:lb6
strong counterplay after 1 o...cxd4 1 1 cxd4 24 1Va4 and White has small advant!lh>c,
..i.xeS!? 1 2 dxeS llk6 1 3 lbf3 f4, for example Bernstein-Marshall, San Sebastian 1 91 1 .
14 .txe4?! dxe4 1 5 ..xd8 l:lxd8 16 tbgs ..i.f5 b) 1 5 g41 (this is even stronger) 1 5...lbxd5
17 ..i.xf4 h6 18 lbh3 - Shafranska-Raetsky, 1 6 .1e61 .tfl 1 7 lbg5 .txL-6 18 lbxe6 ..f6 1 9
Budapest 1 991 - and here 1 8...g5 gives Black li)xfB ltxf8 20 11fd3 and White is much bet
an edge) to...cxd4 1 1 tben (not 1 1 tbdf3? ter, Gipslis-Christiansen, Gausdal 1 992
1fc7 1 2 .if4 g5!, when Black is a1ready al 8 c4
most winning) 1 1...dxc4 1 2 lbxe4 fxe4 1 3 This is the critical try, but 8 l:lc1 is also
.1xe4 d 3 and the position i s complex. reasonable. Now 8... .tf5 9 c4 c6 transposes
b) Playing 7 c4 before castling allows to 8 c4 c6 9 l:le 1 .if5, while 8...l:lc8 9 c4 c6
Black an additional option: 7 ....tb4+ 8 lbbd2 transposes to 8 c4 c6 9 :C l l:le8. This leaves
0-0 9 0-0 .txd2 1 0 tbxd2 (also possible is 1 0 8....tg41? as the independent try: 9 .i.xc4
.i.xd2 .1g4 1 1 .te3 li)c6 1 2 h3 .thS t 3 llc1 dxe4 10 l:lxc4 f5 1 1 l:le1 llk6 1 2 c.1 'iii'hH 1 3
l:le8 14 a3 dxc4 15 .txc4 'irf6 16 .1e2 - lbbd2 11'£6 14 h3 .ths 1 5 ltlc4 l:lae8 1 6 :n
Short-Adam.-;, Wijk aan Zee 2000 - and here (Black also has compensation after 16 ltxe8
Adams suggests 1 6 ...lbg3! 17 fxg3 l:lxc3 1 8 l:lxe8 1 7 ltlxd6 ..xd6 18 .td2 f4) 1 6... £4 1 7
g4 .tg6 1 9 ..d2 lte4! leading to an unclear lbxd6 cxd6 (White was a bit better after
position) to._li)xd2 1 1 .i.xd2 dxc4 12 .txc4 17 .....xd6 1 8 b3 b5 1 9 .tb2 •ds 20 a4 a6 21
.i.e6 (1 2...Wxd4 t 3 .tb4 ..xdt 14 ltaxd 1 axb5 axb5 22 .ta3 l:l£6 23 .1c5 in Kulaots
l:le8 15 l:lft:1 l:lxet+ 16 l:lxel .te6 17 .i.xe6 Rozentalis, Cappelle Ia Grande 2004) 18 b3
fxe6 18 l:lxe6 li)c6 is also equal) t3 .1xe6 g5! 1 9 .1b2 l:lg8 20 g4 (not 20 d5?! .ixf3 21
fxe6 1 4 .g4 1Wd7 1 5 l:lfe 1 l:lf6 1 6 .tg5 l:lf5 'lrxO g4! 22 hxg4 ltle5 23 "irh3 ltxg4, when
17 d5! l:lxd5 1 8 llad1 tbc6 is level. Tiulin White is in serious trouble) 20 ... fxg3 21 fxg3
Raetsky, correspondence 1985. l:le3 22 g4 ..£41 23 lbe 1 l:lxh31? 24 l:lx£4 gxf4
7 0-0
••• 25 Wg2 l:lh4 26 �f2 l:lgxg4 27 ltlo l:lh3
with an unclear position (Kulaots).
8 c6 9 1tc2 tl)a&
..•
48
3 ltl xe5: Black Pla ys 6 . . . .i.d6
1 0 J.xe4
1 0 a3 is studied in Game 20.
10 dxe4 1 1 •xe4 lilli4
..•
1 7 a3
Black is also much better after other tries:
a) 17 lbxd6 1fxd6 18 �hl 1Wxd4.
This is strong but Black has a good alter b) 17 ltlg3 llae8 1 8 .ie3 (Black has as
native in 1 t ...l:le8 12 1Wd3 (or 1 2 1Wc2 lbb4 strong attack after 1 8 1Wxe8 llxe8 19 llxe8+
1 3 -.o3 .i.f5 14 lba3 aS and Black has excel ..tf8 20 lba3 1Wf7 21 lle3 hS) 1 8...1Wt71 19
lent compensation) 12....i.g4 1 3 .i.gS 1Wd7 14 l0a3 lbd31 20 lleb1 lbf4 21 ...<12 �! and
ltlbd2 h6 1 5 i.e3 (White should probably White is in terrible trouble.
prefer 15 .i.h4, but Black stiU has good play 1 7 ... ..bh2+ 1 8 �1 lOa& 1 9 ll\g5 .if5
after 1 5...lbc5 1 6 1Wc2 ..tf5 1 7 'l'c3 ltld3) 20 ll\c3
49
The Petroff Defenc e
20 tik6? leads t o a quick loss foUowing 31 ..g2 wouW have aUowed a neat finish:
20...W'f7 21 cJilxh2 D.aeS 22 d5 'WbS+ 23 cJilgt 3t...llh3+ 32 Wgt .i.f3!.
11fg6+ 24 cJilh2 .i.xe6 25 dxe6 llf5. 31 ...•f5 32 1t'e5 llh3+ 33 �g2 1Vf3+
20. . h6. 34 Wf1 llh1 35 .g3 .th3+ 36 �.1
Black can also tty 20... .i.f4!? 21 tlk6 .i.xe6 1t'xg3 0-1
22 ..xe6+ Wh8, when his attack continues.
2 1 lilge4 Game 20
Kasparov-Shirov
Wijk aan Zee 2001
denly winning for White in Khramov moves but 10..lle8 is also reasonable. Play
Raetsky, correspondence 1985) 25 l:lht .i.f5 continues 11 o!L!c3 and now:
and White has no hope of defending. a) 1t....i.f5 is a solid line: 12l:le1 h6 13 cS
22 .f1 i.e? 14 .i.d2 .i.a5?1 Onstead Anand assesses
White could try 22 .i.xh6!? but Black still 14...o!L!xd2 15 l:lxc8+ Wxe8 16 'it'xd2 .i.xd3
has compensation after 22..lle6 23 .i.e3 17 11'xd3 as cquaQ 15 .i.f4 .i.xc3 16 bxc3
.i.f4. �c7 17 h3 l£le6 18 .i.h2 (Anand-Kasimdz
22 ...1t'f7 23 1t'g2? hanov, Hyderabad 2002) and here White is
Now White is completely losing. After 23 slightly better after 18....i.h71 19 llab1 b6 20
Wxh2 W'hS+ 24 Wg1 .t.h3 25 1l'e2 Wg6+ 26 lDes.
.i.g5 o!L!c71 27 ..e3 lbe6 at least White has b) tt....i.g4 121£lxc4 dxc4 13 .i.xe4 .i.xf3
chances to fend off the attack. 14 .i.xf3 'iVh4 15 g3 11'xd4 16 .i.g5 1i'e5 17
23...i.b8 24 i.d2 ..d2 11'f5 18 �2 o!L!cs 19 l:ad1 (According
Or 24 f4 :C6 25 dS cxd5 26 cxd5 llg6 to Nataf, 19 'irxd6 is also equal after
and the attack is too strong. 19...1i'xg5 20 £4 ..f5 21 b4 l£lc6 22 1i'd7
24...•5+ 25 �g1 i.h3 26 lilg3 i.xg3 l:ab8 23 l:ladt lled8 24 11'e7 :C8 25 1Vh4
27 •xg3 :Xe1 + 28 :Xe1 :Xf3 29 •2 lLif8) 19 . 1£le6 20 .i.e3 l:ad8 21 ..c3 .i.cS 22
. .
50
3 f0xe5: Bisek Plays 6 ... �d6
Black had equalised in Nataf-Vo17.hin, Stock b2) 13 f3 �xd3 14 1i'xc.I3 �cS ! 15 1i'd4
holm 2001. �b3 16 Wxg4 �xal 17 .i.h6 g6 (if Black
1 1 �5 inserts the moves 17 ...'1Fb6+ 18 Whl before
11 cS .i.b8 12 l£les gives Black more playing 18...g6 then 19 Wf4f6 20 .i.xf8 llxf8
chances: 12...Le5 13 dxeS �exeS (but not 21 cxdS cxdS 22 Wd2 promises White a
l3...lt)axc5?! 14f3 �xd3 15•xd3, winning a strong initiative) 18 llk3 (an important
piece) 14 .i.xh7+ 'iPh8 15 b4 1Wh4 16 .i.d3 choice) 18-.'1Fb6+ t91lf21lfe8 20 Wf41i'c7
�xd3 17 •xd3 Whs 18 lle1 .i.fS 19 Wg3 (in Shirov-Leko, Linares 2000 White was a
ftk7 20 �2lbe6 21 �b2 b6 and Black has bit better after 20...f5 21 cxdS lDb3 22 e6
sttung counterplay, Jobava-Mamedyarov, cxdS 23 �xdS 1i'xe6 24 �c7 1i'c6 25 lDxe8
Plovdiv 2003. llxe8 26 g4Q 21 lle2 lle6, which Shirov as
1 1 .....th& sesses as unclear. Instead of 18 ru, 18 cxdS
This is a major decision for Black as the is very messy after 18...:C8 19 �2 Wb6+ 20
alternatives are very interesting: Wh1 llxeSI 21 llxa1 1i'xb2 22 lld1 .ll xdS.
a) 11...�f5 12 �e3 �acS 13 c.lxcS �xeS Perhaps the best move is 18 �xf8 as White
14 cxdS 'ii'xdS 15 llk3 �xc3 16 bxc31lad8 has an edge after 18...9xf8 19 cxdS 'ii' cS+ 20
17 .i.d4 �c3! lH �xfS (after 18 .xc3? Wh l cxdS 21 e6 f5 22 1i'a4.
Motylev assesses 18...�xd3 19 �xg7 �xft 1 2 cxd& cxd& 13 �c3
20 llxft llfe8 21 .i.h8 f6 22 'ti'xf6 lld7 as A good, simple developing move. Taking
clearly advantageous for Black) 18. ..Wxd4 19 the pawn with 13.i.xc4dxc414Wxe4aUows
lLct gp 20 1i'xc31i'xc321 llxc3gxfS is very Black sufficient countcrplay: 14..lte8 15.i.e3
llrawish, Smimov-Motylev, Russia 2004. (or 15 �f4�xeS 16 dxeS llks 171We3l0d3
b) 1t....i.xe5 leads to heavily analysed and Black has excellent compensation)
complications after 12 dxeSlbacS 15...lDcs 16 1i'c2 t:&6 17 'ii'rs Wh4!? 18
�c.l2 �g6 19 1i'g4 1i'xg4 20 �xg4f5 21 �c4
�f8 22 dS fxbr4 23 dxe6 .llxe6 with an I.:<Jual
position, Potkin-Rozentalis, Bad Wiessee
2003.
and now:
b1) Black has an easy position after 13
cxdS ?! 'ii'xdS Qess exact is 13... cxd5 14 f3
llc8 15�e3- not 15 fxe4? dxe4!with a clear
advantage - 15... d4 16 �xe4 .i.e6!? 17 1 3 ...�xc3
.i.xh7+ ¢'h8 with an unclear position) 14 This may seem natural, but Black should
.i.xe4 �xe4 15 llk3 (if 15 f3 �cS 16 lldl seriously consider 13...Le5. For example,
1i'b3! Black has tremendous play) 15...lthc3 14dxcS lDacS 15�e4lDxd3 16 'iixd3�gp
16 Wxc3llfdK. 17 'ilb3.i.xe4 18 Wxb71Wh4 19 1ib3 d4 20
51
Th e Petroff Defence
52
3 liJ.xe5: Blac k Plays 6 . .. �d6
53
Th e Petroff Defence
�xeS 23 dxe5 Wg4+ 24 �f1 11t'h3+ when 1i'xh3+ 26 �gl e3, while 24 ltaz?! e3! 25
White should accept a draw (Gelfand). :xe3 1fc4+ is not much better.
21 dxe5 .*.c6 24...Wxh3+
Now the game must end in a draw. Black
could play on with 24...:0!? but after 25 b51
�xbS + 26 �gl .llxh3 27 �f4 :d3 28 �g3
Wd8 29 1fg4 l:txd5 30 1fxe4 the position is '
simply unclear.
25 �1 hd5 26 'irxd5 'irg4+ 27 Wf1
1ih3+ 28 Wg1 1tg4+ 29 wn % - %
Game 22
Shirov-Piket
Wijk aan Zee 2001
9 -*.fS
••.
54
3 li)xe5: Bisek Plsys 6... i.d6
.!l:Xt7 13 �xe4 fxe4 14 .i.xe4 J.c7 15 .i.g5 b1) 12...J.c6 l3 a3 lilc7 14 ...c2 f5 15
lilf6 16 .tc2 11'd7 17 1l'd3 �4 and Black liJc5 ltc8 16 .if4 g51 17 .td2 (also worth
was obviously better in Yemelin-Ractsky, considering is 17 J.e3 lila6 18 f.3 lL!xc3 19
Rostov on Don 1993. bxc3 .txe5 20 dxe5 �5 21 .td4 with coun
b2) The simple 10 .i.xe4 dxt.-4 11 Le4 is terplay - Gelfand) 17...�6 18 1i'd1 .i.xeS 19
best. Following 11... f5 White has two op dxeS �acS 20 .ic2 �xd2 21 9xd2 d4 22
tions: �4lilb3 23 .txb3 .txb3 24 c6! was unclear
b21) 12 l:e6 gives Black chances after in Frcssinet-Gclfand, Cannes 2002
12...11'd7 13 :Ct lL!a6 14 11'b3 (or 14 lilc3 b2) Also interesting is 12...�b4 13 .i.b11?
l:ae8 15 l:xe8 l:xe8 16 .i.d2 J.b8 with b>t>Od (Black has enough compensation after 13
attacking possibilities) 14...J.xf.3 15 11fxf.3 .i.xe4 dxe4 14 .ig5 1ih6 15 �xe4 .i.c6 16
ltae8 16 J.e3? {White can hang on to equal 'ild l �LacS) 13....ie6 14 a3 �6 15 1Wxb7
ity after 16 J.d2 J.b4) 16...�b4 17 c5? (after �5 16 ...a6lL!xc31 (not 16...lilb3?117lilxe4
17 lLla.3 lDd3 18 :n J.xa3 19 bxa3 f4 20 dxe4 18 .ixe4 �xa1 19 .i.xa8 �c2 20 l:te2
'ii'd1 •f5 21 J.d211'e4 Black is much better, �xd4 21 �d4 J.cS 22 .ie3 .txd4 23 lld2
but there is no forced win) 17.. .ru 18 cxd6 and White was comfortably better in Palac
f4 19 J.d2lilxe1 20 11'b3+ *h8 21 � f.3 M.Ivanov, Cannes 2004) 17 bxc3 �b3 18
0-1 Kosteniuk-Raetsky, Bicl 2004. .ig5! J.xh2+ 19 '1Ph1 ...c7 20ltxe6!? fxd) 21
b22) 12 l:e 1 (again the simple move is the 11fxe6+11'f71 22 9xf7+ :Xf7 23 .i.a2!?lilxa1
best) 12...J.xf.3 13 11'xf.3 11'h4 14 h3 ..xcl4 24 .ixdS and after all the complications the
1511'b31ib6 16 11'xb6 axb6 17lL!c3lbd7 18 position remains unclear.
ltdt l:f6 and White had a tiny edge in c) 10 �3 �c3 11 bxc3 .txd3 12 1l'xd3
Ehlvest-Mamedyarov, Moscow 2002 dxc4 13 'lfxc4lLld7 with another branch:
1 0 1rc2 ct) 14 .igSI? 11'c7 15 .te7!? (15 11'b3 is
This is probably the best of a wide choice: less enterprising: 15.. 1lfe8 16 .lab1 h6 17
a) 10 cS .tc7 11 � l:eR 12 1l'c2lbd7 .i.e3 b6 with a dull equality) 15...�b6 16
13lL!xe4 dxe4 14 .i.xe4 .e7 15 �5lL!f6 16 .i.xd6 9xd6 17 'iib3 l:tacS 18 c4 'flc7 19 a4
f3 (Socko-Skato;hkov, Cappelle Ia Grande lLld7 20 l:tab1 b6 21 g3 1l'd8 22 aS 11'f6 23
2004) and hert: Skatshkuv suggests that axb6 axb6 24 *g2 and White had a marginal
16 ...lilxe4 17 fxe4 h6 18 ...f2 .tg6 is slightly edge in lvanchuk-Gelfand, Lvov 2000.
bcrtcr for Black. c2) 14 1tb3 has also bet."' tried: 14...1i'c7
b) t O eb3 �6 t1 cxd5 cxd5 12 �3 15 c4 :rea 16 .ib2 h6 17 g3l:tad8 18 �4
.i£8 19 11'c2 (perhaps a bertcr try is t9l:tad1
1Wb6 20 .ic3 with unclear play) 19...1fa5! 20
lL!g2ltxe1+ 21 l:txe1 bS 22 .ic3 b4 23 .i.a1
�b6 24 91>3 cS 25 dS Wa6 26 �e3 lte8 27
l:tct 11fc8 and Black was a bit better in Ad
ams-Morozevich, Dortmund 2002.
1 0 ...i.g6
Solidly defending the bishop. 1o.. .lLxl7 is
also playable: 11 �3 lLldf6 12 cS (12 cxdS
cxdS 13 �3 :Cs 14 "xb7 ltb8 15 'W'xa7
l:e7 16 1i'a6 lte6 17 1l'a7 l:e7 is simply a
draw- Piket) 12....ic7 13 b4 J.g6 14 .i.b2
�c3 15 11fxc3 .ixd3 16 1i'xd3 �e4 17 l:te2
and now: l:e8 18 l:taet l:e6 19 �cS 11'h4 20 J.c1 (20
55
The Petroff D e fence
f3? loses after 20..llh6 21 fxe4 'lfxh2+ 22 lies 22 �d6 tnxc3 23 llc3 lDe4 24 :th3 h6
wn Wht+ 23 Wf2 llf6+ 24 tnn �g3+! 25 25 :tft with reasonable play for the pawn.
Wxg3 �>6+ - 1-Jaba) 20...f6 21 ti)f3 'iVh5 22 20. . .�xg5 21 •xg5 h6
g3 llaeS 23 Wg2 g5 was unclear in Lanka Trying to win a piece with 2t...f6? loses to
Haba, Hamburg 2003. 22lDg6+! hxg6 23 1Wxg6lDxc5 24 :te7.
1 1 c5 J..c7 1 2 l0c3 22 -.,5 �f6 23 �xf7+ �h7 24 �g5+
Wg8 25 •s6 hxg5 26 l:le7!
1 2 l0f6
•••
Alternatives are not so good: Black defends easily after 26 fxg5? lbe4 27
a) 12...tnxc3 13 bxc3 tnd7 14 �xg6 hxhr6 :th3.f5.
15 �g5 lDf6 16 &5 and White has a pleas 26 ...�8 27 l:lae1 1tg4
ant initiative - Pikct. Black would be mated after 27.-gxf4? 28
b) 12...f5 13 'lfb3!? and now 13...:tt7 al llxg7+! tnxg7 29 l:te7.
lows a trick: 14 ti)g5! b6 (the point is 28 fxg51th4
14...tnxg5? 15 �xg5 Wxg5 16 :teS+ J:tf8 17 The safest line is 28. .1Wf4! 29 h3 'ifg3 30
.
l:lxfB+ �xf8 18 'ifxb7 and White wins) 15 �h1 'ifxc3 31 :XeS Lc8 321lxc8 'ifct+
�xe4 fxe4 16 lDxt7 �xt7 17 cxb6 axb6 18 33 Wh2 1Wf4+ with a draw (Piket).
lDe2 .f6 19 �e3 tnd7 20 llac I with a clear 29 g3 .3 30 llxb7 .,5 31 ••6+ 1
advanmge, Anand-Morozevich, Monte Carlo 11rxe6 32 l:lxe6 l:ln?!
(blindfold) 2003. Instead, 13... b6 14 cxb6 32...J:tf31? 33 llxc6 J:txc3 is still unclear.
axb6 15 tnxe4 fxe4 16 �xe4 �xc4 17 :Xe4 Now Black is in trouble.
gives White a small advantage. 33 llxn �xn 34 llxc6 l:lb8 35 l:la6 l:lb7
1 3 J..g5 lbbd7 1 4 r!Lle5 .bd3 15 1txd3 36 h4 fi:jc7 37 l:ld6 �5 38 g6+ h7 39
1tc8 1 6 f4 llxd5 �c3 40l:le5+ � 41 c61 l:lb1 +
If 16 b4 lDxe5 17 dxe5 tnd7 18 f4 f6 19 There is no way to save the game: 41.. llc7
exf6 tnxf6 20 g3 �dB Black's stands well. 42 h5 l:lxc6 43 g4 llc6 44 gS+! �c7 45 h6
1 6 .. .i.a5! 1 7 1tg3 �8 18 e.4 .i.xc3 gxh6 (45...lbe2+ 46 :txe2 :txe2 47 hxg7 is a
Preparing ...tne4
. The immediate beautiful finish) 46 g7 l:lxc5 47 dxe5 �t7 48
18...lDe4!? is also possible: 19 lle3 tnxe5 20 e6+ Wxg7 49 !,rxh6+ and a pawn will queen
fxe5 ..i.xc3 21 bxc3 f6 22 exf6 gxf6 23 �h6 (Piket).
llt7 24 c4 1Ve6 25 cxd5 cxd5 26 :lfl l:g8 27 42 �g2 l:lb2+
h3 with an unbalanced position - Piker. 42...tnb5 doesn't help: 43 J:ld5 :tb4 44 a3
1 9 bxc3 �e4 20 l:le3 :tb2+ 45 �h3 <iii'e6 46 lld7 and White wins.
White couJd try 20 �c7 tnxc5 21 fxc5 43 Wf3 1-0
56
3 l'Dxe5: Bla ck Plays 6 ... �d6
57
The Petroff D efence
desperate trouble; instead t6...liJa6 17 'lfbt This is too weakening. Black should play
Zle8 18 g3 Wc8 19 f3 .i.h3 20 IllES is only 17...f6, for example 18 fS fxgS 19 fxc6 �xeS
slightly better for White) 17 f4l .i.c8 18 Wg3 20 dxcS .i.xc5+ 21 Whl gxh4 22 1i'xh4 g6 23
(surprisingly White is already winning by %let and White obviously has compensation
force) 18...b6 (after 18...1i'f8 Anand analyses for the pawn but Black can hope to defend.
a beautiful winning line: 19 %let .i.d8 20 1 8 f5lt:lxc5
lt\g6! fxg6 21 1i'xg6 .i.e? 22 �7+ Wf7 23 Black prefers to sacrifice a piece rather
.i.g6+ Wf6 24 .i.hs .i.fS 25 g4l .i.xh7 26 g5+ than take one: 18...hxg5 19 fxe6 Wxc6 (the
hxg5 27 fxg5+ WfS 28 Zlc:S mate) 19 Zlc:t problem is that 19...gxh4? loses to 20 cxf7+
bxc5 20 dxc5 'iff8 21 .i.c:31lla6 22 .i.d4 g5 'ii?xf7 21 Wxh4, when the king is doomed) 20
23 ...f2 and Black is defenceless and resigned 'lfxe6 fxe6 21 �f3 .i.e7 22 Zlb1 and White
in Anand-Piket, Wijk aan Zce 2001. has a clear advantage - Rozentalis.
1 4 lDh4 1 9 dxc5 hxg5 20 fxe6 .i.xc5+ 21 �1
This prepares a kingside advance. Whi.te 1be6
has tried two other possibilities: If Black tries to recapture the piece Whi.te
a) 14 .i.e7 .i.f4! 15 .i.h4 �7 16 .i.g3 gains a d<.-adly attack: 21...gxh4 22 Wxh4 g6
.A.xg3 17 hxg3 .i.xf3 18 gxf31£lf6 (Black has 23 exf7+ Wg7 24 %let 'ifd7 25 J:.xg6! Wxg6
equalised) 19 WeS dxc4 20 .i.xc4 Wd7 21 26lle5.
Wf4 ZleB 22 11fd2 hS 23 %let :Xet+ 24 'lfxel
bS Vz-112 Movsesian-Rozentalis, Neum 2000.
b) Perhaps White's strongest line is 14 cS,
for example 1 4....i.f8 15 'ltb 1! h6 (after
15 ..g6 16 �eS .i.e6 17 f4 �7 18 �f3
.
58
3 fi)xe5: Black Plays 6 ... .i.d6
:c17 gives good chances. .i.xc6 llad8 21 .i.xd7 l:lxd7 22 .i.f4 l:lfd8
40...J:lg8 41 J.f5 d2 42 J.g4 e37! with a level position (Khalifman).
Now it is only a draw. 42...1lxg7? 43 :Xg7 1 1 J.xc4 �g4 1 2 h3
.i.xg7 44li:)gS, leaving White clearly better, is 12 Wd3 is the subject of the next game.
even worse, but Black could have kept de 1 2....i..h5 1 3 J:le1 l0d7
cent winning chances after 42....i.f6! 43 ¢'g2
cS.
43 lt)h& lbg7 44 :Xg7+ J.xg7 45 lOts
.i.e5 46 �2 c5 47 lrule3 c4 48 J.e2
'>Pd& 49 g4 .i.f4 50 Wt3 WaS
Of course not 50....i.xc3? 51 Wxe3 c3 52
'1Pd3 b4 53 .i.dt when White wins as all
Black's pawns arc worthless.
51 �c2 J.g5 52 J.d1 % - %
Game 24
Firman-Bick
Stratton Mountain 2003
1 4 g4
1 e4 e5 2 l0f3 �f& 3 lL!xe5 d& 4 liJf3 White can also play Jess aggressive])' with
lllxe4 5 d4 d5 6 J.d3 J.d& 7 0-0 0-0 8 14 .i.ft. Now 1 4...l:lc8 1 5 llxe8+ Wxe8 1 6
c4 c& 9 liJc3l0xc3 10 bxc3 dxc4 i.e3 b5 1 7 a4 a6 (17 ...b4 1 8 cxb4 .*.xb4 1 9
Black can also try the immediate 10....i.g4 J:lct li:)b6 is also equal) 18 axb5 axb5 1 9
59
Th e Petroff Defence
60
3 li:Jxe5: Black Plays 6 ... .i.d6
17...:e8 18 .i.e3 .i.c7 19 llafl 1Wd6 20 .i.f2 tnkes over the attack after 21 gxf7+ .i.xf7 22
:C4 21 .i.h4 ltae8 22 �xf6 11'xf6 23 �ht :Xf7 ltxf7 23 1Wg6 .f6 24 .i.xf7+ 1Wxf7 25
Act 24 a4 hS 25 g3 a6, Kamsky-Bareev, 'iPxd6 ltf8) 2t....i.xb3 (21....e8? allows a
Linares 1993. neat forced win: 22 'trh3! :Xf7- the trick is
22_..i.xh3? 23 ltfS+I- 23 Wh7+ Wf8 24 .i.g5
l1f6 25 ..ixf6 gxf6 26 g7+ and Black wiU be
mated -Nunn) 22 'iPh3 l:lxf7 231Wh7+ �£8
24 WhB+ We7 25 .i.g5+ J:l£6 26 .xg7+ We6
27 .i.xf6 'iPg8 28 axb3 .xg7 29 .i.xg7 WfS.
After all the tactics White has a slight edge in
the ending.
1 6...b5
It is also posl>ible to take the piece imme
diately: 16...hxg5 17 fxgS lLlxg4 18 hxg4 1Wd7
(18...�xg4? is a mistake: 19 1i'e4 1i'd7 20 g6
..ie6 21 .i.xc6 fxe6- Capablanca-Northrop,
New York 1909 - and here the simple 22
Wh4 is winning) 19 �3!? (the independent
try; 19 gxhS 9g4+ 20 �f2 ltaeB 21 ltg1 20 ..1671
'1Vh4+ 22 �g2 b5 23 .i.b3 trans�cs to Now Black has an edge. The correct way
16...b5) and now: to hold equality is 20 gxh5 .g4+ 21 Wf2
a) 19 ...'.xg4+ is safest: 20 1i'xg4 �xg4 21 l:aes 221tg1 'trh4+ 23 wg2 "W'h2+1 24 wn
llxf7 b5 22 .i.b3 �h7 with equality (Byrne, .i.f4 25 1i'f3 (not 25 .i.x£4? .xf4+ 26 ..t.>g2
Mednis). llc3 and White must give up his queen)
b) 19 ....i.xg4 is interesting but also risky: 25..llc1+ 26 �xel 'ii'xgl+ 27 We2 .i.x.cl 28
20 Wh4 .i.fS (or 20....i.e6 21 .i.d3 g6 22 :Xcl .xcl 29 g6 lle8+ 30 Wd3 'iWb 1 +
l:t£6!? c5 23 .i.e3 and White has a dangerous (30 .lle7 31 gxf7+ wm 32 �3 'ilrbt+ 33
.
attack) 21 .i.£4 .i.xf4 22 llxf4 bS 23 .i.b3 and .i.c2 .xa2 34 1i'c8+ on - also draws) 31
-
White certainly has compensation. Now in Wd2 .el+ 32 Wd3 'ilrbt+ 1/z-'12 A.Sokolov
the game Moro7.evich-Ippolito, New York OU, Odessa 1989.
1997 Dlack erred with 23...ltae8?, after which 20 ...hg4 21 •xd7 .i.xd7 22 lbf7 lbf7
White could have won with 24 Wf2! 11fe7 25 23 g& .i.e8 24 .i.g5
l:txfS 1t'e2+ 26 Wg1 :C4 2711fh3 (Ippolito). White is also slightly worse after 24 �e3
1 7 .i.b3 hxg5 1 8 fxg5 li:Jxg4 1 9 hxg4 aS 25 .i.e6 Wf8 26 gxf7 .i.xf7 27 llft lla7,
..d7 Nunn-Salov, Brussels 1988
Another hugely complicated line is 24 85 25 �g2
•..
19....i.xg4 20 g6 (or 20 1We4 .i.h3 21 g6 Black also has an edge after 25 l:lfl llaa7
.i.h2+!- not 21....i.xft? 22 1Wht! and Black 26 llf3 a4 27 .i.e6 llac7 28 ..i.d8 l:lb7 29
l"lln resign - 22 Wxh2 .i.xfl with a very .i.gS b4 30 cxb4 �xb4.
complex position) 20....i.e6 21 llxf7!? (Black 25 84 26 .i.e& �8 27 l:lh1 l:lf& 28
...
61
Th e Petroff Defence
hf6 gxf6 29 .*.f5 Wg8?1 rer but Black still has an edge) 18 . .1lxc3 19
.
Black fails to make the most of his llxb6 Axd3 20 .i.e3 1i'g6+ 21 �h1 'WfSI (in
chances. Better is 29.. 1[b8 30 'itf3 (White Capablanca-Marshall, New York 1909 Black
achieves nothing after 30 llh8+ 'itc7 31 missed his chance with 21...'1Fe6 22 Wg2
llh7+ Wd8) 30 ..b4 31 cxb4 llxb4 with a
. 1i'g6+ 23 �h1 1i'e6 when a draw was a&rreed)
definite edge for Black. 22 Wg2 Qf 22 llxd6 Black's idea is
30 l:.h7 22.. '1Fxh3+ 21 �g1 ltxe3! 24 fxe3 'Wg3+ 25
.
Now the threat of 31 .i.e6+ forces Black Whl 1i'xd6 with a clear advantage because of
to bail out to a drawn ending. White's exposed king) 22. ..1la3! 23 1rb7
30 hg6 31 ..i.xg6 .*.f4 32 .llb7 .*.d2
•.. 1i'g6+ 24 Ciilh1 llxe3! 25 :Xd6 Wh5 26 fxe3
33 ..i.e4 l:.c8 34 llb6 ..i.xc3 35 :.Xc6 'irxh3+ 27 Wg1 1i'g3+ 28 Wh1 1i'xd6 and
:.Xc6 36 ..i.xc6 % - % Black has an obvious advantage.
1 2...b6
Game 26 12...�7 is studied in Games 27-28.
V .Gurevich-Meijers 1 3 .llb5 ..i.c7
Germaf!Y 1999 The d-pawn was genuinely threatened. For
example, 13...a6? 14 %hd5! wins a clear pawn
1 e4 e5 2 lbf3 lbf6 3 lbxe5 d6 4 l/)f3 because 14...i.xh2+? 15 �xh2 i.xdt 16
�e4 5 d4 d5 6 ..i.d3 ..i.d6 7 0-0 0-0 8 llxd8 ltxd8 17 llxdl is winning for White.
c4 c6 9 cxd5 cxd5 1 0 lt!c3 ltlxc3
This is the automatic answer to 10 lbc3,
but the rarely played 10...1le8 is also a con
sideration. Now I I lle1 transposes to note
'c' to White's 10th move in Game 23, 11
.ixc4 dxe4 12 �g5 .ifS 13 f3 c3 14 �ge4
.if4 was unclear in Morozevich-Gelfand,
Cannes 2002, while 11 h3 and 11 ..c2 are
also serious possibilities for White.
1 1 bxc3 ..i.g4 1 2 llb1
1 4 h3
14 c4 dxc4! leads to very sharp play after
15 i.e4 lbc6 16 llg5 (if 16 .i.xc6?! Black's
idea is 16...'1Fd6 17 ..LaB i.xf3 18 i.f4 1i'xf4
19 g3 i.xd1 20 gxf4 i.c2, when the queen
side pawns will be very dangerous) 16...i.xt1
17 1i'xt1 '1Fd6 18 llg3 (18 g3 eventually pe
ters out to equality after 18... �d4 19 1i'e3
llfe8 20 :d5 1i'xd5 21 .i.xd5 llxe3 22 fxe3
tbe2+ 23 �g2 :ds 24 i.xt7+ Wh8 25 i.xc4
Instead 12 h3 .i.h5 13 1rb3 is too ambi �xcl 26 llxct) 18...�xd4 19 1i'g4 g6 20
tious: 13....i.xf3 14 ..xb7 �7 15 gxt1 �b6 .i.xa8 f5 21 'Wh4 l:lxa8 (the tempting
16 ltbt 'it'f6 17 �g2 llac8 18 1i'xa7? (18 2L.l&2+? fails to 22 Wht �xg3+ 23 hxg3-
llb31lc7 19 Wa6 .i.f4 20 .i.xf4 Wxf4 is bet- White is clearly better because 23..1l . xa8?
62
3 liJxe5: Black Plays 6. .. �d6
Game 27
Pavlovic-Raetsky
Bie/ 1999
1 e4 e5 2 liJf3 liJf6 3 liJxe5 d6 4 liJf3
liJxe4 5 d4 d5 6 �d3 �d6 7 0-0 0-0 8
c4 c6 9 cxd5 cxd5 10 o!tlc3 liJxc3 1 1
bxc3 �g4 1 2 lib 1 liJd7
1 7 �h6
This seems the natural try. White has also
played 17 .i.xbS 'ird6 18 g3lLlc6 19 Wg2 f5
20 .if4 'ird7 21 gS i.xf4 22 gxf4 'irc7 23
.ixc6 'irxc6 24 �5 'ird6 (Black should
avoid chasing material with 24...'1'a4 because
;tfter 25 Wd3 Wxa2 26 l:r.ht White has a
promising attack - Nikcevic) 25 a4 %laS 26
l:tbt JibS with an unclear position,
i\.Sokolov-Nikcevic, Vmjacka Banja 1998.
17 ...Ilea 1 8 �xb5 Jle4 1 9 c4 lbg4 20
.
cxd5 liJd71?
Completing development is logical but
20....i.f4 is also worth considering: 21 lLleS 1 3 h3
lth4?! (Black has to play 2t ...l%xg2+1 22 Wxg2 Forcing the bishop to the kingside. In
..ixh6 with an unclear position) 22 .i.xf4 stead 13 l:r.b5 lLlb6 14 h3 gives Black the
llxf4 23 'ird2l:[f6 24 cli)g4l:[d6 (Kotronias option of 14.. ..i.d71?. For example, 15 l:r.bt
Roi'.entali�, Debrecen 1992) and here White i.a4 16 'ire2 lieS 17 .i.e3 %lc8 18 cli)gs (the
could have gained a dear advantag.: with 25 game is level after 18 �5 .i.xeS 19 dxe5
llc I lLld7 26 1Wf4 cli)f6 27l:r.c61 (Lepeshkin). l:.xe5 20 .i.xb6 :Xc2 21 Ld8 Zld2) 18...g6
63
Th e Petroff Defence
64
3 /i)xe5: Bla ck Plays 6 . • . �d6
Petursson, Reykjavik 1988; or 18...f6 19 'ii'hS with unclear play) 25 9f3 � 26 L3 �b4
h6 20 .i.xh6! axbS 21 Lg71 f5 22 .i.xfB 27 h5 �3 28 hxg6 hxg6 (Ponomariov-Safin,
.i.xfB 23 i.xf5 and Black's king is in trouble) Yerevan 2001) and here White should force a
19 .i.f61 g6 (19...axbS? nms into a mating draw with 29 'Wf6! .i.xa3 30 llhSI gxhS 31
attack after 20 .i.xh7+! Wxh7 21 'ifh5+ Wg8 'Wg5+.
22 9gS g6 23 1th6) 20 llb2 and White stiU 1 9 �e3
has a strong initiative. Once again there is a choice:
b) 17...1i'c7 18 a4 a6 19 llb2 llab8 20 a) 19 .i.b2 should be equal after 19..ic7
.i.h6! 9c6 (the point is if 20... gxh6 then 21 20 J:lcS l:fd8? (this is a fatal slip; 20....i.d6 2 1
9f5 Wg7 22 'lfxh7+ Wf6 23 .xh6+ Wc7 24 l:b5 .i.c7 is level) 21 llh5 g6 22 d511fd6 23
lie 1+ Wd7 2S .i.f5+ Wc6 26 .i.t->4+ Wd7 27 g3 gxh5 (If 23...f6 then 24 llxh71 and the
llcbl and the attack breaks through) 21 9f5 attack crashes through) 24 9xh5 and White
g6 22 9a5 �d7 23 9c3 1fdS 24 .LfB �xfB was winning in Zagrebelny-Chctverik, Gyula
2S llet and \Vh.ite had an edge in Yandar 1992.
biev-Skatchkov, St Petersburg 2001. b) 19 .i.g5!? is a promising alternative to
18 a4 19 i.e3. For example, 19.. .1lfc8 20 l:lfbt g6
21 .i.e3 Wc6 22 a5 'lfxf3 23 gxf3 �8 24
l:xb7 l:lxb7 25 l:xb7 lle7 26 l:b1 with an
edge for \Vh.ite, De Firmian-Kosebay, Co
penhagen 1996.
1 9 /i)cS 20 llfb1 b6
...
18 llab8
•..
65
The Petroff Defen ce
26...Wh6 27 llhS+I Wxh5 28 'iff41 and mate which gave White an edge in Svidler
is unavoidable. Akopian, Yerevan 1996) 23 axb6 .i.e7! 24
25 •xd4 bxa7 b6 25 i.e4! 11'xa7 26 i.dS i.d6 27
After 25 Jl..g7 'ifxf6 26 .i.xf6 ltd6 27 ::lh4 :Xb6 :bSI 28 :Xb8 llxb8 29 llxb8+ i.xb8
llxf6 28 ltxc4 �6 Black has won back the 30 Jl..xc4 tli'at+ 31 .tn tli'xd4 with a drawn
piece and keeps a clear extta pawn. position (Svidler).
25 ...:txd4 26 .i.c5 ltd2 27 :te5 :txc2 28 b) 19 i.d2 c3 20 i.xc3 :ac8 21 i.e4 :c4
:tea+ �7 29 .i.d4+ 22 :bb1 :xa4 (Black has a promising alter
29 .i.fB+ is le::;s clear: 29. ..'itf6 30 :dt c3 native in 22...:fc81? 2'\ .tat :xa4 24 i.xb7
3t:dd8 with an ob::;cure position. :c7 25 i.e4 :a3 with counterplay - Barlov)
29 f6 30 .i.e37
••. 23 i.xb7 :V 24 :ret (or 24 i.c6 11fc7 25
This spoils White's advantage. 30 ltdt c3 :at :b3 26 :fb1 :xbt+ 27 :xb1 rlcs 28
3t Jl..e3 Wf7 32 ::ldd8 lta2 33 ltxc8:Xc8 34 dS liJxd5 29 :b7 1Wxb7 30 i.xb7 :Xc3 31
llxcS :Xa4 35 llxc3 promises some winning 'IVdt i.f4! heading for a drawn opposite
chances. coloured bishop ending - Kaspamv)
30 . :tc3 31 ltd1 ltd3 32 :C1 ltd7 33
.. 24...tli'c7! 25 ltat %lb8 26 i.e4 ltb3 was
¢1 1 7 equal in Kasparov-Shirov, .Iinares 2000.
Instead 3 3 Jl..f4 :laS 34 llxc4 tbd6 35 1 9...:ac8
:c7! holds on to equality.
33...g5 34 �e2 ¢17 35 llh8 Wg7 36
lte8 lte7 37 :td8 :C7 38 h4 h6 39 f47
The final mistake. 39 :d4 :bb7 40 hxgS
hxgS 41 ltcxc4 liJe7 still leaves White some
hopes of saving the draw.
39 ...gxh4 40 f5 :tbb7 41 .i.f4 ltc6 42
ltcd1 a6 43 lt1d4 h5 44 .i.c1 fi::Je7 45
:t8d7 ltcc7 46 :Xc7 :Xc7 47 .i.f4il)xf5
48 .i.xc7 il)xd4+ 49 �e3 lLlf5+ 50 ¢14
il)e7 5 1 We4 b5 52 a5 b4 53 � b3
0-1
Game 28 20 ltfb1
Anand-Shirov After 20 i.h6 :res White has an unusual,
Unarr:s 2000 way to force a draw: 21 i.fS! gxfS 22 '1Vg4+11
fxg4 23 ltgS+ Wh8 Va-lfz Zaw Win Lay
'
c4 c6 9 cxd5 cxd5 1 0 il)c3 fi::Jxc3 1 1 This is promising but 22 .i.h6!? is also in
bxc3 .i.g4 1 2 :tb1 il)d7 1 3 h3 .i.h5 1 4 teresting: 22.. %lfe8 23 :xb7 'ifc6 (not
.
ltb5 fi::Jb6 1 5 c4 .i.xf3 1 6 •xf3 dxc4 1 7 23 ..11t'xb7? 24 '1Vxb7 %lb8 25 a6! and White
.
66
3 f.i)xe5: Bla ck Plays 6... J.. d6
22.......6 30 lla2?
More solid than 22...:C7 23 .ih6lle8 24 Now White loses control of the position.
Axc7 'flxc7 25 'flxc3 when White has a dan Even worse is 30 'ifxa7? llxe6 31 dxe6 .i.d4
gerous initiative. For example, 25...loa3? 26 32 l:xh7+ WgS 33 'flb7 Wxf2+ 34 Wht c2!
llet! llb8 27 'flxc7 .i.xc7 281le7 ltlxc2 29 35 :t? .i.xal 361lxf2 ct'fl+ 37 Wg2 :xl2+
:Xc7 ltlxd4 30 llxa7 gave White a winning 38 Wxl2 .td4+ when Black has decem win
ending in Motylev-Ristic, Novi Sad 2000. ning chances. White could have maintained
23 1la1 an edge with 30 :b4!? llb8 31 :c4 c2 32
Less threatening is 23 .ib3 'fit'S 24 11fxf5 llct llbl 33 ll4xc2 .td4 34 Wf41lxc1+ 35
gxfS 25 llal (Grischuk-Shirov, FIDE World llxct .i.xf2+ 36 Wg2 (Rogers). However, the
Championship, New Delhi 2000 ) and here best option is Shirov's later suggestion: 30
Grischuk suggested the equalising 25...ltld2!? lla41 c2 31 llc4 .td4 32 1i'f4 Wxf4 33 gxf4
26 g3ltlxb3 27 :Xb3 c2 281lct llc4. .i.xf2+ 34 Wg2 with good winning chances.
23...J..b8?1 30....id4 31 ...,
Now Black is in real trouble. 23...Ac7 24 31 1td3!? .txf2+ 32 Wg2 .i.d4 33 llc2
a6 lDxe3 25 fxe3 'fle7 26 l:.abt is only mar with an unclear position is better.
ginally better for White. 31 .....f3 32 Wh2 •xd5 33 .ixd5 llxe1
24 i.b3 .d6
If 24 ... c2 25 Act Wd6 26 :Xc2 9hz+ 27
Wft ltlxe3+ 28 fxe3 lLc2 29 .i.xc2 White is
clearly better (Shirov).
25 g3 f.i)xe3 26 J..xf7 + �h8 27 •xe3
.,6
Pushing the pawn again achieves nothing:
27...c2 28 llct llc7 29 llxc7 .ixc7 30 .ic6
i.xa5 31 d5 (Shirov).
28 .i.e& llceB 29 d5 .ie5
34 �2?
Now White is lost. 34 f4 :res 35 .to
llct still leaves White with good drawing
chances.
34 ...i.xf21 35 llf7
The point is 35 llxf2 llxf2+ 36 Wxf2 c2
and the pawn queens.
35 1lxf7 36 .ixf7 .ic5 37 .ib3 ¢>g7 38
•..
67
The Petroff Defence
Summary
The variation 6...J.d6 7 0-0 0-0 8 c4 c6 has become fashionable and is likely to remain popular
because of the huge number of complicated, untypical and insufficiently studied positions.
After 9 .c2, defending the knight with 9... f5, as in the historic Wtll.iams-Staunton game
(London 1 851) isn't had because it's not that easy for White to take advantage of the weakened
a2-g8 diagonal. After 9 .c2 �a6 10 a3 Black has a few effective responses: 1 O J:e8 1 1 llk3
•.
J.f5; t t ...J.g4; or the immediate 1 0....i.g4, responding to 1 1 lDe5 with 1 1 .....tf5, 1 t ..J.hS or
1 1 .....txe5 12 dxe5 �c5. The fashionable 1 o...f5 appc..-ars to give White the better chances after
an accurate attack on the queenside with 1 1 �c3 �c7 1 2 lDe2 � 1 3 b4.
In the event of 9 1let. the pawn sacrifice 9.....tg41? deserves careful consideration. After 10
Le4 dxe4 1 1 llxe4 f5 1 2 lle6 1id7 Black has full compensation, while he has prospects of
equalising after 12 1let ..txf3 13 Lf3 1t'h4. In the case of 9 ... J.f5 White can claim an advan
tage with 1 0 llk3 �xc3 1 1 bxc3 ..txd3 1 2 Wxd3 dxc4 1 3 11'xc4 �7 14 ..tgS!?, while after
9...1le8 10 ru �xc3 1 1 bxc3 the position appears to be favourable for White.
To us, the move 9 �c3 seems less accurate than 9 cxd5 cxd5 1 0 llk3 because Black can
proceed with 9...�xc3 10 bxc3 dxc4 1 1 Lc4 ..tg4. We can't find any white's advantage in
either the calm variations or the sharp 1 2 1id3 �7 1 3 clOgs �f6 1 4 h3 .i.h5 1 5 f4 h6 1 6 g4.
Recently 9 cxd5 cxd5 1 0 ru �c3 1 1 bxc3 .i.g4 has not been so popular. White's position
seems to be more promising after 1 2 1lb1 b6 but. if Black defends accurately. White will not
gain an advantage after 12 ..�7. Nowadays there are fewer fans of this complex line, which
requires knowledge of variations developed earlier and where t.-very moment is critical.
9 1le l
9....if5 - Game 22; 9...1le8 - Gan1e 23
9 llk3 �xc3 1 0 bxc3 dxc4 1 1 .lxc4 .i.g4 (D): 12 h3 - GttnJe 24; 1 2 .d3 - G011Je 2;
9 cxd5 10 �c3 lnxc3 1 1 bxc3 .i.g4 1 2 .Rb1 (D) �7
••.
12 ...b6 - Gm11e 26
1 3 h3 .i.h5 14 .Rb5 �6 1 5 c4 .i.xf3 1 6 11xf3 dxc4 1 7 .i.c2 ...d7 1 8 a4
1 8. 1lab8 Game 27; 18...g6 - Gm11e 28
.. -
68
CHAPTER FOUR I
3 ltJxe5: Deviations
from the Main Line
1 e4 e5 2 t0f3 �f& 3 o!bxe5 d& 4 �f3 After S ... dS 6 ..id3 i.e7 7 0-0 the idea
o!bxe4 5 d4 7 ...ll:)d6 is played very seldom, with Bilguer
A fter 1 e4 cS 2 lLlf3 lL!f6 3 lLlxeS d6 4 declaring this move a 'motiveless retreat'. In
ll:)f3 lLlxe4 5 d4 we've already identified two fact, there arc some motives, i.e. a counter
principal trends: S...dS 6 �d3 �e7 7 0-0 c!bc6 measure to c2-c4 and a preparation of .....i£5.
(< :hapters 1 -2) and 6 ....i.d6 (Chapter 3). It is However, a fter 8 i. f4! Black fails to develop
reasonable to devote a separate chapter to his bishop to f5 and is forced into passive
deviations from these �nes. Black rarely dc defence.
c�nes to play the main continuation S...dS in If Blacks wishes to move his bishop to f5,
favour of the passive s...i.e7 (Game 32). he should do so immediately after 5 ... d5 6
After 6 �d3 the knight retreats to f6 rather �d3 i.e7 7 0-0 �f5. Then 8 c4 can be met
than to gS (in view of 7 li:)xgS ..ixg5 8 by 8...dxc4 since the rook is not yet attacking
'ii'c2+). l11en White normaUy proceeds with the knight on c4, and the arising position
7 h3 in order to restrict the c8-bishop. with an isolated pawn gives approximately
After 5 d4 dS 6 ..td3 Marshall suggested equal chances. 8 :Ct lL!c6 (Games 29-30) is
6 .....tg4, which failed the test of time in view more common, after which the position
of 7 "iVe2+1 'fle7 8 0-0 li:)c6 9 �bS, when it is branches. Game 29 reviews the fashionable 9
difficult for Black to complete his develop lbbd2 as weU as 9 lbc3 and the pin 9 i.bS,
ment. In Game 31 we discuss MarshaU's idea taking control of the eS-squarc; Game 30 is
of an early .....tg4 improved by the insertion devoted to the more fundamcnral 9 c4.
of the moves 6 ...li:)c6 7 0-0. The two critical
lines arc 6 ...c!bc6 7 IJ-0 ..ig4 8 c4 lLlf6 9 cxdS Game 29
..txf3 10 'tixf3 'it'xdS 1 1 'it'xdS (or 11 "iVe2+; Kasparov-Karpov
I I l:cl+ �c7 reaches a tabiya from Chapter Mo.rco»' 1981
2) and 9 li:)c3 i.xf3 10 "iVxf3 li:)xd4 1 1 'it'c3+
(or 1 1 'it'h3; again 1 1 l:e l + ..ic7 rt-aches 1 e4 e5 2 t0f3 �f& 3 lDxe5 d& 4 �f3
Chapter 2). l r should be said, however, that �xe4 5 d4 JJ..e7 6 .i.d3 d5 7 0-0 JJ..f5
in modem practice Black usually avoids giv Retreating voluntarily with 7...li:)d6 makes
ing his opponent a variety of possibilities and White's task easier: 8 i. f4 0-0 9 :let ..ie6 (or
often develops his bishop to e7 early on. 9.....ig4 10 h3 ..ihS I I l:leS!? i.xf3 12 1i'xf3
69
Th e Petroff Defence
70
3 li:lxe5: D e viations from the Msin L ine
i.e3 illustrates the potential weakness of d4: Playing directly for exchanges with
1 5...lnxd41 1 6 ttlxf7 ttlf3+! 1 7 Wxf3 lLf7 1 8 1 9...l:te7 leads to some difficulty: 20 tL\b3
'lfh3 llf5 and BL1ck has equalised) 1 5...cxd6 llxe2 21 Wxe2 i.e? 22 'l'g41 Wxg4 23 hxg4
1 6 ttlf3 1Vb6 1 7 l:tab 1 ttlxd4 1 8 ttlxd4 Wxd4 i..d6 24 i..xd6 cxd6 25 f4 and Black's dou
1 9 Wd7 and White's activity certainly com bled d-pawns ensure White of a slight edge in
pensates for the missing pawn but is proba the ending.
bly not enough to force an advantage. 20 ltlf1 .i.f8 21 1t'f3 J:le7 22 ltle3
1 2...1t'd7 1 3 .i.f4 a& 1 4 J:le3 J:lae8 1 5 The assessment is typical of this variation:
J:lae1 White has a nagging pull.
White is more active but Black is very 22...�8 23 .i.xc71?
solid and simplifying exchanges seem likely. An interesting way to unbalance the play.
1 5.....td8 1 6 h3 23 ...1t'xc7 24 ltlxd5
Safeguanling the back rank and keeping
the tension. 1 6 �5 eases Black's task after
1 6...ttlxe5 17 dxe5 'Wbs 1 8 b3 Wxd3 1 9
l:lxd3 c6 20 c4 i.a5 2 1 llbt dxc4 22 bxc4
lld8 23 lla3 (Kavalck-Smyslov, Amsterdam
1 981) and here the simplest path to equality
is 23. ..i.d2 24 i.xd2 llxd2.
.
1 6 J:lxe3 1 7 llxe3
••.
24 1t'd6?1
•.•
1i'c8 21 ltxe6 'l'xe6 22 'irxd) fxe6 23 tl\g5! Now Black manages to untangle. 28 h4!?
and White has a clear advantage (Kasparov). Wg7 29 h5 keeps an edge.
18 J:le2 28 ...Wg7 29 b4 1t'c7 30 J:le3 li:lf7 3 1
This simply improves coordination while 1t'e& •d8 32 a5 h5 33 ee4 ed7 34 ••6
awaiting developments. In a later game 1 8 1t'd8 35 �1 li:lh6 36 g4 hxg4 37 hxg4
ll:\12 was tried and following 1 8...ttle7 1 9 li:lf7 38 We2
tL'lbJ 'irf5! 20 'l'xf5 ttlxf5 2 1 l:tc2 b6 22 tl\c1 Or 38 f4 'ire? 39 llf3 i..d6 40 1i'e4 Wc8
<tlf7 23 ttld3 aS 24 J,>4 ttle7 a draw was and Black still holds on to CI.Juality.
a�-,rreed in Leko-Kramnik, Dorunund 1 999. a8 ltlg5 39 as •d7 40 Wc�3 ..td& 41
.••
71
The Petroff Defence
72
3 li:Jxe5: Deviations from the Main L ine
aggressively) 20....i.e6 (or 20....i.xe5 21 .i.xe5 the wayward knight wiU soon be coUected.
.i.d7 22 .i.xc7 lieS 23 'ifd2 'ifxa3 24 .:.xeS+ 14 i.xd7 /l)xc1 1 5 �c3
.i.xe8 25 •g5! and White keeps an edge) 21
g5 .i.d8 and White was a bit better in Korn
ec.'V-Ro.Percz, Albacete 2000.
1 0 ...lnxf217
This is the critical try. 10...�xd3 1 1 '1Vxd3
'ifxd5 transposes to 7...�6 8 c4 �b4 9 cxdS
�d3 10 9xd3 'IVxdS 1 1 .:.e1 .i.t5 (sec
Games 9-1 0).
1 1 .84+
The only good move. If 1 1 .tb5+?! c6 1 2
.a4 0-0 1 3 dxc6 bxc6 14 Lc6 �fd3 Black
has a clear advantage. 1 2 dxc6? is even worse:
1 2...�xd1 13 c7+ '1Vd7 14 .i.xd7+ .i.xd7 1 5
�a3 �f2! and Black i s winning. 1 5... lDcd3
1 1 ....id7 Black can also try 1 S...ltkxa2, after which
White's best is 1 6 .:.xa2!? �xa2 1 7 �a2 a6
1 8 .i.t5 with an edge. Instead after 16 �xa2
�xd5 Black hits c3 and thn:atens ...�b6. 17
:act �xe3 1 8 .:.xe3 .i.d6 leading to an un
clear position.
1 6 a3 lDc5
Ro.Percz as.'ICsses 16.Jnxb2 17 9b3 lilc4
1 8 llxe7 'flxe7 19 axb4 as a bir better for
White.
1 7 dxc5
This aiJows White to keep some control
lnstc.-ad 17 '1Vxb4 aS 18 'it'bs �xd7 1 9 :act
.i.d6 20 1Wxb7 .:.b8 21 Wa7 .:.xb2 22 '1Vxa5
1 2 i.b5 �b6 is very messy.
Again White has no choice. After 12 1 7 ....hc5 1 8 axb4 i.xe3+ 1 9 �1
1i'xb4? �xd3 1 3 .:.xc7+ '1Vxe7 14 '1Vxb7 0-0 White's two minor pieces are preferable to
1 5 .i.d2 Wc2 Black wins easily. Black's rook and pawn.
1 2.../l)fd3 1 9...1tf6
Advancing further into White's position t 9... a6 20 .:.et .i.b6 21 .i.£5 is also mar
with 1 2...�bd3?! leads only to trouble: 1 3 ginaUy in White's favour.
.:.e2 a6 (or 1 3...�xc1 14 .:.x.£2 �3 1 5 .:.d2 20 i.b5 1lad8 21 11rc2 •f4 22 .i.d3 g6
�f4 16 �5 .i.xb5 1 7 9xb5+ <i'fB 1 8 .:.f2 The other way to cover h7 is 22...h6, but
and White is much better) 1 4 .i.xd7+ '1Vxd7 White keeps the attack going with 23 bS
1 5 'ti'xd7+ <i'xd7 1 6 .:.xf2 �xf2 17 <i'xf2 .:.res 24 .:.n .
.i.d6 1 8 � and White's material advantage 23 b5 lltea 24 :n ts 25 a3 •de 26
is almost decisive. 1i'c4 26 ...1tc57
1 3 1le3 0-0 Overlooking a nasty trick. 26.. Wh8 27
.
73
The Petroff Defence
27 1i'xc5?1
White misses it as weU: 27 d6+! q;g7 28
dxc7 :d7 29 b6 is effectively winning.
27 i.xc5 28 i.c4 h6?1
••.
Game 3 1 8 ...lbf6
Kupreichik-Vusupov Trying to grab material with S. ...i.x£3 9
.
74
3 fi:lxe5: De viations from th e Main L ine
c4 lt.)f6 9 lt.)c3 ..tg4 (sec Game 1 1). greedy: 1 5 We2 'lrg4 - 1 S...lt.)xa1? ill refuted
10 1i'xf3 li:lxd4 1 1 Wh3 by 1 6 ..ixf6 gxf6 1 7 l0ds - 1 6 ..xg4 �g4
1 7 Jl.xe7 Wxe7 1 8 lfitS+ Wf8 1 9 :act l0d4
20 lt.)xc7 l:d8 21 J:lfe1 and Black's awkward
king delays his development, giving White a
pleasant edge) 1 5 Jl.xf6 .i.xf6 1 6 lDd5 1Wd8
1 7 1tb3 i.d4 18 •xb7 0-0 1 9 llab1 ltb8 20
1fa6 c6 21 1fxc6 i.xb2 and Black has es
caped to �::quality, Bologan-Koch, Belfon
2002.
1 4 llad1 c5
If 14 ... h6 White should play calmly with
IS ..tf4 cS 16 lt.)bS when his initiative is very
threatening. 1 5 .i.xh6?! ill premature:
I S...gxh6 1 6 J:ld3 Wh7 17 llg3 �8! and
This is probably the most dangerous op Black defendll successfully.
r:ion but 1 1 'ire3+ is also worth considering: 1 5 llfe1
t l ...c!De6 12 cxdS lOxdS 1 3 lt.)xdS ..xdS 1 4 White's lead in development and the vari
..ic4 'irbs I S a4 Wa6 1 6 llcJt Jl.c7 (the seem ous potential pins constitute good compensa
ingly more active 1 6....i.c5 simply helps tion for the pawn.
White to open lines: 1 7 1i'f3 c6 1 8 b4!? i.xb4 1 5 h6 1 6 hh&l
•••
19 lib I Jl.cS 20 :Xb7 and White has the This is even stronger than the tempting 1 6
initiative) 17 b4! 0-0 (1 7 ...Jl.xb4 is too greedy: llxe7!? hxgS (Black can give up the queen
1 8 .i.b2 0-0 1 9 11h3 �5 20 Wg4 .i.e7 21 h4 with 1 6...1fxe7 1 7 lOds lt.)xdS 1 8 .i.xe7
and White has a strong attack) 1 8 1Vh3 br6 1 9 lt.)xc7, but after 1 9 b4 b6 20 bxcS bxc5 21
'ifc3 (keeping Black'll queen out o f the game; 'ifhs White still has an edge) 17 llxb7 llb8
instead 19 Jl.b2 allows 19 ...Wc41 20 l:d7 1 8 llxb8 ..xb8 1 9 b3 'ifeS when the position
J:lae8 21 i.dS 'ifxb4 22 Jl.c3 lt.)f4 23 .i.xb4 is unclear according to Yusupov.
lt.)xh3+ 24 gxh3 i.xb4 25 :Xc7 b6 26 llxa7 1 6...gxh6 1 7 1i'xh6 fi:lh7
�g7 when Black has cleverly equalised, Kas
parov-Karpov, World Championship (Game
6], London/Leningrad 1 986) 19 ...lt.)g5 20
..ixgS (20 i.b2 may seem more natural but
20 ...i.f6 21 1i'xf6 1fxf6 22 ..txf6 lt.)xe4 23
.
75
The Petroff D e fence
21 Wh8?
..• Trying to equalise by seeking exchanges
Black cracks under the pressure. 21 ...q.,g7 with 6 ...�g5 just falls short: 7 ��5 i.. xgS 8
is essential, when the fight continues. Follow 1i'c2+ Wffi (or 8....te7 9 0-0 0-11 l O �c3 /0c6
ing 22 �xf6 (Black has counterplay after 22 1 l �5 lieS 1 2 .ie3 and White also has a
h4 'ti'g(i 23 11fxg6+ fxg6 24 hxgS llf4 - Yu slight edge, while after 8. .i..e6 9 f4 i.. h4+ t O
.
76
3 l:iJxe5: De viations from th e Main L ine
12 �2 '*g7 13 �1 .i.d7 14 f4 �8 15 'iff3 a) to. .ll)fB 11 :C1 ll)g6 (White has a clear
.
11fh4 16 g3 'ifg4 17 11ff2 and White was a bit plus after 11...d5 1 2 �5 c6 13 cxdS li)xdS
better in Palac-Mun.1', Pula 2002 14 li)xd5 cxdS 15 .i.b5) 12 11fc2 .i.d7 13
7 h3 .i.g5 .i.c6 14 d5 i.d7 15 l:lad1 a6 16 �2
White plays to dominate his opponent's li)hS 17 .tel •cs 18 �d4 cS 19 dxc6 bxc6
queen bishop. 7 0-0 is a sensible alternative, 20 li)5 .i.x£5 21 .i.x5 11fc7 22 1Wa4 li)f6 23
t(,r example 7.....i.g4 8 �bd2 0-0 9 �1 c5 1 0 l2XI4 c5 24 li)c6 and White was much better
h3 ..i.hS 11 ll)ft llk6 and now: in Luther-Bellin, Catalan Bay 2003.
a) Breaking the pin with 12 g4 is weaken b) 10. c6 11 J:le1 ll)f8 12 .i.f4 a6 13
..
ing: t 2 ... .i.g6 13 li)g3 d51? (White has a small 'flb3?! (this allows cnunterplay; 13 d5!? li)g6
advantage after 13...l:le8 14 c3 .c7 15 li)5 14 .i.h2 cS 15 'li'd2, with a slight edge, is
i.f8 16 l:lxe8 l:lxc8 17 .i.f4 cxd4 18 li)3xd4 better, while also promising is 13 b41? �6
a6 19 .a4 �5 20 .i.g3 li)b6 21 .c2, Ste 14 .i.h2 a5 15 b5) l 3H,�6 14 .i.h2 .i.f8
f.'tnsson-Kholmov, Pardubice 2001) 14 dxc5 (Fischer assessed 14... b5!? 15 a4 bxc4 16
.i.xc5 15 .i.e3 .i.xe3 16 l:lxe3 d4 17 J:lc1 .i..xc4 d5 17 ..i.ft c5 as unclear) 1 5 :C2 b5 16
� with complicated play. 1ic2 ..i.b7 17 J:lae1 g6 with a complex posi
b) 12 li)g3 is better: 1 2....i.xf3 (or tion, Fischer-Petrosian, 5th matchgame,
12....i.g6 13 .i.xg6 hxh>6 1 4 dS li)b4 15 c4 Buenos Aires 1971.
and White has a pleasant space advantage) 13 1 0 l:iJc3 e6 1 1 b4 l:iJbd7 1 2 .i.e3 t'iJf8 1 3
11fxf3 li)xd4 14 11fxb7 �8 15 ll)f5 li)x5 16 e4
.i.x£5 .i.fB 17 l:lxe8 'ifxe8 18 .i.d2 J!ll 1 9
.i.d3 d5 20 c4 and White has a strong initia
tive, Ljubojevic-Smyslov, London 1984.
7 0-0 8 0-0 lieS
...
77
The Petroff Defence
lf23...cxd5 then 24 l:c1 ..d8 25 �b5 and This is a good move but White could also
White keeps an edge. cash in immediately with 38 I.Oa6 I.Oe4 39
24 llc1 cxd5 25 tDb6 WbB+ 26 Wg1 I.Oxb8 'l'xb8 40 i.d3.
lle7?! 38 . . .'iFd6 39 e5?1
This aUows White's advantage to grow. This allows Black to complicate. 39 I.Ob7?1
Black can limit the damage with 26 ..ltc8 27 J:lxb7 40 .:Xb7 I.Oxc4 with unclear play is also
lDc1 .:Xct 28 1Wxct 'lrd8 29 l2X!3, when unconvincing, but the simple solution is to
White is only A bit better. prepare e4-e5 by supporting the knight with
27 tDe5! i.d7 39 'l'c3 - White wins easily after 39 ...i.h2+
The tactical justification is seen after 40 �ht 1Wf4 41 e5.
27...I.Oxe5 28 dxe5 '1Vxe5 29 i.d4 'lre4 30 39....be5 40dxe5 'iFxc5+ 41 Wh1?!
�d3 Wf4 3 1 i.xfS 'lrxfS 32 .J:f3 •e6 33 Now White's advantage disappears en
�xf6 gxf6 34 llg3+ �£8 35 1Wc2, when tirely. 41 Wh2 I.Oe4 42 l:c7 'iraS 43 i.b5
Black's weakened kingside causes terrible maintains a slight edge.
trouble. 41 ltla4 42 Wh2 1le6 43 1lc7 'iFa5
...
28 tDd3 ..txb5 29 axb5 ..td& 30 �5! 43..ltc6 44 .:Xc6 ...xc6 45 �d3 1Wg6 is
also good enough for equality.
44 ..i.d3
78
3 li:lxe5: De via tions from the Main L ine
Summary
Firsdy, the passive continuations covered in this chapter (5 ... .i.e7 and a quick ...l0d6 after
5... d5) are not of great interest. The key position of the chapter is the one arising after S...dS 6
.Ll3 .i.e7 7 0-0 .i.fS 8 %let lesc6 9 c4. In response to 9...0-0, White successfuUy develops his
knight to c3 either immediately or after 10 cxdS ..xdS. In the case of 9...lesb4 White maintains
110 opening initiative by retreating with 10 .i.fl, but 1 0 cxdS also appears to be a good response
- the tactical blow 10 ..lC!xf21? is not a refutation. Probably Black should transpose to the line
.
6.. .i.e7 7 0-0 �6 8 c4 l0b4 9 cxdS lC!xd3 10 ..xd3 'tixdS 1 1 :C t -*..fS (see Chapter 2) by
.
9 lesbd2 Ga111t 29
-
5 d4 6 .i.d3 B. . . lilc6
79
CHAPTER FIVE I
3 ltJxe5: Fourth and Fifth
Move Alternatives
1 e4 e5 2 .!Llf3 �f6 3 �xe5 d6 4 �f3 idea. In the event of 5 ... �5 White retreats
.!Llxe4 his bishop to e2 and makes up for lost time
After 1 e4 eS 2 �0 �f6 3 �xeS d6 4 by attacking the knight with d2-d4. After
�0 �xe4 White sometimes declines to play S. . dS
. White continues the attack on the t..-4-
5 d4 in favour of the other continuations, knight with 6 1i'e2, while if S...�f6 White
and these fifth move alternatives are the sub clears the way for the d-pawn by means of
ject of the fll'St part of this chapter. c2-c3 and .ld3-c2.
5 �c3 (Games 33-34) was played as fur 5 d3 (Game 39) is the most modest of
back as thc 1 9th century, while later on White's 5th move options. White offers
Nimzowitsch focused on it. Black the opportunity to transpose to the
5 1i'c2 (Games 35-36) was introduced by a Exchange Variation of the French Defence
1 3-ycar old Paul Morphy against LOwenthal after S.)t�f6 6 d4 dS where, as it is known,
(Nt:w Orleans, 1 850), and the line found an extra tempo in this symmetrical line
worthy adherents in the form of Emanuel promises White just a microscopic advan
L'lsker and Jose Raul Capablanca. Owing to tage. However,. Black can equalise after
the success of such celebrities (sec, for ex 6....le7 foUowed up by ...c7-c5.
ample, Game 35) 5 'i6'e2 was considered to Moving onto 4th move alternatives, be
be nearly a refutation of the Petroff Defence, sides 4 lDO White has two interesting op
but now we know sevcral ways for Black to tions.
equalise. The fmt major game to witness 4 �c4
Kaufmann, a chess player from Vienna, was l hPaulscn-Schallopp, Frankfurt 1 887.
suggested 5 c4 (Game 37) as a way to pre The idea to transfer the knight to c3 is not
vent Black from supporting the e4-knight particularly impressive. In particular, Black's
with ... d6-d5. In principle it's still possible to attack on f2 after 4...lbe4 5 d4 dS 6 lDe3 1i'f6
play S ... dS, but an attack on Black's central- (Game 40) deserves serious consideration for
ist.-d forces with 6 �c3 gives White the better Black.
chances. Instead Black should continue his The Cochranc Gambit 4 lbxf7!? (Games
development with either s ...�c6 or s ...i.e7
. 41 -42), originated by John Cochrane in the
to avoid any problems. 1 840s, stands in rotal contrast to the other
5 .i.d3 (Game 38) is a comparatively new lines in the Petroff Defence. Staunton also
80
3 li::J x e5: Fo urth and Fifth Mo ve A lterna tives
81
Th e Petroff De fen ce
1 9 .i.b2 and White's chances were preferable bring the queen home. 17 c3! is stronger.
in Acs-Haba, Bled 2002. 1 7...'ifxd2 1 8 'iih8+ �e7 1 9 1Wxg7+ �d8 20
6 ..e2 �e7 7 lillle4 dxe4 8 •xe4 0-0 9 l:adt 1Wf4 21 g3 ..f8 22 �xe6+ ami White
�c4 wins (Aiekhine); or 1 7....i.e5 1 8 Wh5 �g8 1 9
Black obtains good play after 9 .i.d3 g6 1 0 Wf7+ �hs 20 1Wg6! �gS 2 1 d4 .i.f6 22
0-0 lbc:6 1 1 .tbS .tf5 1 2 •e2 :Cs 13 .txc6 'ifb7+ �f8 23 �xc6+ and White has a very
bxc6 1 4 d4 l:b8, when the bishops and the large plus.
development compensate for the pawn defi 1 7 ......f6 1 8 �5
cit. 1 8 'ifh8+?1 is weaker. after 18...�e7 1 9
9 ...�d6 10 0..() l:xe6+ Wd7 20 l:xf6 l:txh8 2 1 l:£7+ �7 22
White is aiming for rapid development. h3 g6 Black's chances in the endgame are
Alekhine gave the following line: 1 0 d4 l:e8 preferable.
1 1 liJcS .i.xeS 12 dxeS l'Dc6 1 3 .i.f4 �4 14 1 8 ...�8 1 9 lle3 �f4?
0-0-0 l:xeS 1S l:d8+! 'ifxd8 1 6 .i.xeS ..e7 17 Now I know we should not talk badly of
l:ct �xeS 1 8 "ffxeS ..xeS 1 9 llxeS �fB, the dead, and Rabinovich was truly one of
when the draw is nt:ar. the great fathers of Russian Chess, but here
1 0...lle8 1 1 •d3 1Llc6 1 2 b3! his defence was not particularly good. White
White needs to get his dark-S(]Uarcd is also very happy after 1 9 �4?! 20 llh3 g6
...
bishop into play. After something like 12 21 Wh7+ WfB 22 11i'd7 :e7 (22...�g8? 23
..c3 .i.g4 1 3 .i.dS .i.xf3 14 .i.xf3 �d41 l:h8+1 would be an unpleasant surprise) 23
Black would have good play for the pawn. �h7+ �g7 24 l'Dxf6 llxd7 25 �xd7 l:d8 26
1 2...•f6 13 �b211 %ld3 l'De2+ 27 �h1 l:lxd7 28 g3 (AJekhine).
However, Black can play 1 9 ...�eSI 20 'ifb7+
WfB 21 f4 �£7 22 'ifhs �g8. probably keep
ing the position level.
20 Wh7+ Wf8 21 Wh8+ h7
82
3 liJxe5: Fourth snd Fifth Move A lterna tives
1 -*.f4
White has tried two other means of devel
opment
a) 7 .i.d3 � 8 .i.e3 .i.g4 9 ..ie4 doesn't 7 . . .li1d7
I<K)k particularly appealing, even though this 7. ..li1c6 also looks sensible, for example 8
worked well in the following example: 'ifd2 .i.g4 9 .i.e2 'ird7 1 0 h3 .i.fS 1 1 0-0-0
9 ...'ird7 t O 'ird2 0-0-0 1 t 0-0-0 llhe8 1 2 0-0-0, avoiding the danger of opposite side
lLld4 d S (1 2. ...Ldt ? 1 3 ltlxc6 bxc6? 1 4 'ird3 casding. Now after 1 2 .l:lhe 1 lZ.he8 1 3 g4 .i.g6
and White wins) 13 lbxc6 Wxc6 1 4 .i.xdS 14 .i.bS a6 I S ..ta4 bS 16 .i.b3 ltlaS t 7 c4
'iVa6 1 5 'ii'd3 .i.xd1 (this is necessary; �xc4 1 8 ..txc4 bxc4 19 WdS Wbs 20 WaS+
I S...'iraS 16 f3 c6 1 7 'irc4 .l:lxdS was played 'ifb8 21 'ifxa6+ Wb7 22 Wxb7+ �b7 Black
83
The Petroff Defence
has equal chances in the ending. In this line 1 9....txg6? is bad on account of 20 lDeS
17...lDxb3+ is weaker: 1 8 axb3 1i'c6 19 lDd4 .tf5 21 .tbS :t£8 22 lDd7, when White has a
1i'b7 20 cxbS axbS 21 1i'a5 with an attack for clear advantage. Now Black loses after
White, Boricsev-Raetsky, Fribourg 2000. 22...lle8 23 ..xd5, so he is forced to play
8 •d2 22 ... d4, which is hardly ideal.
20 ll:le5 .tf& 21 ll:lg4
8 �5
•••
84
3 liJxe5: Fourth and Fifth Mo ve Alterna tives
24 lt!xf6+ gxf6 25 bxc3 •c7 26 'ifg4! gave White the chance to irritate Black in
11Ve5 Hodgson-Barua, l .ondon 1 986.
7 .i.g5
27 lld5!1 1bc3
White wins in all lines: 27 ... i.f5 28 'ifxf5 7 �e6
..
or 27....ixd5 28 1i"xg6+ Wffi 29 i.xc5+! For some reason 131ack insists on keeping
when Black will get mated. the queens on, which is not necessary be
28 'ifxe4 1fa1 + 29 �2 •xg1 30 lld7l cause the endgame promises Black guod
1 -0 chances. 7...1i"xe2+ is studied in the next
r----- game, while 7.../i)bd7 8 /i)c3 1rxe2+ (8...h6?!
Gatne 35 9 .ic3 /i)b6 1 0 0.0-0 .id7 1 1 1i"d2 U-0-0 1 2
Em.Lasker-Marshall llc1 is a tad uncomfortable for Black) 9
St Petersbutg 1912 .i.xe2 transposes to 7...9xe2+ 8 .ixe2 /i)bd7
.________________. 9 /i)c3 (see the notes to the next game).
1 e4 e5 2 lt!f3 etlf6 3 lt!xe5 d6 4 lt!f3 8 ltlc3 ltlbd7
lbxe4 5 'ife2 8 .../i)c6 is answered by the natural 9 /i)e4
This line is usually played with the hope of 0-0-0 10 /i)xf6 gxf6 1 I .ie3 d5 1 2 d4 .ibr4 13
a draw. However, many White players have 0-0-0 when White's position is preferable.
found that achieving the draw from an even 8. .. h6 looks dubious after 9 .ixf6 9xf6 1 0
ending is not always so easy. d4 i.e7 1 1 'Wb5+ /i)d7 1 2 .i.d3 (1 2 'iVxb7
5 .. -•e7 6 d3 0-0 1 3 'ifxc7 i!Db6 appears very risky for
The score from this position in our data White) 1 2...g5 1 3 h3 (or 13 ll:le4 'ti"g7 14 h3
base is actually 51°/co for White and 49% for with a slight edge - Keres) 13 ...0-0 1 4 'iVxb7
Black, which suggests complete equality. llab8 1 5 1re4 'iVg7 16 b3 (we think White is
However, it also shows that people play on better here) 1 6 ...cS? (1 6.../i)cS 17 9c3 .if6
from here... 1 8 0-0 llfe8 1 9 1rd2 g4 - Tanasch - with a
6 ...etlf6 messy position is necessary) 17 0-0 cxd4 18
The most common move, but not the /i)JS i.d8 19 .ic4 .!DeS 20 ..xd4 ..xd4 21
1 1nly one. For example, 6...lL!c5 7 lL!c3 i.g4 8 /i)xd4 LdS 22 .ixdS .if6 23 J:ladt and
.i.e3 (8 ll:lds Lf3!? 9 'ifxe7+ .ixc7 10 White enjoyed a large plus in Capablanca
ltlxc7+ Wd7 1 1 .!Dxa8 .ic6 is not clear, but Marshall, St Petersburg 1 914.
should be fine for Black) 8...c6 9 h3 .i.h5 lO 9 0-0-0
hr4 .ig6 1 1 .tg2 lL!bd7 12 lL!d4 ltle6 13 f4 9 d4 leads to equality after 9...h6 1 0 i.h4
li:lxd4 1 4 .i.xd4 f5 1 S 0-0-0 'ii'xe2 16 ltlxe2 g5 1 l i.g3 �IS 12 /i)xdS .ixdS 13 0-0-0
85
The Petroff Defence
'ifxe2 14 .i.xe2 .i.g7. More interesting is 9 sors.20...�c4 is the only move, although after
�bS!? .!llb6 1 0 .Lf6 gxf6 1 1 g3 dS 1 2 0-0-0 21 l:tb3! 'ifgS+ 22 �b 1 (also strong is 22 f4!?
.i.h6+ 13 �b 1 with an unbalanced position, gxf3+ 2..� �b1 �2+ 24 'lt>at �xb3+ 25 cxb3
something that no one would have thought 'ifxg3 26 .!l:\x£3 .id6 27 llct .i.f4 28 �eSI
possible just a few moves ago. .ixeS 29 Wa7+! �cS 30 dxeS �d7 31 �xc7
9 . . .h6 10 .th4 g5 with a clear plus for White) 22...�2+ 23
1 0... 0-0-0 seems to be more accurate. Af �a1 �xb3+ 24 cxb3 .i.d6 25 Wa7+ �cS 26
ter 1 1 d4 gS 1 2 .ig3 �b6 1 3 1i'b5 a6 1 4 'it'aS �xd6+ l:txd6 (26...cxd6 27 f4! brxf3 28 �xf3
i.g7 the position is unclear. 1fe3 29 l:te1 and Black can no longer prevent
1 1 .tg3 .!Dh5 the deadly l:tcl+) 27 'ifa8+ �d7 28 'ifxh8
This looks like a loss of time, but Black White has every chance of winning even if
does not have it easy. 1 l ..ig7 12 �4 0-0 Black can put up some resistance.
1 3 h4 g4 1 4 �£5 ..d8 1 5 �xg7 �xg7 1 6 hS 21 c;h,1 .td6
is not a serious alternative. Black also loses after 21...'ife7 22 l:lb3
1 2 d4 �xg3 13 hxg3 g4?! l:d6 23 a4 :c6 24 aS, when White's attack is
This move deprives the bishop of the £5- conclusive.
square, which proves to be a problem in the 22 1lb3
later tactics. 1 3... 0-0-0 1 4 We3!? with a slight Or 22 l:tc3!? .!llc4 23 Wa7+ �cS 24
edge for White was better. �xd6+ cxd6 25 �a1 ! .:r.de8 26 llb3 and
1 4 �h4 d5 1 5 1rb5! 0-0-0 Black is busted.
1 5...Wb4? loses to the pretty 1 6 �xdS!. 22. . .1lhe8 23 a4! .tf5
1 6 'fra5
White should avoid 1 6 .!llxdS? .ixdS 17
..xdS 'it'gS+ 1 8 ..xgS hxgS, winning a piece.
16 ••• a6 17 .txa6! bxa6
Or 17...Wb4 18 Wxb4 .ixb4 1 9 .i.d3
.ixc3 20 bxc3 when White is simply just a
pawn up (Kasparov).
1 8 •••6+ Wb8 1 9 �b5 o!Db6 20 lld3
24 flJ.7
Or 24 aS .i.xc2+ 25 �xc2 l:le2+ 26 q;,b1
and it is aU over.
24....td7 25 a5 1i'd2 26 axb6 lle1 +
27 �a2 c6 28 �b5 cxb5 29 1ta7+ 1 -0
GaHJe 36
Aronian-Akopian
20 1i'g5+
••.
Ohrid 2001
After this we cannot find a defence for
Black. The last chance for was given by Kas 1 e4 e5 2 �f3 �f6 3 lbxe5 d6 4 lilf3
parov in his series of books MY GTMI Prr:deces- lbxe4 5 1i'e2 1i'e7 6 d3 lilf6 7 .tg5
86
3 liJxe5: Fourth and Fifth Mo ve A lterna tives
8. . ..te7
8...lbbd7 is equaUy good after 9 lbc3 h6
and now:
a) to .i.h4 gS 1 1 .i.g3 lbhS 12 lbd4 (12
�iS �d8 13 d4 lbb6 is totaUy level too)
1 2...lbxg3 1 .3 hxg3 �f6 1 4 0-0-0 .i.d7 15
.l:ldel 0-0-0 16 lbd1 cS 17 lbf3 .i.g7 1 8 �e3
.i.c6 with equal chances, ApiceUa-Nikcevic,
France 2000.
b) to .i.d2 br6 (1 0...lbb6 1 1 lObS!? i!LlbdS - 10 ..th4
1 1 ...�d8!? - 12 c4 .i.d7 13 a4 c6 14 i!Llbd4 10 .i.f4 does not look very dangerous ei
lbc7 1 S 0-0 .i.e7 1 6 b4 1ead to a white advan ther: what exactly is the bishop meant to be
tnge in Spassky-Yusupov, Moscow 19B1) 1 1 doing here? Anyway, one game continued
0-0-0 (1 1 lObs �dB 1 2 c4 .i.g7 1 3 0-0 :CB to...lbc6 1 1 0-0-0 .i.e6 1 2 d4 .!i)b4 1 3 a3
14 J:lfcl a6 I S lbc3 lbcS 16 b4 .!Lle6 17 d4 aS .!i)bds 14 tt:'lxdS .!i)xdS 1 s .i.d2 0-0-0 1 6
18 bS gS is another example of Black obtain J:lhe1 J:lhe8 1 7 h3 .!i)b6 1 8 b3 .i.dS 1 9 i:e3
ing equal play) 1 1 ....i.g7 12 d4 .!Llb6 1 3 h3 0-0 .i.e4 20 c4 li)d7 21 g3 cS 22 b4 .i.£6 23 lbg1
1 4 .i.d3 .i.e6 tS J:lhcl J:lfcB 1 6 b3 a6 1 7 .!Lle4 cxb4 24 axb4 .!i)b6 25 �b2 .i.fS with
lt)xc4 1 B .i.xe4 .i.dS 1 9 .i.xdS lt)xdS 20 c4 chances for both sides, McShane
lt)f6 with complete equality, Westerinen Mamedyarov, Lausanne 2004.
Ractsky, Hafnarfjonlur 1 999. 1 0...liJbd7
9 liJc3 10 ...lbc6 is also fine, for example 1 1 0-0-0
Or 9 c4 h6 10 .i.f4 lbc6 1 1 lbc3 .i.fS 1 2 .i.d7 12 h.1 0-0-0 1 3 d4 J:lde8 14 .i.c4 J:lhfB
0-0-0 0-0-0 1 3 J:lhc1 gS 14 .i.c3 �g4 1 S lbdS l S .i.g3 .i.d8 16 dS .!Lle7 17 .i.h2 a6 1B a3
�xe3 16 lt)xe7+ �xe7 1 7 fxe3 J:lde8 with a .!i)g6 19 J:lhc1 J:lxe1 20 J:lxcl J:lc8 21 J:lxe8
level position, Spassky-Karpov, Hamburg .!i)xe8 22 Wd2 lt)h4 23 lllih4 1/z-1/z Shott
1982. Anand, Wijk aan Zee 2000.
9 ...h6 1 1 liJd4 lOb& 1 2 a4 a6 13 a5 liJbd5 1 4
Another option is 9...c6 and now: o!Oxd5 liJxd5 1 5 .txe7 liJxe7 1 6 .tf3 c5
a) 1 0 0-0-0 lba6 1 1 J:lhc1 lbc7 1 2 llk4 1 7 .!082 liJc6 1 8 .txc&+ bxc6
87
The Petroff Defen c e
88
3 lDxe5: Fo urth Bnd Fifth Mo ve A lternB tives
position is better) 1 S...J:lxe1 1 6 'irxe1 •e8 17 We consider ...lDd7-cS to be the most reli
'irdt ?l ..c4 18 b3 aS 19 .i.b2 .i.xb2 20 llxb2 able way to develop, but there arc other op
l:le8 21 llket and here, with Black slightly tions: 7 ...llk6 8 .td3 lbeS 9 lbxcS dxeS 1 0
better, the players agreed a draw, Brodsky 'ifc2 .i.gS 1 1 0-0 .te6 12 J:let .i.xct 1 3
Ractsky, Cappclle Ia Grande 2000. l:axct •gs 14 lle3 0-0-0 1 5 J:lce1 f6 1 6 b4
6 lDc3 lld7 and Black has solved his problems.
6 d4 is also interesting, for example: Maroczy-Marshall, San Sebastian 1 91 1 .
a) 6...0-0 7 .i.d3 lbgS!? leads to wild play 8 -*.d3 l0d7 9 1ic2 g6
after 8 lbxg5 hgS 9 1ths l:e8+ t o �d1 9...h6 seems to be less harmonious. After
Let 1 1 'irxh7+ �f8 1 2 1th8+ �e7 13 10 .i.e3 .i.f6 11 0-0-0 (the unambitious 1 1
1!rh4+ �f8 1 4 1Wxd8 J:lxd8 l S Wxc1 dS and 0-0 J:le8 1 2 l:ad 1 lbeS 13 ltlxe5 .ixeS 14
we end up in an endgame where Black has llfe 1 9h4 1S g3 'iVhs gives equal chances)
enough counterplay for the pawn, Slobodjan 1 t ...lbc:S t 2 lbxeS .i.xeS 13 f4 .i.f6 1 4 Wf2
Van Der Sterren, Bundesliga 1 999. lle8 1 S h3 .i.d7 16 :he1 a6 1 7 J:le2 b5 1 8 c5
b) 6...dS 7 cxdS [1 llk3 lbxc3 8 bxc3 0-0 9 White was a bit better in Velimirovic
cxdS 1WxdS transposes to 7 cxdS) 7...1Wxd5 8 Motylcv, Novi Sad 2000.
llk3 lbxc3 9 bxc3 0-0 10 .i.d3 cS 1 1 0-0 10 -*.e3
cxd4 1 2 lle1 .tf6 1 3 J:lb1 lbc6 1 4 .i.e4 ..xa2 1 0 .i.h6 J:le8 1 1 0-0 looks less aggressive:
1 5 9d3 h6 1 6 cxd4 .tg41? (or 1 6...J:ld8 17 1 1 ...lbcs 1 2 l:fc1 .tg4 1 3 lDd4 .tf8 14 .i.e3
.i.a3 .i.e6 18 llb2 ..c4 with an unclear posi (14 hf8 :xet+ 1 S l:xe1 'irxf8 1 6 lbbS
tion) 1 7 .i.a3 iLxO 1 8 gxf3 tLlxd4 1 9 .txf8 'ird8 is simply levd) 14 ...tLlxd3 1 5 'irxd3
l:lxf8 20 llxb7 1faS 21 1ffl .i.eS 22 IIebl f5 .tg7 16 h3 .i.d7 11 9d2 'irh4 1 8 .tgs 9hs
23 .c4+ �h7 24 .i.dS 'ifd2 and Black had 19 lbe2 h6 20 .te3 .ic6 21 lbf4 1ff5 22
good play for the exchange in Gajewsky .td4 a6 23 .ixg7 Wxg7 24 ltlds .ixdS 2S
Mista, Poland 2003. cxdS and a draw was agreed in G.Kuzmin
6 lDxc3
.•• Munl7.ia, Cappelle Ia Grande 2003.
6 ...lbg5 is worse: 7 .i.e2 0-0 8 0-0 lbc6 9 10 ...l0c5 1 1 0-0-0
d4 J:le8 to lbds lbxf3+ 1 1 .i.xf3 .i.f6 12 1 1 .i.xcS dxcS 12 0-0-0 is interesting: fol
.i.e3 .t£5 1 3 b4 a6 1 4 a4 h6 1 S .d2 and lowing 12...i.d6 13 h4 .tg4 1 4 bS!? .ixhS 15
White was better in Nisipeanu-Motylev, Bu l:xhS!? gxhS 16 .i.xh7+ �g7 the position is
charest 2001. highly unclear.
7 dxc3 1 1 ...lDxd3+ 1 2 •xd3 -*.f5 1 3 •d2 c5
1 3....i.e6!? 14 b3 aS with counterplay is
also Logical.
14 -*.g6 -*.xg5 1 5 lDxg5 1fa5 1 6 b3 b51?
Energetically played. White has an edge af
ter l6 ...'1fa3+ t 7 1ib2 .xb2+ 18 �b2 llae8
19 l:d2.
1 7 cxb5 c4
The most precise. After l7 ...a6 1 8 b6
llab8 19 Wb2 :txb6 20 1Wf4 White enjoys a
slight plus because Black has no c..-asy way to
weaken White's Icing's position.
1 8 �2 lZab8 1 9 b4
19 1Wxd6?! is very risky. After cxb3 20
89
The Petro ff Defence
axb3 WxbS 21 b4 aS Black's attack has every 26 Wxd4 bxa2+ 27 �xa2 'lla6+ is equal.
chance of being successful.
1 9 .Axb5
.•
Game 38
1 9 ...'i'xb5 is another attractive option. Naiditsch-Timman
Following 20 WdS 1rxd5 (20...Wa4!?) 21 Dortmund 2002
.ILdS l:lfe8 22 lld2 the endgame is probably
equal, though not necessarily drawn. 1 e4 e5 2 �f3 llJf& 3 ofu:e5 d6 4 �f3
�xe4 5 .i.d3
20 �f3
After 20 ..xd6 ..a4 21 c;tlat 9a3 22 l:lc1 This somewhat anti-positional move has
l:lb6 23 ..e7 Da6 24 ..e2 �d3 25 'i'b2 ..a4 found some popularity over the last few
Black has good play for the pawn. years. However, it shouldn't seriously
20 .:fbs 21 �4 Lb4+ ! 22 cxb4
•.. threaten Black.
Lb4+ 23 ltlb3 5 ltlf6
...
90
3 liJxe5: Fourth a nd Fifth Mo ve A lterna tives
(1 1 ...'1'xc5?! 1 2 lLlg5 'l'c4 1 3 lDxe6 'l'xe2 1 4 Here we recommend that Black should
llxe2 fxe6 1 5 lixe6+ � fl 16 l:le5 is awful for deviate with 1 2 ...1fb6!? 1 3 b3 d5 1 4 lLld2
Black) 1 2 .i.e3 a6 1 3 a4 b4 14 c3 a5 1 5 cxb4 i.e6 leading to a position with equal chances.
axb4 1 6 li)bd2 and White enjoys a plus, 1 3 lLld2 'tlfd7 14 J:lfe1 J:lfe8 1 5 �h2
Morozevich-Shirov, Sarajevo 2000. We suggest 1 5 l:le2, for example 1 5.. ..i.f5
6 h3 �e7 7 0-0 0-0 8 c3 liJc6 1 6 i.xf5 Wxf5 17 l:lae1 .tf8 1 8 lDe4 lL!xe4
Alternatively: 1 9 l:lxe4 and White's control of the e-tile
a) 8 ... d5 9 .ic2 lieS to d4 .id6 1 1 i.g5 annoys Black; or 1 5...c5 1 6 dxc5 dxc5 1 7
lLlbd7 1 2 li)1x12 h6 (12... lLlfB t 3 1lc1 l:lxel+ l:laet and White has a slight plus.
1 4 1fxel h6 1 5 .i.h4 lDg6 1 6 .i.g3 i.xg3 17 1 5 g6 1 6 c4 d5 1 7 c5 liJh5
•.•
fxg3 lL!fB 1 8 g4 wa s better for White in Shi 1 7 ...lL!e4 leads to trouble after 18 li)xe4
ruv-Anand, Leon 2001 - Black's pieces can dxe4 19 :Xe4 .i.d5 20 i.b3!, when the attack
not really get into the game) 1 3 .ih4 li)b6 1 4 against fl leads to a position with an extra
a4 a S and Black has nothing to fear. pawn for White.
b) 8...c5 9 i.c2 lbc6 to d4 i.e6 1 1 l:le1 h6 1 8 'irb3 f6 1 9 liJf1 �f8 20 liJe3 �h6
91
The Petroff Defence
92
3 ltlxe5: Fourth and Fifth Mo ve Alternatives
1i'xc2+ 9 Wxe2 i.g7 10 ltk3 ltlc6 1 1 ltlb5 i.e2 f5 1 6 f4 ltld7 1 7 i.d3 ltlc7 1 8 1i'f3 i.f6
Wd8 and the endgame should of course end 19 ltld1 b5 20 c3 g6 21 i.d2 ltlb6 22 lbe3
in a draw. 'ii'd7 both sides have chances, Shtyrenkov
7 ..i.d3 ..i.g4 Raersky, Voronezh 1989. lnstead of 1 2 ltk3,
7 ...0-0 allows 8 h3 when White may be 1 2 g4 i.g6 1 3 i.xg6 hxg6 14 ltlc3 ltlh7!? 1 5
able to gain a slight pull: 8...c51? 9 0-0 ltk6 1 0 a4 ltlg5 is interesting - maybe White i s a bit
%let h6 t t ltlc3 i.c6 1 2 i.b5 lbd5 t3 ltlxd5 better here
i.xd5 1 4 dxc5 (this is better than 1 4 c3 a6 1 5 1 1 ...dxc5 1 2 g4 ..i.g6 13 ..ixg6 hxg6 1 4
i.xc6 i.xc6 1 6 dxc5 dxc5 17 1i'xd8 i.xd8 1 8 •xd8 ..ixd8 1 5 ..if4 ..ia5!
i.c3, Kveinys-Schandorff, Germany 2001 ; Accurate play. After 15 ...ltlbd7? 1 6 gS
here Black can play 1 R...i.xf3 1 9 gxf3 b6 ltlh5 17 i.d6 i.aS 18 l:le7! White has gained
with a slightly more comfortable position, a serious advantage.
even though it is probably not enough for an 1 6 c3
advantage) 14...dxc5 1 5 i.f4 i.f6 1 6 ltlc5 Or 16 lld1 ltla6 1 7 ltla3 i.e? 1 8 i.xc7
and as so often in the Petroff the position is ltlxc7 1 9 lld2 l:lfd8 20 :lad1 llxd2 21 l:lxd2
more or less level but Black is a bit passive. (Panchenko-Tolstikh, Cheliabinsk 1993).
8 0-0 Now Black is no worse after 21 .-ltle6 22
8 ltlbd2 isn't intimidating: 8...0-0 9 ltln lObS b6.
cS!? 1 0 ltle3 i.h5 1 1 0-0 �c6 12 ltlf5 l:le8 1 3 1 6...Ad8
Jxc5 (1 3 c3 cxd4 1 4 cxd4 i. ffi 1 5 i.g5 h6 1 6
i.c3 ltld5 17 ltlg3 i.g4 1 8 h3 i.e6 equalised
fully in Strzelecki-Raetsky, Katowice 1 990)
1 3...dxc5 14 l:le1 i.ffi 1 5 l:lxe8 1i'xe8 1 6 i.g5
ttlc5 and Black has enough countcrplay.
93
Th e Petroff Defence
Or:
a) 5 'ife2 'ife7 6 c!De3 c!Df6 7 b3 c!Dc6 8
.i.b2 .J.e6 9 g3 0-0-0 10 .i.g2 d5 1 1 d4 (1 1
0-0 h5 12 h4 'ifd7 and Black has good coun
terplay on the kingside) t t ...1Wd7 1 2 tZkl2 g6
1 3 h3 .i.h6 14 0-0-0 l:r.he8 with even chances,
Murey-Harikrishna, Pardubice 2002.
b) 5 c!Dc3 shouldn't lead to an advantage:
bl) 5 ...c!Dxc3 6 bxc3 (6 dxc3 is not very
dangerous: 6 ....J.e7 7 .J.f4 0-0 8 1i'd2 lDc6 9 lbis move seems to be the strongest -
0-0-0 .i.e6 1 0 .i.d3 .i.f6 and Black can't be Black develops quickly and annoys White in
worse) 6...g6 (also fme is 6...lDd7 7 d4 d5 8 the process. Alternatively:
c!De3 �£6 9 .i.d3 .i.d6 1 0 0-0 0-0 1 1 1Vf3 a) After 6... c5?! White gains a plus with 7
Aes 1 2 l:r.b1 c6 1 3 c4 dxc4 14 .J.xc4 1Vc7 1 5 .i.bS+ llk6 8 0-0 a6 (8....te7 was met by 9
h3 b5 1 6 .i.d3 .i.e6 1 7 c4 bxc4 1 8 .i.xc4 c4! dxc4 10 d5 lDJ6 1 1 .i.a4 bS 12 dxc6 bxa4
llad8 and Black has equalised, Gallagher 13 llM15 with some advantage in Jansa
Delchev, Batumi 1999) 7 d4 .i.g7 8 .id3 0-0 Miralles, Paris 1 989) 9 .i.xc6+ bxc6 10 c4!
9 0-0 tZkl7 10 f4 �b6 1 1 �b6 axb6 1 2 f5 .i.e6 1 1 cxd5 cxd5 12 lDc3 lDf6 (1 2...lDxc3
lla51 and the inclusion of the queenside rook 13 bxc3 .i.d6 1 4 .ta3 leads to a rigid struc
into the struggle guarantees Black decent ture with better chances for White) 13 f4
chances in this complicated position. 1 1 cxd4 1 4 'ifxd4 l:r.c8 1 5 �a4 .i.e7 1 6 b3 0-0
c!De3?! i'> weaker after l l ...c!Da41 1 2 .i.d2 c5! 17 .i.b2 'ifa5 18 l:r.ad1, Sanakocv-Raetsky,
(12...lt)b2?! 1 3 9£3 �d3 14 cxd3 f5 and Voronezh 1 987.
White had a slight edge in Ovetchkin b) 6...c6 7 tZk12 .id6 should be playable
Motylev, Russia 2004) 13 dxc5 lDxc3 14 'iff3 too, but Black was unlucky in the following
dxc5, when we rhink Black has the better game: 8 .i.d3 f5 9 0-0 0-0 1 0 f4 (10 c4 f4 1 1
chances. lDc2 lt)xd2 1 2 .txd2 n 1 3 g3 dxc4 14
b2) Even 5...lt)f6 is possible: 6 d4 .J.e7 7 .i.xc4+ Wh8 is complete!)' unclear, but
.i.d3 c!Dc6 8 d5 lDe5 9 lDxe5 dxe5 10 0-0 0-0 probably more unpleasant for White than for
1 1 llet lt)d7 12 .t£S lDb6 1 3 .i.xc8 llxc8 1 4 Black) 10....i.e6 1 1 g3 c5 12 c4!? lDxd2 (a
llxe5 .i.f6 1 5 lle3 l:r.e8 1 6 l:r.d3 .i.xc3 17 reasonable alternative is 1 2...cxd4 13 lDxd5
94
3 li)xe5: Fourth and Fifth Move A lternatives
llX5 14 �f3 lilid3 1 5 ..xd3 �c6 with a ll\d6 10 .ia4 ll\d7 Black's chances are no
complex position offering level chances) 1 3 worse.
.ixd2 dxc4 14 .ixc4 .ixc4 1 5 �xc4 .i.e7?1 9 l0xd5 •dB 1 0 Jle1 1rxd5 1 1 &3 •dB
(1 5...cxd4 was necessary, although 16 1Wf3 1 2 lhe4+
:t7 17 1i'd5 would have been a little uncom 1 2 .ig5 is inferior. White's main idea is to
fortable) 16 1i'b3 1i'd5 (maybe this is also not reach a murky position after 1 2...1ixg5 13
the best, but Black completely missed �xe4 1Vg6 14 �f6+ �d8 1 5 :C8+ �c7 16
White's next move) 17 �b6!! ..xb3 18 axb3 ll\dS+ �d7 17 'lre2 �6 18 :C1 (here Black
with a clear edge, Lukjanenko-Raetsky, Vo is mated after 1 8...b6?? 19 lld8+!). Instead
ronezh 2000. Black can play 12 ... .ie7 1 3 .ixe7 �xe7 14
7 .ibS+I? .D.xe4+ WfB when, according to Makarychev
A very ambitious, but not nece.'ISarily fool· and Smagin, the compensation is insufficient
ish. piece sacrifice. 7 ..e2 is less adventurous; and Black has slighdy better chances.
following 7....i:e6 8 c3 llX6 9 �d2 0-0-0 1 0 1 2....ie7 1 3 11re2
g3 (10 �xd5 .i.xd5 1 1 �xc4 ..g6 t 2 lDg3 13 .tg5 llX6 14 •e2 f6 1 5 d5 �5
.id6 gives Black excellent compensation for doesn't seem to cause Black the same kind of
his pawn) 1 0.. 1Wg6 (an improvement over
. problems.
10 ... �xd2 1 1 .ixd2 h5 1 2 .i.g2 1Wg6 1 3 0-0-0 1 3 ...�c6
�7 14 .i.f3 with a slight edge for White,
Smagin-Makarychev, Moscow 1 990) 1 1 .i.g2
f5 12 �f3 .id6 13 0-0 1Wf7 Black has good
counterchances.
7 c6 8 0-0!?
•••
95
The Petroff D efen ce
15 .i.xe7 lnxe7 16 .:.e5 0-0 a) 5 .i.c4+?! was the original idea, but it is
1 6...�£8 1 7 'ifxb5 'ifd7 18 ltac:1 would considered more or less refuted these days:
ensure White an initiative. S ... dSI (this is stronger than 5...�e6 6 .i.xe6+
1 7 llxe7 % -% �e6 7 0-0 Wd7 8 d4 WeB 9 c4 /l}bd7 1 0
/l}c3 - White has compensation here as
Black has no easy way to coordinate his
forces - 10...Wb8 1 1 f4 a6 12 a4 Wa7 1 3 b4
lieS 1 4 .i.eJ with a continued attack, Coch
rane-Mohishundcr, Calcutta 1 848) 6 �b3 (6
exd5 .i.d6 7 0-0 l:lfB 8 d4 �g8 and Black is
close to winning - Bisguier) 6...�e6 7 e5
/l}e4 8 d4 c5 9 'iVf3+ WeB and Black has a
large advantage, as stated in JrhadJ�IIIng as
far back in 1 8611
b) 5 �3 is more inventive: 5...c51?
(5...1We8 6 d4 transposes to 5 d4 '1Ve8 6 /l}c3)
6 .i.c4+ .i.e6 7 .i.xe6+ Wxe6 8 d4 Wfl 9
After 1 7...b4 1 8 /l}d1 ..xd4 1 9 c3 bxc3 20 dxcS �6 10 'iVe2 1Wd7 1 1 .i.c3 dxcS 1 2 f4
/l}xc3 .i.g4! (Makarychev/Smagin) the posi lle8 1 3 e5 lDg4 (1 3.../l}d4 14 1Wc4+ 'ifdi 1 5
tion is level. 'iVxe6+ llxe6 1 6 0-0-0 is unclear) 1 4 lldt
r-------. 'irES 1 5 0-0 h5 1 6 .i.ct /l}d4 17 'ifc4+ Wg6
Game 4 1 and the advantage was probably with Black
Vitolinsh-Anikaev in Topalov-Kramnik, unares 1 999.
Riga 1982 5 . . .g6
The most important alternative, 5...c5, can
1 e4 e5 2 lLlf3 lnf& 3 lnxe5 d& 4 lL!xf7 be seen in the next game. Other tries include:
The Cochrane Gambit is probably not en- a) 5.../l}xe4? is bad because of 6 'ifh5+ g6
tirely correct, but it is also not that easy to 7 'IVdS+ Wg7 8 ..xe4 �6 9 dS .i.fS 10 'ifa4
refute. Vitolinsh was a great expert on this /l}eS 1 1 .i.e2 .i.e7 1 2 0-0 when White's posi-
� � � ���
4 �xf7
.•. b) 5 ... �e7 6 /l}c3 l:te8 is a very natural
way to develop, although after 7 .i.c4+ .i.e6
8 .i.xe6+ Wxe6 9 f4 Wfl 10 e5 dxe5 1 1 fxe5
White has a genuine attack. Now Black
should play 1 t .../l}c6!, for example 12 0-0 (12
exf6?! .i.b4+ 13 Wf2 ..xf6+ does not work,
while the slow 1 2 �c3?! �g8 1 3 .i.f2 /l}d5
14 /l}e4 .i.h4 15 0-0 �xf2+ 1 6 llxf2 .J:lfB
gave Black a clear edge for Black in Petrik
Kujovic, Slovakia 2000) 1 2...'ifxd4+ 1 3 '1Vxd4
/l}xd4 1 4 exf6 .i.xf6 1 5 �g5 /l}xc2 1 6 l:lad 1
�3 17 .i.xe3 l:lxe3 1 8 lld7+ lle7 1 9 ltxe7+
Wxe7 20 /l}dS+ Wfl 21 /l}xc7 ltd8 and
White has some practical problems holding
5 d4 this endgame.
Or. c) 5...'ife8, taking prophylactic measures
96
3 IC.xe5: Fourth and Fifth Move A lterna tives
aWtinst .i.c4+, is a very recent idea: 6 liX:3 (6 dxeS 1i'xe5 1 1 Jle1 'IreS 1 2 .i.e3 1Wb4 1 3 a3
.i.d3 is punished by 6... c5! 7 dxcS dSI 8 liX3 '1Vg4 14 0) 10 h3 ltlh6 1 1 exd6 �g8 (White
dxt.-4 9 ltlxe4 .i.f5 10 0 .ixcS 1 1 1i'e2 .ixe4 wins after 1 1 ...cxd6? t 2 llet '1Vc6 13 .ie4) 12
1 2 .i.xe4 1i'a4, when Black is clt.-arly better) dxc7 (1 2 ltld5!? cxd6 1 3 ltlc7 1lc6 1 4 ltlxa8
6 ... d5 (6 ... c5 also looks fine) 7 eS .i.b4 b6 1 5 ltlxb6 axb6 1 6 c3 with a slight edge is
also pos!>ible) 12...ltlc6 1 3 d5 ltle5 (Vitolinsh
Domuls, Riga 1983). Now after 1 4 .ixh6!?
.i.xh6 1 5 d6 Wf7 1 6 .i.e4 we're leaning to
wards giving White a clear edge. This of
course is not all due to 7 .i.d3, but it does
confirm that it's a viable option for White.
7 . . .i.e& 8 d5 i.cB
Here we assess 8 ... .td71? 9 0-0 bS 1 0 .i.d3
.i.g7 as a bit better for Black.
9 0-0 i.g7 1 0 Ae1 llfB
and now:
c 1) 8 .i.e2 ltle4 9 0-0 .i.xc3 1 0 bxc3 ltlxc3
(J O_.'tlc6 1 1 c4!? also gives White play) 1 1
.ihS+ g6 1 2 'IVO+ �g8 1 3 1i'xc3 gxhS 14
:1b1 and the position was a complete mess in
11crnandez Romero-Andres, Seville 1999.
c2) 8 11'0 �g8 9 .id3 ltle4!? 1 0 0-0 (this
is stronger than 1 0 .ixe4 dxe4 1 1 'lfxe4
'iVc6! with an advantage for Black - Re
inderrnan) 1 0....i.xc3 1 1 bxc3 ltlc6 12 .ixe4
dxc4 1 3 1i'xe4 1i'g6 with a complex position, 1 1 e5
Ucrgcz-Grimbcrg, Clichy 1 998. 1 1 f4 regains the piece but then the attack
6 lbc3 11t'e8 dissipates after 1 l ...�g8 1 2 eS dxeS 1 3 fxeS
Alternatively: ltlg4 14 d6+ �h8 1 5 e6 ltles 16 d7 (1 6 e7
a) 6 ...'j;g7 7 f4 We8 8 cS looks very dan ltlxc4 17 cxfB1i'+ 1i'xf8 18 dxc7 ltlc6 is very
�crous for Black, for example 8...dxe5 9 fxeS unclear play but we do not think Black
.li.b4 10 .i.c4!? .i.xc3+ 1 1 bxc3 with the idea should fear this) 1 6 ...ltlbxd7 1 7 exd7 Wxd7
l l ....i.g4? 12 0-011 .i.xd1 13 exf6+ �fB 1 4 1 8 1i'xd7 .i.xd7 with a drawish endgame,
.li.h6 mate! Popov-Grodzensky, correspondence 1 983.
b) 6....i.g7 7 .i.c4+ .i.e6 8 .i.xell+ �e6 9 1 1 ...dxe5 1 2 d6+ i.e& 1 3 llxe5 .bc4
f4 cllfl 1 0 e5 lieS 1 1 0-0 was played in Vito 1 4 1lxe8 llxeB 1 5 dxc7 lila6
linsh-Anikaev, Frunze 1 979. Now after After 15 ...ltlc6 16 .i.gS llac8 17 1i'O
l l ...ltlfd7 12 f5 dxeS 13 fxg6+ �g8 1 4 Wo llxc7 18 W£4 li:)eS 19 li:)e4 llc6 20 l:ld1
White would have a strong attack. White manages to generate threats.
7 .i.c4 + 1 6 .*.f4 llecB?I
7 .td3!? is another possibility, for example 1\ sad decision - Anikaev was probably
7 ... .i.g7 8 0-0 llfB 9 eS ltlg4 (the pawn afraid of ghosts such as ltlb5-d6. It would
w-:tbbing 9 ... dxe5? leads to a disaster after 10 have been better to leave a rook on e8, e.g.
97
Th e Petroff Defence
ltX:eS 26 •e5 h5
Also strong is 26...b6!? 27 h5 :c5 28 1Ve4 6 11fa5+
.••
.1Lc8 29 h6 lDf6 30 1Vf3 lbges with clear After 6...c!L!c6 7 i.c4+ we have two possi
advantage for Black. bilities:
27 g4!? hxg4 28 h5 lDxh5 a) 7...d5 8 .i.xdS+ i.d) (not the only op-
28...gxh5?? 29 I?Jf5 would revt:rsc the tion; 8...1?Jxd5!? 9 cxd5 lL!b4 looks very
trend! strong for Black) 9 i.xt:6+ ci>xc6 10 'ife2
98
3 l0xe5: Fourth a nd Fifth Move A lterna tives
1Wa5+ 1 1 ltlc3 .i.xcS 12 0-0 (12 1rc4+ 9ile7 was close to losing in Short-Shirov, Dubai
U 0-0 llad8 is unclear) 1 2...�f7 1 3 11'c4+ (rapid) 2002.
b2) 9 ... d5 is also interesting:
b21) 10 e5?! � 1 1 1i'g4+ �f7 12 'l'f5+
�e8 and we do not think there is enough
compensation, for example 13 �c3 llXI4 1 4
Wh3 ..d7 1 5 e6 'iltxe6 1 6 ..d3 �xc3 1 7
bxc3 ltle2+ 1 8 �h 1 �ct 1 9 l:axct � f7 20
llcct 'irc6 with a clear edge for Black, Vi[O
linsh-Dautov, Minsk 1 988.
b22) 10 ltlc3!? d4 1 1 ltla4 (1 1 1re2?1 is
awarded a '!' by Forintos and Haag, but after
1 t ...dxc3! 1 2 Wc4+ �e7 n es 1Wd4! 1 4 cxf6+
*xf6 White has no compensation) 1 1 ...�f7
with chances for both sides.
and now:
at) 1 3...�g6?1 t4 ltld5 .i.d6 1 5 .i.f4 llad8
1 6 llad1 11'c5 1 7 11'b3 �xe4? (a mistake,
although Black was already in ttnuble) 1 8
.tcJ Was 1 9 •xb7 :Cs 20 b41? (20 1i'd7!
nlso win.<�) 20...�xb4 21 �f4+ .i.xf4 22
11fxc4+ 1Wf5 23 9xb4 and White was win
ning. Vitolinsh-Raetsky, Naberczhnye Chelny
1 9R8.
a2) 1 3 ..'�f8 14 ltlclS ltleS 1 5 Wb3 ltlxc4
.
99
The Petroff Defence
1 9 f6
'This leads nowhere. White can struggle o�
a while with 1 9 lbds llxd5 20 exdS ..xdS+;
21 'iff3 Wxf3+ 22 llxf3 llxg7, or t 91ih �
�g7 20 f6+ �h8. Although Black is dearl�
better, the g:1mc is not totally decided. j
19...dxc3 20 1i'e2 1i'c5 I
14 f5+?1 Also strong is 20...cxb2 21 llae1 'ifc5
:;j
This eliminates White's chances to invade Wxd ..xeJ 23 llxe3 lld61 24 eS lle6 25j
100
3 lllx e5: Fourth and Fifth Mo ve Alternatives
tlb3 b6, when Black is a knight up for noth 'ii'd2 'ifb6 181i'h6 1-0 'Iimofuev-Phoobalan,
ing. Goa 2002.
21 l:tf5 cxb2 22 l:taf1 � 23 1ih5+ 4 1We2 1We7 5 1fxe4 d6 6 d4 dxe5 7 dxe5
'tte6 24 1ih3 Wi'xe4+ 25 l:t5f3+ Wd6 26 7 1i'xeS?! is a weaker choice: 7.....xe5+ 8
*g2 �d4 27 l:txe3 1Wxe3 28 1Wxb7 l:tbS dxe5 .i.fS 9 c3 (91£lc31? .txc2 10 lDdS �d7
29 f7 1Wf3+ 0-1 I 1 .i.f4 l£lc6 12 .tbS l:ld8 13 llct is still
White resigned because of 30l:lf3 btW' 31 probably better for White) 9...1£Jd7 10 f4 (1 0
�1l:lbt32Wg2l:lg7. .tf4 0-0-0 11 1£Jd2 I£Jxe5 12 �xeS l:le8 and
Black is at least equal - Yusupov) t0...0-0-0
Game 43 1 1 .i.e3 f61? t2 I£Jd2 (12 .i.xa 7 fxe5 13 fxe5
Kholmov-Belousov I£Jxe5 14 .i.d4 :C8 15 �d2 g6 gives Black
Gorky 1974 serious threats) 12... fxe5 13 fxe5 I£Jxe5 14
0-0-0 I£Jct3+ 15 .i.xd3 :Xd3 and Black's posi
1 e4 e5 2 �3 �6 3 lllxe5 lll xe4?1 tion is preferable, Kos-Potapov, Ceske Bude
jovice 19tJS.
7 ...lllc6
.ixf6 gxf6 1S ll'le4 .i.e6 16 lbg3 WaS? 17 After 9...1i'b4?! 10 .i.c4 �£laS It .i.d3
101
The Petroff Defence
Whi et is a pawn up for nothin g. es sary . Afte r 1 3 Wf4 <l.>b 8 1 4 lta dt White has
a b ig advanta ge. StiU , it is very diffic ult to
defe nd again st suc h a sac rifice . Th is, alo ng
wi th the ae sthe tic be auty of the idea,e nc our
age d K holm ov take the r isk.
1 3...�7 14 1ta4 l0c6 15 •as+ �7
1 6 :ad 1 + We& 17 1txb7 1te6 1 8 Ade1
.te7
1 8..1lb
. 8 lose s to 19 .xc 7 llc 8 20 .g3
.g6 21 i.c S+ .i.e? 22 11 a3 1lg5 23 lle 3
f oU owe dby lif e 1.
1 9 .tc5 Ab8
19...1i d7? is me t wi thbye el gant 20 .xc 6!
11V xc 6 211l xe 7+ �f8 221l xc 7+, winnin g.
10 0-0 20 1ta6 ..d5?1
A sim pler way to play is to .i.f4 gS 1 1 Th is os
l esby force. A fter 20.. .Wf6 21 11V a4
i. g3 ( 1 1 e 6 fxe 6 1 2 i.e 3 i. g7 is unc el ar) �f8 221lxe 7 �xe 7 23 :C 1 White hase very
1 1 ...i. g7 1 20-00 - h5 1 3 h4! with ac el ar ad c hance of winnin g,b ut the game is stiU on.
vantage. 21 1ta3 1tg5 22 .:e3 Wf8 23 .txe7+
10.. .lt)xe5 rtJxe7 24 f4 1tf6 25 lZ.fe1 :ea
10...a6 1 1 i.c 4 lilie S 1 2 /lhls 11Ve 6 1 3
.i.gS 6 f 1 4 .i. f4 give s White good a ta t ckin g
c hance s, while 1 1 i. xc 6i. xc 6 12 Wf5+ '1Vd7
1 3 Wf4 should be be tter for White too.
11 .te3
1 1 .i.xd7+ �xd7 12 '1Va4 a6 13 �S We S
14 WaS also pro vide s White wi th an edge.
1 1 ....txb5
Or 1 1 ...a6 12 /lhls ( 12 .i.xa6 .i.c 6 13
Wf5+ .i.d7 14 .e4 i.c 6 with a re pe titi on
Kh olm ov) 12...We 6 1 3 �f4 'lieS 14 .i.xd7+
�xd7 1S W£5 and White retain s the i nitia
ti ve .
1 2lilllb5 &6?1 26 :Xe71
Missing ac hance . A re f r 12...a6 13 �4 g6 F orcin g a winn inge ndgame .
1 41l ef 1 i. g7 1 Sila d1 llhe8 B lac k would be 26 ....xe7 27 :Xe7 Axe7 28 g4 f5 29
aU ri ght. •c5 g6 30 gxf5 gxf5 31 1txf5+ Af7 32
1 3lillla7+ !7 1tc5+ �8 33 Wg2 h5 34 �3 Ah6 35
Th is et m ptin g sacrifice is not str ci tly nee- 1tg5+ Wh7 36 a4 1-0
102
3 l0xe 5 : Fourth and Fifth Mo ve Alterna tives
Summary
S ll:lc3 has been underestimated by theory, and Black should act carefully here. In Game 33
Alckhinc played fantastically to defeat Rabinovich, but the pawn sacrifice with 5...d5 is dubi
ous. 5...ll:lf6 is quite passive, but Black appears to be on the verge of equalLo;ing here. The more
popular and sharper S...ll:lxc3 6 dxc3 �c7 normally results in opposite side castling and a fight
with mutual chances.
The easiest way for Black to cqualL<ie against 5 ...e2 is with S.....e7 6 d3ll:lf6 7 �g5 ..xc2+
8 J..xe2 ll:lbd7 and 9...h6, preparing the simple ...g7-g6 or the extended fianchetto (...g7-g5).
The continuation 5 c4 is more challenging than 5 d3, but Black's play in Game 37 demon
Mrates that it's also fairly harmless. The modern 5 J..d3 is pretty safe for Black. However, the
pc 1sitions are quite untypical and have not been studied very much.
The sacrifice 3ll:lxc5 d6 4ll:lxf7 deserves consideration because it has been shown dmt it is
not easy for Black despite the extra piece. 5 d4 ..e8, in conjunction with either ... g7-g6 or ...d6-
d5, seems to be promising for Black, while 5...c5 and in particular 6 dxc5 ltk6 7 J.. c4 ..tc6 8
J..xe6+ Wxe6 has not lost its favourable reputation. Of course, the presence of the black king
in the centre puts higher demands on Black but, objectively speaking, Black's chances are pref
erable in the Cochrane Gambit.
After 3...ll:lxe4 4 '1Ve2 '1Ve7 5 ..xc4 d6 6 d4 dxcS 7 dxe5llk6 White has better prospects.
lie has a pleasant choice between 8llk3 and 8 ..ib5.
4 ttlt3
4llk4 - Game 40
4ll:lxf7 Wxf7 5 d4 (D)
5... c5- Gatne42; 5... g6- Game41
6 ttlxe4 5 tDc3 (D)
.••
5 ...ttlxc3 - Game34
5...d5- Ga111e 33
5 d4 5 ttlc3 7 �g 5
103
CHAPTER SIX I
3 d4: The Main Line
104
3 d4: The Main L ine
days of 5...�7. Black has easy play after 8...9114 transposes to 7 ...'ifh4 8 c4 .i.d6-
6...li)xe5 7 dxeS �5. see the notes to Game 49, while tL0-0 9
Finally, the unexpected sacrifice 6 li)xf7!? cxdS %le8 10 li)c3 9114 11 g3 li)xc3 12 bxc3
105
The Petroff Defence
11'h3 13l%b1 favours White - Black has no 14 i.xc4+�hS 15 i.gs •es t61Wh41ooks a
easy way to get his pawn back. bit better for White) 14 l%bl!? 1Wxd4 15
9 cxd5 llxb7 dxc4 16 .i.xc4+ 9xc4. Here the play
Another option is 9 ltX3 and nuw: ers agreed a draw in Velicka-Rabiega, Berlin,
a) 9...0-0 10 'iihs lC!f6 11 -..,4 dxc4 12 1999. After 17l%xd7 1t'c6 White is forced to
.i.xc4 :Cs 13 .i.gS h6 14 .i..xf6 1Wxf6 t 5 play the adventurous 1S llxg7+1 Wxg7 19
1Wxf6 gxf6 is ftnc for Black. After 16 llfel i.b2+ i.eS (19...WgS 20 1Wg5+ Wfl 21 1Wf6+
.i.e6 17 lbc4 ll..c7 HI .L.c6 fxe6 19 lbcs WeB 22 We6+ W<.l8 23l%dt looks very dan
.ll.xc5 20 dxc5 l%adR 21l%ad1 Wfl the play gerous for Black) 20 i.xe5+ WgS 21 l%<11
ers agreed a draw in Sveshnikov when he has substantial compensation for
Mikhalchishin, Kuibyshev 1986. the exchange.
b) 9...lbxc3 10 bxc3 dxc4?! (Black could 1 3.. :ec7 1 4 i.d2 :aea
still transpose to the main game with 10 ...0-0)
11 l%et+ ll..c7 12 .i.:u .ie6 13 .L.e7 1Vxe7
14 .ixc4 0-0 15 .lieS 1Wa3 16 .i.xe6 fxt.-6 17
11Ve1 and White had some pressure, Shott
Hubner, Wijk aan Zce 1986.
9. .cxd5 10�3
.
1 5 -*.c2
Maybe White has a brighter future after 15
g3, for instance:
a) lS .. .:Xet+ 16 :Xet WaS 17 1r f3 'it'xa2
1Sll..f4 i.xf4 19l%e7l%f7 20l%xf7 Wal+ 21
Wg2 Wxfl 22 1Vxf4 when White has consid
The main alternative is 10 Whs - sec erable activity for his pawn, and Black needs
Games 46-47. to defend precisely to survive.
1 O . .l2�xc3 1 1 bxc3 0-0 1 2 Wh5 f5
. b) 15. g6 16 Wf3 l%e4!? (this compelling
..
Ths is the most solid move. The pawn idea crops up more than once) 17 :Cct bS
sacrifice with 12..g6 is considered in the next 18 a4 a6 19 .ih6 :Cs 20 .i.xe4 dxe4
game. (20... fxe4?! 21 1Wf6 .i.fB 22 .i.xfB .:.XfB 23
1 3lZe1 1Wxa6 bxa4 24 c4 and there is no compensa
Alternatively: tion for the exchange) 21 'lfe3 1Vc4 22 axbS
a) 13 11V f3 Wh8! is fine for Black, for ex ll..xbS 23 i.f4 .ifB 24 i.eS and it looks like
ample 14 ll..f4 ..c7 15 ll..xd6 1Wxd6 16 i.c2 White has a slight edge, Ponomariov
:CeS 17 :ret l%ac8 18 h4 g6 19 h5 Wg7 20 Moro?.evich, Moscow (rapid) 2002.
g3 ..c6 21l%xeS :XeS 22 Wg2 1Wd6 23l%h1 1 5 ... g6 1 6 1rf3lZe4! 1 7 g3
i.bS 24 i.b3 l%e4 with equal chances, 17 ll..xe4? fxe4 1S 1We2 .i.xh2+ 19 �h1
Galkin-Macicja, Istanbul 2003. l%5 (Golubev) gives Black a strong attack.
b) More interesting is 13 c41Wf6 (13...dxc4 1 7 ...-*.b5 1 8 i.f4
106
3 d4: The Main Line
1 H .i.xe4 fxe4 19 'iVg4 .i.d7 offers Black 36 l:l.4b3 'iVxu4 37 Jle2 WgS is probably
compensation, while 18 .i.h6 l:lf7 1 9 l:lnb I better, though Black has all the chances.
'iVc6 is equally unclear. 36 ...Wg5 37 lle2 Wf4 38 llb31Vxd4+ 39
18 ....txf4 19 gxf41Vd6 Wg2 1l'c4 40 llf2+
19 ...'itxf4 20 .i.xe4 fxe4 21 'itxf4 l:lxf4 is Or 40 llbe3 'iVct 4 1 Wf2 1Wht and Black
also fine for Black, but probably just a dmw. wins.
20 .txe4 40 �g5 41 llf7 d4 42 h4+ Wxh4 43
•••
White no longer has a choice. Even llxh7+ Wxg4 441lg3+ Wf5 0-1
though objectively the position is probably
equal, White's game is harder to play. Game 45
20...fxe4 21 1Vg3 llxf4 22 :Sb1 .td7 23 Rublevsky-Vallejo
f3 Ohrid 2001
23 h3 b6 24 c4 'iff6 25 cxd5 l:lf5 26 Wb8+
c.tr>g7 27 llb3 'tWxd4 28 llbe3 (Baklan) is an 1 e4 e5 2 ll'lf3 ll'!f& 3 d4 ll'lxe4 4 .td3 d5
interesting possibility for White, who has 5 ll'lxe5 ll'ld7 6 ll'lxd7 .i.xd7 7 0-0 .td& 8
some countcrplay of his own. c4 c& 9 cxd5 cxd5 10 ll'lc3 lL!xc3 1 1
23 ... b6 bxc3 0-0 1 21Vh5 g6!7 13 Wxd51Vc7
A weaker option is 23...exf3?1 24 'it>f2 b6
25 lZ.e3 .i.f5 26 l:te5 Wa3 27 'iVxf4 'ilfxa2+,
when Black has not completely e'lualised.
24 lle31Vf6 25 Af17
25 l:lb2 'irfB 26 llf2 1Wh6 is equal accord
ing to Golubev.
25 . . ..tg4!
This 'sacrifice' destroys White's position.
261Vxg4
26 h3 .i.xf3 27 �h2 llf5 28 a4 llgs 29
1i'h8+ Wg7 30 l:tf2 Wh6 is also discouraging
for White.
26...1lxg4+ 27 fxg4 1Vg5 28 llg3 b5 29
Af2 Wg7 30 Wg2 a5 31 llb2 b4 32 cxb4 141Vf3
axb4 33 h3 1Vc1 34 1lgb3 �h& 35 1lxb4 It seems that the text move is the most
1Vd1 36 Wf2?1 dangerous for Black, although it's not the
107
The Petroff Defence
only attractive option. In fact, White ha.., tried strongest is 16....i.eS! 17 f4 (17 dxc5? %lxc5
a few other moves here: 18 .i.f4l:l.xd5 19 .i.xc7 ll.xd3 favours Black
a) 14 h3 does not appear to be dangerous slighdy) 17....i.c6 18 ..c4 b5 19 Wb3 ..d71?
for Black: 14..ie6 1S Wf3 ...xc3 16 ll.bt 20 51? .i.g3 21 Wg1 with complex play,
(after 16 .i.h6 Black plays 16..llfd8 17 .i.gs LepeUetier-Marciano, Auxcrre 1996.
ll.e8 18 .i.f6 .i.fS 19 .i.xfS ..xf3 20 gxf3 gxfS c2) 15 h4 .i.e6 16 •o •xc3 17 •f6 .i.f8
with equal chances) 16.....xd4 (16....i.xa2 17 18 .i.xfB llxfB with a fmal split:
llxb7 ..xd4 18 .i.b2 ..f4 19 •xf4 .i.xf4 20 c21) 19 .i.e4 19....1:lad8 20 %lfd1 ..td5 21
ll.a1 .l:lfd8 21 .i.ft .i.e6 22 ll.axa7 lL'lc8 is ll.act 1fa3 22 hS ..d6 23 ..xd6 %lxd6 24
level according to Belikov) 17 .i.e4 We5 18 .i.xd5 llxd5 25 ll.c7 ll.xh5 26 d5 ILls 27
g3 .i.xh3 19 .i.b2 'fle7 20 ll.fe11lae8 21 .i.c2 ll.xb7 llhxd5 with a draw in Sutovsky
(or 21 bb7 1ixet+ 221lxe1l:l.xet+ 23 ..th2 foridman, Medellin 1996.
.i.fS 24 •c3 :Cs 25 f4 ll.e2+ 26 .i.g2 .i.eS c22) 19 llfdt WaS (19..llad8!? should be
27 fxeS .i.e4 28 e6 ll.xg2+ 29 Wh3 llxb2 30 fine; maybe White continues with 20 .i.e4, as
exf7+ ll.xf7 31 ..xb2 with a draw- Belikov) 20 .i.xg6!? hxg6 21 hS leads to a draw after
21.....gS 22 J..b3 (G.Guseinov-Belikov, 21...gxh5 22 'lfg5+ �h7 23 ..xh5+ Wg7 24
Alushta 2001). Now Bclikov gives the 1t'g5+ Wh7 25 ll.act �) 20 .i.e4 Wds 21
following line as the most natural way to end ..xd8 ll.axd8 22 d5 (White should have a
the game: 22...• 51 23 1t'c3 .i.cS 24 :XeS puU here) 22....i.f5 23 .i.xfS gxfS 24 d6 ll.d7
..xbt+ 2s Wh2 W'fl 26 Lf7+1 Wxf7 27 zs Ads ll.rds 26 :tad1 Wg7 21 ll1d3 f6 2s
:fS+! �e611 28 ll.e5+ Wf71 29 .l:lfS+! with %lg3+ 1/z-1/z liviakov-Yusupov, Groningcn
perpetual check - a very attractive variation. 1994.
b) 14 h4 .i.e6 1S 9f3 ..xc3 16 .i.h6 14 . .ixh2+
..
ll.fd8!? 17 .i.gS (17 9f6?! .i.fB 18 .i.xfB 14.....xc3 15 .i.h6 ll.fe8 16 .f6 J..fB 17
..xd4! and Black is better) 17...:d7 18 .l:lfd1 .i.xf8 %lxf8 18 .i.e4 is preferable for White as
.i.f8 (18.....xd4 19 .i.e4 •es 20 g3 ll.c8 also Black cannor free himself easily. Black has
seems line) 19 ll.act •as 20 .i.f6 a6 21 .i.e4 compensation for the exchange after
.i.d5 22 "lfg4 .i.e6 23 •f4 J..d5 24 ll.c5 l5...•xd41? 16 J..xfB .i.c6 17 'ffh3 .l:lxf8 but
Wxa2 25 .i.eS with compensation for the he still has much to prove.
pawn, Nedev-Urban, Elista 1998. 1 5 Wh1 .*. d& 16 c4
c) 14 .i.h6 ll.fe8
108
3 d4: The Main L ine
56 ...llc3 57 .i.b4 llc4 58 .i.d2 llc5+ 59 and it is White who has to prove C(JUality.
ot>f&?l b2) 11 .i.e3 ..g6 12 'li'xd5 .ic6 looks like
59 We4 .l:lhS 60 l:lxh5 gxh5 61 Wf3 would enough compensation for the pawn to us.
109
The Petroff Defence
After 13 11t'a5 b6 14 'lra6 Black is of course ..i.c4 �6 17llxt2ibxc4 18 �xc4 •c8 with
fine, but now the simplest solution is some compensation) 16 ..f3 ..i.c6 17 .i.c4
14 ...llk51 15 .ixg6 hxg6 16 .if4!? (otherwise (here White should consider 17 i.xe4!? ll'te4
it's perpetual check) 16...t'tlxa6 17 i.xd6 18 ..d1 .i.d6 19 lbc3 Wh4 20 h3 ltxe3 21
0-0-0 18 .tg3 ltxd4. fxc3 .i.xg21 22 ..g4 1Wxg4 23 hxg4 .t.xft 24
c) 11 ltlc3 (the most challenging) Jlxfl with an edge - Y.Gonza.lez) 17...�6
t t ...11t'xd4 12 1Wxd5 ..xdS 13 lll xdS ltlcs 18 .tdS .txdS 19 'ifxdS � 20 �2 'irxb2
(after 13... 5 14 .lf4 .lxf4 15 t'tlxf4 0-0-0 16 21 llab1 11t'c2 22 llfct 1Wd3 23 ll!f3 ltac8
ltfe1 .lc6 17 lladl g6 18 f3 1l!f6 Black was and a draw was agreed in De Ia Paz-Y. Gon
close to being level in Konguvel-Barua, San zalez, Santa Clara 2004.
gli 2000) 14 :C1+ llle6 15 .tgS 0-0 (Black is
not much worse after 15 ...h6 16 i.e3 .i.c6 17
lllf4 We7) 16 i.e7 .i.xe7 171llxe7+ Wh8 18
llad1 lL.dS 19 .le4 with a plus for White,
Svcshnikov-1. Zaitsev, Moscow 1991.
1 3...J.b4
Apart from 13...lbg5 (see the next game)
Black has a couple of enticing alternatives:
a) 13...ltc8?! is punished by 14 .txe41 Qess
strong is 14 ..i.M :Cs 15 li:kl2 .td7 16 1i'f3
1 1-.xdS 1i'h4 17 g3 ibxd2 18 .i.xd2 'lrxd4 19 i.c3
11 W k-ads to a draw after 11...g6 12 .g4 20 .xg4?1 120 1Wf6 ltxc3 21 .xc3 .i.c6
'irxdS .i.c6 13 �3 �xc3 14 bxc3 Lh2+ 15 22 f31 could give White a slight L-<ige; it is not
Wxh2 1Wh4+ 16 Wg1 .i.xg2 (lvanchuk). easy to tcU, but not too relevant either]
11 -*.c& 1 2 �5 g6
••• 20....i.xg4 with a draw, lvanchuk-Yusupov,
12.-1if6?1 13 llk3 ..xd4 14 .i.xe4 .t.xe4 Linares 1993) 14....i.xe4 15 llk3 lte8 (or
15 Ad1 .td3 16 'tth3 :t'd8 17 ltxd3 1Wxd3 t5.....i.f5 16 •o Wh4 17 g3 •xd4 18 Ad1
18 'lfxd3 .i.xh2+ 19 Wxh2 llxd3 20 i.e3 •c5 19 .i.e3 1Wc6 20 �5 with some advan
gives White good winning chances, while tage) 16 .i.e3 (also strong is 16 .i.g51? ltxc3
13...g6 141th6 ltfe8 I S dS! (Yusupov) is also 17 Wh4! llh3 18 .i.xd8 .t.xh2+ 19 Wh1
good for White. However, 12. ..i.b4 is play l:lxh4 20 .i.xh4 .td6 21 .i.g5 with good win
able, 13 .te3 :CS 14 a3 g6 t 51th3 transpos ning chances) 16....tf5 17 1Wf3 .i.b8 18 ltfe1
ing to the main game. bS 19 g4 (Iimman-Yusupov, Unarcs 1993)
1 3 trh3 and now following 19...b4 20 gxf5 bxc3 21
This move reaches the most critical posi bxc3 ltxc3 22 ltact llxcl 23 ltxc1 White
tion in the 7 ....td6 1ine. simply has an extra pawn.
t 31th6 Des 14 .i.e3 .tfB is an improve b) 13...1Wh6?! is probably not particularly
ment for Black: 15 1th3 .i.d7 (1 S...h5!? 16 good: 14W 1Wxd4 15 .i.xe4 ..i.xe4 16 11h4
1 10
3 d4: The Main Line
after 22..Wc6 23 b31le8. However, 21 lld8+ This seems to be the best option. Yusu
l:lxJK 22 'irxd8+ Wt7 23 'IVJ7+ We7 24 pov docs not believe in Black's position after
'IVliS+ ..e6 25 'ifxb7+ We7 26 ..d5+ 'ire6 27 18...i.xd4?! 19 i.xe4 i.xe4 20 lldl i.xe3 21
'IVnH would have given Palac a laq,re advan 'irxe3 Wb6 22 'irxb6 axb6 23 lld4, and we
I;�Jt:. follow our bruru all the way here.
1 9 .bf2 ..txd4 20 lld1
The strongest move. Aftt:t 20 et)d 1??
:Ct+ 21 i.ft .i.b5 22 tbcJ .i.xf2+ 23 �xf2
1f'f6+1 Black has a decisive attack, while 20
Whl?! ..i.xf2+ 21 �f2 ..d4+ 22 Wft .i.d7
23 11'f31le3 is also unappealing.
20 .i.xf2+ 21 Wxf2 1tb6+ 22 Wf1
.•.
111
The Petroff Defence
crazy) 18 'lf'd1? 1fh4 19 h3? (although 19 g3 '1Vxc4 30 JlfB+ llxf8 31 'ilfxf8+ ..g8 32
1th3 20 f3 h4 is great for Black anyway) 1Wf6+ ifg7 33 1l'd8+ with a draw.
t9. ..lng5 20 d5 �h3+! 21 gxh3 •xh3 22 b322) Black can also try 16....i.f4 17 .Lf4
lte1 .i.h2+ 2.1 Wh1 .i.xdS+ 24 f3 .i.f4+ 25 11'xd4 18 .i.e4 f5 (or 18...�xf4 19 'ilfxf4 .i.xe4
Wg1 �xf3 0-1 Lautier-Gelfand, fo'JDE 20llfel .i.b1! 21 'IVct �f5 22 11'd1 1Vxd1 23
World Championship, Las Vegas 1999. llexd1 LiB 24 f3 �g7 1/z-112 Bologan·
b2) 15 �3 h5 16 'ii'h3 �xd4 looks fine Onischuk, Biel 1999) 19 'iVd1 (19 .txf5 leads
for Black, for example 17 �4 �2+ 18 to a draw after 19...�xf4 20 llad1 WeS 21 g3
1 12
3 d4: The Main Lin e
1 13
The Petroff Defence
Macicja must have either missed or mis b1 ) 9 11'h4 .i.e6 (or 9 ...ll'lc4 10 'iVxd8+
evaluatl-d this sacrifice. llxd8 1 1 n- 1 1 llcl 0-0 1 2 .ixe4 dxe4 1 3
28 l:txf2 Wxd5+ 29 Wf4 1td4+ 30 ¢>95 llxe4 llfe8 1 4 llxc8+ llxe8 1 5 .i.e3 .if4 is
30 �11 :e3+ is winning for Black. equal- 1 t ...ll'lf6 12 .i.g5 c6 1 3ll'lc3 �f8 1 4
30 1txf21
••• llae1 h6 1 5 .i.xf6 gxf6 16ltld1 fj;}g7 and it is
No perpetual- Black is winning! not clear whether White has any advantage at
31 1tf4 all, 1iviakov-P.Nielsen, Bergen 2000) 10 ltlc3
White cannot escape, for example 31 �h4 .i.c7 1 1 lle1 (1 1 'irg3 0-0 1 2 ll'lb5 lC!cs 1 3
:e5 32 WeB+ �g7 33 'lfc3 f6 34 'lfc7+ �h6 .i.f4 c6 1 4ll'lc7lllxc7 1 5 .i.xc7 Wd7 is level
and there is no defence; or 31 �h6 'iVf6 32 - 1iviakov) 1 t.. 'it'd7
. 12 .i.ft a6 13 n ll'le4
1i'f4 :e5l!, winning instantly. 1 4 11h5 yfJ 1 5 'it'h6 .iffi 16 11'h4 .i.e7 17
31 1tc5+ 32 .tf5 1te7+ 33 �h6
•.• 11'h6 .if8 1/z-1/z Tiviakov-Lcko, Wijk aan Zee
Or 33 �g4 1i'e2+ and Black wins. 2001 .
33...f6 34 ..i.xg6 1tg7+ 0-1 b2) 91le1 + is a worthy alternative.
Game 48
Shirov-Kramnik
Be lgrade 1997
Karpov, Buenos Aires 2001 . 8. ..'it'h4 has also been tried, e.g. 9 g3 i!bxc3
b) 8...lnf6 and now: to bxc3 tlg4 (10 ...11'h3 1 1 llc 1+ �fB 12 'iVn
1 14
3 d4: The Main Lin e
l:6 13 .i.f4 Jl.e7 14 J:lab1 b6 15 c4 favours 1Wb3 J:lad8 1 6 .i.d2 'if g6 with compensation
White) 1 t J:lcl+ *d8 (11...*£8 12 .i.e2 9£5 - Mikhalchishin) 14....J:lae8 15 J:lxe8 .i.xe8 16
1.� J:lb1 J:lb8 14 c4 dxc4 15 .i.xc4 h5 16 h4 is lle1 .i.h5 17 9g.l (17 'ifxh5 .i.xf4 18 9xf5
also a bit better for White) 12 .i.e2 95 13 9xf5 19 .i.x£5 .id2 is drawing) 17..Jle8 18
Ab1 b6 14 c4 dxc4 15 .i.xc4 J:le8 16 .i.e3 J:.xe8+ .i.xc8 19 Jl.xd6 cxd6 20 11'0 g6 21
.i.c6 17 d5 .i.d7 18 .i.ft h6 t 9 c4 with a 9xd5 9e7 and Black has play for the pawn.
sli�ht plus for White, Kamsky-Karpov, Pavlovic-Mikhalchishin., l..enk 1999.
FIDE World Championship (Game 6), Elista 11 .•• b6
1996.
9 bxc3 0-0
A recent game in this line continued
IJ ...Jl.c6 10 llb1 :bs 11 9h5 c6 (1t...g6 12
1irh6 *d7 13 .i.gS 9£8 14 1Vh4 h6 15 .i.f6
.i.c7 16 llfe1 looks slightly better for White)
12 .i.g5 .i.c7 13 .i.xe7 9xc7 14 f4 'iff6 15
l:.bc1 *d8 16 g4 g6 17 9h6
1 2 'irf3
Again White has other options:
a) 12 c4 dxc4 13 Jl.xc4+ *h8 14 ..igS
9e8 15 'iVh4 1i'g6 16 .i.d3 J:lae8 17 c4 h6 18
.i.e3 9f7 was level in Palac-Fridman, Pula
1997.
b) 12 :C1 c6 (this is even better than
12...1i'f6 13 •o c6 14 Jl.f4 b5 15 11fg3 .i.xf4
and now: 16 1fxf4 J:lf7 17 *rt l:ta£8 18 J:lc2 ..d8 19
a) 17...*c7?! 18 f5 ..i.d7 19 fxg6 'ifd6 20 J:lbet � when Black is close to equality,
l:txf7 hxg6 21 'iff4 'if'xf4 22 l:txf4 and White Shirov-Anand, Groningen 1997) 13 .i.gS
cvcntunUy converted his pawn in Shirov ..c7 14 c4 .i.e8 15 9h3 dxc4 16 .i.xc4+ (16
Bologan, Sarajevo 2004. .i.x£5 .i.xh2+ 17 *h1 J:lx£5 18 9xf5 .i.d6
b) Shirov described the position after 17 gives Black sufficient compensation - Salov)
'irh6 as the most fascinating he had ever 16....i.f7 17 .i.e6 Le6 18 l:txe6 J:lae8 with
analysed. The foUowing variation is a sum level prospecrs, Shirov-Kramnik, 2nd match
mery of several page of his analysh.-: game, Cazorla 1998.
17 .. .i.xg4 18 51 *d7 19 h3 .i.h5 20 fxg6
. 1 2...c6 13 ..tf4 •c7 14 ..txd6 'irxd6 1 5
.if3! 21 J:le5! hxg6 22 J:.£5! 9xf5 23 .i.x£5+ Afe1 l:lae8
�-,rxfS 24 1i'f6 J:.bg8+ 25 *f2 .i.e4 26 1fxf7+ Black has completely equalised. In fact, it
�c8 27 9e6+ *c7 28 9e7+ �c8 29 h4 and is White who should be careful. For example,
White is a bit better. 16 c4 J:.xel+ 17 :Xe1 9b4 and White is
10 Wh5 f5 1 1 lZb1 alrt."lldy starting to encounter problems.
11 J:lc1 is interesting: 11...1i'f6 12 •o 1 6 AxeS :XeS 1 7 c4 •a31 18 h4?1
WhK 13 :b1 b6 14 ..if4 (or 14 'ifxdS .i.c6 15 After this move Black is probably slightly
1 15
The Petroff Defence
Game 49
Anand-lvanchuk
Linares 1993
and now:
cl) 8...'ifh3? looks natural but White can
play 9 f3 llld6 (9...lllf6?! is even worse after
10 llet+ .i.e6 11 :CS! gS 12 .i.f5 g4 13 .tgs
.i.e7 14lllc3!, when Black is in trouble, Pav
lovic-Konguvel, Bcnasque 1998) 10 :CI+
i&.c7 11 lleS .i.c6 12 g4 ..h4 13 i&.gS, win
ning the queen fur insufficient compensa
tion.
c2) R flf6! 9 .i.xe4 dxe4 10 lllc3 1fg6 11
..
This is the principal move. The alterna Other moves arc weaker: �
1 16
3 d4: The Main L in e
hxd dxc4 12 ltc1+ Wf8 13 i.xc4 i.g4 14 the attack i n Rublcvsky-Ngucn Anh Dung,
'it'll3 should favour White. Moscow 2001) 16 l:l£2 hS 17 l:lb2 h4 with
b) 8 ...lt�f6 can be met by 9 llkl21?, for reasonable countcrplay.
example 9.....xd4 10 .!DO 'irb6 11 cxdS il.e7 9...g5 10�3
(11....!Dxd5 12 o!Dgsl h6 13 .!Dxf7 Wxf7 14 Other options will be discussed in Games
'it'h5+ WgH 15 ..xd5+ i.e6 16 ..e4 is great 51-52.
for White) 12 l:lcl 'i'd6 (12....!Dxd5 13 .ic4 1 0....ig7
c6 14 .i.g5 gives White a very strong attack)
13 i.gS h6 14 i.h4 'ifxd5 15 l:lct i.c6 16
'it'c2 'ild8? (Z.Almasi-Cs.Horvath, Hungary
1997) and now 17 l:lxc6! bxc6 18 .i.xf6 gxf6
19 o!Dd4 is a nice win. 16.. .'1Ve6 is stronger,
hut Black's position is still unenviable.
9 c5
9 cxdS is less critical. Black continues
IJ...i.d6 10 g3 (10 h3? .i.xh3! is simply too
dnngerous!) 10... .!Dxg3! 11 fxg3 i.xg3 12 'iVc2
'it'xd4+ 13 Wg2 i.xh2 14 Wxh2 9h4+ 15
'wt>gl 'iVg3+ 16 'ilg2 Wxd3 with great coun
tcrplay for the piece.
9 o!Dc3 is also not to be feared. The best reply. The alternatives have not
been successful:
a) 10....!Df6?1 11 g3!? 'iVh3 (1t ...'iVxd4?1 12
c61 is a well-known trick) 12 'ilf3 ltlg4 13
Wg2 'i'xg2+ 14 Wxg2 h6 15 f41 .i.g7 (or
15...l:e8 '1 6 fxg5 hxg5 17 h3 o!De3+ 18 .i.xe3
:Xe3 19 i.f5 with a clear edge) 16 h3 .!Df6
17 fxg5 hxgS 18 g4 and White has a cll:ar
advantage, Tseshkovsky-Gagloshvili, Kras
nodar 1997.
b) 10...f5?! 11 .!Dxd5 and now:
b1) 11....!Dxf2l? 12 l:lx£2 'lfxd4 13 .i.xg5
(we think that White should play 13 c6!?
..xdS 14 cxd7+ l:lxd7 15 l:ld2 .i.b4 16 .i.c4
Black can pL1y the following: ..c5+ 17 Wh1 i.xd2 18 i.e6 with a clear
a) 9....!Df6 10 cS .i.e7 11 .ie3 .!Dg4 12 .if4 edge) 13 ...i.xc5 14 �3 (14 WOI? i.c6 15
looks good for White. i.xd8 l:lxd8 16 i.xfS+ Wb8 17 lltll 'it'x£2+
b) 9...i.d6 10 g3 .!Dxg3!? 11 fxg3 i.xg3 12 18 'ifx£2 i.x£2+ 19 Wx£2 i.xdS 20 .i.xh7
'irc.J2 'iVxd4+ 13 ..e3 i.xh2+ 14 Wxh2 'iVh4+ looks great for White) 14...f4 15 i.xf4 l:lhf8
15 c;t>g1 ..g4+ 16 W£2 d4 17 'lfg3 dxc3 18 16 Wft ..xe3 17 i.xc3 .i.xe3 18 ..e2 i.x£2+
bxc3 i.e6 is equal according to Yusupov and 19 �ht ltde8 20 ..c2 and White had decent
lliibner, but to f4 may be better for White. chances to win the endgame, Wetlberg
c) 9....!Dxc3 (the easiest) to bxc3 dxc4 1 1 Schneider, Torshavn, 1987. Even so, White
�xc4 i.d6 12 f4 f 6 13 .i.dS c6 14 g3 Wg415 should punish Black further with deviations
.if3 ..g61? (but not 15...'iVe6 16 'iVa4 a6 17 on move 13 or 14.
J:tbt hS 18 .i.d2 1t'f5 19 l:lb2 and White had b2) tt...i.c6 also does not work after 12
117
The Petro ff Defence
ltk3l:lxd4 (or 12...i.g7 13 g31i'h6 14 olllxe4 unclear) 20 1ib6 9e7 21 -'.&6+ Wa8 22'
fxe4 15 .i.xe4 l:lxd4 16 1i'f.3 and Black is in 9xc6+ Wb8 23 'ifh6+ Wa8 and now taking;
trouble) 13 .i.c3l:ld8 14 'Vc2 olllxcS 15 Lf5 the draw by perpetual check with 24 1i'c6+
.i.d6 16 g3 1fc4 17 .i.xgS l:ldfB 18 .i.xe6+ might make sense as after 24 Lc4 24...-.,7
o!ilxe6 19 .i.c3 and Black is simply a pawn the game could go either way.
down, Glck-Raetsky, ZeD am Ziller 1993. b) 13... 15 14 ..tg2 with a further branch:
1 1 g3 b1) 14....i.c6 15 .i.xc6 (15 d5 .i.xd5 16
This is the most critical line for Black. ..txdS c6 17 'ifa4l:lxdS 18 11xa7 f4 is unclear
Other options arc featured in the next game. according to Vladimirov - White has prob- ·
1 18
3 d4: The Main L in e
1 19
The Petroff Defence
1 6...c6 1 7 Wa4 looks good for White: game via 22...1lxJ4! 23 .i.xd4 1Wxd4 24 Wb3
1 7... h5?! 1 8 1i'xa7 llxd4 19 llel l lld7 :C7 25 .i.fl 1i'xc5, with chances fur both
(19...1ldd8 20 .i.a6! and winsQ 20 b4 1Wg4 21 sides.
b5 and White has a very strong attack. 22 .hd4?
...
Game 50
J.Howeii-Makarychev
Fnmze 1989
21 ...1rg4?1
Soon after this game Black's play was re
paired to some extent with 21 ...1Wd5 22
.i.xa6 .i.xd4 23 .i.xb7 .i.xc5! 24 .i.xc6 .dl +
25 Wg2 Wxa4 26 llxa4 llxe3 27 fxe3 lld2+
28 �f3 llxb2 (Hemandez-J.HoweU, Matan·
zas 1 993). This position should lx: within the
120
3 d4; Th e Main L ine
1 3 .llxe21
•.
An idea of Akopian's.
14 'it'xe2
and now it's White who has to make a de 14 .i.xe2 is strongly met by 14...�g41
ci:;ion:
b1) 1 3 fxe4?! (this reckless move hasn't
been tried) 13 ... fxe4 14 l:lxf8 (not 14 .i.c2?
J:lxfl+ 1 5 1rxf1 l:tf8 and the game is already
over) 1 4 ...l:txf8 1 5 .i.e3 exd3 16 1Vxd3 l:tf6
and Black has the advantage.
b2) After 1 3 'ii'e 1 Black should exchange
with 1 3 ...1Vxet. Instead 13 ...'ii'h5?1 proved
futile after 14 fxe4 dxe4 1 5 .i.c4 f4 1 6 d5
'it>b8 17 c6 .i.c8 1 8 cxb7 .i.g4 t 9 lbxf4!? gxf4
20 .i.xf4 and White was virtually winning,
Mannion-Kobese, Yerevan Olympiad 1996.
b3) 1 3 a4 l:tde8 with a final split:
b31) 14 g3 lllxg31? 1 5 hxg3 (15 lbxg3 and now:
..ixd4+ 16 �h 1 f4 1 7 lbe2 .i.h3 gives Black a a) 1 5 fxg4? ..i.xd4+ 1 6 �hl .i.e5 17 .i.f4
good attack) 1 5...'ii'h3 1 6 c6! (16 l:lf2?! .i.xd4! (even worse is 1 7 h3? 1lg3 18 �g1 'ifh2+ 1 9
17 lllxd4 1Vxg3+ 18 Wfl f4 is very bad for � f2 .i.g3+! 20 �e3 - 2 0 W f3 .i.h4! -
White) 16....i.xc6 17 l:tf2 .i.xd4 1 8 lbxd4 20...1le8+ 21 �d3 .i.bS+ and Black wins)
'irxg3+ 19 �fl 'ii'h3+ with a draw. 1 7 ....i.xf4 1 8 l:lxf4 gxf4 19 1rxd5 l:lcHI 20
b32) 1 4 .et 'ii'xet 1 5 Axel f41 16 fxe4 •d 1 .i.c6 21 1i'f1 :C3 22 :C1 1re7 23 .to
tlxc4 1 7 .i.c4 f3 18 .i.e3 fxe2 1 9 l:lxe2 c6 20 .i.xf3 24 gx£3 lle2 and Black had an edge in
dS cxd5 21 .i.xd5 h6 22 l:td2 l:td8 with an Sherzer-Halasz, Budapest 1990.
level ending, Sax-Salov, Brussels 1 988. b) 15 .i.f4 lbf2!? 16 .i.g3?! (16 J..xg5
1 2 f3 li)f6 1 3 .td2 lbh3+! 1 7 gxh3 1Vxg5+ 18 �h1 1Vf4 with
121
The Petroff Defence
unclear play is stronger) 16...llhd1 17 ..i.xh4 lle8. Although Black is better, the game is
..i.xd4+ 18 �ht lt)e3 (we prefer Black's posi still open.
tion after the stronger 18...lbxb2!? 1 9 ..i.xgS 21 .. Jlf8 22 ..txh7?1
l:le8) 19 ..i.xgS l:le8 20 ..i.xc:3 llxe3 21 l:lfe1 Now it is all over. 22 c6 bxc6 23 llee1
.i.xb2 22 ..i.fl .id4 with an unclear endgame, .ie6 was the last chance.
Movsesian-Raetsky, Pardubicc; 1 992. 22.....te6 23 Ae1 ..te3 24 g4 AhB 25
.i.f5 �dB 26 Le6 fxe& 27 ..txe6 Lh2
28 Ad1 Ah1 + 0-1
White resigned because of 29 �e2 l:lxd1
30 Wxd1 c6 31 gS ..i.xc5.
Game 51
Kasparov-lvanchuk
Debrecen 1992
1 4...c!bh51?
14...Wxd4+ 1 5 .te3 WeS 16 Wd2 is !>imply
great for White.
1 5 1if2 •xf2+ 1 6 �xf2
1 6 :Xt2 ltlf4 17 .i.xf4 gxf4 1 8 c6 .i.c6 1 9
cxb7+ Wb8 with unclear play (Yusupov)
is probably a better option.
1 6. . .1i�f4 1 7 ..txf4 gxf4 18 Afe1 .txd4+
1 9 �1 ..txb2
19 ... ..i.xc5 20 J:leS ..i.e6 21 J:lhS allows
counterplay.
20 Aab1 ..td4
1 0 ..te3
White has a couple of alternatives here:
a) 10 g3 Wh3 11 lbc3 t5!? 12 .ie2 l:Zg8 13
..i.f3 .ie8 14 a4 �6 15 lla3 1i'f6 16 b4 ..i.g7
17 lbe2 hS with chances to both sides,
Pilipovic-Nikcevic, Tivat 1995.
b) 10 lbd2 lbxd2 (also possible is
lO ... i.g7 11 ltlf3 �5 12 lbcs Wxd1 13
l:lxdt .i.e6 14 i.c2 f6 1S ltl.n t5 16 lbe5 f4
when Black is no worse) 11 i.xd2 llg8 12
.i.c3 (or 12 l:lct !? Wxd4 13 ..i.c3 1i'h4 14
..tf6 l:le8 15 f4 i.e7 16 c6!? with unclear
play, Pinkas-Kuczynski, Wroclaw 1987; one
21 Ae7?! continuation is 16...i.xc6 17 .it5+ Wb8 18
More resilient is 2 1 l:lect .ic3 22 l:lc2 llxc6! bxc6 19 1ib3+ �a8 20 i.d7 .icS+ 21
122
3 d4: The Main L in e
and now:
ct) n...Whs t4 lbbs Wb8 1 s 11'et !? g4 1 6
'ireS 'irxeS (Black should play 1 6....i.xbS 1 7
1i'xhS lbxhS 1 8 .i.xbS gx f3 1 9 gx f3 .i.e7 20
:ret ..if6 with a slighdy worse position) 1 7
dxeS lbhS 1 8 fxg4 .i.xg4 (Khait-Raetsky,
Upetsk 1 993) and now after 19 c61? bxc6 20
lbd4 Wb7 21 lbb3 Black's position looks
10 Ae8
... uncomfortable.
Black should probably look carefuUy at c2) 1 3...'1'h61? is stronger: 14 l:tet?! g4
rhc alternatives to this: (Black now has the initiative) 1 5 c6 .i.xc6 1 6
a) 10....i.g7 1 1 f3 lbf6 transposes to 10 f3, .i. f5+ Wb8 1 7 fxg4 .i.d6 1 8 •o 1i'g7 1 9
while I t ...lbxcS?! 1 2 dxcS .i.xb2 1 3 lbd2 l:te2?1 ( 19 .i.h4 is better, although Black's
.1Ka1 14 'lrxa1 is fantastic for White. position remains preferable after 1 9 .:d£B 20
..
b) 10...lbf6 is certainly a playable altcma gS lbd7) 1 9 ... hS 20 gxhS J:lxhS 21 .i.h3 l:tdh8
rive: 1 1 lbd2 J:lgs (the main point) 1 2 lbf3 22 Wf1 &4 and Black was close to winning
WhS 13 lDeS 11fxd1 1 4 :axd1 .i.e6 l S f4 in Werner-Raetsky, Cappelle Ia Grande 1 999.
gxf4 16 .i.xf4 (1 6 :xf4 .i.h6 1 7 :o .i.xe3+ Instead of 14 l:te1?1, White shouW play 14
i H :xe3 lbe4 1 9 :n with a slight edge,
, lbbS Wb8 1 S 11'd2 g4 16 11'aS .i.xbS 17
�hould be considered) 1 6 ...lDe4 17 c6 f6 1 8 'iVxbS (but not 1 7 .i.xbS?? gxf3 and Black
cxb7+ 'iPxb7 1 9 lbc4 f5 (l.Gurevich-Barua, has a winning attack) 17 ...c6 1 8 11'a5 .tg7
Hastings 1 993/94). Now White could with chances for both sides.
probably have been slighdy better after 20 1 1 .!i)d2 .i.g7 1 2 �ta -.s 1 3 lt)xg5
li)e3 .i.g7 21 .i.eS .i.xcS 22 dxeS .rigS 23 This leads to a slighdy dull position, but
J:lr4. one that is preferable for Black. 1 3 .i.xe4
c) lO .. f5 is also enticing - we see no fault
. :Xe4 (13 ...dxe4 1 4 lbxgS irg6 1 S dS h6 16
with this move. For example, 1 1 f3 lbf6 (not c6 is also reaUy messy) 1 4 lbxgS l:lg4 1 S f4
l l ...f4?! 12 fxe4 fxe3 1 3 g3 Wh3 1 4 cSI hS?I (IS lbm llxg2+1 16 Wxg2 .i.h3+ 1 7 Wh 1
1 5 c6! and Black was in deep trouble, .i.g4 and Black wins) 1 S...h6 1 6 lbf3 J:lg8 is
Emcste-Goldmane, Riga 1 989) 1 2 lDcJ (1 2 certainly very unclear.
c6 .i.xc6 13 ..ixf5+ i.d7 1 4 g3 'lfhs 1 S 13 •xd1
•••
i.xd7+ :xd7 1 6 lbc3 .i.g7 gives both sides 1 3....i.g4?! is strongly met by 1 4 f3 lbxgS
chances) 12 ... f4 1 3 .i.f2 1 S fxg4 1Wh6 1 6 .i.f5+ Wb8 1 7 .i.f4 and
123
The Petroff Defence
45 ...�6?1
This loses and should be avoided, though
4S....ia2 46 llc4 Cit>f8 47 lle3 is also nice for
White.
46 llh8 'ite5 47 .Z:c8 �d4 48 llxc7
llh6 49 l:ld7 llh2+ 50 �1 1 -0
Game 52
Lastin-Najer
Elista 2000
124
3 d4: The Main L in e
125
The Petroff Defence
21 .. .i.b7
21 ...1la8? would be met by 22 1Wa51 .i.b7
This makes no sense - Black probably 23 llact!! l:lxct+ 24 ..i.fl+ 'it>b8 25 'ii'c7+
overlooked the note to White's 20th move. <tla7 26 llc3 and White wins.
After 18 ....i.c6 1 9 .i.xb7 .i.a4 (even 22 .txb7 + �b7 23 111b4 + � 24 'lra5 +
1 9....ixb7 20 c6 Axel+ 21 llxel l:ld7 22 �7 25 M»5 + �7 26 '1ra5 + %-%
cxd7 �xd7 23 1Wc3 a6 is comfortable for
Black) 20 1i'b4 (20 ..xa4? /l)g4 and Black Gan1e 53
wins) 20 ..a5 21 'irb6 li)d7 22 1Wxa5 llxel+
. Anand-Hubner
23 l:txe1 Wxb7 24 11fxd8 dxc5 Black has the Dortmund 1992
advantage.
1 9 cxb& i.e& 1 e4 e5 2 ltlf3 ltlf& 3 d4 ltlxe4 4 J..d3 d5
5 lZ':ixe5 lbd7 6 ltlxd7 J..xd7 7 0-0 1Wh4 8
c4 0-0-0 9 c5 g61?
20 bxa7+
Much stronger is 20 1Wc3t lbg4 21 h4 tbe5
22 dxe5 dxe5 23 llact and Black is under This move keeps the g5-square vacant as a
attack. possible retreat square for the knight.
126
3 d4: The Main Lin e
1 2 i.e3
and now: White has alw tried 1 2 lL!xe4 dxe4 1 3
a l ) 1 2 a4 /.l)g5!? 1 3 :a3 h6! 1 4 J.c2 (1 4 .ixe4 i.h3 1 4 1Wb3 11fa6 (or 1 4...c6!? 1 5 d5
i.c3 �6 1 5 1i'd2 with unclear play is proba cxd5 16 J.xdS Ld5 17 "tlfxd5 :dB 18 "tlfb3
hly better) 14 ...llxe2! 1 5 1i'xe2 �6 1 6 1Wd1 9c6 19 f3 J.xfl 20 Wxfl :d 5 with unclear
li:)xd4 and Black had enonnous play for the play) 1 5 :et :xd4 16 .i.f4 J.e6 17 9f3 c6
exchange, I vanchuk-Rozentalis, Dcbrecen 1 8 .id6 l:.d2 19 Wf4 l:r.d4 and the position
1 992. was very messy, Sax-Skembris, Burgas 1992
n2) 1 2 .i.e3 J.h6 1 3 .i.xh6 9xh6 1 4 1Wct
and White might have a very slight edge.
b) 1 1 '*.e3 lLixc5 1 2 g3 1i'h3 13 .i.e2 (1 3
dxcS d4 14 lLld5 dxe3 1 5 fxe3 - 15 c6? exf2+
I (l llxf2 .i.d4 17 cxd7+ Ld7 is bad for
White - 1 5...-*.e5 1 6 Wc2 �b8 leads to an
unclear position) 13 ...h5 14 lL!xd5 J.a4! 1 5
b.l .D.xd5 1 6 bxa4 :es (but not l 6 ...:hd8?!
17 i.f3 llxd4 1 8 Wc21 with a clear edge for
White, J.Polgar-Skembris, Moscow 1 994) 17
'i'c2 lL!e4 18 :ret Wd7 and the position is
very complex.
1 1 ..f&
...
127
The Petroff Defence
128
3 d4: The Main Lin e
b) 8 c4 is very aggressive, but not clear at 'ife3 lld6 19 .xa7 l:txc6 20 .aS+ �c7 21
nil. Black plays 8...1Wxd4 llab 1 .C.b6 22 :Xb6 �b6 23 .C.b1+ �c5 24
1i'a3+ �d4 25 .e3+ �e5 26 i.xg6+ 1-0
Kotronias-Atalik, Pucarevo 1 987.
8 dxe4 9�3 0-0-0 10 lllxe4 1tg6
...
and now:
b1) 9 ltk3 �c5 1 0 Xlel+ i.e6?1 (1 0...�d8
iN necessary; after 1 1 i.e2 1Wxd1 1 2 l:txdt
dxc4 13 i.xc4 White has compensation, but 1 1 f3
nothing more) 1 1 .C.xe6+! �xe6 12 cxd5 i.c5 The alternative 1 1 /i)g5 is probably best
1 3 i.e3 •e5 1 4 dxe6 i.xe3 1 5 i.bS+ �f8 1 6 met by 1 1...£6, for example 1 2 �f3 i.h3!?
.f3 .xe6 17 fx.e3 c 6 1 8 .L4 �18 1 9 i.b3 (1 2... h5 13 c4 h4 14 l&1 h3 15 g3 i.g4 16 f3
'fkc7 20 .C.ft f6 21 llk2 and Wh_ite had a clear i.cS 1 7 i.e3 llhe8 1 8 i.f2 i.f5 1 9 d5 i.d6
nlgc in the game Tseshkovsky-Bareev, Kiev, 20 1i'd2 gives White a slight edge, S7jebcrt
1 1)R6. Raetsky, CappeUe Ia Grande 1 999) 1 3 �e1
b2) 9 cxd5 is probably less dangerous: i.c5 14 d l:the8 15 i.e3 i.d6 with compen
11...0-0-0 1 0 .c2?1 (10 i.xe4!? .xe4 1 1 �c3 sation for the pawn.
'fkh4 1 2 i.e3 i.d6 13 g3 with unclear play
was necessary) 10 ... �5 1 1 i.c4 i.d6 1 2
'fkc6 1 1 i.d2 g6 1 2 'iff3 f6 13 l:tfel 'lff7 1 4 Black ha.c; no path to absolute equality, for
c4 dxc4 1 5 i.e4 c6 1 6 d5 i.e8?1 (16...i. f5 1 7 example:
i.x£5+ gx f5 1 8 .xf5+ 'ird7 19 1i'xd7+ �xd7 a) I 1 ...f5 1 2 �f2 i.bS 13 .C.e1 .i.d6 14
211 .C.ab1 is better for White, but the game �h3 .C.dc8 1 5 i.f4 :Xel+ 1 6 1i'xe1 Xle8 17
WI 1uld still be undecided) 17 dxc6! l:txd2 1 8 'ifg3 1i'xg3 1 8 hxg3 is a bit better for White.
129
The Petroff Defence
b) 11 ...-*.5 12 c3 h5 (or 12...i.xe4 1 3 fxe4 lion for dte pawn) t 8 .!Dc3 'ifb4 19 l:la4 'ii'e7
...xe4 1 4 :Xf7 i.d6 1 5 .*.g5 lldf8 16 llxf8+ 20 d5 g5 21 .ig3 f5 22 gxf5 .tf7 23 ltc4 with
ltx£8 1 7 .d2 and we prefer White due to the a dose-to-winning advantage for White,
extra pawn) 13 'ifa4 'ii1b8 14 .*.f4 h4 1 5 llae1 A.Ivanov-Kochicv, Kosttoma 1 985.
h3 16 g4 i.d7 (Klovans-Ro?.entalis, USSR b3) 16 ...a5!? seems to be necessary, al
1 985). Now after 1 7 'ifc4 llc8 18 �5 i.d6 though after 1 7 'lrb5!? llxd4 1 8 1ixb6
19 .*.xd6 ..xg5 20 �e7 i.b5 21 .*.xg5 .*.xc4 l'lxdl+ 19 l:lxd1 cxb6 20 .i.e5 dte endgame is
22 llf2 .*.xa2 Black is worse but he has &JOOd more pleasant for White.
drawing chances in the endgame. 12 f6 1 3 ..th4 �
•••
1 2 ..tg5
12 .*.f4 is also strong after 1 2..h4 13 'ifd3
exd3 1 9 bxa7 �d7 20 a8'ii' lLa8 21 llxa8 Not 19 ...fxg4 20 .txh8 (20 l'lf6!? 1Whs 21
dxc2 22 llct i.b3 23 .*.eS c5 24 lL.3 c4 25 c5 .ieS 22 c61 is equally strong) 20...'ifxh8 21
l:la8 and the advantage is definitely widt llk4 .ixh2+ 22 �xh2 'ifeS+ 23 �g1 Wxe4
White) 1 8 gxf5 i.xf5 19 a6 b6 20 .i.g3 ..g6 24 Wd2 and White the advantage, and a sub
21 'ii'c4 lld7 22 d5 Le4 23 'ii'xc4 l'lh6 24 stantial one at dtat.
lla4 .i.cS+ 25 'ii1f1 and White was clearly 20 gxf5?
better in Raetsky-Sivets, correspondence. In For some reason White decides to assist
fact, following 2S...:C7? 26 d6! it was already Black's attack with his next two moves. Here
rime to resign. he should play 20 c51 .i.cS 21 .i.xe5 11fc3+ 22
b2) 16 ...a6?! 17 aS 'ii'xb2?! (17 ...'ii'c6 is bet l'lf2 'irxe5 23 'iffl fxg4 24 c6 bxc6 25 ...a6+
ter, even dtough after 1 8 ...c3 'ii'xc3 1 9 �b8 26 dxc6 .i.c8 and Black is only a bit
.!Dxc3 .i.b4 20 ltk4 Black has no compen.<�a- worse in this messy position.
130
3 d4: The Main L ine
20. . . c5!
131ack grabs his chance instantly. l11e key
Idea is 21 .i.f2 .f41 22 .i.g3 ..g5 23 lbe4
Wg3+ and the queen's dance wins a piece.
21l... .i.xh2+1? 21 Wxh2 .d6+ 22 Wh1 ll.xg2! 25 ..i.f6
h:;tds to a draw as 23 .i.g1 ?? Wg3! 24 ll.f3 After 25 .i.f2 ..g5 26 11t'f3 :lxc4 White is
l:.xgt+ 25 ..xgt 1i'xf3+ 26 ..g2 ..xf5 is finished.
clc;trly better for BL1ck. 25 ...1lg1 + 1 26 Wxg1
21 dxc67 Or 26 :xgt ..c4+.
A terrible mistake, and suddenly Black has 26 ..tc5+ 27 �1 •xc4 0-1
.••
131
The Pe tro ff Defence
and Black is struggling, possibly in vain) t 0 hxgS. Now 19 d5!? should en.or;ure some ad
'il'd2 0-0 t t li:lc3 c6 12 o!ik2 ...d7 13 lbg3 vantage. Black has to be careful, for example
li:lh5 1 4 .!fuh5 hh5 1 5 .i.f5 ...d8 (not 1 9..Jife8 20 lZxe8+ lZxe8 21 dxc6 .ixc6 22
1 5......xf5?! 16 :Xc7 lZfc8 17 lZact lZxc7 1 8 h41 1i:lb5 23 li:lxb5 .i.xb5 24 hxgS with a clear
:Xe7 with a clear edge, a s t 8...b6? t 9 llc5 edge for White.
...g6 20 llg5 is winning) 1 6 :C3 i.g6 17 10 c5 0-0 1 1 .i.f4 c& 1 2 b4
Lg6 fxg6 1 8 1lac1 lZf7 1 9 g3 with a slight
edge for White, Palac-Zaja, Pula 2000.
8 c4
8 lZe1 doesn't really work: 8...0-0 (8...1i:lf6
transposes to 7...lbf6 above) 9 .i.xe4 dxe4 10
D.x'--4 .tf6 1 1 :C t i.c6 1 2 .i.c3 ...d5 1 3 0
l:tfe8 1 4 �3 (14 c3?! i.M! 1 5 .i.f2 .i.xf2+
1 6 Wxf2 :Xet 17 ...xe1 :Cs 18 ...d1 .tb5
with a strong initiative) 14......d6 1 5 Wh1
ltad8 1 6 Wd2 i.xd4 1 7 .i.xd4 1i'xd4 1 8
:XeS+ .ixe8 19 ...xd4 lZxd4 and Black is
ever-so-slightly better (Skatchkov).
8 lLif6
•.•
132
3 d4: The Main L ine
37 'fre57
•••
Game 56
Sorokin-Raetsky
Kra.modar 1984
1 9...i.d8?!
19 ... f5 is stiU correct. 1 982, Vorone1.h State University, De
20 ltxe8+ 'lfxe8 21 aS a6 22 bxa6 bxa6 partment of Mathematics, 23 years before the
23 'lfd2 f5 24 ll:la4 .i.f6 25 lte1 '1Vd8 26 publication of J>ehvjf Defmce by Everyman
ll'lb6 Chess. Wasting no time during a physics
Or 26 ..tn ..tg7 27 l:tbl Vf6 28 l:tb4, lecture, the future author makes an impormm
when White is more comformble. discovery on his pocket chess computer (a
26 ...1ta7 27 ..if1 �g7 28 ltb1 h5 29 version of the computer software Fritz 0.03):
11rc3 'lfha 30 ltb4 �7 the Petroff Defence is stiU a]jvc!
Black should probably play 30... h41? 31 He has just found a beautiful combina
1,rxh4 Vxh4 32 g3 Vhs, although White does tional refutation of ECO'r critical assessment.
keep an edge. The bible of that time insists on '14 �12
31 'lfd2 'lfe8 32 lba4 '1Vd8 33 ltb6 'lfh8?! with a slight edge' as in the game Mortensen
33...Vc8 is a better option, although after Borik played at the t 980 Chess Olympiad,
14 1ib4 White has a strong pressure. but his intuition whispers 'it's not that sim
34 La6 ple' and 'Eureka!'...
133
The Petroff Defence
and now:
a) 9 WI? 1fd4 1 0 'irbS+ c6 1 1 ..xb7
ltc8 1 2 .ie3 11'xe5 t 3 .xa7 .ib4 1 4 'iVd4 ·
134
3 d4: The Msin Lin e
..xeS 1tb4 with a slight edge for White) sides, Krakops-Raetsky, Apolda 1994.
1 4... .i.d7 1 S lbf3 .i.c6 16 1i'c4+ Wt7 17 1 0 ....tb4+
..xt7+ �xt7 1 8 lbxe5+ .i.xeS 19 .i.xeS and 10...0-0-0 1 1 ..a5
White was a pawn up in Suetin-Bcx, Biel
1 99S.
9 1rxe5 ..d7
135
The Petroff Defence
here. Instead White should play 18 b3 .i.d3 Other moves played here include:
1 9 c4 ..d7, although Black retains excellent a) 16 :ret b6 1 7 •a6 .i.xh2+ 1 8 Wxh2
compensation. :xe3 1 9 Wgt 1le6 20 "lffl :ea 21 :xe3
b) 1 2 ..xg7 0-0-0 1 3 �2 was played in ..xc3+ 22 Wf2 Wd3 and Black was slightly
00-Khalifman, Sochi, 1 984. Now after better, Barcenilla-Ye Rongguang, Beijing
1 3 .. .'Vc6!? 1 4 f3 :hg8 1 5 1fxh7 :Xg2 Black 1 992.
has excellent play for the pawns. b) 1 6 .i.f2 b6 1 7 1fa6 Lh2+ 1 8 Wxh2
1 2...•c6 13 f3 1fh6+ 1 9 Wgl 1fxd2 looks good for Black,
bast..-d on 20 c4?! .i.xf3! when White is in
trouble.
18 Jie2 1 7 c4
.•
Game 57
Dolmatov-Mamedyarov
Moscow 2002
136
3 d4: The Main L in e
0-0 10 f4 f6!? 1 1 cxf6 l:xf6 12 c4 .le6 1 3 White hasn't quite got the most from his
CKdS 'ii'xdS 1 4 lbc3 1i'xd1 1 5 l:axdl c6 1 6 position.
:02 aS 1 7 b3 l:ff8 1 8 l: fd l l:ae8 lead to a2) 9 llb1 .lxd4?! (9. ..c!Dxe5, transposing
even chances in Wedberg-Rozenailis, Vasby to 6...�xe5, is better) 1 0 �xd7 J.xd7! 1 1
2000) :Xb7 0-0 ( 1 1 ....lc6? loses to 1 2 .la31 .lxb7
13 llel+ �d7 1 4 .lf5+ �c6 1 5 Wxd4 and
mate is imminent) 1 2 .lxh7+ �xh7 1 3 Wxd4
and White has a dear edge.
b) 6...�xe5 7 dxeS .lb4 (7...�4?! 8
.lxe4 dxe4 9 �5 'ild8 10 .lgS 'ilc.l7 1 1 e6
fxe6 12 �5+ g6 1 3 lbf6+ �f7 14 �4 gives
Black serious problems, while 7...lbc5?1 R
�xd5! �xd3+ 9 1i'xd3 J.e6 10 �f4 obvi
ously favours White) 8 0-0
and now:
a) IL.c6 (the solid choice) 9 f4 f51? 10 llk2
( 1 0 exf6?1 is weaker: 1 0...1Wxf6 1 1 f5 - 11
llc l+ .lc7 1 2 ..c2 �xd3 1 3 cxd3 �fl and
us White has no �5+ coming, Black is bet
ter - 1 1 ...�xd3 1 2 Wxd3 .le7 13 �2 0-0 1 4
ltlg3 'ilf7 1 5 .le3 b6 and Black was slightly
hettcr, Rozentalis-Turov, Montr'->al 2001)
I O .le7 11 .le3 0-0 with chances for both
•••
137
The Petroff Defence
b2) 8...lt�xc3 9 bxc3 .i.xc3 10 llb1 0-0 Or 10 llb1 ..d7 1 1 .i.gS 0-0-0 12 0-0 h6
(1 0...1t'e7 1 1 llb3 .i.xe5? - Sax gives 1 3 Le7 11'xe7 1 4 11'e2 (14 l:lb3 cS 1 5 l:lbS
1 t ....i.b4 12 f4 with unclear play - 1 2 lle1 c4 16 .i.fS llhe8 with equal) 1 4...1ic5 1 5
leaves Black in trouble; Sax-Nunn, Brussels 1id2 d4 1 6 cxd4 llxd4 17 .e3 llhd8 and
1985 concluded 12 ..0-0? 13 1rh5 f5 14 .i.f4 Black was no worse in Kremcnietsky-Pripis,
t -O) I t .i.xh7+ Wxh7 12 .d3+ Wg8 1 3 Moscow 1 977.
'lfxc.l d4 1 4 'iWg3 ..d7!? 1 5 c3 d3 16 lld1 10 g6 1 1 •t3 f5 1 2 exf6
•..
11'g4 17 ..xg4 (17 llxd3 .xg3 18 llxg3 .i.fS 1 2 llbt 'ireS 1 3 .i.e3 0-0 14 0-0 cS also
19 lla 1 llfd8 gives Black enough play for the gives Black decent counterplay.
pawn) 1 7 ....i.xg4 1 8 f3 .i.c8 19 llb3 cS is 1 2 -*.xf& 1 3 0-0 0-0 1 4 J.a3
.•.
b) 8....i.c5 9 0-0 .i.e6 (9...1ih4?! loses a fxg6 hxg6 with unclear play was a more chal
pawn to 10 .i.bS+ c6 11 1Wxd5) 10 D.bt .i.b6 lenging try.
1 1 Whs h6 12 Wh1 •e7 1 3 f4 J!fl 14 1t'e2 24 l:lxd7 •xd7 25 l:lxd5 •a4 26 f5
0-0-0 1 5 a4 •cs 16 llbS 11t'c6 (or 16. ..xd.. •••2 27 h3 l:lxf5 28 l:ld8+ Wg7 29
17 .i.d2 •c6 1 8 a5 .i.cS 19 l:lfb1 with an •xb7+ Wh& 30 •e4 ••1 +
attack) 1 7 aS!? a6 18 axb6 axbS 1 9 bxc7 with Or 30...1lxc5 31 •f4+ llg5 32 •f8+ 'it'hS
compensation for the exchange, Reefat 33 ..f3+ with perpetual check.
Vakhidov, Dhaka 2003. 31 Wh2 ••5 + 32 1rxe5 :XeS 33 l:lc8 a5
9 'Wh5 .i.e& 1 0 f4 34 Wg3 a4 35 Wt4 l:le1 36 :as l:lc1 %-%
138
3 d4: The Msin Line
and now:
ct) 7...�f7?! 8 'lfhS+ �f6 (after B �e6..•
loses to 13...'1fg5+ 14 �e2 'ifxct 15 .ixe6+ 'ifxd4 J:lxd4 1 5 .i.e3 .i.c5 with level chances)
Wc6 16 .i.xcl5+ �b6 - Yusupov) 13 ...lDc4+ 9 ...'1fh5 10 fxe4 ..tg4 1 1 'ife3 dxe4 t 2 lDxe4
1 4 �e1 (14 �d3 is met by 1 4...llk5+1! 15 0-0-0 with a further split:
clxc5 .i.f5+! 16 :Xf5 'lfe4+ and 81ack wins) c21) 13 lDxf6 gxf6 1 4 0-0 .i.d6 1 5 g.� (or
1 4...'ifh4+ 1 5 g3 (Smerclon-Solomon, Gold 15 h3 llhg8!? 16 'lfh6 - but not 1 6 hxg4
< :oast 1999) and now Black can play Wh2+ 17 �f2 'lfh4+ 1 8 �e2 J:lde8 1 9 h4
139
The Petroff Defence
:lxg4 20 J.£5+ �d8 21 .Lg4 Wxg4+ 22 .13+ Wc6 could still transpose to the game,
�d3 :xe}t 23 J.xc3 f5 and White has a lot though it looks risky with the king on f6) 9
of defending to do - 1 6..ixh3! 17 Wxh5 Wh5+ with a repetition.
:Xg2+ 1 8 �h 1 l:th2+ with a draw) 15 ....ZZ.he8
(15...Ade8 16 '1Vh6 would benefit White).
After 1 5 ...:he8 Black has enough play, for
example 1 6 Wh6 J.c5! 17 dxc5 (17 'Wxh5
.i.xd4+ 1 8 �g2 hh5 is equaQ 17 ...Wxc5+
1 8 �h 1 •d5+ and Black delivers perpetual
check.
c22) 1 3 0-0 is stronger: 1 3...lnd5 (after
13...tbxe4?! 14 .i.xe4 .i.d6 1 5 h3 l:[hf8 16
J.d2 White is simply a pawn ahead) 14 9g5
li)b4 1 5 h3 (15 tb£2 tbxd3 1 6 tbxd3 D.xd4
W"dS played in Zhao Zhong Yuan-Solomon,
140
3 d4: The Msin L ine
141
The Petroff Defence
Summary
In the main line Black has very much switched to 7 ....i.d6 recendy. However, the annotated
games prove that 7 ..11rh4 remains interesting and gives Black excellent counterplay after 8 c4
.
0-0-0 9 c5 g6, and especially after 9...g5. Nevertheless, the truth is that 7 ...'1fh4 leads to ex
tremely sharp positions that are not to everyone's taste.
As Game 54 shows, 6 ltlxd7 .i.xd7 7 0-0 ..f6 allows White to win a pawn after 8 .i.xe4
dxe4 9 llk3 1i'g6 10 ltlxe4; Black has some compen.ution, but no more than that. However, if
we insert the moves 7...'1fh4 8 g3 and now 8...'1rf61 9 .i.xe4 dxe4 1 0 ltlc3 'lrg6 t 1 lLlxe4 0-0-0,
Black has a full compensation due to the weakened light squares on the kingside.
The brave 6 ltlxf7 (Game 58) has not been played much rccendy. Black should accept the
'gift' with 6...'1Pxf7 and following 7 'lfhS+ he should move forward with 7 ...We6! - Black•s
chances arc preferable in the arising complications. On the other hand, White has a plus aftet
6.....e7 7 'lrc2! 'lrxf7 8 f3 ltlf6 9 lLld2 1i'h5 1 0 fxe4 .i.g4 1 1 ..e3. The assessment of the Zait
sev-Karpov game has changed and become more precise. At present the best line is considered
to be 6 ...'1rc7 7 ltlxh8 lLlc3+ 8 Wd2 ltlxd1 9 l:te1 ltlxf2 10 .i.xh7 ltle5 1 1 1lxc5 J.c6 1 2 .i.gS
1i'h4 l3 .i.f7+ �d8 14 :xe6 'lrg5+ 1 5 �e2 'lrxg2, which is 'a hell of a mess'.
10 Wbs 0--0 1 1 1i'xd5 .i.c6 12 1ih5 g6 1 3 11'h3: 1 3....tb4 - Gt����t 46; 1 3. lbg5 - Game47
..
5 .ltid7
.. 9 c5 9... cxd5
142
CHAPTER SEVEN I
3 d4: 5 .i.d6 and
. . .
1 e4 e5 2 �f3 tl)f6 3 d4 ltlxe4 4 i.d3 d5 the comer is the main question of this key
5 �xe5 position, one which is studied in Game 59.
In the previous chapter we looked at the 5 ...lbc6 attacks both White's centralised
main line with 1 e4 e5 2 lbf3 lbf6 3 d4 lbxc4 knight and the d4-pawn. l lowever, this is not
4 .id3 d5 5 lbxe5 lbd7. Now it is time to tJ1c best idea as White can simply swap tl1c
:;rudy other 5th move options for Black. knights to inflict doubled pawns on his op
Until quite recently the symmetrical re ponent. Now 6 lbxc6 bxc6 7 0-0 �e7 trans
sponse 5...�d6 (Games 59-62) was as popu poses to 5 ....i.e7 6 0-0 lbc6 7 lbxc6 bxc6 (sec
lar as 5 ..lLld7. After mutual castling White
. Game 63), while 6 li)xc6 bxc6 7 0-0 .i.d6
:macks the centre, but this can be done in transposes to 5...�d6 6 0-0 lbc6 7 lbxc6
more than one way. If White develops his bxc6 (sec Game 61). However, Bil!,ruer
lJUeen's knight for this purpose, he should demonstrated that White docs not have to
prefer 7 lbd2 (Game 61) to 7 lbc3 (Game castle; 7 'iWe2 creates problems for Black
62) because it avoids doubled pawns. along the e-file, and that's why 5 ..lbc6 is
.
143
The Petroff Defence
144
3 d4: 5 . . �d6 and Fifth Mo ve Alterna tives
.
wch6 20 IZ.xd4 �h7 21 lld7 and White has a should accept an inferior position with
preferable rook endgame. However, whether 1 6...bxc4.
this is enough to win is hard to tell. 14 i.f4
1 1 �xe4 �xc2 1 2 i.xd5 i.f5 14 .i.e4 �xa1 1 5 �3 is the alternative,
t2...lt�xa1 gives White a slight advantage but probably less critical (1 5 .tf4 simply
after 1 3 i.e4 (13 c6 lbc2 1 4 exf7+ llxf7 1 5 transposes): 15 ... £5 16 exf6 .i.h3 17 llet
lldl .t f5 1 6 .i.xf7+ *xf7 17 ltk3 c6 1eads to llae8 18 .L12 (1 8 .i.e3 IZ.xe4 19 �xe4 �2
immediate equality) 1 3..l[c8 14 l&3 llxc5 20 llct �xe3 21 fxe3 c6 22 �gS .tf5 23 f7+
1 5 .i.d2 llxe4 1 6 �xe4 �c2 1 7 llct i.f5 1 8 *h8 with equality, Tal-1imman, Reykjavik
f3 li)d4 1 9 llxc7 .lxe4 20 fxe4 b6 21 *f2. 1 987) 18..l[xe4 1 9 �xc4 �c2 20 :C1 li)d4
The endgame is uncomfortable for Black, 21 llxc7 �f3+ (21 ...gxf6?! 22 .i.e3 �2+ 23
lhough not necessarily impossible to defend. Whl looks better; 21 ..1lf71? 22 llxt7 �f3+
Note that after 21 ...�6 White retains the 23 c;i;>ht �xf7 24 fxg7 *xg7 25 .i.e3 b6 26
pressure with 22 lle7! *£8 23 .i.b4, when �12 �xd2 27 .i.xd2 aS 28 f3 a4 29 ¢>g1
23 ...�5 loses to 24 llc7. .i.t.-6 30 a3 '12-1/z Tai-Karpov, Milan 1 975 -
13 g41 Black will place his king on f7 and White can
never make any progress) 22 �h1 �xd2 23
llxg7+ c;i;>h8 24 lbgs .i.f51 (the most precise;
24.. .:Xf6 25 IZ.xh7+ Wg8 26 1Z.xh3 IZ.xf2 still
gives White chances - Gipslis) 25 �f7+ llxf7
26 llxf7 .i.e4+ 27 f3 .lxf3+ 28 c;i;>gt Wg8 29
lld7 lDc4 30 Wf2 .i.c6 3t llg7+ wm 32
llxh7 li)d6 33 We3 �e8 and the endgame is
drawn, Raetsky-Bclomestnykh, correspon
dence 1 982
14 ...�xa1 1 5 .i.e4
1 3 . i.xg4
..
end up with material losses. Instead Black �3 fxcS 1 8 .lxeS llae8 1 9 f4 gS 20 llxa1
145
The Petroff Defence
gxf4 21 .i.d4 a6 with compk-tc equality) 17 Oddly enough the bishop is best placed
cxf6 :Xf6 1 8 .i.d (Sveshnikov-Tischbierck, here, where it seemingly has little influence.
Budapest 1 988). Now Black could have con The reason for this is found in the foUowing
tinued 1 8 .. .:C:8 1 9 lL!c3 lL!b3 (1 9...bS 20 .i.g2 line: 1 8 .i.e6 g5! 1 9 ..LgS l:lae8 20 .i.c4 (or
llg6 21 �h1 b4 22 lL!e4 .i.fS 23 l0g3 :xg3 20 .i.h6 :xc6 21 .i.xf8 .i.h3 22 :C3 %Zg6+ 23
24 hxg3 lLlc2 25 i.xc6 looks pnmusing for llg3 �g8 24 .i.e7 f4 25 ltxg6+ hxg6 26 lL!a3
White - two bishops and a weakened black f3 27 .i.c5 b6 28 .i.c.l4 .i.£5 29 b4 ru 30
queenside) 20 axb3 a6 21 i.d3 .i.fS and lL!xc2 .i.xc2 '/z-'/z OU-Khalifman, Vilnius
Black's position is at worst marginally infe 1988) 20...b5 21 .i.ft f4 (Rm:entalis
rior; in fact we think it's equal. Ivanchuk, Minsk 1986). Mter 22 .i.f6+ llxf6
1 5...f5?1 23 exf6 .i.fS 24 f7 llf8 Black is no worse.
After this move we cannot find a route to 1 8 .. .llfd8 1 9 �d2
fuU equality. Probably Black should try This is the critical move. 1 9 f3 .i.hS 20
lS ...f61? 16 lL!c3 fx�:5, when his results have lLhl :U4 21 .i.e3 lZ.b4 22 lL!c4 (22 llxa1 f4
been encouraging: 23 .i.f2 llxb2 24 lL!c4 llc2 2S lL!d6 b6 is not
a) 1 7 .i.g3 :ad8 1 8 l:lxa1 .i.£3 1 9 .ixe5 clear) 22...l:la4 23 lL!a3 leads to a repetition,
J:td2 gives good compensation (Yusupov). Sax-Yusupov, Thessaloniki 1 984.
b) l7 .i.e3 .i.f3 1 8 l:lxa1 .i.x�:4 1 9 lL!xe4
b6 20 b4 aS 21 b5 :ad8 22 l:.ct ll£7 23 a4 h6
24 Wg2 J:td3 and Black had sufficient
counterplay in Kasparov-1imman, Paris
(rapid) 1 991 .
1 9 ...h6
1 9...llxd2?1 has an awful score. FoUowing
20 .i.xd2 :dB 21 .i.c3 :d i+ 22 llxd1 .i.xd l
White should play the foUowing plan: 23
.i.fl I (23 f4 lLlc2 24 Wf2 �PB 25 a4 aS 26
1 6 -td5+ .i.xa5 lL!d4 27 .i.ft .i.b3 1/z-1/.t Kasparov
Much weaker is 16 .i.xb7?! l:lab8 1 7 .i.dS+ Anand, Unares 1991) 23 ...g6 24 .i.c4 �g7 25
�h8 1 8 f3 (18 lL!a3?1 l:lfd8 19 .i.c4 J:[d4 20 b4! (25 �:6+ Wf8 26 .i.f6 'iPe8 27 e7 lLlc2 is
.i.e3 lL!c21 21 .i.xd4 lL!xd4 and Black's more unclear - Yusupov) 25...W 26 .i.b3 when it
active pieces mean that he enjoys an advan looks impossible for Black to improve his
tage in the endgame) 1 8. ..th3 19 lld1 Lb2
. position. Black seemed to lose th�: foUowing
20 e6 l:.g2+ 21 Wh 1 l:.c8 22 .i.c4 llg6 23 endgame without any real chance: 26 ...�f8
.i.g3 f4 24 e7 l:ld6 25 llxd6 cxd6 26 .i.xf4 27 .i.d2 We7 28 f3 �e8 2<J �f2 b6 30 .i.f4
.i.d7, when the responsibility of holding this cS 31 bxcS bxcS 32 c6 c4 33 i..xc4 lL!a3 34
endgame lies entirely with White. .i.dS .i.a4 3S .i.e3 a6 36 .i.c5 lL!c2 37 i..c4
1 6...�8 1 7 llc1 c6 18 ..tg2 aS 38 �g3 lLle 1 39 Wf4 i..c6 40 .i.e2 h5 41
146
3 d4: 5 . . . �d6 and Fifth Move Alterna tives
147
The Petroff Defence
b) 8 cxdS cxdS 9 'irc2 .IL:8 10 f3 (10 R.xe4 R.d7 13 .i.d2 lr.b8 14 l:tab1 .i.£6 15 �h1 h5
dxc4 1 1 .xe4 .f6 1 2 .i.£4 �c6 looks fine t6 �2 h4 17 h3 'l'c8 18 J..a6 'Wd8 t9 .i.d3
for Black) to ...�f6 1 1 .i.gS h6 1 2 .i.h4 �a6 ..c8 20 .i.a6 the game finished with a draw
1 3 a3 -*.xeS 14 dxeS lr.xeS t S R.f2 and White in Kaspamv-Yusupov, Horgen 1995. 1 8
has compensation for his pawn, although it is �f41? � 19 lr.fet is interesting, but we
difficult to see how Black should ever end up believe that Black is okay after the foUowing
being worse. line: 19...-*.gS!? 20 .i.a6 ..dB 21 �xe6 lr.xe6
c) 8 llXJI seems to give White the edge af 22 :Xe6 R.xe6 23 .i.xgS ..xgS 24 ..xc6
ter 8 ...�xc3 9 bxc3 and now: .i.xh3! 25 R.n ! .i.xg2+! 26 .i.xg2 ..5 27
c1) 9 ... R.e6 10 £4 .i.xeS 1 1 fxeS dxc4 1 2 l:lg1 h3 - the game will end in a draw.
-*.xh7+!? Wxh7 1 3 1Wh5+ �g8 t4 R.gs •as b) 10 l0d2!? looks stronger: 1 0.. ..i.f6 1 1
1 5 .J:r.£3 �7 1 6 lr.g3 looks very dangerous �xc4 dxe4 1 2 .i.xe4 .i.a6 (t2....i.xd4 1 3
for Black. .d3 �h8 14 .i.g51? .i.xf2+ 1 5 lr.xf2 .xgS
c2) 9...lbd7 1 0 £4!? (tO �xd7 R.xd7 trans 16 Lh7 looks slightly better for White) 1 3
poses to S...llkl7 6 �xd7 R.xd7 7 0-0 .i.d6 8 :C 1 .i.xd4 1 4 1i'c2 lr.eR 1 5 .i.e3 .i.xb2 1 6
c4 c6 9 W �xc3 10 bxc3 0-0) 10 ...�f6 1 1 .i.xh7+ �h8 (Sanche-.r.-Morgado, correspon
cS J..e7 t 2 5 J..d7 1 3 g4 gave White an at dence 1 978) and now after 1 7 •xb2 �xh7
tack in Korneev-Y.Hernandez, Mondariz 1 8 l:ad1 .e7 19 1i'c3 we think White has
1 997. some pressure.
c3) 9....i.xe5 10 dxeS dxc4 l t R.xc4 fle7
(1 t ...'Wxd1 1 2 lr.xd1 J.. 5 1 3 .L3 lr.eB 1 4 f4
is clearly better for White; it gets even better
after 14 ...�7? 1 5 e6! when White was win
ning in Maroczy-Marshall, Paris 1900) 1 2 a4
lr.d8 1 3 1Wh5 l:te8?! (13...g6 was sadly neces
sary although after 14 J..gS gxhS 1 5 .lxe7
:CB 1 6 J..d6 Black is in a bad state) 14 J..a3
'l'd7 1 5 l:adt '1'5 16 W'h4 .xeS 17 f4 'Wf6
(17 ...1i'e3+ 1 8 �h1 .i.e6 1 9 .i.d3 h6 20 aS!
and Black's queen is trapped) 18 .x£6 gxf6
t9 .J:r.£3 J..e6 20 lr.g3+ �h8 21 J..e7! hS 22
J..x f6+ t -0 Chigorin-Lebedev, Moscow 1900.
S lOxc& 1 0. . .f5
This is the right path to an opening advan Black seems to be unable to equalise here,
tage. Mter 8 cxdS �xd4 9 .i.xc4 (9 �c4 and it's not due to a lack of trying:
1Wh4 1 0 �xd6 t'Oxd6 t 1 �c3 .i.5 and Black a) 10....i.5 1 1 f3 �xc3 1 2 bxc3 .i.xd3 13
has equalised - Euwe) 9 ....ixe5 1 0 �3 .IL:8 'l'xd3 aS (13....i.g5?! 14 ..a6! .i.xct 1 5
1 1 lr.e1 eo5 12 •n g6 1 3 -*.£4 ..i.x£4 1 4 Let .d7 16 lr.fel lr.fe8 17 lr.xc8+ l:r.xe8 18
'ffx f4 �6 1 5 :C3 J..d7 1 6 lr.act 5 1 7 .LI3 h3 h5 19 .xa7 and Black did not have
..f6 t8 a4 lr.xe3 1 9 lr.xe3 lr.e8 Black is very enough for the pawn Shirov-Hiibner, Frnnk·
close to equality, Zelcic-Pavasovic, Nova furt [rapid[ 1996) 1 4 l:tb1 lr.e8 1 5 .i.d2 and
Gorica t 997. White is a bit better.
s. . .bxc& 9 c5 j.e7 1 0 &3 b) to....i.f6 1 1 .c2 (1 1 �xe41? dxe4 1 2
This seems to be the soundest way of .i.xe4 transposes to 10 �d2 .i. f6 l 1 �xe4
playing for White, but it is not the only way: dxc4 1 2 .i.xe4) t l ....i.xd4 1 2 �c4 dxe4 13
a) After tO f3 �gS t t �c3 :Cs t 2 •a4 R.xe4 W'h4 (1 3...R.a6?! 14 .J:r.dt with a plus;
148
3 d4: 5 . �d6 and Fifth Mo ve A lternatives
. .
�d5 f4?
Black should play 20 .:ae8 21 �7+ �h8
..
149
The Petroff Defence
The last chance was probably 24....i.xf5 25 0-0-0 cS with a very complex position in
lbxf5 wrg6 26 'itc2 'iig4 27 'ii'd1, when which White might have an edge) 9 wrxd3 c6
White retains a large advantage but Black can 10 0-0 .i.g4 11 lbg5 (the less obvious 11
struggle still. .i.d2 llXI7 12lbd4 � 13 h3 .i.e6 14lbxe6
25 hxg3 g6 26 i.xd7 1rd2+ 27 �h3 fxe6 is not easy to evaluate) 11...h6 12 'itd4
l:lxf1 28 i.xc6 l:laf8 29 i.g2 .i.e6 13lLlxe6 fxe6 14 c4 1fe7 15 .i.d2 c5 16
Also winning is 29 .lie7 l:te1 30 Axe 1 Wg4 d4 and Black has no obvious worries in
'itxe1 31 'itb3 Wg7 32 g4, but the game con this unclear position, Raetsky-Y.Aleksandrov,
tinuation looks safer despite the complex correspondence 1983.
lines. d) 6 c4
29.. Jl1 f2
Black also loses after 29...l:t1 fl 30 l:le8 gS
31 .i.d5 llf6 32l:.xf8+ llx£8 33lbf3, but not
32 lbm g4+ 33 Wxg4 l:lg6+ 34 Wh3 :h6+
35 lbh4 l:lxh4+! 36 gxh4 'itd3+ whc::n Black
escapes with a draw.
30 1rd7 g5 31 l:le7 1rc2 32 ..i..e4 l:lh2+
33 �Lh4+
Black's attack also runs out of steam after
33...'fle2+ 34lbf3 34...h5+ 35 Wxg5 We3+ 36
Wg6 ltg8+ 37 ll.g7 1fxe4+ 38 Wh6 'ife3+ 39
lDgs, when despite his extra exchange Black
is 'out of bullets'.
34 gxh4 'ife2+ 35 �g5 l:lg8+ 36 This allows Black to act quickly with
�6 1-0 6 ....i.xe5 7 dxeS lbc6 8 0-0 lbcS 9 cxdS
r------. 'itxdS 10 .i.bS 'ifxdl (weaker is 10....i.e6 11
Game6 1 li'xdS .i.xdS 12 lbc3 0-0-0 13 .i.e3 lbe4 14
Timoscenko-Yusupov lbxcl5l:.xd5 15 .i.xc6 bxc6, when White had
F'mn� 1979 an endgame edge in Lindoerfcr-Meijers,
Schwabisch Gmund 1998) 11 llxd1 .i.d7 12
1 e4 e5 2 ll:!f3 lDf6 3 d4 lbxe4 4 i.d3 d5 lbc3 lbc6 13 f4lbcd4 14 .i.a4 0-0-0 and here
5 llJxe5 i.d6 6 0-0 we can see no argument supporting a white
White has a extensive list of alternatives: edge.
a) 6 lbc3 lbxc3 7 bxc3 0-0 8 0-0 ll:X17 6 ...0-0
transposes to Game 62. Or:
b) 6 'irf3 0-0 7 0-0 cS!? 8 .i.xe4 dxe4 9 a) 6 ....i.xe5 7 dxcS lbc5 is a slightly dubi-
Wxe4 'fle7 10 lle1 lle8 11 lbd2 cxd4 12 ous favourite of the solid GM from Uthua-
lbet1 (12lbec4i.b4 13 'iixe71lxe7 14 :xe7 nia, F�uardas Rozentalis. Now 8 .i.e2! is
.i.xe7 1Slbc4lbc6 would give Black a pref- probably the right path for White here, for
crable endgame) 12 ...lbc6 13 1fxe7 llxe7 14 example 8.. 0-0 9 b31? lbc6 (9...c6 10 .i.a3
.
lbe4 .i.c7 15i.d2 .i.b>416lbc5 .i.d6 17lbe4 'iie7 11 f4 aS 12 lbc3 lblxl7 13 wrd4 gives
.i.c7 with a draw, Raetsky-Mironov, corre- White a powerful initiative) 10 L3 b6 11 f4
spondencc 1985. lte8 t2lbc3 d4 13i.f3 .i.b7 14 .i.xcS bxcS
c) 6lbd2 .i.xeS 7 dxe5lbcs 8lbf3lbxd3+ 15 lba4 wrc7 16 We2 ll:XI8 17 'ifb5 and
(or 8....i.g4!? 9 h3 lbxd3+ 10 Wxd3 .i.xf3 11 White has strong prt:ssure, C.l-lanscn-
'irxf3 0-0 12 'irg3 Wh8 13 i.gS 'ii'd7 14 Ro:r.entalis, Malmo 1997.
150
3 d4: 5 .. . �d6 and Fifth Move Alternatives
b) 6 ...Qk6 7 lfu.c6 bxc6 8 c4 Wlt4?! equality but Black needs to prove it! In Slo
(fL0-0, transposing to 6...0-0, is preferable) 9 bodjan-Forintos, Germany 1992, Black failed
g.� 1i'h3 (9 ...'ii'f6 is refutoo by to i..xe4 dxe4 to do so and foUowing 12...ltld7 13ltla3 i..£6
l tltlc3 i..f5 12 f31 ..g6 13 fxe4 i..h3 14 e51 14llk2ltle5 15ltlxd4ltlxf3+ 16ltlxf3 b6 17
i..xft 15 exd6 i..h3 16 'ii'e2+ 'l'e6 17 'ii'xe6+ i..g5 i..b7 18 i..xf6 .llx£6 19 :C5 White held
i..xe6 18 dxc7 i..xc4 19 i..f4 with a clear a clear advantage.
advantage for White - analysed by the very c) Lltld7 8 Le4 (8 ltlxd7 leads to an
reliable duo Yusupov and Dvoretsky) to c5 immediate draw after 8...i..xh2+ 9 Wxh2
i..g4 11 i..e2 i..xe2 12 1i'xe2 i..e7 13 f3ltlf6 1i'h4+ 10 �g1 1i'xf2+) 8...dxe4 9 ltlc4 ltlb6
14 i..f4 and Black has achieved little witlt his 10 ltlxd6 1i'xd6 11 llxe4 i..f5 12 l:leS Wg6
kingside actions - White has a slight advan and Black has good counterplay for the
laJ.,>e. pawn.
c) 6...c51? allows White to gain an edge 7 �xeS 8 dxe5 �c5 9 l0b3
••.
with 7 i..bS+ltld7 8 dxc5 i..xeS 9 9xd5 0-0! 9 'ii'h5 does not look dangerous here,
(9...1i'e7?1 10 c6! would be an unpleasant 9...ltlxd3 10 cxd3 c5 being the must natural
surprise) to ..xe4 'ii'c7 11 i..d3 g6 12 'ii'h4 reaction. Now 1 1 b4?! looks unjustified:
ltlxc5 13 i..e 2 i..f5 when Black has some 11...cxb4 12 a3 bxa3 13 i..xa3 l:le8 14 i..d6
compensation for the pawn, but hardly ltlc6 15 l:la4 'ii'd7 16 l:lf4 1Ve6 17 d4 'ii'g6 18
enough. White could even consider 14 Qk3!? 1i'f3 i..c6 with a clear edge for Black, J. Pol
here. gar-Kamsky, Groningen 1993. After the
7 lDd2 more sensible 11 ltlb3 'ii'c7 12 i..f 4 i..e6
7 ltc 1 is a respectable alternative, after Black shouldn't be worse.
which we have the foUowing possibilities: 9 l0xd3 1 0 '1Fxd3 tOe&
•.•
757
The Petroff Defence
will have to defend, but rtln defcm.l with ac 23 :ata 24 .i..e3 1tg6 25 �2 lL!f5 26
...
Game62
Anand-Mishra
India 1988
1 e4 e5 2 lLJf3 lL!f6 3 d4 �xe4 4 i.d3 d5
5 lL!xe5 i.d6 6 0-0 0-0 7 lL!c3 lL!xc3
The colourful 7...f5!? 8 f3 .ixeS 9 dxcS
�xc3 10 bxc3 11c7 1 I .D.ct .ic6 12 a4 cS
lead to unclear play in Geller-Yusupov, Vil
nius 1980.
8 bxc3 &iJJJ7
152
3 d4: 5 ... �d6 and Fifth Move Alternatives
1 7 •c6?
...
1 53
The Petroff Defence
still favourable to the game. 6 lbd2 might be a good way to tight for
1 9 f5 11t'c5 an advant:tf.,re. Now Black has two main ways
Or 19....txf6 20 exf6 .i.d7 21 l:tftl and togo:
the double threat of fxg6 followed by .i.xg6 a) 6...lbd6 7 'iff3 (l 111h5 g6 8 .i.xg6?1
combined with 'ti'h6 ends aU speculation does not work on account of 8 ... fxg6 9lbxg6·
about the result. .i.g4! to 'ifxg4 l:tg8 II 'ilh5 Axg6 1 2 ltl£3
20 -*.xg7 'it>xg7 21 •f6+ �g8 22 e6 1 -0 lbd7, when White doesn't have enough for
the piece) 7 ...c6 8 ltlftl? (Kapcngut) 8... 0-0 9
Game63 lbg3 lbd7 10 ltlxd7 'ifxd7 1 1 0-0 1tg4 1 2
Aagaard-Legky 'ifxbr4 .i.xg4 1 3 .i.f4 and White might have a
Budapest 1996 very tiny edge.
b) 6... ltlxd2 7 .i.xd2 ltlc6 (1...0-0 8 'ifh5
1 e4 e5 2 �f3 �f6 3 d4 �e4 4 .i.d3 d5 g6 9 1i'h6 ltlc6 1 0 lfuc6 bxc6 1 1 0-0-0 l:te8
5 �xeS -*.e7 1 2l%dc1 :bs 13 :CS looks better for White)
8 lbxc6 bxc6 9 0-0 0-0 10 'iVhs g6 1 1 'ti'h6
llb8 1 2 b3 .i.f6 1 3 c3:c8 t 4:fe1 :xe l+ 1 5
:xe1 .i.e6 1 6 ..f 4 .i.e7 1 7 'ii'gJ .i.d6 1 8 .i.f4
.i.xf4 19 1txf4 'ifd6 20 1if6 and White had a
bit of pressure in Hort-Spassky, Reykjavik
1 977.
6 ...0-0
The saf est choice. Others include:
a) 6...lLld7 7 .i.f4 (l c4 is less dangerous:
7...lbxc5 8 dxcS c6 9 cxd5 1txd5 10 Wn
.i.fS! 1 1 ltlc3 t£lxc3 12 'ifxf5 br6 1 3 'iVh3 :d8
with unclear play- Yusupov; 7lbxd7 .i.xd7
ttansposes to s. .ltld7 6 t£lxd7 .i.xd7 7 0-0
.
This is a bit passive and Black can easily .i.e7) 7 ...t£lxe5 8 .i.xeS 0-0 9 c4 c6 10 1tc2
end up in a slightly worse position. and White has an edge according to Euwc.
Another sideline that is no longer popular b) 6...t£!c6 also doesn't fully equalise: 7
is 5 ...ltlc6 6 ltlxc6 bxc6 7 'ife2!? (l 0-0 .i.d6 ltlxc6 bxc6 8 c4 0-0 9 t£lc3 ltlxc3 10 bxc3
ttansposes to 5....i.d6 6 0-0 ltlc6 7 ltlxc6 dxc4 1 1 .i.xc4
bxc6, and 7 0-0 JJ..e7 transposes to 5....i.e7 6
0-0 ltlc6 7 lbxc6 bxc6) 7...'ife7 8 0-0 g6
(8...lLld6 may be safer though after 9 :c 1
1txe2 1 0 ltxe2+ .i.e6 1 1 ltld2 �d7 1 2 ltlf3
f6 1 3 c3 .ifS 14 ltlc1 g5 1 5 .i.xfS+ lDxfS 16
lLld3 White retains a very slight edge) 9 .i.xe4
..xc4 1 0 1td2!? (this looks more logical than
lO 1txt:4+ dxc4 1 1 :e1 f5 1 2 f3 .i.g7 1 3 c3
0-0 1 4 .i.f4 c51? 1 5 dxc5 :b8, which gave
Black considerable compensation in PiUs
bury-Schlechter, Munich 1900) IO....tc7 1 1
:e1 1tf5 12 'iff4 and White has St.-vera1
threats.
6 0-0 and now:
154
3 d4: 5 . . . .J.d6 and Fifth Move Alternatives
bt) 11....LI6 12 1fd3 lZ.b8 13 h3 gives b) 8...c6 9 lZ.e1 ll'lbd7 10 .i.g5 dxc4 11
White a s(jght edge. Holzke-Yusupov, .i.xc4 ll'lb6 (11...ll'ld5?! is punished by 12
Hundesliga 2000 continued 13...c5?! 14 dxc5 .ixd5 .ixg5 13 .ixt7+1 :Xt7 14ll'lxt7 �xt7
.lxc5, an(l here White can play 15 .ixt7+! 15 1Vh5+ *fB 16 li'xh7 ll'lf6 17 'lt'h8+ll'lg8
<bxt7 16 'tl'c4+ .i.e6 17 'tl'xc5 'ird6 18 'ifxd6 18 d5 .id7 19 llad1 with a strong attack,
cxd6 19 .ia3 with good chances of convert Yurtaev-D.Frolov, Tomsk 1998) 12 .ib3
ing the extra pawn into a full point. ll'lbd5 13 li'f3 .ic6 14 llad1 WaS and Black
b2) 1t ....tf5 12 .if4 .td6 13 .tg3 .JZ.e8 14 is only slighdy worse (Yusupov).
'ira4 .ie4 15 lLe1 .ixg3 (15 ... h5?! is reck 8&3
less: 16 f3 .i.d5 17 llxe8+ Wxe8 18 lle1 1t'd8 8 cxdS cxd5 9 .i.xc4 dxe4 10ll'lc3 .ifS 11
19 .i.xd5 cxd5 20 1Vc6 .i.xg3 21 hxg3 lZ.b8 l:le1 ll'ld7 12 ll'lxd7 1Vxd7 13 ll'lxe4 lZ.ad8
22 :C5 with a clear plus for White, Ivan gives Black excellent compensation for the
chuk-Yusupov, Novgorod 1995) 16 fxg3 pawn, a draw being the likely result.
i.d5 17 .i.d3 'irg5 18 'ifc2 h6 19 Wf2 and 8 1fc2 ll'lf6 9 c5 is less forcing and there
here we believe there is no argument about fore also more ambitious: 9 ...ll'lbd7 toll'lc3!?
White's edge. ll'lxe5 11 dxe5ll'ld7 12 .ixh7+ *h8 13 .if5
7 c4 ll'lxe5 14 lle1 .i.f6 (14...f6 is wt:aker; after 15
Maybe the best idea for White is 7 llel!?, .i.e3 .i.xfS 16 1fxf5 d4 17 :Cd1 i.xc5 18
as first suggested by Steinitz. After 7 ... f6 8 lL\.14 ..e7 19 1Vh3+ *g8 20 ll'lxc5 dxe3 21
ll'ID f5 9 c4 .i.c6 10 cxd5 .ixd5 11 ll'lc3 1fxe3 White enjoys slight pressure) 15 .if4
ll'lxc3 12 bxc3 ll'lc6 13 ll'lc5 ll'lxe5 14 llxe5 lZ.eS 16 lZ.e3 .i.xfS 17 1fxf5 lL\g6 18 llh3+
White was a bit better in Ed.Lasker-Kupchik, *g8 19 1Vh5ll'lh4 20 .i.g3 1fe7 with a com
New York, 1915. plex batde, Mi.Tseidin-Karasev, Leningrad
7 ...c6 1970.
B �c3 9 bxc3 dxc4
...
155
The Petroff Defence
The only move, but good enough to f�ght White also achieves a draw after 31.../lkl7
on equal terms. 32lla6 'i'bt+ 33 �h2 'i'e1 34 'ii'f4+ ltJ£6 3S
1 6 c4 lla7+.
lb.is might be too rash a decision. After 32 llxf6+ �f6 33 •xd7 ft1 + 34 �2
lS .i.a3!? .i.xa3 16 'ii'xa3+ cS 17 c4 the posi � 35 .d8+ � 36 .g8+ �6 37"
tion is less clear, though it is difficult to be •d8+ �c6 38 11'c8+ �5 39 •d7 + Wc4
lieve that Black should be worse. 40 •xg7 1lrd6+ 41 g3 a4 42 •xh7 a3:
1 5 ...lL!f6 1 6 .tb2 .i.f5 1 7 d5 ..d7 1 8 43 •c2+ Wb4 44 �1 + �a5 45 ft3·
Jlad1 Ilea 1 9 dxc6 •xc6 20 .ta3 .te6 �%-%
156
3 d4: 5 ... Jl.d6 and Fifth Move Alternatives
Summary
After 1 e4 eS 2 �£3 �f6 3 d4 �xe4 4 .i.d3 dS 5 �xeS, the move S...i0c6 is unattractive in
vit:w of the strategic problems Black faces after the doubling of the pawns. In our view, an
interest in S....i.e7 will be maintained; this continuation is not ambitious and docs not have any
positional flaws.
After S....i.d6 interest in the forced line 6 0-0 0-0 7 c4 .i.xeS 8 dxeS �c6 9 cxdS 'ifxdS 10
'tl'c2 �b4 t 1 .i.xe4 �xc2 12 .i.xdS will die away as it is thoroughly analysed and docs not
appear to be more promising for White in terms of gaining an advantage than, for instance, 7
ltlc.l2!?.
7 Jl.xe5
...
7...�c6 Game 60
-
157
CHAPTER EIGHT I
3 d4: Fourth Move
Alternatives
158
3 d4: Fourth Move Alternatives
This move was a shocking novelty when it t2.!Dc3 c6 13 :C1 ..c7 14 g3lDe6 15 1Wd3
first appeared in 1993. Nowadays, of course., i..e5 16 .i.e3 d5 (Sadvakasov-Koneru, Jodh
ir has lost its surprise value. pur 2003) and now 17 .i.d2 l:lab8 18 l:labt
6 ..txe4 .i.f6, with a level position, is a possible con-
White has a large number of options: tinuation.
a) 5 .!Dxe5 .!Dxe5 (5....!Dc5 6 .i.c4 .!Dxe5 7 5 ...d5
dxe5 d6 8 1i'f3 1Wd7 9 .i.e3 c6 10 Lc5 dxc5
II 0-0 favours White, whilst 5...d5 transposes
In 4 .i.d3 d5 5 .!Dxe5 .!Dc6) 6 .i.xe4 d5! 7
dxe5 dxe4 8 11'xd8+ �cl8 9 .!Dc3 .i.b4 10
.ig5+ �e8 11 0-0-0 .i.g4 12 l:d4 .i.xc3 13
hxc3 h6 14 llxe4 .i.h3! (a typical Shirov
move, here played against him) 15 gxh3 hxg5
16 l:tg1 llxh3 17 llxg5 �fB 18 .J:leg4 :C8 19
llxg7ltxe5 20 llg8+ �c7 with a likely draw,
Shirov-Timman, Wijk aan Zee 1998.
b) 5 dxe5 and now:
b1) 5...d5 6 exd61Dxd6 7 0-0 .i.e7 8.!Dc3
Jtg4 9 .!Dd5 .!Dd4 10 i..e2 (10 :C1 i..x£3 11
hrxf3lbc6 12 f4 f!f.l13 f5.!Dxf5 14 i..xf5 gxf5 6 ..txh7
15 'ti'£3 leads to a wildly complex position 6 .!Dxe5 dxe4 7 .!Dxc6 bxc6 8 0-0 .i.d6 9
with chances for both sides; note 15....!Dd4?? :Ct 0-0 10.!Dd2 (10 l:lxe4 i..f5 11 :C1 1i'h4
16.!Df6+ �fB 17 i..h6 mate!) to...lbc6 11 h3 12 g3 'ti'h3 looks dangerous) 10...f5 11 .!Dc4
.i.h5 12 :C1 0-0 13 .!Dxc7+ 11'xc7 14 .!Dg5 .i.di 12 b3 Wh413.!Dxd6 cxd6 14 1fd2 gives
..i..xe2 15 1i'xe2 Wf6 16.!Dxe6 .J:lfe8 17 Wg4 a dynamicaUy balanced position, Hracek
Axc6 18 l:lxe6 fxe6 19 c3 l:lf8 with level Barua, Moscow 1994.
chances, Yakovich-Makarychev, Elista 1995. 6 i..g5 leaves Black with a wide range of
b2) Also fine is s....!Dc5 6 0-0 .!Dxd3 7 choices:
'iixd3 d6 8 ..i..f4 dxe5 9 '1Vxd8+ 1Dxd8 10 a) 6...f6 7 .!Dxe5 dxe4 8 11Vh5+ (8 .!Dxc6?!
i.xe5 lbc6 11 :C1 i..d7 12 i..xc7 .J:lc8 13 bxc6 9 .i.e3 .id6 10 c4 0-0 11 c5 .i.e7 12
..i..e5 llxc2 14 lDbd2 i..c6 15.!Db3 .i.x£3 (an .!Dc3 f5 13 Wb3+ �h8 14 g3 f41? 15 gxf4- or
improvement over 15...i..b4?! 16 .!Dfd4 .J:lc4 15 .ixf4 ..xd4 16 .i.e3 We5 with attacking
17 :Ccl l:lxct+ 18 .J:lxcl �d7 19 a3 i..e7 20 chances according to Frolyanov - 15..ig4
.!Dxc6 bxc6 21 l:ldt+ �c8 22lDa5 c5 23lDc6 16 llg1 .i.f3 and Black had an attack in Yur
with a clear edge for White in Bezgodov taev-Frolyanov, Moscow 2003) 8...b>6 9.!Dxg6
Cs.Horvath, Ljubljana 1995) 16 gxf3 i..b4 17 hxf!f.l10 1fxg6+ �d7 11 i..xf6 (11 1Vf5+ �e8
l:tc4 i..c5 18.!Dxc5 l:lxc5 with an <.>tJual posi with a draw is of course possible) 11..lth6 12
tion. Wxh6 i..xh6 13 i..xd8 �xd8 14.!Da3 and we
c) 5 d5 was meant to give White an edge evaluate this endgame as slighdy favourable
once upon a time, but we cannot see how: for White.
5....!Dc5 6 dxc6 e4 7 cxb7 (7 .i.c4 ex£3 8 b) 6...1Vd6 7 dxe5 Wh4+ 8lDc3 dxc4 9 a3
cxd7+ i..xd7 9 Wxf3 1Ve7+ 10 .i.e3 .ic6 11 WaS?! 10 .!Dd4 .!Dxe5 11 0-0 .i.d7 12 .!Dxe4
i.d5 .i.xd5 12 WxdS 1i'e6 13 1i'xe6+ .!Dxe6 li)g6? (Black was in a bad way, but resistance
14 .!Dd2 0-0-0 15 0-0-0 was drawn in V. Ko was still possible) 13 :Ct .i.e7 14 b41i'b6 15
rnv-Frolyanov, Russia 2003) 7..ixb7 8 .i.e2 ..txe71Dxe7 t6.!Dc5 l:ld8 17 We2 1-0 Palac
cxf3 9 .i.x£3 i..x£3 10 1fxf3 .i.d6 11 0-0 0-0 Kos. Fcldbach 1997. Instead of 9...'ifas.
159
The Petroff Defence
.
Black can try 9.. 1Wxb2!? 10 ll:kls .tcs 11 llc1 (or 9 dxc5? 'iVh4 10 h3 .txh3! and Black:
llb1 1Wxa3 12lbxc7+ �fB 13lbxa8 exf3 14 wins) 9...lllf3+ 10 gxf3 .te6 1 I 'iVd3 llh3
'iVxf3 'ttxf3 15 gxf3 lbxe5 with a very un with a very strong attack for Black.
clear position. 7 ...i.g4 8 i.f4 11d7
c) 6...1Wd7 (the normal move) 7 .td3 e4 8 8... g51? is also enticing: 9 .tg3 f5 10 exf6
0-0 f6 9lle1 1Wxf6 11 lbc3 .i.xf3 12 'iVxf3 .xf3 13 gxf3
.i.b4 14 0-0 .i.xc3 15 bxc3 0-0-0 with an
equal ending; or 9 1t'd3 .i.xf3 to gxf3 (Aiek•
seev-Bezgodov, Hoogcvecn 2002) and aftCJ:
10..llh6 11 .tg3 .i.g7 12 f4 gxf4 13 .i.xf4
llc6 Black should have sufficient counter
play.
9 �bd2 11f5 1 0 i.g3 0..()-0 1 1 0-0 i.c5
1 2 a3
12 llct lZ\ct4 13 lllxd4 .i.xd4 14 lZIO
i.xb2 15 llb1 .i.c3 161lc3 d4 17lld3 11'g6
18 h3 lldh8 gave Black a strong attack in'
Svcshnikov-Pavasovic, Nova Gorica 1996.
1 2...i.b6
with another branch: Or 12...�51? 13 c3 (13 b4? does not
.
ct) 9. .fxg5 to c4 .tb4 11 lbc3 (11 cxd5 work on account of 13...i.d4 , with a clear
.txel 12 'lrxe1 lbe7 13 .txe4 g4 14lbc5 advantage for Black) l3...d4 14 1Wa4 d3 with
'iVd6 15 lba3 a6 with an unclear game is substantial counterplay.
given by Murey) 11...0-0 12 cxd5 lbxd4 13 1 3 b4 �5 14 l:.e1 �d4 1 5 a4 a6 1 8
lbxd4 exd3 14 llX.o6 llt7 15 1fxd3 favours l:.a3 �f51
White.
.
c2) 9. ..te7 10 .tf4 cxd3 11 'ilfxd3 0-0 12
lbc3 .tb4 13 llc2 (Tunman-Hiibner, France
1993) and after 13...1Wt7 14lbb5 .taS the
chances are level (Hubner).
6 ...Axh7
1 7 �f1
17 .tf4 would be met by 17...g51 18 .i.xgS
J:lgs with a strong attack (Forintos).
1 7 ...trucg3 1 8 �xg3 i.xf3 19 •xf3,
'W'xh2+ 20 �1 g6 21 l:.d3 I
Or 21 We2 1Wh4 22 llh1 11'c4+ 23 '1Vd3
7 dxe5 llxh1 24 1Wxc4 dxc4 25 lbxh1 llc8 and th ,
7lbxc5?! is weak: 7...'ttc7 8 0-0 lllxc5 9 endgame favours Black.
J
1 60
3 d4: Fourth Move Alternatives
21 ...1rh41 22 llxd5 irc4+ 23 lld3 llxd3 ..t.gS d6 10 1i'e3 1i'd7 11 �c3 h6 12 ..t.h4 gS
24 •xd3 1rxb4 25 e671 13 �S �ffi and Black is under attack, but
'Ibis accelerates an uncomfortable posi nothing is clear) 7 'iVdSlbxh1 8 .i.hS fle7 9
tion into more trouble. After 2S l:te4 1WcS 26 ..t.gS .i.f2+ 10 We2 'l'e6 11 �c3 h6 12
..Wd2 :J,8 White is still only 'somewhat' ..t.d8!? lbc6 13 1Wxe6 dxe6 14 .i.xc7 with a
worse. position almost impossible to understand
25...1rf4 26 l:le2? without dedicating days or weeks of your life
to it!
b) 5 1i'dS ..t.xf2+ (S...�xf2? 6 WxcS �xht
7 .i.gS f6 8 cxf6 hrxf6 9 ..t.h4 is probably los
ing for Black)
�xeS d6 4 �f3lbxe4 S d4 and 4...dS S ..t.d3 .z:tf6 (1t...'.tc6?! 12 ..t.xd8 l1xd8 13 �gS+
1ransposes back to 4 ..t.d3 dS S �xeS. �xeS 14 �dn+ Wd4 1S �xd8 �xc4 16 JZ.ft
4 . d5
. . and White has the advantage) 12 .i.xdS �c6
4...d6 transposes to the Philidor Defence 13 exf6+ gxf6 14 �xfS+ �e8 15 �d6+ �e7
(I c4 eS 2 �f3 d6 3 d4 �f6 4 dxeSlbxe4), 16 .i.f4 with very unclear play. Of course
which is outside the scope of this book. We White could take a draw by perperual with 16
will say, however, that White is generally �fS+.
1hought to have a slight advantage. b2) 6 <.te2 f5 7 �3 (Black would get
4.....t.cS!? leads to very sharp play: away too easily after 7 exf6 �xf6 8 WeS+
a) S .i.c4 �f2 (S....i.xf2+!? 6 �e2 ..e7 7 �f8 9 .i.gS ..t.b6 10 �3 �6 It 1i'f4 h6 12
WJ3 f5 8 �3 c6 9 ..t.c3 .i.xe3 10 1i'xe3 with .i.xf6 'iVxf6 13 'ilrxf6+ gxf6, after which
unclear play is also possible) 6 ..t.xn+ �f8 White cannot fully justify the loss of a pawn)
(or 6...�n 7 ..dS+ �e8 8 'ifxcS �xh 1 9 7...c6 (7...�xc3+ 8 bxc3 .i.h4 9 �xh4 'iVxh4
161
The Petroff Defence
162
3 d4: Fourth Move Alternatives
Indeed for Black) 13 g4 .i.g6 14 .i.xg6 fxg6 A valid alternative is 18...lla41? 19 .i.b5
15 llxd5 'fle7 was called for, reaching a posi lla3 20 'irxd5 :Xc3, when it is probably
tion with chances for both sides. more pleasant to be Black.
1 1 .....tg6 1 2 .IZ.e1 ..tb6 1 3 ..tg3 19 �xf2?
An understandable mistake as White is
scared of allowing his king to go too far into
the open. However, this is the worst of the
two evils and Black now has a very strong
attack.
The alternative is 19 Wxf4 ...ffi+ 20 .te;
'fld6+ and now:
a) 21 lle5lilie5 22 .L.f2 �xg4+ 23 Wxg4
h5+ 24 Wh3 .i.x£5+ 25 Wg2 .ie4 offers
Black compensation, but it is very likely that
White wiU survive the attack.
b) 21 WgS! suggests to Black that he
should settle for a draw, as after 2L..'flc7+ 22
1 3 ...f5 Wf4 .i.xh4?! (22.. .1t'd6+ is perpetual check)
13...d4 is weaker: 14 cxd4lilid4 15 �xd4 23 �xh4 1fxh4 24 llht 1ff6 25 1fxd5 l:l.d8
Wxd4 16 .i.xg6 hxg6 17 'flxd4 .i.xd4 18 26 1t'c5 it is not so obvious that Black has
%lad1 .i.b6 and White has many ways to se sufficient compensation for the exchange.
cure the advantage in the endgame. The White king looks strange, but Black's
14 �2 ¢'h8 1 5 .i.h4 •ea 1 6 e6 txg4 1 7 piccc:s are not that active.
hxg4.1Z.f4 19 .. ...te4 20 ..tg5
Black can also play the safer 17 ... .taS! 18 20 .i.e2 1i'xe6 21 .i.g3 llf7, with ... l:laffi
11td2 d4 19 �xd4 (not 19 e7? llxf31 20 .ixg6 to follow, does not help White at aU.
hxg6 21 Wxf3 .i.xc3 22 'flf4 .ixat 23 llxa1 20 ....1Z.xg4
�xe7 24 9xc7 �6 and Black has the ad Or 20... llxf3+!? 21 Wxf3 .i.xf3 22 Wxf3
vantage) 19....ixd3 20 11Vxd3 �xd4 21 1Vxd4 d4 and Black has goo d chances. However,
1rc6+ 22 'fle4 .ixc3 �' 1rxc6 bxc6 when the game continuation makes more sense.
nlthough White will have enough compensa 21 llg1 •xe6 22 .IZ.xg4
lion for a draw, he will never be able to win.
18 �g3 ..txt2+ !?
163
The Petroff Defence
22...1rxg4 23 .i.e2 llf8 24 1Wg1 .i.xf3 0-1 should play 10...'1Ve71? 11 lbbS lbba6 12
..xdS lidS 13 'ffc 4 1Wxe5 with a messy posi
Game66 tion) 11 lbb5lbe6 12 c4 a6 13 cxdS axbS 141
Smirin-Aiterman dxe6 ..xe6 15 .i.xbS+ c6 112-lfz Volokitin�
Haifa 1995 Mikhalchishin, Portomz 2001. In the final:
position 16 .i.e2 lZXI7 17 f4 l:lxa2 18 l:lxa.Z
1 e4 e5 2 li)f3 ll)f& 3 d4 li)xe4 4 dxe5 d5 11'xa2 19 0-0 appears to offer White some
5 li)bd2 chances.
c2) 6lbxe4 .i.xe4 7 .i.d3lbc6 8 0-0 .i.e1
9 l:le1 .i.xd3 10 1i'xd3 '1Vc.l7 11 .i.f4 0-0-0 12
a3 Wg4 13 .i.d2 f6 14 .i.c3 d4 15 ..tb4 l:lhe�
(t s....i.xb4?! 16 axb4 lbxb4 17 Wc4 lbc6 t!i
b4! with a clear plus for White, e.g.J
IS...c;ilbS?I 19 bS lbxeS 20 lbxeS fxeS 21
l:lxa7! 1-0 Glek-Mikhalchishin, Zurich 2001)�
16 .i.xe7 :Xe7 17 exf6 gxf6 18 l:lxe7 lbxe"f!
19 fle1 .!Og6 20 h3 'iWd7 21 l:le4 with a sligh�
edge for White (Giek).
d) s....!Ocs 6 .!Ob3
5 . .i.e7
..
1 64
3 d4: Fourth Move Alternatives
..id2 �e4 1/2-'12 Svidler-Kramnik, Dos Her bly play 12....!tlxd41? 13 .ixd7+ �d7 14
manas 1999) 7...0-0 8 'A'g3 �f5 (8 ...f6?! 9 ltxd4+ We6 15 h3 (15 h4!?) 15 ...h4 16 .lh2
..th6 :n 1o 0-0-0 ltlc6 11 �c4 �e6 12 c3!? 17 fxe3 .i.e? with compensation for the
.i.xdSI �xdS 13 e6! �xc6 14 .llxd8+ :Xd8 pawn through the safe, yet active, king.
15 .i.f4 gives White a clear edge) 9 c3 lieS 10 6 c6
...
165
The Petroff Defence
king is not really safe. weaker due to 23 .i.a3+ We6 24 f5+! l:lxf5 25
14 Wxg7 tng& 1 5 tLlf3 Wd7 1i'xc6+ Wfl 26 l:lfl !l0f4 27 l:lxf4! l:lxf4 28
After 15...1i'd6 16 lOgs .i.xgS 17 .i.xgS 1i'c7 and Black is lost) 23 cxdS 1i'c8 24 �
..d7 18 ..c3 White is slighdy better accord .i.d8 25 .i.a3+ WxdS with unclear play. Our
ing to Greenfeld. computer prefers Black, but that assessment
16 c4 can change in an instant!
22 .i.a3+
1 6. .••g8?1
Black should probably play 16...Wf8 17 22 .•• c5?
11fc3 llgs 18 :C1 .tb4 19 �3 dxc4 20 Black is better off playing 22. .We6. After
.
:dt+ Wc7 21 Wc2 .i.cS with a very unclear 23 cxdS+ l:lxdS 24 Wxc6+ .i.d6 25 l:lad1
position (but not 19....i.xe1? 20 ..xb7+ Wd6 l0e7 26 l:lxeS+ WxeS 27 .i.xd6+ Wf5 Black
21 cS+ WxcS 22 .i.e3+ Wd6 2.1 .i.c5+1 and seems to survive miraculously. However, the
mate is imminent). stronger 23 .i.xe7l0xe7 24lLe5+ WxeS 25
1 7.d4 llf81? 1i'xe7+ Wd4 26 l:le1 gives White a very dan
17....i.d6 18 cxdS exdS 19 :C1 looks to gerous attack.
give White a slight edge. 23 .i.xc5+ ! wxc5 24 llac1 dxc4
1 8 Wxa7 .:Xf3 1 9 1i'xb7+ Wd& 24... Wd6 25 cS+ We6 26 11fa6+ would pick
After 19 . We8? 20 ..c8+ .i.d8 21 .xc6+
. . up the rook.
Wfl 22 cxdS White's position is crushing. 25 .:Xc4+ 1 -0
1 66
3 d4: Fourth Mo ve A lternatives
Summary
I�xpcrience has proven that in the case of 4 dxe5 d5 5 ltlbc12 (and 5 i.d3) Black maintains
e<.Juilibrium in various ways. However, like in some other systems of the Petroff Defence, it is
difficult for Black to move from a solid equalisation to seizing the initiative. That's why we
recommend trying the sharp 4....tc51? - we can't fmd any advantage for White after either 5
1Wd5 or 5 .tc4.
Murey's brilliant discovery (4...ltlc61?) looks artificial and poor at ftrst sight. In fact, it is un
clear how White can gain <.-ven a slight advantage. In any case, 5 i.xe4 d5 6 .txh7 (as in Game
64) cannot be recommended for White since it gives Black active counterplay.
167
CHAPTER NINE I
3 d4: Black Plays 3 ... exd4
Bishop Opening (1 e4 eS 2 .i.c4 l£lf6 3 d4 studied in Game 67. The arrangement wid1
cxd4 4lL\f3), or the Two Knights Defence Of ...JJ..e7 and ....i.e6 is effectively met by the
4 ...l£lc6 is played). Instead the games in this queenside castling plan. The attack on the c2-
chapter arc devoted to the tabiya after 4 eS pawn by ....i.f5 is not popular in view of 9
llk4. .i.b5, but White doesn't seem to have a sub
White's main choice is the obvious 5 stantial advanl:3b>c. The main part of Game
'i6'xd4 (Game 67-69). After S...dS 6 exd6 67 concentrates on the surprising 8. ..t£1f5!?
t£lxd6 Black faces a dreary prospect of fight· the knight moves for the fourth time our of
ing for equality with only a slight chance of eight to prepare attacking the enemy queen
sei;dng the initiative. This is why one of the with ....i.d6. This idea was rriet.l for the ftrst
168
3 d4: Bisek Plsys 3 . . . exd4
time during the Kasparov-Karpov match c3 dxc3 9 1t'xd8+ �xd8 10 l0xc3 �e7 11 0-0
(New York, 1990). A fourth option, the very lld8 (Stein-Bronstein, Thilisi 1966) and here
fashionable fianchetto of dte black bishop Stein could have obtained a sb.eable advan
with 8 . g6, is covered in Game 68. Here cas
. . tage with 12ll)a4 .i.b6 13 b3 .i.c7 14 .i.e3.
ding long looks more risky for White than 4 e5
against 8....i.e7, but White still stands better. 4 .i.c4 transposes to the Bishop's Open
The paradoxical S .i.b5 attracted our at ing (2 .i.c4ll)f6 3 d4 exd4 4ll)o).
tention (fal employed it a few times) and is 4 �
...
covered in the annotations to Game 70. The only sensible move. If 4...ll)d5?! then
Black can turn the game into the Berlin De 5 '1Vxd4 c6 6 .i.c4lt!b4 7 0-0 lt!xc2 8 .i.xf7+!
fence of the Ruy Lopez by means of 5 .lik6,.. �f7 9 ..c4+ �e8 10 .xc2 and White's
transposing to a line that is harmless for advantage is clear-cut.
Black (1 e4 e5 2 ll)f3 ll)c6 3 .i.b5 ll)f6 4 d4 5 1Vxd4
exd4 5 e5 ll)c4). A wormy alternative is 5 ...c6
6 1Wd4 'tWaS 7 c3 ll)xf2 (Keres) 8 0-0 ll)h3!
(F.uwe) - anomer in-between move that de
stroys White's pawn chain.
The main part of Game 70 concentrat<..'S
on me interesting move 5 1t'e2. Steinitz dis
cussed dtis move in his book Modmt Chess
lflslnlctor (1889). White attacks the centralised
knight from the e-ftle in order to prevent
Black from strengthening the knight with
either the d- or the f-pawn. However, after
s....i.b4+ 6 �dl !? (Steinitz) it is possible to
improve support the knight with 6...d5 7
exd6 f5, when Black has the initiative. That's 5 d5
•..
why White prefers the calmer 6 ll)bd2, while Morozevich has risked the remarkable
Black often retreats with 5 ...ll)c5 after 5 1t'e2. 5 ... f5!? here:
Black's play in the opening in Game 70 is still
considered to be a perfect model.
Game67
Cabrera-Collas
Malaga2003
1 69
Th e Pe troff De fen ce
b) 6 .i.c4 .i.cS 7 'ifxcSI? (the queen sacri �xc2 1 5 :tel .i.d3 1 6 tnds 'ird8 17 1i'd4
fice is, of course, only temporary; instead 7 �a6 (after 17....tg6 Sax gave 1 8 0-0 lDfS 19
1i'd5 .i.xt2+ 8 We2 1i'e7 9 lDbd2 lDf6 10 'ireS lDxe3 20 lDxe3 as clearly better for
1i'd3 dS 1 1 .i.bS+ c6 1 2 Wxf2 /0g4+ 13 We2 White) 18 lDxc7 l:tb8 (Sax-Yusupov, Rotter
cxbS is very complicated) 7 lDxcS 8 .i.gS
..• dam t 988). Now White should play 1 9
1i'xgS 9 lDxgS lDc6 10 0-0 lDe6 1 1 ltlxe6 (1 1 lDxa6! WaS+ 20 ltlb4 ltlf5 (not 20...l:txb4? 21
ll)f3 b6 1 2 .ids .i.b7 13 c4 0-0-0 14 lDc3 h6 'ifxd6! lld4+ 22 b4, winning immediately) 21
1S l:tad1 gS gave Black good countetplay in l:tcSI lDxd4 22 LaS l:txb4 23 0-0 with good
Bacrot-Moro7..evich, Biel 2003) 1 l ...dxe6 12 winning chances.
f4 a6 1 3 lDd2 and Black's position is slightly b4) 9...9e7+
cramped.
6 exd6 lLlxd6 7 lL!c3 �6 8 11f4 lLlf5
Other than 8 ...g6 (sec Game 68), Black
has two inferior options:
a} 8 ....i.e7 9 .i.d3 (also good enough is 9
.i.c3 .i.e6 to 0-0-0 0-0 1 1 lDgS .i.xgS 1 2
1i'xgS 1i'e7 1 3 9xe7 tDxe7 1 4 .i.cS l:tfd8 1 5
�e2, when White's bishop pair will be use
ful) 9 ...�e6 10 .i.e3 �f6 1 1 0-0-0!? .i.xc3 12
bxc3 11'£6 (after 1 2...�xa2 13 c4 'if£6 14
:bet White keeps the initiative) 1 3 ...xf6
gxf6 1 4 ltld4 �d4 1 5 .i.xd4 Wc7 16 l:thc1
b6 17 f4 c5 1 8 .tf2 f5 1 9 g41? and Black was
under severe pressure, Parkanyi-Krivolapov, with a further split:
Gyongyos 1 998. b41) 10 Wf1 .te4 1 1 .i.xc6+ .i.xc6 1 2
b) 8. .. �f5 9 .i.bS lDe s 'ire6 1 3 lDxc6 bxc6 1 4 'ir f3 'ifc4+ 1S
Wg1 .i.e7 1 6 b3 1i'a6 17 �b2 0-0 and White
has a slight edge. Instead Klovans-Harman,
correspondence 1 967 continued 1 2... 0-0-0?!
1 3 lDxc6 bxc6 14 'ifa4 lDbS 1 S 1i'a6+ Wb8
1 6 .tc3 9b4 17 'ifxc6 tnd4 1 8 11fa6 .i.cS 19
a3 'ifb7 20 1i'xb7+ Wxb7 21 l:tct �b6 22 g3
when White was simply a pawn up.
b42) The simple 10 .i.e3 is also promising:
l O...lDxbS 1 1 lDxbS 9b4+ 1 2 'iVxb4 �xb4+
1 3 c3 .td6 (not 1 3 .. i.aS?! 14 b4 �d3 1 S a4
a6 16 lDbd4! .tb6 1 7 lDxc6 bxc6 1 8 .txb6
cxb6 19 lDeS 0-0-0 20 lDxc6 l:the8+ 21 Wdl
lldS 22 tnd4 and White is a pawn up) 1 4
and now: lDxd6+ cxd6 1 S 0-0-0 .i.e6 1 6 :Xd6 .txa2 1 7
bl) lf 9 ...�xc2? then 1 0 lDcs wins in .tcS! 0-0 1 8 l:txc6 bxc6 19 .i.xf8 Wxf8 20
stantly. tnd2 .tdS 21 8 l:te8 22 Wc2 and White's
b2} 9 ...lDxbS? 10 lDxbS 1i'e7+ 1 1 �fl l better structure gave him the edge in Matu
and c7 drops. lovic-Kholmov, Sochi 1 968.
b3) 9..i.e7 10 .i.xc6+ bxc6 1 1 llks 0-0 9 .i.d2
1 2 lDxc6 1i'e8 1 3 lDxe7+ 1i'xe7+ 14 .i.e3 White has a range of options, but Black
1 70
3 d4 : Black Pla ys 3 . . . exd4
should be fine in all cases: A bold and unclear <:JUeen sacrifice. l11e
a) 9 lDbS .i.b4+ 10 c3 .laS 1 1 1i'e4+ ..e7 simple 1 3 .i.bS leads to equality after
12 1i'xe7+ lDcxe7 1 3 .lf4 lDdS 1 4 .ics f6 1 S 1 3 ... lDxb5 1 4 lDxbS 0-0 1 5 llhe1 1i'f6 1 6
0-0-0 a6 1 6 llxd5 axbS 1 7 .lf4 c6 and the �3 .i.c6.
JX)Sition is level. 1 3....i.xa4 1 4 �xf5 1t'd7 1 5 �xg7 + �dB
b) 9 .i.c4 .id6 1 0 ..e4+ .le7 (1 0.....e7 is Instead 1 S ...We7?1 16 llhet Wf6?1 walks
what White wants: 1 1 .id2 ..xt.-4+ 12 �xe4 into trouble: 1 7 �hS+ �g6 1 8 g4 and Black
0-0 1 3 0-0-0 J.e6 1 4 .i.xe6 fxt.-6 1 5 l:hel will lose too much material in escaping from
l:r.ae8 1 6 �fgS eS 1 7 .lc3 a5 1 8 a4 b6 1 9 f3 the mating net.
h6 20 lDxd6 lDxd6 2t lt)c4 lDf5 22 b3 with a 1 6 .tg5+ �c8 1 7 ZZ.d41 ZZ.g8?1
pleasant advantage, Bc:rzinsh-Nciksans, Riga
2003) 1 1 .ld2 0-0 1 2 .idS .if6 13 0-0-0
lDcd4 1 4 g4 lle8 1 5 ..f4 lDxf3 1 6 'Wxf3 lDd4
17 1i'g2 c5 (the more passive 17...c6 leaves
White with the initiative after 1 8 .le4 .le6
1 9 gS .leS 20 .i.e3) 1 8 .i.e3 'Wb6 1 9 lDe4
i..e7 20 c3 .le6 21 .ixe6 lDxe6 was unclear
in Tiviakov-Ye Rongguang, Groningen 1 997.
c) 9 .ibS .i.d6 1 0 ..e4+ ..e7 1 1 .ig5 f6
12 .id2 .id7 1 3 0-0-0 1i'xe4 1 4 lDxe4 ole7
1 5 g4 a6?! (more accurate is 1 5...lDd6 16
lDxd6+ .lxd6 1 7 :del+ �fB 1 8 l:btg1 lle8
1 9 llxe8+ ¢>xe8 with equality - A7mai
parashvili) 16 .lxc6! Qnstead 1 6 .i.c4 lDd6 17 Now White regains the sacrificed material
lDxd6+ J.xd6 1 8 l:lde 1+ was agreed drawn in with interest. Black should play 17 .....c6 and
Kasparov-Karpov, World Championship now:
!Game 1 01, New York 1 990) 1 6 ....i.xc6 1 7 a) Not 1 8 l:txa4? ..xg2 1 9 .ig4+ fSI and
llhe1 .lxc4 1 8 llxe4 lDd6 1 9 lle2 � f7 20 Black wins more material.
�4 lDc4 21 .if4 and White has an edge. b) Also weak is 18 �xa4?! 1i'xg2 1 9 .i.g4+
9....td6 1 0 1t'a4 •e7+ Wb8 (19... 5?! 20 lDxfS ..xht+ 21 Jldt
..xh2 22 lDxd6+ �b8 23 lDfl is unclear) 20
:Ct hS! 21 lDxh5 bSI 22 lDc3 llxh5! and
Black is much better.
c) Best is 1 8 :hd ll 1 8..ltg8 1 9 .i.g4+
Wb8 20 .i.h6 'Wxg2 with a massively compli
cated position.
18 .tg4 ZZ.xg7 1 9 .txd7+ .txd7 20 .tf4
ZZ.xg2 21 �d6 cxd6 22 �d6 �7 23
ZZ.hd1
The game has simplified leaving White
with a useful extra pawn.
23 .tc6 24 ZZ.f6 ZZ.xh2
••.
171
The Petroff Defenc e
take the knight. even though the rook ending a) Getting n."3dy for kingside castling with
is probably lost. 9 J..bS is also reasonable: 9 ...J..g7 10 0-0 0-0
27 a3 :h8 28 b4+ Wb5 29 :d4 1 1 J..xc6 bxc6 12 J..e3 l:.b8 1 3 l:lab1 aS 1 4 a."
(it's unclear after 1 4 J..cs llc8 1 5 llkl4 J..e S
16 'irf3 J..b7 17 .J:lfd1 •gS) 14...l:.e8 15 J..a7
l:.b7 1 6 J..cS ( 1 6 J..d4 gives Black the option
of 1 6...g5!? 17 .cl f6 HI h3 .i.f5 with a
sharp position) 1 6...J..f5 1 7 1i'a4 ll:k4 1 8
l:.lxlt ti'c8 19 J..d4 li)xc3 20 .i.xc3 J.. xc3 21
bxc3 l:.bS and White may have a tiny advan
tage, Romanishin-Smyslov, Leningrad 1 977.
b) 9 .i.d2, like 9 J..e3, prepares castling
queenside, but this time preventing the pos
sibility of doubled c-pawns, although of
course the bishop is slightly less active:
9...'..c7+ 10 J..e2 .i.e6 1 1 0-0-0 J..g7 1 2 h4
29 ••• 85 h6 1 3 l:.he1 0-0-0 14 .i.d3 ti'f6 1 5 1i'xf6
The king is also caught after 29...a6 30 c4+ .i.xf6 1 6 .i.xg6 .J:.dg8 (this is the right rook;
Wa4 31 �b2 .i.xd5 32 llxdS b6 33 .J:.ff5 b5 instead 1 6...l:.hg8 17 hS J..g4 18 J..h7 l:.h8 19
34 1lf6 liaS 35 cxb5 axb5 36 .J:.dd6. ..td3 J..xh5 20 .i.e2 J..g6 2·1 lbd5 ..tg7 22
30 a4+ Wxa4 31 �c3+ ¢'a3 32 b5 .i.c3 .i.xc3 23 li)xc3 l:.he8 24 li)h4 gave
:h1 + 33 ¢'d2 1 -0 White an edge in Geller-Smyslov, Moscow
r------. 1 991) 17 hS ll:k7 1 8 .i.f4 J.. xc3 19 bxc3
Game 68 li)xg6 20 hxg6 l:.xg6 21 li)h4 llhr4 22 g3 h5
Stefansson-Yusupov and the position is level.
Eupen 1994 9 ....i.g7 1 0 0-0-0
1 0 .i.d3 is a worthwhile alternative. For
1 e4 e5 2 �f3 �f6 3 d4 exd4 4 e5 � example, lO....i.e6 1 1 0-0-0 1i'f6 1 2 �
5 11xd4 d5 6 exd6 �xd6 7 .!003 .!006 8 ti'xf4 1 3 J..x f4 0-0-0 (Yusupov assessed
1rf4 g& 9 ..ie3 1 3....i.xc3 1 4 bxc3 0-0-0 1 5 li)xe6 fxe6 1 6
l:.het l:.hfB 1 7 .i.g3 .J:.de8 1 8 f3 as a bit bet
ter for White) 1 4 l:.he1 .i.xc3 1 5 bxc3 .i.xa2
16 c4 (Black has an edge after the weaker 16
J..xd6 l:.xd6 1 7 c4 .J:.d4 1 8 lDxfl llg8 19 �5
lDxcS 20 .J:lxeS .J:.gu8) 16 ...h6 1 7 lbe4 lDxe4
(less accurate is 1 7 ...lba5 1 8lbxd6+ cxd6 19
cS! d5 20 .J:.e7 .J:.d7 21 llxd7 �xd7 22 J..bS+,
as in Raetsky-Rodionov, correspondence
1 982; now 22...�c8 23 lt:t .i.c4 24 l:.c8+
.l:lxe8 25 .i.xe8 would have been marginally
in White's fitvour) 1 8 .l:lxe4 l:r.he8 1 9 llde 1
:Xc4 20 llxe4 g5 21 .i.d2 ..ib3 22 h4 and
White has reasonable compensation for the
Preparing to castle long without worrying pawn.
about the prospect of doubled pawns. Alter- 1 0 . . ...ie6
natively: This is more flexible than 1 (J...0-0, which
1 72
3 d4: Bisek Pla ys 3 . . . exd4
perhaps commits the king a move too early: a winning attack for White.
1 1 h4 h6 12 .i.cS!? .i.e6 13 .i.bS a6 14 .i.xc6 28 lle2 •xf3 29 1i'xf3 llxf3 30 �2 c6?
bxc6 1 5 .id4! (the greedy 1 5 .ixd6 cxd6 1 6 Now White wins by force. 30...gxh5 31
l:lxd6 allows Black dangerous counterplay �13 l:.xd3 32 cxd3 �18 woukl have given
after 1 6 ...'iWb6 l7 lDd4 ltab8) 1 5...f6 1 6 .icS Black some drawing chances.
l%f7 1 7 llhel ..Ll7 1 8 �4 WeB 19 'ifg3 31 lbe4 lbb5 32 c4 lbc7 33 lld7 lba6 34
�b7 20 �a4 g5 21 'ifb3 �aS 22 'ifd3 and hxg6 l:lef8 35 g7 Jl8f7 36 lld8+ �xg7
White had a powerful bind in lvanchuk 37 llg2+ �h7 38 lbg5+ 1 -0
Akopian, Lucerne 1997.
1 1 lbg5 Game 69
Bonch-Osmolovsky - Baranov
Moscow 1954
1 73
The Petroff Defence
� was level in Konstantinopolsky ..xg7 1i'd5 20 .i.xg6 and the king is in real
Smyslov, SvenUovsk 1 993. danger) 1 1 11xf6 .i.xf6 12 .i.c3 �xc3 1 3
b2) Black should seriously consider the di c!bxc3 and White may have a tiny advantage.
rect 7 ... f6. For example, 8 .i.f4 c!lk6 9 11t"d2 8 i..e3 .ltlf5?1
�fS 10 .i.e2 11t"e7 1 1 0-0 0-0-0 1 2 lte1 lfr4 1bis wastes too much time - instead
1 3 11t"ct gS 14 .i.d3 �h6!? 1 5 llk3 (the wild Black should develop with 8.. ..i.f5. For ex
t S li)h4!? leads to unclear play after 1 5...gxh4 ample, 9 ill c!bc6 1 0 11'f4 .i.xd3 1 1 cxd3
16 �xh6 llhg8; note that White must avoid 'iWe6 12 0-0-0 .i.e7?1 (instead 12 ...1fg6 1 3
1 7 1i'f4? ltxg2+! 1 8 Wxg2 :gS+ 1 9 �1 lthe 1 0-0-0 is unclear) 1 3 d4 11'f5 14 d 5 c!bb8
/bg3+) 15 ...gxf4 16 .i.xc4 .i.xe4 17 l:lxc4 1 5 Wxf5 c!bxf5 16 .i.f4 c!ba6 1 7 g4! c!bh4 18
flg7 1 8 11'f1 .l:thg8 1 9 .1Le1 1i'g4 20 �h1 c!bxh4 Lh4 1 9 d6! 0-0-0 20 c!bbs c6 21 d7+!
�f8 was unclear in Stcinitz-PiUsbury, St Pe :Xd7 22 c!bd6+ l:lxd6 23 .i.xd6 and White
tersburg 1895. However, after 12 ..'iWf7!? 1 3 was much better in Spassky-Kholmov,
.i.d3 .i.g4 Black has a slight plus. Rostov on Don 1 960.
9 i..xf517 .i.xf5 1 0&3 Wb4
White's lead in development is significant
whichever way Black plays:
a) 10....i.xc2 1 1 :ct llk6 1 2 'ff£4 c!bb4 13
0-0 and the initiative continues. For example,
1 3...c!bd3?! 14 'ti'c4 c!bxct 1 5 Let and, with
c!bd5 also threatened, White is winning.
b) 1 0...c!bc6 1 1 .f4 1lt"b4 1 2 0-0-0 and c7
is again impossible to defend.
1 1 fke5+ i..e6 1 2 0-0-0 cltlc6 1 3 fkxc7
Ac8 1 4 fkf4 1ra5
7 fke7+
•••
1 74
3 d4: Black Pla ys 3 . . . exd4
1 8 i.gS! 0-0 1 9 i.xe7 lhxc7 20 lhd4 is un This simple move is more reliable th:m
pleasant but at least Bl.1ck can play on. 5...j.b4+
1 7 c!i)d4 Zlxc3
and now:
Allowing a beautiful finish. 17 ...f6 1 8 a) Steinitz suggested the strange 6 Wd1,
'irh5+ f!!l 19 11b5+ 11'xb5 20 �lxbS would for example 6...d5 7 exd6 f5 8 tbgs (after 8
have lost slowly and painfully. dxc7 11'xc7 9 lhxd4 lhc6 10 c3 lhxd4 1 1
1 8 1Vd8+1! cxd4 i.d7 Black has a tremendous initiative)
A mating combination d1at exploits the 8. . .0-0 9 lhxe4 fxe4 10 11'c4+ Wh8 1 1 dxc7
power of discovered checks. (Lipschutz-Showalter, USA 1 896 continued
1 8 Wxd8 1 9 lDxe6+ We7
..• 1 1 11'xb4 .i.g4+ 1 2 i.e2 i.xe2+ 1 3 'itxc2
Or 1 9 ...We8 20 lhxg7+! i.xg7 21 i.g5+ lhc6 14 ..el?! Wfxd6 1 5 �d1 L8 1 6 b3 c3
�fB 22 :ld8 mate. 1 7 .ta3 'lrf4 and Black had a wonderful posi-
20 �g5+ f6 21 lDd8+1 1 -0 tion) 1 1 ...'ifc7 1 2 cxb811' .l:txb8 and Black
.------., obviously has a powerful initiative - Porreca.
Game 70 b) 6 lhlx12 lhxd2 7 i.xd2 'il'e7 8 0-0-0
Tal-Kholmov lhc6 9 i.xb4 Wxb4 10 11'c4 b6 11 lhxd4
AlmaAla 1968 i.b7 1 2 i.bS 0-0-0 1 .3 a3 lhxd4 (or 1 3 .. ."ti'c5
1 4 lhxc6 i.xc6 1 5 i.xc6 dxc6 1 6 f4 and
1 e4 e5 2 tDf3 tDf& 3 d4 exd4 4 e5 tDe4 White has a defmite cdbrc) 1 4 axb4 lhb3+ 1 5
5 1i'e2 cxb3 i.xc4 1 6 .l:td4 i. £5 1 7 f4 and despite his
A rare alternative to 5 ..xd4. White has bi?.arrc quecnside structure White was a litde
another unusual try in 5 .tbS c6 (5...lhc6 better in Rodrique?. Andrcs-Ginzburg, San
transposes to the Ruy l .opez: 1 e4 c5 2 lht'3 Martin 1 995.
�c6 3 i.b5 lhf6 4 d4 ed4 5 e5 lhc4) 6 'ifxd4 6 c!i)xd4 lDc6 7 �e3
'ifa5+ Qess active is 6...lhc5 7 i.c4 lhe6 8 This is more testing than 7 lbxc6 dxc6 (the
'1Ve4 dS 9 exd6 i.xd6 1 0 0-0 when White has com.'Ct recapture; instead 7 ... bxc6 8 .i.c3 'irh4
a tiny edge) 7 c3 lhxf2!? 8 0-0 (White should 9 1i'c4 'lrxc4 1 0 i.xc4 lha4 1 1 i.b3 lhb6 12
consider 8 'ifxf2 1fxb5 9 a4 11'd5 1 0 i.e3 lhd2 aS 1.3 a3 a4 1 4 .ta2 i.n6 15 .txb6 cxb6
.te7 1 1 0-0 0-0 1 2 lt:la3 with good play - 1 6 llk4 left White a touch better in Spassky
Yusupov) 8...lhh3+! 9 b'Xh3 'lrxb5 10 11'f4 Vistiniet7.ki, Tallinn 1 959) 8 lhc3 i.£5 9 i.e3
11'd3 t l lhd4 'ifg6+ 12 �hl d5 with a com h51? 1 0 f4 l&.:4 1 1 lhxe4 i.xe4 1 2 1if2 ..d5
plex posicion, Zapata-Castro, Colombia 1 999. and Black has good counterplay.
1 75
The Petroff Defence
7 �xd4
.•. Instead the slow 1 7...0-0 is marginaUy in
Safer than 7...l£lxeS?I 8 f4 � 9 l£JbS l£la6 White's favour after 18 .i.d3 .i.g4 1 9 .:dfl
to .i.d4+ 'ile7 1 1 f5 1i'xc2+ 12 .i.xc2 l£lh4 h4 20 l£ld2.
13 f6 g6 14 0-0 when, despite the exchange of 1 8 ..td3
queens, White has a dangerous initiative.
8 hd4 1ih4
A more active approach than 8...l£le6 9
.i.c3 dS 10 exd6 'Wxd6 1 1 l£ld2 .i.d? 1 2 l£lc4
'ife7 (or 12 ...Wcs 1 3 l£les with a promising
attack, e.g. t3..J:d8? 1 4 1i'f3 'We? 1 S 0-0-0 c6
1 6 .i.c4 hS 1 7 l:the1 l:th6 1 8 ..i.b4 was already
winning for White in Rac::tsky-Varlamov,
correspondence 1 983) 1 3 l£leS 0-0-0 1 4 W'e3
and White has a pleasant edge.
9 ..te3 itb4+ 1 0 c3
1 8 ...f6?!
Black should continue the qucensidc pawn
storm with 1 8...a61? 1 9 l:tdf1 c5 when he has
a strong attack.
1 9 exf6 ..txf6 20 ..tc5
20 l£ld41? .i.d7 21 f5 with the initiative is
also good.
20-....tf5 21 llge1 <i>t7 22 �2
And here the simple 22 l:lc3 aS 23 l:ldc1 is
promising.
22 . . �xc5
.
1 76
3 d4 : Bla ck Pla ys 3 . . . exd4
Summary
Statistically 3...l0xe4 has not scored overwhelmingly better than 3 ... cxd4. However, the �"Xam
ples shown here after 3...cxd4 were mainly played more than a decade ago.
The continuations 4 e5 lbe4 5 .ib5 and 5 '1Ve2 do not worry Black any longer, but with 5
11xd4 the situation is more serious. It appears that after 5 ...d5 6 exd6 l0xd6 7 c!lk3 c!lk6 8 1i'f4
itlf5!? Black maintains equilibrium. However, with the symmetrical pawn structure and easier
development, White has many aggressive possibilities that Black needs to be ready for. Conse
quently, the general interest in 3 ...cxd4 has dropped, but Morozevich's experiment with 5 ..f5 is
.
interesting.
4. . . l0e4 6. . . l0xd6
1 77
CHAPTER TEN I
Third Move Alternatives
For White
1 78
Third Mo ve A lterna tives for White
Black docs not have any difficulties and often and Black is fmc) 1 0...a6 1 1 .i.e2 b5 12 1if4
maintains a small space advantage. .i.b7 1 3 .i.f3 lLldf6 14 tlJcs lLlxcS I S .i.xcS
.i.xf3 16 9xf3 lL!e4 1 7 .i.e3 lle6 1 8 9f5
Game 71 lieS 1 9 .g4 'it'cS 20 .i.d4 'lfxg4 21 hxg4
Svidler-Akopian lle6 and a draw was agreed. [f anyone should
World Team Ch., Lucerne 1997 feel relieved in tlte final position, it is White.
7 d5
...
White was slightly berter in Alapin-Aiekhinc, 1 5...c6 1 6 Wd4 l%e6 17 h4 'ifg6 1 8 l:ae1
Carlsbad 191 1 . llaeS 1 9 lle3 and White is slightly better,
5 ll:ld3 .txc3 6 dxc3 ll:lxe4 7 .te2 Chudinovskikh-Raetsky, Orel 1 992.
This is the critical line (although there is a 8 0-0 c& 9 ll:lf4 Ae8
limit to how critical such a position can be). Black is also doing okay after 9 ... .i.f5, for
Instead Galdunts-Ractsky, Aachen 1994 con example l 0 c4 d4 1 1 .i.d3 lieS 1 2 f3 lL!d6 1 3
tinued 7 .i.c3 lieS 8 1Wf3 d6 9 0-0-0 lLld7 1 0 c 5 i.xd3 1 4 Wxd3 tiJbs 1 5 a4 lL!c7 1 6 c 3 (1 6
h3 (or 1 0 lLlf4 lL!df6 I t h3 .i.d7 1 2 .i.c4 .i.c6 b4 aS 17 bS cxbS 1 8 axbS lL!ct7 gives Black
179
The Petroff D e fence
l."XccUcnt countcrplay) 16... dxc3 17 'irxc3 a5 i.xg4 16 'irxg4 c5 17 i.h6 rfl 1 8 Wft may
18 �h5 f6 19 i.h6 liXI5 20 'irc4 g6 21 �3 be slightly better for White) 14 i.xe6 i.xe6
�7 22 J:tfet f5 23 .i.d2 :Xet+ 24 :Xel 15 �xe6 llxc6 16 Axe6 fxc6 17 i.d2 'irf6
..fB and Black has equalised, AJams the game is level, Adams-Hubner, Dortmund
Rozentalis, Copenhagen 1997. 1996.
1 1 . ..1tla6 12 c3
12 0 �c5 13 lle l li)xd3 14 cxd3 i.fS
might even preferable for Black.
1 2....if5!?
1 0 c4
to i.c3 is best met by 10...�16 (10...�d7
11 c4 dxc4 12 i.xc4 �5 13 'irxd8 :Xd8 1 4
.i.c2 i.fS 15 g4 i.d7 1 6 f3 �6 17 l:tad1
li:XIc4 18 i.c1 i.e8 19 b3 �b6 20 .lb2 gave An interesting pawn sacrifice. After 12..c5
White a bit of pressure in Biro-Chetvt:rik, 13 ..c2 �f6 14 i.d2 White has a pull.
Nagykanizsa 1 995) 1 1 i.d3 i.fS 12 'ifh5 g6 1 3 g4 .tg& 14 f3 �5 1 5 .txg& hxg& 1 6
13 'ifh6 .lxd3 14 cxd3 �fS 15 'iFh3 � 16 cxd4
:ae1 'irf6 17 .i.d2 �5 18 d4 llk4 19 Act Black can answer the alternative capture
Ae4 with level chances, Benjamin-Yusupov, 16 ..xJ4 with 16...'irc71?.
Munich, 1994. 1 6... b5
10 d4
.•• The alternative 16...1if6!? 17 d5 llad8 18
h4 �h7 19 g5 1ie5, with play for the pawn,
is also strong.
1 7 cxb5
17 ..d3 is powerfully met by 17...'1fd6 18
cxb5 �b4 19 1Wb3 'ifxd# 20 Wht cxb5 21
a3 �a6 22 1ixb5 llk5 when Black is very
active.
1 7...cxb5 1 8 d5
Or 18 h4 �6 19 �xe6 Axe6 20 i.g5
1id6 21 ..d2 llaeB 22 i.f4 1id5 with suffi
cient play for the pawn.
1 8 llc8 1 9 d&
.•.
180
Third Mo ve A lterna tives for White
Game 72
Lev-Alterman
Ramal-Can 1992
181
The Petroff Defence
.i.e3 with level chances) 9 h3 .i.hS 10 d3 One line continues 7...c6 8 llc 1 dS 9 l:lxe4!
dxe4 11 dxe4 .!LXI7 12 We2 'flc7 with an dxc4 (9...dxe4? to J..xf7+ �h8 1 1 "ifhs J.. f5
equal position, Spiclm:mn-Marshall, Buda 1 2 lL\g6+! J..xg6 1 3 J..xg6 leads to mate) 10
pest 1 928. :04 1ie7 t 1 lDxc4 and White was a pawn up
in LGuliev-Smougalev, Moscow 199S.
7 i.g5
Practice has also seen 7 1ie2 lDbd7 8 .i.gS
h6 9 J..h4 lDcs to l:ad1 We7 I 1 lbd2 gS 12
J..g3 Wh8 13 h4 (dlis docs not seem logical)
1 3...J..d7 1 4 hxgS lucgS I S f3 Wg7 16 Wf2
llh8 1 7 l:l.h1 lDhS 1 8 �fl 1if6 and Black
had the initiative in Kofidis-Alterm:m, Ko
motini 1992.
7 . . . h6 8 i.h4 i.g4
8. ..g5 c:m be met by 9 �xgS!? hxgS 10
.i.xgS with the idea of meeting 10 ... J..e6 with
1 1 .i.d3 �bd7 12 f4!, after which White has
5 ....i.xc3 a very strong attack - the absence of Black's
Black usually uses this chance to inflict dark-squared bishop is really felt.
double pawns on White, but he can also try 9 h3 .i.h5 1 0 'ird3 �7 1 1 b4 'ireS
s ... d6!?, for example 6 d3 .i.g4 7 h3 .i.e6 8
lbc!S .i.xdS 9 exdS h6 10 c3 .i.aS 1 1 d4 exd4
(our improvement over 1 1 ...e4 1 2 �h4 c6 1 3
dxc6 �xc6 1 4 � f5 d S 1 S .i.b3 .i.c7 1 6 f3
when we prefer White, Priehoda-Chetvcrik,
Martin 1 996) 12 �xd4 �bd7 1 3 �f5 �S
14 .i.b3 J..b6 with more or less level chances.
6 dxc3
182
Third Mo ve A l ternatives for White
Black has escaped from the opening with a) S...d6? loses to 6 lbgS .i.e6 7 .i.xc6 fxc6
adequate play for the exchange. 8 11ff3 (Bilgucr).
22 c4 l:lf8 23 .ig3 h5 24 f3 Wh7 25 b) S ....i.e7 is met strongly by 6 lbxe5 0-0 7
.i.xf4 llxf4 26 �3 1ih6 27 Aae1 c5 28 •n i.. f6 8 tDg-4 i..e7 9 lbh6+l? gxh6 10
b5 ltlf8?1 .i.xh6 with a powerful attack.
28...lbf6!?, with good play, would have c) S ...c6!? leads to equality after 6 lbxeS dS
been better. 7 Wc2 i..e6 8 .i.d3 llkl7 9 f4 lbxeS 10 ..xeS
29 ltle4 .ixe4 30 fxe4 ltle6 31 gxh5 'ild6 1 1 1Wc2 .i.c7 1 2 0-0, as in San Claudio
'lrxh5 32 Wg2 'lrf7 33 1i'd3 'lrg6+ 34 Bonari, Mislata 2001 .
'lrg3 1i'h6 35 llxf4 exf4 36 1i'g4 t£!d4 37
h4?!
6 0-0
6 lbh4 g6 7 f4 is probably less dangerous
37 �h2! would have kept an edge for (I 0-0 transposes to 6 0-0 g6 7 lbh4). Play
White. continues 7 ... c6! 8 f5 dS 9 fxg6 (9 .i.b3 �f7
37 . . .f3+? to c4 d4 is better for Black) 9 . ..dxc4 1 0 'ifhs
Black misses his chance. After 37...lbc2! Wd7 1 1 g7 .i.xg7 12 ..g4+ Wd6! (12...Wc7?!
the game would most likely have ended in a 1 3 11'xg7+ lbd7 1 4 .i.h6 'iVe8! I S 0-0-0 l:r.g8
draw. 1 6 ...xh7 %lh8 is 'only' equal) 1 3 'iVxg7 ...f8
38 Wh3 Wh7 39 llg1 f2 40 llf1 1i'e3+ 14 ...g3 (or 1 4 ..xfB+ :Xf8 1 S .i.e3 .i.c6 1 6
41 Wg2 l002 42 llxf2 ltle1 + 43 Wt1 0-0-0+ .i.dS and Black dominates th e board)
ltld3 44 llf7 1i'e1 + 45 Wg2 ...d2+ 46 14....i.e6 1 S .i.c3 lbd7 16 0-0-0+ Wc7 17
Wg1 •c1 + 47 Wh2 'lrb2+ 48 1tg2 llhfl llg8 1 8 •rz b6 t 9 lbf5 J:ld8 and Black
•e5+ 49 1tg3 1tb2+ 50 Wh3 Wg8 51 has a small plus, Tribushevsky-Raetsky, cor
1tf3 1 -0 respondence 1982.
,.-------------.. 6 . . .l006
GaHJe 73 Or:
Morphy-Barnes a) 6...c6? is now strongly met by 7 lbxeSI
umdon 1858 d5 8 'ifhs+ We7 9 .i.d3 with a winning at-
_______________. tack.
1 e4 e5 2 ltlf3 ltlf6 3 .ic4!? b) 6...g6 7 lbh4 ..e7 8 �h I c6 9 f4 dS 1 0
A romantic gambit. .i.b3 e4 i s worth thinking about - we prefer
3 ...ltlxe4 4 t£Jc3 ltlxc3 5 dxc3 f61 BL1ck here.
This should lead to an advanta�;,>e for c) 6 ...'ile7 7 lle1 d6! (l...c6?! is met by 8
Black. Altc�tivcly: lbxcS!? fxcS 9 'fibS+ g6 10 .xeS dS I 1
183
Th e Petroff Defence
1txe7+ .i.xe7 12 .i.gS 0-0 13 il..xe7 dxc4 14 1 1 ... .i.e6 1 2 11h5+ g&
.i.xf8 �fB t 5 llad1 and the endbrame looks 12 ...i.f7 13 'li'g4 g6, with a slight edge,
good for White) 8 �4 c6 9 f4 g6 1 0 P.i dS was also poNSible.
1 1 .i.d3 'ilg7 also looks better for Black.
d) 6...d6 7 l0h4 g6 8 f4 'ile7 9 f5 'flg7
should be better for Black. Now 10 .i.e3 c6
1 1 1te2 dS 12 .i.b3 gS 13 c4!? gxh4?1 14 cxdS
cS 1 5 d6! .i.xd6?! 1 6 lladt gave White tre
mendous compensation in Crepan-Rezonja,
Bled 2000. However, stronger is 1 3...d4! 1 4
�xd4 gx.h4 when Black has a ck"al' plus.
7 � '11fe7
Or:
a) 7.../8:7?! is very weU met by 8 .i.d3 g6 9
f4, when White has a powerful initiative. One
game continued 9._.i.g7 10 fxeS fxcS 1 1 �gS
c6? 1 2 lLlf5! gxf5 1 3 1Whs+ WfB 1 4 :X f5+ 1 -0 1 3 ltlxg6?!
Saburov-Lutze, correspondence 1 906. Risky. After 13 'ile2 0-0-0 14 f4 .i.cS 1 5
b) 7...g6 8 f4 f5 9 l0f.3 e4 10 lOgs .i.cS+ 1tf2 .i.xe3 1 6 'li'xe3 White i s maybe only
1 1 Wh1 'ti'f6 12 'ti'dS (12 .i.£7+ We7 13 .i.ds slighdy worse.
d6 14 :C1 WEB 1 5 g4 lDe71? 1 6 gxf5 lLlxdS 1 3. . ..i.f7 1 4 11h4 hg& 1 5 •xf& llgB 1 6
17 1txd5 gxf5 is l.'VCO worse, Black having a %lad 1 .i.e7
clear plus in Schlcchter-Marco, Berlin 1 897) Not 1 6...l0e7? 17 l:lxdS! lLlxdS?l 1 8
12...d6 13 lLlxh7 1te7 1 4 ltlg5 lDds and we 1txe5+ .i.e7 1 9 .i.xdS when White wins be
prefer Black even if it is not entirely clear. cause of the threat of .i.£7+.
1 7 .e6 .i.f7 1 8 11h3 �
8 �f5?!
8 lthS+ Wd8 9 lLlf5 (9 c!l\g6? 1te8 and 1 8...l:td8!? 19 'ilxh7 D.d6 looks good,
Black wins) 9...g6 10 l0xe7 hrxhS 1 1 lLlxc8 White not having <.:nough for the piece. After
l:lxc8 1 2 .i.e3 is better, giving White good 1 8...ltld8 the trend starts to chaf1bre.
compensation fi>r the pawn. 1 9 f4 e4 20 llxd5
8 ......c5 9 .i.b3 d5 1 0 .i.e3 1t'a5 1 1 �h4 20 .i.xdS!? is an interesting option:
Black's position is also more pleasant after 20 ....i.xd5 21 1ibS+ l0£7 22 :XdS ...xa2 23
1 1 l0g3 .i.e6 1 2 1th5+ g6 1 3 1th4 .i.e7. :Cs and White has an auack.
184
Third Mo ve A lternatives for White
20 .. �xd5 21 'lh15+
Game 74
Svetushkin-Miles
Alushta 1999
1 e4 tnf6 2 d3 e5 3 tnf3
Por our purposes, the Petroff move order
would have been 1 e4 e5 2 ltlf3 ltlf6 3 cB.
3 . . .tnc6
21 .•. �8?1
Much stronger is 21 . ..l:g6 22 .ixd5 .lc5
23 ..thl 'ifa6 (23 ...ixe3 24 .if7+ ltlxf7 25
.
23 b4 ...86 24 f5! 4 c3
4 g3 is also harmless: 4...d5 5 exd5 ltlxd 5 6
.ig2 .i.cS 7 0-0 0-0 8 :Ct %le8 9 h3?! (9 ltlc3
ltlxc3 I 0 bxc3 .i.g4 would transpose to the
so-called Glek Variation of the Pour Knights
Opening - the chances are equal) 9 .. .i.f5 10
.
24...'irf6 25 ..id4 'ifg5 26 .i.xg7+ ��7 27 Black has also tried 6...g6 7 b3 (7 0-0 i..g7
•es was necessary, when White's attack is 8 .l:lcl 0-0 9 .ift J:leR 1 0 b3 b6 I I .la3 .la6
fear.;omc. but the game goes on. 12 'it'c2 'ii'd7 1 3 llad t llad� looks pretty
25 f6! .ixf6 26 b51 1Wd6 27 .ixf7 + b6 standard and pretty level) 7...i..g7 8 .la3
p
27 ...l:lxn s .icS is the end. il)h5 9 0-0 ltlf4 10 J:lc 1 ltlxe2+ I I 'it'xe2
28 .ih6 W87 29 .i.xg7 .txg7 30 ..tb3 :m .ie6 (or 1 1 ...d4!? 1 2 cxd4 exd4 13 .l:lacl ltlb4
31 :n + :xn 32 1Wxf7 + �dB 33 1Wxg7 with equality - Makarychev) 1 2 exd5 'ii'xd5
•d1 + 34 �2 1Wd2 + 35 Wg3 e3 36 13 c4!? tlfd7 14 .i.b2 .ig4 15 d4 0-0-0!? 16
irf6 + lilc8 37 i..e6 + �7 38 1Wf3 + 1 .0 dxeS llhe8 17 'We3 £6 and Black had good
185
The Petroff Defence
0-0 8 a4
7 ()..()
Other moves to consider are:
a) 8 "irc2 a4!? 9 llb 1 lle8 to b4 axb3 1 1
axb3 liJh5 1 2 llet liJf4 1 3 .i.fl wiili equal
chances. 9 exd5
b) 8 b3 lle8 9 a3 .i.a7 (or 9...d4 1 0 .i.b2 White has also played 9 .c2 .i.g4 to lle1
dxc3 1 1 .i.xc3 .i.g4 with equality) to llbt h6 1 1 tl'ln 'iVd7 1 2 .i.e..'� .i.f8 (or 1 2...d4 1 3
dxe4 1 1 dxe4 liJh5 1 2 c!2k4 "irf6 13 b4 axb4 cxd4 .i.xO 14 d5!? .i.xc2 I S dxc6 "irxc6 1 6
1 4 axb4 tl'lf4 1 5 .i.xf4 1fxf4 1 6 b5 with un llxe2 .i.xe3 1 7 tl'lxe3 1Wd7 with level pms
ck'ar play, Gclashvili-Hellsten, Korinthos pects) 1 3 llad1 llad8 1 4 liJg3 d4 1 5 .i.cl
2002. (after 15 cxd4 .i.xf3 16 .i.xf3 liJxd4 17 .i.xd4
"irxd4 1 8 "irxc7 .i.c5 1 9 llf1 b6 Black has
fantastic play for the pawn) 1 5....i.c5 16 liJh4
.i.e6 17 liJhf5 �h7 18 .i.f3 liJg8 1 9 llk2
.i.f8 and Black was a bit better, Vt.>sclovsky
Scrgeev, Ceske Bude�>Vice 1997.
9 �xd5 1 0 lDe4
•.•
(14 ... 1lad8 1 5 .i.a3 llfe8 is a sound alterna 13 'iVbS was better, though after 13 ...'iVc8
tive; we prefer Black) 1 5 .i.e2 f5 1 6 ltJg3 14 .i.ft lla6 White is still struggling.
186
Third Move A lternatives for Wh ite
1 3...li)b6!?
Maybe even stronger is 1 3. ..d71? 14 �5
.. 3 1 f4?1
:Cb8 1 5 li)c5 1t'e8 1 6 li)xe6 l:txb5 17 axb5 The only chance was 31 l:te3 f5 32 llk3
llkb4! (the hidden point) 18 cxb4 ..txb4 with �5 although Black still has all the fun.
a clear plus for Black. 31 ...1lxd3
1 4 1Wxc6 Simpler was 3t .....txc4! 32 l:txe4 l:txd3 33
Forced. After 1 4 ..ie3? ..id71 1 5 ..ixb6 fxe5 li)g5 34 :C1 Wb7 and Black wins.
cxb6 followed by Jla7 Black wins the queen
.. 32 lDf2 lidS 33 fxe5 f5! 34 .*.c1 .*.aS!
for a rook. After this there is no defence.
1 4... -*.d5 1 5 'ttxeS 35 lla2 1Wb7 36 Wf1 1fg2+ 37 ¢>e2
The queen is trapped, and after 1 5 'irb5 c6 .i.f3+ 3S �3 .*.xd1 0-1
it would be sold more cheaply.
1 5 1Wxe8;1'6 .i.d1 c5 1 7 c4
.••
187
The Petroff D e fence
Summary
In practice White's deviations from the main continuations on move three quite often trans
pose to other openings. The least justifiable is 3 d3, since White gives away his advantage of
the first move. Those who enjoy attacking the king should study 3 .i.c4 l£lxe4 4 l£lc3 l£lxc3 5
dxc3, which is especially recommendable fi>r rapid and blitz tournaments. Unfortunately,
4...l£lf6 cools down White's ardour.
3 l£lc3 .i.b4 leads to l)uite another type of position with a slow, manoeuvring fight and
White having the advantage of two bishops (often in the endgame). Like in many other basic
variations of the Petroff Defence, Black has to neutralise White's pressure with careful play.
188
I INDEX OF COMPLETE GAMES I
1 89
The Pet roff Defence
190
Index o f Complete Games
19 1