Contents

Bibliography 5
Introduction 7

1 The Tartakower Variation 9

2 The Lasker Defence 65
3 Alternatives to the Main Line 83
4 The 5 Íf4 Variation 121
5 The Exchange Variation 152
6 The Catalan 179

Index of Variations 205

Index of Complete Games 207
 Introduction

This is a repertoire book for Black against 1 d4, based on the Queen’s Gambit De-
clined (QGD). It deliberately concentrates more on theory than on strategical ex-
planations; there is no shortage of the latter in the QGD, but the literature has
been missing detailed theoretical coverage of the Tartakower, for instance, for an
extraordinarily long time. Although these systems are some of the oldest in chess,
nothing stands still, and the variations at the cutting edge today are not those
which concerned the classical masters of the 1930’s, nor even those which were in
fashion in the 1980’s.
I don’t cover lines without c4 such as the pseudo-Tromp, Veresov, Colle and
Torre for various reasons. First, they aren’t so common or significant after 1 d4 d5.
Second, I did that before, in Dealing with d4 Deviations. Third, they’re dealt with
well in other books on the market. And, lastly, there’s only so much space in a book
and I wanted to be sure the main repertoire was presented in sufficient detail.
One of the appeals of the QGD is that against flank openings Black can pretty
much set up in the same way as he does against 1 d4; the only thing he needs to
know something about is the neo-Catalan or Réti line which arises after 1 c4 e6 2
Ìf3 d5 3 g3 Ìf6 4 b3 Íe7 5 Íg2 0-0 6 0-0 c5 7 e3 Ìc6 8 Íb2, or some such se-
quence. I’m afraid I haven’t covered that either, so let me just recommend either
4...dxc4 5 bxc4 e5 or 4...d4 as irritants for White if you don’t want to play the main
line, and move on. The Catalan, however, I have counted as part of White-main-
lines-with c4, and given a repertoire for Black against it.
The repertoire is based around the Tartakower, which forms Chapter One. I
give an alternative system in the form of the ancient and modern Lasker Defence,
in Chapter Two. Chapter Three deals with White systems which avoid the Tarta-
kower and Lasker proper, but without falling into any of the subsequent three
parts. The most significant of these is ‘Íxf6’ systems, but also a couple of early
Îc1 lines, and 5 Ëc2 or 5 e3 are covered. Chapter Four is devoted to the fashion-
able 5 Íf4 0-0 6 e3 Ìbd7, popularized by Kramnik among others. Chapter Five
covers the Exchange Variation, which I have preferred to deal with by the sidestep

7
Declining the Queen’s Gambit

3 Ìc3 Ìf6 4 cxd5 exd5 5 Íg5 c6, and now either 6 e3 Íf5 (Nigel Short’s recipe), or
6 Ëc2 Íe7 7 e3 Ìbd7 8 Íd3 Ìh5 (a very solid system favoured in his time by Ulf
Andersson, and also by Lars Bo Hansen and the English GM Jonathan Parker).
Chapter Six covers the Catalan, and the system I have chosen is the fashionable
4...dxc4 5 Íg2 (or 5 Ëa4+ c6) 5...Íb4+ 6 Íd2 a5, which was reinvented by Topalov
and since then has produced three of the epic contests of modern chess (Kramnik-
Topalov, Game 1 of the 2006 Elista match; Anand-Topalov, Game 4 of the 2010
Sofia match; and Carlsen-Kramnik, Wijk aan Zee 2010).
It’s handy to have a hero when studying an opening; someone whose games
you look out for in the database and make an effort to grasp. Of course with the
QGD there’s no shortage of world champions to play this role, but sometimes it’s
better to aim slightly lower, and I have no hesitation in naming Rafael Vaganian
as the man to follow. I don’t know how he’s played so many brilliant combina-
tional attacks from the opening, but he has. Nigel Short, of course, is the modern-
day knight of the Tartakower and has contributed hugely to its theory, and I would
also mention the solid German GMs Uwe Boensch and Fabien Doettling as people
to keep an eye on.
As usual I would like to thank John Emms, Richard Palliser and Byron Jacobs for
their courtesy and patience as contractual deadlines, polite-request deadlines,
drop-dead deadlines, final deadlines and the rest flew by. I had the misfortune to
be paired with John in the 4NCL somewhere around the point where any possible
excuse for my tardiness expired, and his accusing stare at my back as I fled the
scene after a draw was agreed remains with me still...

John Cox,
Ruislip,
June 2011

8
 Chapter Two
The Lasker Defence

1 d4 d5 2 c4 e6 3 Ìc3 Ìf6 4 Ìf3 Íe7 5

Íg5 h6 6 Íh4 0-0 7 e3 Ìe4 Game 16
W________W B.Jobava-D.Shengelia
[rhb1W4kD] European Championship,
[0p0Wgp0W] Batumi 2002
[WDWDpDW0]
1 d4 d5 2 c4 e6 3 Ìc3 Ìf6 4 Íg5 Íe7 5
[DWDpDWDW] e3 0-0 6 Ìf3 h6 7 Íh4 Ìe4 8 Íxe7
[WDP)nDWG] Ëxe7 9 cxd5
[DWHW)NDW] Unfashionable, but equally danger-
[P)WDW)P)] ous, leading to less technical positions
[$WDQIBDR] than the very much more popular 9
W--------W Îc1, which is covered in the next game.
This, our alternative repertoire sys- By forcing Black’s next White gains a
tem, is named after the second World considerable central majority.
Champion (‘of course Lasker was the 9...Ìxc3 10 bxc3 exd5
greatest of the World Champions’ – W________W
Tal), and is Black’s main alternative to [rhbDW4kD]
the Tartakower. As one can guess from [0p0W1p0W]
the name, it is a very old line, but in
[WDWDWDW0]
very recent times it has seen a revival,
so much so that the fifteenth World
[DWDpDWDW]
Champion retained his title with a
[WDW)WDWD]
black win with it in the final game of [DW)W)NDW]
the last title match. Its idea is simply to [PDWDW)P)]
break the pin and free Black’s game. [$WDQIBDR]
W--------W
65
Declining the Queen’s Gambit

11 Ëb3 dxc5 Îab8 15 Ëa4 Ëxc5 with perhaps

This move has been played by nu- a tiny advantage for White in
merous World Champions and just V.Eingorn-Li Wenliang, World Team
about everyone who has reached this Championship, Lucerne 1993. Black
position. White’s idea is that by putting held the draw easily against his much
sufficient pressure on d5 (and b7), he higher-rated opponent; in these posi-
will force Black into playing ...dxc4 be- tions the mutual weaknesses on c3 and
fore White has to move his king’s d5 and the strong posts on d4 and c5
bishop, a version of the perennial ‘fight tend to balance each other out.
for the tempo’. Since Black’s main de- 11...Îd8
fence to it (see the note to Black’s 12th, This move is usually chosen today
below) is based on the fork ...Ìa5, rather than the older 11...Ëd6. The rook
though, one does get the idea that the looks a bit prospectless, but Black an-
immediate 11 c4 might be worth a ticipates White’s next and calculates
thought. 11...Íe6 (11...dxc4 12 Íxc4 c5 that sooner or later he will be able to
is an alternative, more nihilistic, try) achieve ...c5, and that when he does his
often leads to the same position as the rooks will belong on d8 and c8.
game after 12 Ëb3 Îd8, but White 12 c4
might also contemplate the untried 12 W________W
Îc1. I suspect Black’s best answer is the [rhb4WDkD]
sharp 12...Ìd7 13 cxd5 (13 Ëb3 can be [0p0W1p0W]
met with 13...c6, since 14 Ëxb7 Ëa3 15
[WDWDWDW0]
Ëb1 Îab8 gives Black excellent com-
pensation for his pawn) 13...Íxd5 14
[DWDpDWDW]
Îxc7 Îac8 15 Îxc8 Îxc8 16 Íd3 Ëb4+
[WDP)WDWD]
17 Êe2 Íc4, when Black will win back [DQDW)NDW]
the a-pawn with an unbalanced but [PDWDW)P)]
equal position. [$WDWIBDR]
White’s other moves have the defect W--------W
of not preventing ...c5: for example, 11 12...Íe6!?
Íd3 c5 12 0-0 Ìc6 (if Black wanted to This long-forgotten move was
be more combative then 12...c4 is per- thought to have been refuted forty
fectly possible) 13 Îb1 (Black also held years before this game and is still little
without any real difficulty after 13 Îe1 known. So far it has achieved excellent
Îd8 14 e4 dxe4 15 Íxe4 Ëf6 16 d5 results, but much of what follows is
Ìe7 17 Ìe5 Íf5 18 Íf3 Ìg6 in the original analysis and for that reason I
latest Topalov-Anand Lasker clash, have included a lengthy note about the
Monte Carlo (rapid) 2011) 13...Íe6 14 older 12...dxc4, whose idea is to gain

66
 The Lasker Defence

time after the obvious 13 Íxc4 by nasty bind. It looks to me as though

13...Ìc6 with the threat of ...Ìa5. either 28...Ëxc4 or 27...Ëxd4 28 exd4
White has dealt with this threat in Ìd8 29 Ìb4 Êf8 is preferable, particu-
three ways in practice: larly the latter.
a) 14 Íe2 b6 15 0-0 Íb7, with a fur- a2) 16 Îfc1 Îac8 17 Ëa4 Ìa5 18 Îc3
ther divide: c5 19 Îac1 cxd4 20 Ìxd4 Îxc3 21 Îxc3
W________W Îc8 22 Îxc8+ Íxc8 23 h3 g6 24 Íf3
[rDW4WDkD] Íd7 25 Ëc2 Ëc5 26 Ëe4 was all but
[0b0W1p0W] equal in Deep Fritz-V.Kramnik, 5th
matchgame, Bahrain 2002, but com-
[W0nDWDW0]
puters play this sort of position very,
[DWDWDWDW] very well and Kramnik went down in the
[WDW)WDWD] game; maybe the best at this moment
[DQDW)NDW] was the uncompromising 26...Êg7.
[PDWDB)P)] b) 14 Ëb2 was the favourite move
[$WDWDRIW] of that cunning old fox and QGD per-
W--------W ennial Gideon Stahlberg.
a1) 16 Îac1 has been proven equal W________W
in several games since the stem game [rDb4WDkD]
V.Kramnik-C.Lutz, German League [0p0W1p0W]
1994, which went 16...Ìa5 17 Ëb2
[WDnDWDW0]
Îac8 18 h3 (18 Îc3 c5 19 Ëa3 Êf8 was
given by Kramnik as equal and has also
[DWDWDWDW]
been proved so in various games)
[WDB)WDWD]
18...c5 19 dxc5 Îxc5 20 Îxc5 Ëxc5 21 [DWDW)NDW]
Îc1 Ëe7 22 Ìd4 Ëg5 23 Íg4 Ëd5 24 [P!WDW)P)]
Íf3 Ëd7 25 Íxb7 Ìxb7 26 Ìc6 Îa8 27 [$WDWIWDR]
Ëd4 reaching a position which puzzles W--------W
me. Kramnik won a splendid ending, Black can choose between the Lutz
which Lutz analyses wonderfully in plan of 14...Ìa5 15 Íe2 b6 16 0-0 Íb7
Endgame Secrets and agrees with 17 Îac1 Îac8 (Atalik), and the Yusupov
Kramnik’s suggestion that 27...Ëe6 28 plan of 14...Íg4 15 Íe2 (if 15 Ëxb7?
Ëc4 Ëe8 was the way to defend, ‘fol- Íxf3 16 gxf3 Ìxd4) 15...Íxf3 16 gxf3
lowed by ...Ìc5, when the knight on c6 Îab8, with approximate equality in
has to retreat and White isn’t better at either case.
all’. I don’t understand this; it seems to c) 14 Ëc3 Íg4 15 0-0 (White can
me that White can defend the knight hardly avoid the doubled f-pawns since
by Ëd5 and follow with Îd1 with a 15 Íe2 Íxf3 16 Íxf3 allows 16...Ìxd4)

67
Declining the Queen’s Gambit

15...Íxf3 16 gxf3 Ëf6! (it’s important pov managed to hold.

to hit f3 and cover c6 at the same time) c2) 18 Îac1 Ìe7 19 Êh1 (Babula’s
17 Íe2 Îac8! is ‘ a world-class move’ 19 Ëa5 could perfectly well be met
(Sadler), and was part of Yusupov’s with 19...a6 since 20 Îxc7?? loses to
preparation for his 1989 Candidates’ 20...Ìc6 21 Ëb6 Ìxd4, but as Dautov
match against Karpov. Now: points out Black can also make an im-
W________W mediate draw with 19...Îd5 20 Ëxa7
[WDr4WDkD] Îh5 21 Îc5 Ìd5 22 Ëxb7 Ëg5 23 Êh1
[0p0WDp0W] Îxh2+, and perpetual), recommended
by various people in 1989 and subse-
[WDnDW1W0]
quently, can be met by 19...b6 20 Ëb4
[DWDWDWDW] (or 20 Ëa3 c5 21 dxc5 Îd2) 20...c5,
[WDW)WDWD] since 21 dxc5 bxc5 22 Îxc5?? loses to
[DW!W)PDW] 22...Ìd5 followed by ...Îxc5 and ...Ìc3.
[PDWDB)W)] d) Finally, 14 Íd5 has never been
[$WDWDRIW] played, although the fact disappoints
W--------W my computer; a clean response seems
c1) 18 Îab1 b6 19 Îfc1 (alterna- to be 14...Ìa5 (14...Ìxd4 15 Ìxd4 c6
tively, 19 f4 Ìe7 20 Ía6 was well met 16 Ìxc6 bxc6 17 Íxc6 Ëf6 18 Îc1 Íe6
by 20...Ëg6+ 21 Êh1 Ìd5 22 Ëd3 19 Ëa4 Îd2 20 0-0 is spectacular but
Ëxd3 23 Íxd3 c5 in M.Kobalija- doesn’t quite give enough compensa-
O.Korneev, Krasnodar 1998, and 19 tion) 15 Ëb5 c5 16 Íxf7+ Ëxf7 17
Ía6 Ëxf3 20 Íxc8 Îxc8 21 Îfc1 Ìe7 Ëxa5 b6 18 Ëb5 Íb7 19 dxc5 bxc5 20
gives Black good compensation) Ëxc5 Íxf3 21 gxf3 Îac8 22 Ëe5 Ëxf3
19...Ìe7 20 Êh1 Îd5 21 Ëc2 (accord- 23 Îg1 Ëb7 with good compensation
ing to Yusupov White is already worse for the pawn.
and his best is 21 Ëe1 c5; White’s trou- Returning to 12...Íe6:
ble is that he is not in time for 21 f4 c5 W________W
22 Íf3 in view of 22...cxd4 23 Ëxc8+ [rhW4WDkD]
Ìxc8) 21...Ëh4, and now only the [0p0W1p0W]
much-praised 22 f4 enabled White to
[WDWDbDW0]
hang on in the stem game A.Karpov-
A.Yusupov, 5th matchgame, London
[DWDpDWDW]
1989, and I’m not sure even that would
[WDP)WDWD]
have done the trick if Black had contin- [DQDW)NDW]
ued more calmly with 22...c5, rather [PDWDW)P)]
than the game’s 22...Ëxf2, which led to [$WDWIBDR]
an instructive rook ending which Kar- W--------W

68
 The Lasker Defence

13 Ëxb7 only is he a tempo or so up, but also he

If this move isn’t good for White can get in 15...c5 right away.
then probably Black’s last move was 13...Ëa3!
simply good. Moreover, White’s other 13...dxc4 14 Ëxa8 Ëa3 15 Îb1 Íd5
tries have so far looked like bringing was played quite a lot before White
him more grief than Black: discovered 16 Ëxd5!, which looked as
a) 13 Íe2 c5 14 0-0 Ìc6 was obvi- though it had closed this line down,
ously fine for Black in V.Malakhatko- but Shengelia’s idea of inverting the
T.Sanikidze, Baku 2007. moves has breathed new life into it.
b) 13 Îc1 c5 14 Ëa3 Ìd7 15 cxd5 Having said that, the untried 13...dxc4
Íxd5 16 Íe2 b6 gave White nothing in 14 Ëxa8 Ëb4+ looks quite plausible;
S.Skembris-A.Mastrovasilis, Salonica after 15 Ìd2 c3 White has the choice
2006. between a draw with 16 Îd1 cxd2+ 17
c) 13 c5 b6 14 Îc1 bxc5 15 Ëa3 (15 Îxd2 Ëb1+ 18 Îd1 Ëb4+, and so on, or
Îxc5 Ìd7 16 Îxc7 Îab8 17 Ëc2 Ëa3 a very unclear position with 16 Íd3
18 Íd3 Îb2 saw Black regaining his cxd2+ 17 Êe2 Ëc3. For example:
pawn with equality in P.Nikolic-D.Jojua, W________W
European Championship, Plovdiv 2008) [QhW4WDkD]
15...Ìd7 16 Íb5 Íg4 (‘the point of [0W0WDp0W]
Black’s play’, according to Anand, but
[WDWDbDW0]
16...Îab8 17 Íxd7 Ëxd7 18 Ëxc5 Ëb5
looked fine for Black too in
[DWDWDWDW]
V.Malakhatko-D.Jojua, Arcapita 2009)
[WDW)WDWD]
17 Íxd7 Îxd7 (this ‘works tactically’, [DW1B)WDW]
as Anand says) 18 Ëxc5 (both 18 Îxc5 [PDW0K)P)]
Ëe4 19 Êe2 Îd6 20 h3 Íc8! and the [$WDWDWDR]
cute 18 Ìe5 cxd4! 19 Ëxe7 Îxe7 20 W--------W
Ìxg4 h5! demonstrate Black’s point) a) 18 Îab1 Ìc6 19 Ëxd8+ (or 19
18...Ëe4 19 Îg1 Îe8! was at least equal Ëb7 Íc4 20 Íxc4 Ìxd4+ 21 exd4
for Black in the recent high-profile Îe8+) 19...Ìxd8 20 Îb8 g6 21 Îxd8+
game V.Topalov-V.Anand, Nanjing Êg7 22 Îd1 Ëa5! is better for Black;
2010; Black’s idea is 20 Ìe5 Îxe5 21 ...Ëg5 is on the cards.
dxe5 d4 22 h3 dxe3 23 hxg4 exf2+ 24 b) 18 Ëxa7 Ìc6 19 Ëa4 (not 19 Ëc5
Êxf2 Ëf4+ 25 Êe1 Îd4, and wins. Ìxd4+! 20 exd4 Íg4+) leaves Black
d) 13 cxd5 Íxd5 14 Íc4 Íxf3 15 gxf3 with a great many interesting tries
(½-½ K.Sasikiran-O. Korneev, Corsica which are fun to analyse, or an imme-
(rapid) 2005), is a clear improvement diate draw with 19...Îxd4 20 exd4
for Black on Karpov-Yusupov, since not Ìxd4+ 21 Êe3 Ìf5+ 22 Êe2 Ìd4+.

69
Declining the Queen’s Gambit

c) 18 Ëe4 g6 19 Ëf4 (if 19 Îab1 Íf5 excellent pawn sacrifice 18...c5: for ex-
20 Îb3 Ëc1!) 19...Ìc6 20 f3 Ìb4 21 ample, 19 Íxc4 Íf5 20 Îc3 Îb2 21
Íe4 f5 22 a3 fxe4 23 axb4 exf3+ 24 Íb3 cxd4 22 exd4 Ìf6, maintaining a
gxf3 d1Ë+ 25 Îaxd1 Ëb2+ 26 Êe1 Íc4 definite pull.
27 Îd2 Ëc1+ 28 Îd1 Ëb2 with a draw. 15 cxd5 Íf5 16 Îc1
14 Îb1 16 Îd1 was perhaps a sterner test:
Instead 14 Ìd2 dxc4 15 Íxc4 16...Ìd7 17 Ëa6 Ëb3!? (17...Ëxd5 18
(White settled for 15 Íe2 Íd5 16 Ëb1 Íc4 Ëd6 19 Ëa2 Ëb4+ 20 Ìd2 enables
c5 with equality at least for Black in White to castle and perhaps claim a
A.Lugovoi-A.Alavkin, Russian Team slight advantage) 18 Ëc4 seems to en-
Championship 2003) 15...Íxc4 16 able White to consolidate and claim a
Ëxc7 Ìc6! 17 Ëxc6 Îac8 18 Ëb7 Ëc3! slight edge (by contrast 18 Íd3 Íxd3
ties White in knots, and 14 Ëb3 Ëa5+ 19 Îxd3 Ëxd5 20 Îc3 c5 21 Ëa5 Ìe5!?
15 Êd1 (if 15 Ìd2 c5!) 15...c5 is also at produces some wonderful complica-
least equal for Black (15...Íd7 16 a4 is tions, probably about equal): for in-
not quite so effective). stance, 18...Ëa3 19 Îc1 Îab8 20 Íe2
W________W Ìb6 21 Ëc5 Ëb2 22 Ìe5 Ìd7 23 Ìxd7
[rhW4WDkD] Îxd7 24 Íf3.
[0Q0WDp0W] 16...Ëa3
[WDWDbDW0] W________W
[DWDpDWDW] [rhW4WDkD]
[WDP)WDWD] [0Q0WDp0W]
[1WDW)NDW] [WDWDwDW0]
[PDWDW)P)] [DWDPDbDW]
[DRDWIBDR] [WDw)WDWD]
W--------W [1WDW)NDW]
14...Ëxa2 [wDWDW)P)]
14...Ìd7 is also an interesting move, [Dw$WIBDR]
since 15 Ëb2 (not 15 cxd5 Îab8 16 W--------W
dxe6 Îxb7 17 exf7+ Êf8 18 Îxb7 Ìc5!, 17 Ëxc7?
when Black wins) 15...Ëxb2 16 Îxb2 It’s easy to say once you’ve seen
dxc4 looks a little better for Black and Black’s reply, but this move is a mis-
it’s hard to see what else White can take. It’s no surprise that White didn’t
play. This might be important if White fancy 17 Îxc7, but this was the critical
can indeed do better than Jobava does move: for example, 17...Ëa1+ 18 Êe2
in the text line. It is important that af- (18 Êd2 Ëa2+ 19 Êd1 Ëa4+ 20 Êe1
ter 17 Îc2 Îab8 18 Ìd2 Black has the Ëa1+ repeats) 18...Ìd7 19 Ìd2 Ìb6 20

70
 The Lasker Defence

h3 (20 Îxf7 Íd7 21 Îe7 Îab8 22 Îxd7 hoping to be recommending a promis-

Ìxd7 23 Ëc6 Îb2 is terrible for White) ing and slightly unknown line, but To-
20...Íc8 21 Ëc6 Ìxd5 22 Ëc1, al- palov’s miserable failure in Topalov-
though here too Black has much the Anand has rather brought it into the
easier play. spotlight.
17...Ìc6!!
W________W
[rDW4WDkD] Game 17
[0W!WDp0W] V.Ikonnikov-A.Gasthofer
[WDnDWDW0] Hockenheim 2006
[DWDPDbDW] 1 d4 d5 2 Ìf3 Ìf6 3 c4 e6 4 Ìc3 Íe7 5
[WDW)WDWD] Íg5 h6 6 Íh4 0-0 7 e3 Ìe4 8 Íxe7
[1WDW)NDW] Ëxe7 9 Îc1
[WDWDW)P)] In modern play this is very much
[DW$WIBDR] White’s main try against the Lasker
W--------W Defence.
18 Îd1 9...Ìxc3 10 Îxc3 c6
White is busted. 18 Ëxc6 Îac8 19 W________W
Îc5 Îxc6 20 dxc6 Îb8 21 Ìd2 Íg4 is [rhbDW4kD]
hopeless, and 18 Îxc6 Îac8 19 Ëb7 [0pDW1p0W]
Îb8 20 Ëa6 Îb1+ 21 Êe2 Îb2+ 22
[WDpDpDW0]
Ìd2 Ëb4 23 Êf3 Ëxd2 24 Ëe2 Ëb4! is
also extremely unpleasant.
[DWDpDWDW]
18...Îac8 19 Ëf4 Íc2 20 dxc6 Íxd1 21
[WDP)WDWD]
Êxd1 Îxc6 22 Ìd2 Ëc1+ 23 Êe2 Îc2 [DW$W)NDW]
24 Êf3 Ëxd2 25 Êg3 Îc3 26 Êg4 Ëd1+ [P)WDW)P)]
27 f3 Ëd2 28 e4 Îxd4 29 Ëe5 Îd6 30 [DWDQIBDR]
Ëe8+ Êh7 31 Ëxf7 Ëg5+ 32 Êh3 Îf6 W--------W
33 Ëd5 Îfxf3+ 34 g3 Îxg3+ 0-1 11 Íd3
A natural developing move. Of
Conclusion course White would like to find some
12...Íe6 is a powerful idea which for more waiting moves to play before
the moment has driven White away ...dxc4 happens, but after, say, 11 Ëc2
from this line. Black needs to be confi- Ìd7 he doesn’t really have another
dent in the critical 13 Ëxb7 lines, but one.
otherwise White has had little that’s White can also play the restrained
challenging to offer so far. I was rather 11 Íe2, whose idea is to avoid the in-

71
Declining the Queen’s Gambit

termezzo 11 Íd3 Ìd7 12 0-0 e5 13 12...Îe8 now has less point, since 13
cxd5 e4, discussed in the next note. Ëc2 e5 14 dxe5 dxc4? loses material to
Black can try 11...Ìd7 12 0-0 Îe8, as in 15 Îxc4 Ìxe5 16 Ìxe5 Ëxe5 17 Îe4)
V.Ivanchuk-A.Onischuk, World Team 13 dxe5 (one of Black’s points is that 13
Championship, Beersheba 2005, which cxd5?! loses a piece to 13...e4, although
went 13 Ëc2 dxc4 14 Íxc4 e5 15 d5 White is not actually lost after 14 dxc6
Ìb6 16 dxc6 bxc6 17 Îc1 Ìd5 18 Îb3 bxc6 15 Íe2 exf3 16 Íxf3 Íb7 17
a5 19 Ëe4 a4 20 Îa3, when 20...Ëb7 is Íxc6 Íxc6 18 Îxc6) 13...dxc4! (this
said to give Black good chances to intermezzo is the other main point) 14
equalize, although I think Black would Íxc4 (14 Îxc4 Ìxe5 15 Îe4 Ìxf3+ 16
do better to stick with the line in the Ëxf3 Íe6 17 Íc4 Îad8 18 Íxe6 fxe6
main game. Indeed, 12...dxc4 13 Îxc4 gives White nothing despite the visu-
(13 Íxc4 would be the main line, of ally weak e-pawn, and was quickly
course) 13...e5 14 Ëc2 exd4 15 Ìxd4 drawn in M.Sadler-V.Kramnik, Tilburg
Ìb6 16 Îc5 Îd8 17 Íf3 Íe6 18 Îc1 1998, among other games) 14...Ìxe5
Íd5 was nothing for White in 15 Ìxe5 Ëxe5.
T.Radjabov-V.Kramnik, 1st matchgame, W________W
Kazan 2011. [rDbDW4kD]
W________W [0pDWDp0W]
[rhbDW4kD] [WDpDWDW0]
[0pDW1p0W] [DWDW1WDW]
[WDpDpDW0] [WDBDWDWD]
[DWDpDWDW] [DW$W)WDW]
[WDP)WDWD] [P)WDW)P)]
[DW$B)NDW] [DWDQDRIW]
[P)WDW)P)] W--------W
[DWDQIwDR] This position is the same as that
W--------W reached in the old main line of the Or-
11...dxc4 thodox QGD (after 6...Ìbd7 7 Îc1 c6 8
Obviously before he can play ...e5 Íd3 dxc4 9 Íxc4 Ìd5 – Capablanca’s
Black needs to play this move in order legendary ‘freeing manoeuvre’, exten-
to prevent himself from being left after sively tested in the 1927 Alekhine-
...e5; cxd5 with a weak and isolated d- Capablanca match – 10 Íxe7 Ëxe7 11
pawn, but he does have a clever alter- 0-0 Ìxc3 12 Îxc3 e5 13 dxe5 Ìxe5 14
native in 11...Ìd7. The idea of this Ìxe5 Ëxe5), save that Black’s h-pawn
move is well seen after 12 0-0 e5! (by is on h6, which is a small improvement
contrast to Ivanchuk-Onischuk, for him:

72
 The Lasker Defence

a) 16 Ëb3 b6 (I like this move of when the tragic position of Black’s h-

Oleg Korneev’s – a great Lasker expert – pawn on h6 is meant to mean that he
better than 16...b5, though the latter can’t sensibly evict a knight from e5
has Kramnik’s authority behind it; 17 with f6 because of the weakness of g6,
Íe2 Íe6 18 Ëc2 Íd5 was all very well and thus has a bad version of the typi-
in J.Lautier-V.Kramnik, Monte Carlo cal Carlsbad structure.
(rapid) 1999, but I don’t see how Black W________W
was going to meet 17 Íd3, since obvi- [rDbDW4kD]
ously 17...Íe6 18 Ëc2 Íd5? 19 e4 [0pDn1p0W]
won’t do, and otherwise Black gets a bit
[WDpDWDW0]
penned in) 17 Îd1 Íf5 reveals the
point of 16...b6; White has to waste
[DWDpDWDW]
time with his rook on c3 and this
[WDW)WDWD]
should enable Black to equalize: for [DW$B)NDW]
example, 18 h3 (18 f3 b5 is now fine) [P)WDW)P)]
18...Ëe7 (a good move, stopping Íxf7+ [DWDQDRIW]
tricks) 19 Íd3 Íe6 20 Ëa4 c5 21 Îc2 W--------W
Îad8 22 Îcd2 g6. This is, however, pretty much non-
b) 16 f4 is Rubinstein’s idea in the sense, as Khalifman points out, and
analogous Orthodox line; there was a White hasn’t been able to make much of
great deal of theory on this line 60 his position in high-level games. The
years ago, none of which gave White main reference is still P.Nikolic-
anything much, but according to Sadler A.Yusupov, Horgen 1994: 13...Ìf6 (Yu-
the ‘standard line’ is 16...Ëe4 17 Ëe2 supov actually played 13...Îe8 first in the
when, rather than Sadler’s 17...Íf5, far belief that after 13...Ìf6 14 Ìe5 was
commoner and more efficient is in fact good for White but in fact this is not
17...Îd8, meeting 18 Íd3 with the true; see the next note) 14 Ëb1 (the rea-
ruthless equalizer 18...Íg4. son for Yusupov’s belief was the line 14
c) Recently 16 Ëc2 Íf5 17 Íd3 Ìe5 Ìd7 15 f4 f6 16 Ìg6 Ëxe3+ 17
Íxd3 18 Îxd3 Îad8 19 Îfd1 Îxd3 20 Êh1, when Black loses the exchange in
Îxd3 Ëa5 21 h3 Ëxa2 22 Îa3 Ëd5 23 view of 17...Îe8 18 Ìe5!, winning the
Îxa7 c5, reaching a pretty level end- queen, but after the simple 15...Ìxe5
game, was seen twice in the Lublin GM followed after either recapture by ...f6
event, where both E.Alekseev-B.Grachev Black is fine, as Khalifman pointed out)
and, a round later, R.Wojtaszek- 14...Îe8 (Black can also play 14...a5,
E.Alekseev were shortly drawn. which may be more reliable, when 15
The problem with 11...Ìd7 is some- Îfc1 Îe8 would transpose to L.Ftacnik-
times said to be 12 cxd5 exd5 13 0-0 R.Vaganian, German League 2004, which

73
Declining the Queen’s Gambit

soon levelled out after 16 Ìe5 Ìg4; in- with 12...Ìd7 13 0-0 e5, when the rook
stead 15 Îa3 b6 seems fine for Black to on c4 is simply stupid and the bishop
me, although this is Khalifman’s rec- on d3 also less than ideal for an IQP
ommended line: for example, 16 Îb3 position (it is better on b3 as in the
Ëc7 17 Îc3 c5 18 Ìe5 Ëd6 19 f3 c4 20 game).
Íc2 b5) 15 b4 (15 Ìe5 Ìe4! is similar) 12...Ìd7 13 0-0
15...Ìe4 16 Íxe4 dxe4 17 Ìd2 Íe6 18 Radjabov preferred 13 Ëc2 second
Îfc1 (if 18 b5 Íd5, or 18 Ìxe4 Íf5 19 f3 time around in T.Radjabov-V.Kramnik,
Íxe4 20 fxe4 Ëxe4 21 Ëxe4 Îxe4) 3rd matchgame, Kazan 2011, when
18...Íd5 19 b5 Îad8. Now 20 a4 h5 was there was nothing wrong with 13...e5
unclear in the game, but Yusupov, fol- with a likely transposition after 14 0-0
lowed by Khalifman, gave 20 bxc6 (20 to the notes to White’s 14th, below (in-
Îc5 Îd6 21 bxc6 Íxc6 22 Ìc4 Îg6 is stead 14 d5 cxd5 15 Íxd5 Ìf6 looks
okay for Black) 20...bxc6 21 Îc5 with the like an easy equalizer, although the
idea of Ìc4-e5 (playing Ìc4 before Îc5 bolder 14...Ìb6!? 15 dxc6 Ìxc4 16
allows ...c5) as a little better for White, Îxc4 bxc6 17 Îxc6 Íb7 was preferred
but I don’t really see it after 21...Îb8 22 in M.Hoffman-O.Korneev, Vila Nova de
Ëc2 Îb7: for example, 23 Ìc4 Îeb8 24 Gaia 2010, while 14 Ëe4 Îe8 15 0-0
Ìe5 Îb2 25 Ëd1 Îxa2 26 Ìxc6 Íxc6 27 exd4 16 Ëxe7 Îxe7 17 Ìxd4 wasn’t all
Îxc6 Îbb2, when Black can hardly lose. that impressive for White in M.Carlsen-
It seems to me that 20 Îc5 would be A.Volokitin, Foros 2008; I couldn’t tell
more to the point; I don’t see why White you why Black now avoided 17...Ìe5
needed to open the b-file. and, earlier, 14...Ëb4 15 0-0 exd4 16
Returning to the immediate ex- Ìxd4 Ìf6 deserved definite attention
change on c4: too, but not 16...Ëxb2? 17 Îb3 Ëd2 18
W________W Ìf5 with a virulent attack), but Kram-
[rhbDW4kD] nik preferred 13...b6 14 Íd3 (not
[0pDW1p0W] forced, but 14 0-0 Íb7 followed by ...c5
has long been known to be completely
[WDpDpDW0]
equal) 14...Ìf6! (an idea of
[DWDWDWDW] Kasimdzhanov’s; 14...Íb7 15 Íe4
[WDp)WDWD] would have fallen in with White’s
[DW$B)NDW] plans) 15 Îxc6 Ìd5 16 Ëb3 Ìb4 (this
[P)WDW)P)] manoeuvre is Black’s main point) 17
[DWDQIWDR] Îc1 Ìxd3+ 18 Ëxd3 Íb7 19 0-0 Íxf3
W--------W 20 gxf3 Ëg5+ 21 Êh1 Ëd5, regaining
12 Íxc4 his pawn and simplifying towards the
12 Îxc4 can be met as in the game resulting draw.

74
 The Lasker Defence

W________W 19 e4 got nowhere in V.Kunin-

[rDbDW4kD] M.Prusikin, Griesheim 2004) 16...Îd8
[0pDn1p0W] 17 Îcd3 is ‘+=’ according to Grivas, but
after 17...Íd7 I think Black has only to
[WDpDpDW0]
play a few accurate moves to equalize:
[DWDWDWDW] for example, 18 a4 Îac8 with the idea
[WDB)WDWD] of ...c5, or 18 Ëd2 Îe8 (to prevent Ìf5).
[DW$W)NDW] b) 15 exd4 leads to similar positions
[P)WDW)P)] as those in the notes to Black’s 14th,
[DWDQDRIW] below: 15...Ëd6 16 Îe1 Ìb6 (one point
W--------W of not having the bishop on b3 is that
13...e5 Black can’t manage a ...Îd8 and ...Ìe6
Of course 13...b6 is the main alter- regrouping, because if he moves the
native, which Anand unkindly gave his rook Ëb3 follows) 17 Íb3 Ìd5 18 Ìe5
seal of approval to in his World Cham- Íe6 19 Îg3 Ìc7 is the same critical
pionship match with Topalov just after regrouping; Black’s key idea in these
I had finished this chapter, but I am lines is to get the knight somewhere
recommending the less common text. where it supports ...Íe6.
14 Íb3 W________W
Karpov’s widely praised move, al- [rDWDW4kD]
though White has a reasonable alter- [0phWDp0W]
native in 14 Ëc2, which was warmly
[WDp1bDW0]
recommended by Efstratios Grivas in
New in Chess Yearbook 90, after which
[DWDWHWDW]
14...exd4 (14...Îe8 15 d5 would be Iv-
[WDW)WDWD]
anchuk-Onischuk above, and is per- [DBDWDW$W]
haps slightly better for White; since [P)QDW)P)]
Ëc2 isn’t massively useful in an IQP [DWDW$WIW]
position, it makes sense to make the W--------W
exchange at once, but 14...Îd8 is also a After 20 Îe4 (the less logical 20 Ëe4
sensible move, when Grivas gave 15 was played in E.Zude-M.Becker, Ger-
Íb3 exd4 16 exd4 Ìf6?!, missing the man League 2006, which went on
point of Black’s set-up, which is to play 20...Íxb3 21 Îxb3 Îab8 22 h3 Îfe8 23
16...Ìf8 and transpose to the note to Îee3 Ìb5 24 Ëf5 Îe7 25 Ìf3 Îd7 26
Black’s 14th, below) gives White a Îe4, and now 26...g6 would have
choice: kicked the white queen away from the
a) The feeble 15 Ìxd4 Ìf6 16 Îd1 kingside and left Black at least equal)
(16 f3 Îd8 17 Ëf2 Ìd5 18 Íxd5 Îxd5 20...Îad8 (an excellent move; Black

75
Declining the Queen’s Gambit

should keep the bishop preventing Îg4 W________W

for the moment, and the rook on d8 is [rDb4WDkD]
more valuable than after Grivas’ [0pDnDp0W]
20...Îae8) 21 h3 (21 Îh4 would be
[WDp1WDW0]
more in line with Grivas’ ideas; after
21...Íxb3 22 axb3 Îfe8 or 22 Ëxb3
[DWDWDWDW]
Ìe6 Black will need to play a few accu-
[WDW)WDWD]
rate moves, but equally White can eas- [DB$WDNDW]
ily get his rooks stuck offside: for ex- [P)WDW)P)]
ample, 23 h3 Ëf6 24 Îh5 Ëe6 25 Ìg4 [DWDQ$WIW]
Êh8, and the position is still tense) W--------W
21...Íxb3 22 Ëxb3 Ìe6 23 Ëxb7 Ìxd4 Now:
was equal in A.Grischuk-D.Jakovenko, a) 17 Ëe2 Ìf8 18 Ëe7 Ìe6 19 Ëh4
Khanty-Mansiysk 2009. Ìf8 (this is equal according to Dautov;
Instead 14 dxe5 Ìxe5 15 Ìxe5 19...Ìc7 deserves consideration too,
Ëxe5 would be Lautier-Kramnik above; though not 19...Ëf4 20 Ëxd8+ Ìxd8 21
the point of Black’s move order finesse Îe8+ Êh7 22 Îxd8) 20 Îce3 (20 Ìe5
in that game, with 12...e5 before the Íe6 is fine for Black) 20...Íe6 is
exchange on c4, is to force White into Khalifman’s repertoire suggestion. He
this position by the threat of ...e4. gives White the better of it ‘according
Finally, the enigmatic 14 Ëb1 is also to I.Donev’, but I think Black can be
played, and best met by 14...exd4 15 happy enough; computers tend to
exd4 Ìb6 16 Íb3 Ëd6 17 Îe1 Ìd5. overrate these positions for White, and
14...Îe8 Black can play to win them as well.
I am recommending this move, an- b) 17 d5 cxd5 18 Ëxd5 Ëxd5 19
other clever idea of Yusupov’s, al- Íxd5 Ìf8 20 Íe4 Íe6 doesn’t give
though Black has an excellent alterna- Black too much trouble, as Khalifman
tive too, namely 14...exd4 (this was says.
how Yusupov was slaughtered in the c) 17 Îce3 Ìf8 18 Ìe5 Íe6 19 Ëh5
famous eighth and decisive game of his (after a semi-waiting move, such as 19
Candidates’ final with Karpov in Lon- Îe4, Black’s best next move is usually
don 1989, but he too went for a ...Ìf6 19...Îe8, and certainly not 19...Íxb3?
plan, and Black can do much better) 15 20 Ëxb3 Ëc7 21 Îf4; instead Sadler
exd4 (others are feeble really, but if claimed that 19 Íxe6 ‘was powerful’
Black is bothered about them he can with the idea of 19...Ìxe6? 20 Ëh5,
play first 14...Îd8 and only after seeing presumably overlooking 20 Ìxf7! and
White’s 15th, 15...exd4) 15...Îd8 16 Îe1 in any event missing Black’s point,
Ëd6. which is 19...Ëxe6 followed by taking

76
 The Lasker Defence

the queen to d5, when Black is fine) (see Ivanchuk-Onischuk, above), and
19...Íxb3 20 Îxb3 Ëc7 21 Ëg4, and 15...cxd5 gives Black no particular
now 21...Îd6 22 Îf3 Îe8 is even a little trouble in equalizing: for example, 16
better for Black according to Dautov, Ëxd5 Ìf6 17 Ëc5 Ìe4 18 Ëxe7 Îxe7
while 21...Ìe6 22 Ìxf7 Ëxf7 23 Îxe6 19 Îc4 Ìg5 20 Ìxg5 hxg5 has led to a
Îxd4 was also good enough for equal- number of draws, starting with the
ity in R.Leitao-I.Morovic Fernandez, stem game P.Nikolic-A.Yusupov, Bel-
Mario Covas 2003. grade 1989.
This is a very interesting alternative b) 15 Ëc2 exd4 and then:
and might even be my choice next time b1) Vyzmanavin claimed that White
if I can’t persuade myself that Hebden was better after 16 Ìxd4 Ìf6 17 f3,
and Grischuk’s 15 Îe1 in the next note but I don’t believe it: Black should just
is really as equal as it looks when An- develop quietly, put his rooks on the c-
and is Black. Computers tend to over- and d-files, and wait for a good mo-
rate White’s position in this line – in ment for ....c5: for example, 17...Íd7 18
the fullness of time Black tends to e4 Îac8 19 Îc1 Îed8 20 Ìe2 b6 21 Ìf4
equalize and if White is inaccurate Ëd6, and Black is fine thanks to the
Black can even gradually obtain tactical turn 22 Ìd3 c5 23 e5 Ëd4+ 24
chances against the IQP. Ëf2 Íf5!.
W________W b2) 16 exd4 Ìf8 17 Îe3 (17 d5 is not
[rDbDrDkD] particularly alarming but has to be met
[0pDn1p0W] calmly: 17...Íd7!, and now a sample
continuation might be 18 Ëd2 Îad8!,
[WDpDWDW0]
not fearing 19 d6 Ëf6: for example, 20
[DWDW0WDW] Îe3 Îxe3 21 fxe3 Ìe6 22 Ìd4 Ëe5 23
[WDW)WDWD] Ìf5 Êh8; playing such a position
[DB$W)NDW] against computers reminds me of Kar-
[P)WDW)P)] pov’s bon mot, ‘Black is worse now, but
[DWDQDRIW] soon he will be better’) 17...Íe6 18
W--------W Îfe1 Ëd6 is much the same as Mager-
15 Ìxe5 ramov-Korneev, below.
This was recommended strongly by c) 15 Ëb1 exd4 16 exd4 (16 Ìxd4
both Sadler and Palliser as the refuta- Ìf6 17 f3 c5 18 Ìf5 Íxf5 19 Ëxf5 is
tion of Black’s system, but as we shall equal, as Khalifman points out)
see it isn’t as strong as they believed. 16...Ìf8 17 Îe1 Íe6 18 Îce3
Others: (E.Magerramov-O.Korneev, Podolsk
a) 15 d5 is less effective than in the 1992) shows Black’s basic idea well: to
same position with Ëc2 instead of Íb3 enter IQP positions with his ...Îe8 wait-

77
Declining the Queen’s Gambit

ing move being more useful than 15...Ìxe5 16 dxe5 Ëxe5 17 f4

whatever White plays on move 15. W________W
W________W [rDbDrDkD]
[rDWDrhkD] [0pDWDp0W]
[0pDW1p0W] [WDpDWDW0]
[WDpDbDW0] [DWDW1WDW]
[DWDWDWDW] [WDWDW)WD]
[WDW)WDWD] [DB$W)WDW]
[DBDW$NDW] [P)WDWDP)]
[P)WDW)P)] [DWDQDRIW]
[DQDW$WIW] W--------W
W--------W 17...Ëf6
Black’s best now is the ultra- 17...Ëe4 is also entirely possible;
methodical 18...Ëd6 19 Ëd3 Îe7, fol- both Sadler and Palliser believed that
lowed by doubling on the e-file or White had 18 f5 (18 Íc2 is more sensi-
...Îd8 as appropriate, with a very solid ble, but White doesn’t have much after
position where White can easily find 18...Ëb4: 19 f5 Ëxb2 20 Ëd3 Ëb6 21 f6
himself running out of ideas. g6 is scary, but I don’t think White ob-
d) 15 Îe1 is a clever try, played jectively has enough, or 19 Îb3 Ëc5 20
twice by Grischuk, which prevents Ëd2 f5), because 18...Íxf5 loses ‘mate-
15...exd4 because after 16 exd4 Ëf8 17 rial’ (Sadler) or ‘a piece’ (Palliser) to 19
Îxe8 Ëxe8 18 Îe3 Ëf8 White prevents Íc2, but software instantly points out
the ...Ìf8/...Íe6 manoeuvre and is that this is not true: 19...Íg4 (actually
much better. Instead 15...e4 16 Ìd2 this turn has been known since Rubin-
Ìf6 17 Îc5 (17 Ëb1 was Grischuk’s try stein’s time) 20 Íxe4 Íxd1 21 Íxc6
in a later game, preventing ...Íe6, and bxc6 22 Îxd1 Îad8 is a very drawish
could be well met by 17...Ëc7, renew- rook ending.
ing the threat and meeting 18 Íc2 18 f5
with 18...Ìd5 followed by ...f5) 17...Íe6 This is the point, cramping Black’s
18 Îe5 Íxb3 19 Ëxb3 (19 Îxe7 Íxd1 bishop.
20 Îxe8+ Îxe8 21 Îxd1 looks equal to 18...b6
me, but after beating me with con- An obvious reply, but Black has al-
summate ease from this position Mark ternatives: 18...Îd8? (inexplicably ex-
Hebden assured me that White is changing one of Black’s two developed
slightly better) 19...Ëc7 was A.Gris- pieces instead of developing the other
chuk-V.Anand, Moscow (blitz) 2009, two) 19 Îd3 (Sadler) isn’t one of them,
which was pretty equal. but either 18...Îe5 to develop the

78
 The Lasker Defence

bishop on f5 after all, or (my favourite) b4) 19 a4 is probably best; Black’s

18...a5!? looks entirely possible. For ex- idea is that now 19...Îe5 is an im-
ample: proved version: 20 Íc2 (White meekly
a) 18...Îe5 is possible but fairly defends the pawn; 20 Îd3 is now less
tenuous: 19 Îd3 Íxf5 20 Îd7 Îf8 (a effective since there is no a-pawn en
difficult decision; Black might have bet- prise at the end: 20...Íxf5 21 Îd7 Îf8 –
ter chances to hold after 20...Ëg5, for 21...Ëg5 is also equal – 22 Îxb7 Îxe3
instance 21 Îxf7 Íe6 22 Íxe6 Îxe6 23 23 Ëd2 Îe5 24 Ëd4 Îe1) 20...c5 21 Îd3
Ëd7 Îae8 24 Ëxb7 Ëxe3+ 25 Êh1 Îe8 22 b3 b6 23 Îd6 Ëc3 24 e4 Íb7 25
Ëd3, but it isn’t much fun) 21 Îxb7 Îe1 c4!, and Black holds thanks to
Îxe3 22 Ëd2 favours White (22 g4 some tactics, basically the fact that 26
Îxb3 23 Îxb3 Îd8 followed by ...Ëd4+ Îxb6 Îad8 is very good for him.
was Black’s point). 19 Îf4?!
b) 18...a5 maintains that the bishop This is natural somehow, but I don’t
is developed just fine where it is so think it’s the best move. White’s most
long as the queen’s rook can get out surgical solution is to exchange the
round the side, and also has some nice piece which is preventing him from
tactical points: realizing his idea of f6: 19 Ëd4 Ëxd4 (if
W________W 19...Îe5 20 Îxc6, of course, but
[rDbDrDkD] 19...c5!? 20 Ëxf6 gxf6 might well be the
[DpDWDp0W] best try) 20 exd4 Íb7 (20...Ía6 is no
better: 21 Îff3 Îac8 22 f6 gxf6 23 h3;
[WDpDW1W0]
this idea of doubling on the third to
[0WDWDPDW] enable White to challenge the e-file in
[WDWDWDWD] some lines and continue his attack on
[DB$W)WDW] the kingside in others – 23...Êg7 24
[P)WDWDP)] Îf5! – is very hard for humans to see)
[DWDQDRIW] 21 f6 g6 22 Îff3 Îad8 23 Îce3 c5 24
W--------W Îxe8+ Îxe8 25 Îg3 Êf8 26 dxc5 Îe2
b1) Black’s point is that 19 Îf4 is 27 cxb6 axb6 28 Íc4 Îxb2 29 Îd3, and
well met by 19...a4 when 20 Íxa4 (20 White is much better.
Íc2 a3!) 20...Íxf5 is possible, since 21 19...Ía6 20 Ëh5 Îad8 21 Îa4
g4? runs into 21...Ëg6! 22 Îxf5 Îed8!, White had been relying on this
and 21 Ëf3 is coolly met by 21...Ëe7. move, presumably, but the sight of the
b2) 19 a3 a4 20 Íc2 Îa5 also gives black rooks and the empty spaces
Black’s push with the a-pawn point. around his king and back rank must
b3) 19 Ëd4 Ëxd4 20 exd4 Íd7 21 f6 have given him a somewhat queasy
a4 22 Íc4 Íe6 can’t trouble Black. feeling.

79
Declining the Queen’s Gambit

W________W Conclusion
[WDW4rDkD] Black is as solid in this system as ever, if
[0WDWDp0W] not more than ever.
[b0pDW1W0]
[DWDWDPDQ] Game 18
[RDWDWDWD] M.Gurevich-D.Jakovenko
[DB$W)WDW] Odessa (rapid) 2010
[P)WDWDP)]
[DWDWDWIW] 1 d4 d5 2 c4 e6 3 Ìf3 Ìf6 4 Ìc3 Íe7 5
W--------W Íg5 h6 6 Íh4 0-0 7 e3 Ìe4 8 Íxe7
21...Îe5! 8 Íg3 is best met with the unnatu-
An excellent move which wins the ral 8...Íb4, which reaches a kind of po-
game; presumably White was expect- sition Black would rather like to get
ing only 21...Íb7 22 Îxa7 Îe7 with from the Manhattan Variation, but
reasonable compensation for the can’t manage without throwing in the
pawn. rather undesirable ...g5. Black has good,
22 g4? free play with ...c5: for example, 9 Îc1
Ikonnikov must have lost his bal- c5 10 cxd5 exd5 11 Íd3 Ìc6 12 0-0
ance completely to play such a move; Íxc3 13 bxc3 Íg4 14 dxc5 Ëa5 15 c4
you suspect he hadn’t seen Black’s last Ìxg3 16 hxg3 d4, G.Kamsky-Wang Yue,
at all. Obviously 22 Îxa6?? Îxf5 23 Monte Carlo (rapid) 2009.
Ëe2 Îd2 24 Ëe1 Îff2 wasn’t on, but 8...Ëxe7 9 Ëc2
against a fellow GM you suspect Ikon- 9 Ìxe4 dxe4 10 Ìd2 is a rare bird at
nikov would have gone for 22 Ëxf7+ high level.
Ëxf7 23 Íxf7+ Êxf7 24 Îxa6 Îd1+ 25 W________W
Êf2 Îxf5+ 26 Êe2 Îg1 with an equal [rhbDW4kD]
rook endgame. [0p0W1p0W]
22...Íb7 23 Îxa7 Îe7
[WDWDpDW0]
Black has more than sufficient
compensation now, and White doesn’t
[DWDWDWDW]
manage to find any sort of defence at
[WDP)pDWD]
all; probably there isn’t one. It’s diffi- [DWDW)WDW]
cult to imagine what White was think- [P)WHW)P)]
ing in going in for this. [$WDQIBDR]
24 Ëh3 c5 25 Ëg3 Ëc6 26 Îxb7 Îxb7 W--------W
27 Îc1 Îd2 28 Îf1 Îe7 29 Îf2 Îxf2 30 10...f5 is a perfectly good reply, but
Êxf2 Îd7 0-1 also good is Boensch’s 10...e5 when

80
 The Lasker Defence

White’s only reasonable move is 11 d5 Íh7+ Êh8 16 Íe4 Íb7 with a very
(11 Ìxe4?! exd4 12 Ëxd4?? Îd8 loses a comfortable Semi-Tarrasch-like posi-
piece, while Black was quickly better tion for Black.
after 11 dxe5 Ëxe5 12 Ëc2 Íf5 13 c5?! 10...dxc4
Ìd7 in L.Ftacnik-U.Boensch, German W________W
League 1996), with an unexplored posi- [rhbDW4kD]
tion where Black can choose between [0p0W1p0W]
11...f5 and 11...Íf5. After the latter a
[WDWDpDW0]
couple of possibilities are 12 Íe2 Ìd7
13 0-0 c6 and 12 g4 Íh7 13 Íg2 Ìd7
[DWDWDWDW]
14 Íxe4 Íxe4 15 Ìxe4 Ëb4+, both
[WDp)WDWD]
with interesting play. [DW!W)NDW]
9...Ìxc3 [P)WDW)P)]
W________W [$WDWIBDR]
[rhbDW4kD] W--------W
[0p0W1p0W] Black might as well do this at once,
since he doesn’t have a convenient
[WDWDpDW0]
waiting move – 10...c6 isn’t useful if
[DWDpDWDW] he’s going to play the way he does in
[WDP)wDWD] the game, and after 10...b6 White
[DWhW)NDW] wouldn’t play 11 Íd3?!, allowing Black
[P)QdW)P)] to gain a tempo on the text, but 11
[$WDwIBDR] cxd5.
W--------W 11 Ëxc4
10 Ëxc3 11 Íxc4 doesn’t change the play
After 10 bxc3 the only high-level ex- much; Black is too comfortable with
ample is K.Sasikiran-Z.Azmaiparashvili, ...b6, ...Íb7, ...Ìd7, ...Îfc8, and thereaf-
Pune 2004, which went 10...Ìc6 11 ter either ...a5 or ...c5: for example,
cxd5 exd5 12 Íd3 b6 13 0-0 Íg4 14 D.Navara-V.Korchnoi, Karlsbad 2007,
Ìd2 Ìa5 15 e4 dxe4 16 Ìxe4 Íf5 17 saw 11...b6 12 0-0 Íb7 13 Íe2 Îc8 14
Îfe1 Îae8 18 Îe3 Ëd8 with rough b4 Ìd7 15 Îfc1 c6 (not an obvious
equality; I feel a little more comfortable choice; presumably Korchnoi didn’t
with White, though. It might be me, want to allow 15...a5 16 b5, but after
but I would prefer the play of a player 16...c5 I don’t really understand why
rated 400 points lower in P.Oster- not) 16 Ìd2 a5 17 bxa5 Îxa5 18 Ìc4
meyer-R.Marian, German League 1985, Îa7 19 Ëb2, and at this point 19...b5
which went 10 bxc3 c5 11 Íd3 Ìc6 12 20 Ìd2 e5 would have been a conven-
0-0 dxc4 13 Íxc4 Ìa5 14 Íd3 b6 15 ient equalizer. This example could be

81
Declining the Queen’s Gambit

multiplied by many other master at all with either 14 Ëh4 Ëxh4 15

games stretching back a century or so, Ìxh4 Îc8 16 Íb5 Ìa6 17 c6 Ìb4 18
almost all of them drawn. Êe2 in E.Alekseev-D.Jakovenko, Mos-
11...b6 12 Îc1 cow 2007, or 14 Ëa4 Îc8 15 Íe2 Îxc5
This is the only significant try; oth- 16 Îxc5 Ëxc5 17 0-0 Ìc6 18 Îd1 Ìe5
erwise Black equalizes easily along the in L.Aronian-D.Jakovenko, Sochi 2008.
lines of Navara-Korchnoi. 14...Îc8 15 0-0
12...c5! 15 b4 bxc5 16 b5 is asking to end up
Without this move Black would find worse; for example, after 16...a6 17 a4
himself a little annoyed, needing to axb5 18 axb5 Íd5.
start with 12...c6, although that might 15...Îxc5 16 Ëa4 Ìc6 17 a3 Ìe5 18
well be fine. At the moment, however, Ìxe5 Îxe5 19 Îfd1 Îd5 20 Îxd5 Íxd5
this seems to be an extremely clean 21 Ëf4 e5 22 Ëf5 Ëg5 23 Íg4 Ëxf5 24
equalizer. Íxf5 g6 25 Íg4 f5 26 Íe2 Íe6 27 Îc6
13 dxc5 Íb7 Êf7 28 Îc7+ Êf6 29 f4 Îd8 30 Êf1 exf4
W________W 31 exf4 Îd7 32 Îxd7 Íxd7
[rhWDW4kD] White can point at the f-pawn and
[0bDW1p0W] claim some sort of moral victory, but
that’s all.
[W0WDpDW0]
33 Íc4 g5 34 g3 gxf4 35 gxf4 a5 36 Êf2
[DW)WDWDW] Êe7 37 Êe3 Êd6 38 Êd4 Íc8 39 Íf7
[WDQDWDWD] Íd7 40 Êc4 Íc8 41 Êd4 Íd7 42 Íh5
[DWDW)NDW] Íe6 43 Íg6 Íd7 44 Íf7 Íc8 45 Íg6
[P)WDW)P)] ½-½
[DW$WIBDR]
W--------W Conclusion
14 Íe2 These irregular variations really don’t
It seems that, as Anand would put have anything to offer White, though
it, Black’s last move works tactically; Jakovenko’s 12...c5 is handy to remem-
certainly White demonstrated nothing ber.

82