The Chess Mind

By Dennis Monokroussos.
This is a blog for chess fans by a chess fan, one who loves the beauty of the game and wants to share it with those who are like-minded.
Yet the chess mind is not only a chess mind, and other topics, such as philosophy, may appear from time to time. All material copyrighted.

Friday, April 29, 2005

Responding to 2.Qh5: I Was Right!
A week or so ago, I discussed American GM Hikaru Nakamura's semi-successful trek into the wilderness that is 1.e4 e5 2.Qh5, and suggested that such a cynical opening should be met by the even more derisive 2...Nf6!

Nakamura has since taken to playing this opening in online blitz games - this, despite the fact that I, a player whom he outrates by about 400 points, explicitly warned him about this move. (His response: "lol".) Fortunately, his opponents all played the polite, compliant 2...Nc6 - until yesterday:

Smallville (3194) - nulletokkpokk (3183) [C20]
ICC 3 0 Internet Chess Club, 28.04.2005

1.e4 e5 2.Qh5 Nf6 3.Qxe5+ Be7 4.Be2 Nc6 5.Qf4 d5 6.e5 Ne4 7.Nf3 g5 8.Qe3 g4 9.Nd4 Nxd4 10.Qxd4 Bc5 11.Qa4+ Bd7 12.Qb3 Qh4 13.Qxd5 Qxf2+ 14.Kd1 Qxg2 0-1




Black is always better, according to the computer, and much better after 7...g5! White's moves are forced until the blunder on move 13, but even after the better 13.Qxb7 Qxf2+ 14.Kd1 Qxg2! 15.Rf1 O-O Black has a pretty big advantage.

The moral of the story? Meet junk with junk!
Posted by Dennis Monokroussos on Friday April 29, 2005 at 1:25am. 0 Comments 0 Trackbacks

Saturday, April 23, 2005

Nakamura-Sasikiran and Junk Openings
The chess world was abuzz today due to Hikaru Nakamura's unusual opening choice against Indian GM Krishnan Sasikiran: 1.e4 e5 2.Qh5!? Black won, but White's experiment was a success, as he achieved a significant advantage out of the opening.

Black attempted to meet Nakamura's idea in a "respectful" manner, but I wonder if he should have. Take a look at the previous post, Responding to Junk Openings, then take a careful look at the position after White's second move. Is 2.Qh5 a joke? If so - and maybe even if it isn't - then 2...Nf6 may be the psychologically correct response. What would you do?

Nakamura,Hikaru (2657) - Sasikiran,Krishnan (2642) [C20]
13th Sigeman & Co Copenhagen/Malmoe DEN (7), 22.04.2005

1.e4 e5 2.Qh5




What? Does Nakamura expect to mate his opponent on move 4?? No, not at all - and not because he's hoping for 2...Ke7?? 3.Qxe5#, either. But then what IS the point? For starters, I wouldn't rule out purely psychological motives: Sasikiran is a strong GM, and it would be embarrassing to be known as a player who lost to what we might call the Scholar's Mate Attack - a factor that might put some pressure on him during the game. Second, it might be a bit of showing off, a way of not-so-humbly announcing to the world that "I can play anything and still win". I suspect both motives are at work, but I don't think Nakamura would have played it unless he had some confidence in the move's possibilities. The White queen will end up on g3 or h4, depending on Black's reaction, and both are pretty common squares for a White queen in some middlegames (in Sicilians and French Defenses, for example). What makes the idea unusual is that it seems to get things backward: usually a player stakes out certain squares with pawns, then brings out the minor pieces (knights before bishops, according to textbook wisdom) to help in the battle for the center and to get the king castled, and only then to worry about the queen, rooks and whatever pawn breaks are needed to make further progress. 2.Qh5 might seem less flexible than the normal 2.Nf3, but it forces Black to make committal decisions as well; in particular, Black is more or less forced to fianchetto the king's bishop, and while it isn't horrible there, it's not as active as it would be on the a7-g1 diagonal. 2...Nc6 And now, what ought Black to do? Should he view 2.Qh5 as a junk opening, as a provocation deserving to be one-upped? [2...Nf6!? is the move I always play when facing this in blitz, and it might even be sound! The following is a first attempt at some serious, computer-aided analysis, and surprisingly enough, Black seems to be doing pretty well. 3.Qxe5+ (3.Qh4 Be7 4.Nf3 d5 5.exd5 Nxd5 6.Qe4 Nf6 7.Qxe5 0-0 8.Be2 Nc6 9.Qf4 Nb4 10.Na3 Re8-/+) 3...Be7 4.Nc3 (4.d3 Nc6 5.Qf4 0-0=/+) 4...Nc6 5.Qf4 Nb4 (5...d5 6.e5 Nh5 7.Qf3 Nd4 8.Qxh5 Nxc2+ 9.Kd1 Nxa1 10.d4 Be6 11.Bd3 Qd7 12.h3 0-0-0 13.b3 f6 14.Bb2 fxe5 15.Qxe5 Bd6 16.Qe2 favors White.; 5...0-0 6.e5 Ne8 7.d4 d5 8.Nf3 f6 9.Qg3 Nb4 10.Bd3 Nxd3+ 11.cxd3 Bf5 12.0-0 c6 is approximately equal.) 6.Kd1 d5 (6...0-0 7.a3 Nc6 8.e5 Ne8 9.Qe4 d6 10.f4 g6 11.Nf3 Bf5=/+) 7.a3 (7.e5 Ng4 (7...Ne4 8.Nxe4 dxe4 9.Qxe4 Be6 10.c3 Nc6 11.d4 Qd7 12.Nf3 Bf5 13.Qf4+/=) 8.Bb5+ (8.d4 f6 9.e6 Bxe6 10.a3 Nc6-/+) 8...c6 9.Be2 g5 10.Qg3 Nh6 11.f4 g4-/+) 7...d4 8.axb4 dxc3 9.bxc3 0-0 leaves Black good compensation for the pawns.] 3.Bc4 g6 4.Qf3 Nf6 5.Ne2 Bg7 6.Nbc3 d6 7.d3 Bg4



[We're in virgin territory, as far as I can tell with my databases, and it looks like Black has equalized with all normal, natural moves.] 8.Qg3 Qd7 [8...Nb4 9.Bb3 Be6 might at first seem to win a pawn (or force White to either give up castling with 10.Kd1 or recapture with the c-pawn after 10.O-O Bxb3 11.cxb3), but White can save it with 10.Ba4+ followed by a timely a3.] 9.f3 Be6 10.Bg5 Nh5 11.Qh4 h6 12.Be3



[But now White seems to have an edge! The players' pieces are about equally active, but White is better organized for f4 or d4 than Black is to play d5 or f5.] 12...Na5 13.Bb3 [13.Bxe6 looks even better. It's not so much that Black gains the two bishops after White's choice, as that Black's knights are misplaced and White will gain time while Black returns them to better squares. Further, Black has to recapture with the pawn, which leaves g6 a little weak as well: 13...fxe6 (13...Qxe6 14.Nd5 Qd7 15.g4 c6 16.gxh5 cxd5 17.hxg6+-) 14.0-0-0 Nc6 15.Rhf1 Nf6 16.Qg3 Qf7 17.f4+/-] 13...Nxb3 14.axb3 a6 15.d4 Qe7 16.Qf2 exd4 17.Bxd4 Nf6 [17...0-0 18.Bxg7 Kxg7 19.0-0-0 Qg5++/=] 18.0-0-0 [18.0-0] 18...0-0-0 19.Nf4 Rhg8 20.Rhe1 Kb8 21.Kb1+/- g5 22.Nfe2 [22.Nfd5 Bxd5 23.exd5 Qd7 24.g4+/-] 22...Rge8? [22...c5 23.Be3 Nd7 24.Nd5 Bxd5 25.Rxd5+/-] 23.g4? [23.e5 wins! If the knight moves to a square that doesn't protect the Bg7, then 24.exd6 wins a pawn. 23...Nh5 runs into 24.g4 followed by exd6 (24...dxe5 25.Ba7+); finally, 23...dxe5 24.Ba7+ Ka8 25.Bc5+- wins the exchange.] 23...Qf8 24.Ng3 [24.h4+/=] 24...Nd7 25.Be3 Qh8!=



At this point Sasikiran has (at least) equalized the chances, due to his strong control over the central dark squares and the long diagonal in general. Nevertheless, despite Black's eventual victory in this game, White's experiment was a success: he created an interesting, imbalanced position he had undoubtedly thought about more than his opponent (who researches 2.Qh5?), and it resulted in a clear White advantage. The question now is whether we'll ever see it again at the GM level. 26.Nge2 Be5 27.h4 Qg7 28.Rh1 Nf6 29.Bd4 Nd7 30.Qe3 Qf6 31.hxg5 hxg5 32.Bxe5 Qxe5 33.Rh5 Rg8 34.Nd5 Rde8 35.Qc1 Qg7 36.Ne3 Nf6 37.Rh2 Rh8 38.Rg2 Nd7 39.Nd4 Rh3 40.c4 Qf6 41.Rf2 Reh8 42.b4 Qe5 43.c5 dxc5 44.bxc5 Nxc5 45.Qc3 f6 46.Rc2 Na4 47.Qb4 Bd7-/+ 48.Nb3 Rh1 49.Rxh1 Rxh1+ 50.Ka2 Nb6 51.Qf8+ Qe8 52.Qxe8+ Bxe8 53.Nc5 Nd7 54.Nxd7+ Bxd7 55.Kb3 Re1 56.Rc3 Be6+-+ 57.Kc2 Re2+ 58.Kc1 a5 59.Nc2 Rf2 60.Nd4 Bd7 61.Rc5 b6 62.Rd5 Kc8 63.e5 fxe5 64.Rxe5 c5 65.Nb3 Rf1+ [65...Rxf3? 66.Nxc5! bxc5 67.Rxc5+ Kb7 68.Rxa5 Bxg4 69.Rxg5=] 66.Kd2 a4 67.Nxc5 bxc5 68.Rxc5+ Kb7 69.Rxg5 Rxf3 70.Rd5 Be6 71.Rd3



71...Rf1 [71...Rxd3+?? only draws. 72.Kxd3 Bxg4 (72...Kb6 73.Kc3 Kc5 74.g5 Bf7 75.g6 Bc4 76.g7 Bg8 77.Kc2 Kb4 78.Kb1 Kb3 79.Ka1 just makes the inevitable take longer, but gives Black no more winning chances than before.) 73.Kc2 Kc6 74.b3 a3 75.Kb1 Kb5 76.Ka2 Kb4 77.Ka1 Kxb3 is a well-known theoretical draw: because the bishop is of the wrong color relative to the pawn (this only applies to rook's pawns), there is no way to force mate or promotion - Black's powers do not extend beyond the ability to force stalemate.] 72.Rg3 Rf2+ 73.Kc3 Kb6 74.Kb4 Rf4+ 75.Ka3 Kb5 76.Re3 Bd5 77.Rd3 Bc4 78.Re3 Rd4 79.g5 Rd1 80.b3 [80.Re5+ Bd5 and the threat of 81...Ra1# forces White to give up material.] 80...axb3 81.Re8 Ra1+ 82.Kb2 Ra2+ 83.Kc3 Rc2+ 84.Kd4 b2 85.Rb8+ Ka4 86.g6 Bb5 87.g7 b1Q [87...b1Q 88.g8Q Qd1+ 89.Ke5 Re2+ 90.Kf6 Qf1+ 91.Kg7 Rg2+ 92.Kh8 Qa1+ 93.Kh7 Qh1#] 0-1

Friday, April 22, 2005

Responding to Junk Openings
Chess opening books entitled "Winning with the ..." are distressingly popular: popular because of the ubiquitous desire for the proverbial free lunch; distressing because (1) it often represents a certain lack of respect for the game and (2) because it's entirely unrealistic - if winning could be achieved by virtue of playing opening X, all the top players would play it and win all their games.

This desire for the quick, gimmicky wins goes back, for most of us, to our early experiences with the Scholar's Mate (1.e4 followed by 2.Bc4, 3.Qh5 and, if all goes well, 4.Qxf7#), but the world of junior and beginning chess is rife with this sort of thing.

One such line goes as follows: 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4?!, when Black's big idea is to sucker White into capturing the "free pawn": 4.Nxe5 Qg5 5.Nxf7 (more freebies?) Qxg2 6.Rf1 Qxe4+ 7.Be2 Nf3# (perhaps it wasn't so free after all).

White can achieve an easy edge with moves like 4.c3 or 4.Nxd4, but neither move really punishes Black in the way he or she so richly deserves! Black wants to destroy White by giving up some material in return for an attack; I say we return the favor. Two lines come to mind:

(a) 4.Nxe5 ("falling for it") Qg5 5.Bxf7+ Ke7 6.O-O Qxe5 7.Bxg8 Rxg8 8.c3 followed by 9.d4. White has two pawns, a massive presence in the center and an exposed Black king as compensation for the piece. Objectively, the position may be roughly equal, but as a practical matter White's position is much easier to play.

(b) 4.Bxf7+ Kxf7 5.Nxe5+ Ke7 (the brave 5...Ke6 is interesting, but 5...Ke8 6.Qh5+ should win for White) 6.c3 d6! 7.Nc4 Nc6 8.d4 Nf6 9.O-O Kf7. Here White probably lacks sufficient compensation, though I think practicing such a position in blitz and casual games is both educationally worthwhile and likely to result in a reasonable rate of success.

Objectively, the normal lines are best, but lines (a) and (b) are more fun and, in a certain way, more appropriate responses to 3...Nd4. Speaking generally, one should try to achieve two things in the opening, when possible: reach a type of position one finds comfortable and, when possible, that one's opponent does not. These lines do just that.

Wednesday, April 20, 2005

A Look at the Poisoned Pawn Variation
In a recent post on the previous blog, I presented a quote from Edmar Mednis's How to Beat Bobby Fischer alluding to a number of games in the so-called Poisoned Pawn Variation of the Sicilian Najdorf, mostly by Fischer. Here are the games:

Keres,Paul - Fuderer,Andrija [B97]
Gothenburg Interzonal Gothenburg (16), 1955

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Qb6



8.Qd2 Qxb2 9.Rb1 Qa3 10.e5 Nfd7 [10...dxe5 11.fxe5 Nfd7 is the approved response nowadays. It might look odd for Black to throw in the exchange of pawns, as it opens both the d- and f-files for White. In return, however, Black's Bf8 can develop quickly, facilitating castling, and White no longer has the f5 option.] 11.f5 Nxe5 [11...dxe5 12.fxe6 exd4 13.exf7+ Kxf7 14.Bc4+ Ke8 15.Qe2+ Be7 16.Nd5 Qa5+ 17.Kd1 Nc6 18.Nxe7 Nce5 19.Nf5+- Nunn in The Complete Najdorf: 6.Bg5] 12.fxe6 fxe6 13.Be2 Nbc6 14.Nxc6 bxc6 15.Ne4 d5 16.0-0 Qa4 17.Bh5+ Kd7



18.Rxf8! [18.Rxf8 Rxf8 19.Nc5+] 1-0


Parma,Bruno - Fischer,Robert James [B97]
Bled Bled (5), 09.09.1961

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Qb6 8.Qd2 Qxb2 9.Rb1 Qa3 10.Bxf6 gxf6 11.Be2 Nc6




12.Nb3 [12.Nxc6 bxc6 13.0-0 is more common nowadays.] 12...Bg7 13.f5 0-0 14.0-0 Ne5 15.Nd4 b5 16.Kh1 Bd7 17.Bh5 Rac8 [unclear - Nunn]



18.Rb3 Qc5 19.Nce2 Nc4 20.Qc1 e5? [20...Qe5 21.Rg3 Kh8 22.Rxg7 Kxg7 23.Rf4 Rg8 24.Rg4+ Kh8-+] 21.Rg3 Kh8 22.Rxg7 Kxg7 23.Rf3 Rg8 [23...Rh8 24.Rg3+ Kf8 25.Qh6+ Ke7 26.Bxf7 Rcf8 27.Be6+-] 24.Rg3+



24...Kf8 [24...Kh8 25.Qh6 leads to mate - the pawn on e5 prevents the Black queen from defending the f-pawn. 25...Rxg3 26.Nxg3 Rg8 27.Bg6 fxg6 28.fxg6 Rxg6 (28...Rg7 29.Nh5) 29.Qf8+ Rg8 30.Qxf6+ Rg7 31.Nh5 Qxd4 32.Qxg7#] 25.Qh6+ Ke7 26.Qxh7 exd4 [26...Rxg3 27.Qxf7+ Kd8 28.Ne6+! Bxe6 29.fxe6+-] 27.Qxf7+ Kd8 28.Rxg8+ Kc7 29.Nf4 [29.Rxc8+ Kxc8 30.Qxf6 Ne5 31.h3+-] 29...Qa3 30.Nd5+ Kb8 31.h3 Ne3 32.Rxc8+ [32.Nxe3 dxe3 33.Rxc8+ Kxc8 34.Qxf6 Qc1+ 35.Kh2 Qd2 36.Qg5+-] 32...Kxc8



33.Nxe3? [33.Qf8+ first is a big improvement, keeping the king away from c7 where it would defend the d6-pawn. 33...Kb7 34.Nxe3+-] 33...Qxe3= 34.Bf3 Qc1+ 35.Kh2 Qf4+ 36.Kg1 Qc1+ 37.Kh2 Qf4+ 38.Kg1 Qc1+ 1/2-1/2

Fischer,Robert James - Geller,Efim P [B97]
Monte Carlo Monte Carlo (11), 04.04.1967

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Qb6 8.Qd2 Qxb2 9.Rb1 Qa3 10.f5 Nc6 11.fxe6 fxe6 12.Nxc6 bxc6 13.e5




13...Nd5 [13...dxe5 is both more common and scores better, but I don't think Geller's move is "officially" a mistake.] 14.Nxd5 cxd5 15.Be2 dxe5 16.0-0 Bc5+ [16...Ra7 is better.] 17.Kh1 Rf8 18.c4 Rxf1+ 19.Rxf1 Bb7



20.Bg4? [20.Qc2 e4 21.Bg4 Be7 (21...Qd3 22.Qa4+ Bc6 23.Qxc6#; 21...Bc8 is Black's best, according to Nunn (The Complete Najdorf: 6.Bg5).) 22.Qf2 0-0-0 23.Bf4 Bd6 24.Bxe6+ Kb8 25.Qb6 Bxf4 26.Qxd8+ Ka7 27.Rb1 Qd6 28.Bxd5! Bxd5 (28...Qxd8 29.Rxb7+ Ka8 30.Rd7++-) 29.Qxd6 Bxd6 30.cxd5 1-0, Tal-Bogdanovic, Budva 1967; 20.Bd1 Be7 21.Bxe7 Kxe7 22.Qg5+ Kd6 23.Rf7 Re8 24.c5+! Qxc5 25.Rxb7 Qf2 26.h3 Qf1+ 27.Kh2 Qf4+ 28.Qxf4 exf4 29.Rxg7 e5 30.Rxh7+/- Nunn (same source).] 20...dxc4 21.Bxe6 [21.Qc2 is better here, too, though Black is in control after 21...Bd5] 21...Qd3-+ 22.Qe1 Be4! 23.Bg4 Rb8! 24.Bd1 [Threatening 25.Ba4+] 24...Kd7 25.Rf7+ Ke6



0-1


Spassky,Boris V (2660) - Fischer,Robert James (2785) [B97]
World Championship 28th Reykjavik (7), 25.07.1972

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Qb6 8.Qd2 Qxb2 9.Nb3 Qa3 10.Bd3
[10.Bxf6 is by far the most common move here.] 10...Be7 11.0-0 h6 12.Bh4



12...Nxe4! 13.Nxe4 Bxh4 14.f5 exf5! 15.Bb5+ axb5?! [15...Ke7 is a suggestion of Shredder 9's, but perhaps a little suspicion is in order here!] 16.Nxd6+ Kf8 17.Nxc8 Nc6! 18.Nd6? [18.Qf4 Rxc8 19.Qxh4 Qa4=] 18...Rd8 19.Nxb5 Qe7-/+ Nunn 20.Qf4 g6 21.a4 Bg5-+



22.Qc4 Be3+ [22...Kg7] 23.Kh1 f4 24.g3 g5 25.Rae1 Qb4 26.Qxb4+ Nxb4 27.Re2 Kg7 28.Na5 b6 29.Nc4 Nd5 30.Ncd6 Bc5 31.Nb7 Rc8 32.c4 Ne3 33.Rf3 Nxc4 34.gxf4 g4 35.Rd3 h5 36.h3 Na5 37.N7d6 Bxd6 38.Nxd6 Rc1+ 39.Kg2 Nc4 40.Ne8+ Kg6 41.h4 f6 42.Re6 Rc2+ 43.Kg1



[Black has been better all the way, but the position has been sharp and the edge has drifted away a little at a time. After the move played, the position is drawn, but Black could have kept some edge with 43...Rxe8 44.Rxe8 Nd2 45.Kf2 Nf3+ 46.Re2 Rc4 , but 47.Rxf3 (47.Rb2 Rxf4 48.Ra3 Nxh4+ 49.Kg3 Kg5 50.Rxb6 Nf3-+) 47...gxf3 48.Kxf3 Kf5 49.Rb2 Rxf4+ 50.Kg3 Rxa4 51.Rb5+ Ke4 52.Rxb6 lets White escape into a theoretically drawn rook ending.] 43...Kf5 44.Ng7+ Kxf4 45.Rd4+ Kg3 46.Nf5+ Kf3 47.Ree4 Rc1+ 48.Kh2 Rc2+ 49.Kg1 1/2-1/2


Spassky,Boris V (2660) - Fischer,Robert James (2785) [B97]
World Championship 28th Reykjavik (11), 06.08.1972

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Qb6 8.Qd2 Qxb2 9.Nb3 Qa3 10.Bxf6 gxf6 11.Be2 h5 12.0-0 Nc6 13.Kh1 Bd7 14.Nb1




[This move excited many early commentators, but if so it was only on psychological grounds, as Black can force a repetition with 14...Qb2 15.Nc3 a) 15.a3 Rc8; b) 15.a4 d5 16.exd5 Nb4 17.Na3 Nxd5!-/+ (17...Rc8 is evaluated as clearly better according to Timman (cited in Nunn's The Complete Najdorf: 6.Bg5), but after 18.dxe6 fxe6 19.Bc4 , threatening both 20.Bxe6 and 20.Rfb1, for starters, Black has to play very accurately to equalize: 19...Rxc4 20.Nxc4 Qxc2 21.Nd6+ Bxd6 22.Qxd6 Rg8 (22...Qxb3 23.Qb8++-) 23.Qb8+ Kf7 24.Qxb7 Qc6 25.Qxc6 Bxc6 26.Rf2 Bd5 27.Nc5 Rc8 28.Rb2 a5 29.Nb3 Nd3 30.Rd2 Nxf4 31.Nxa5=) ; 15...Qa3 and White gets to decide all over again.] 14...Qb4 15.Qe3 d5?! [15...Ne7 16.c4 f5 17.a3 Qa4 18.Nc3 Qc6 19.Nd4 Qc5 20.exf5 Bg7 21.fxe6 fxe6 22.Rad1 Bxd4 23.Qxd4 Qxd4 24.Rxd4 Nf5 25.Rd2 Rc8 26.Rf3 Ke7 was better for Black in the game Qi Jingxuan-Karpov (!!!), Hannover 1983, which concluded 27.Kg1 h4 28.Bd3 Rc5 29.Bxf5 Rxf5 30.Rfd3 Rxf4 31.Rxd6 Bc6 32.c5 Rhf8 33.h3 Rf1+ 34.Kh2 Rc1 35.R6d3 Rff1 36.Nd1 e5 37.Nf2 Rg1 38.Nd1 Re1 39.Rd6 e4 40.R2d4 a5 41.Nf2 e3 42.Rxh4 exf2 43.Rh7+ Ke8 44.Rh8+ Kf7 0-1 Qi Jingxuan-Karpov,A/Hannover 1983/MCL] 16.exd5 Ne7 17.c4 Nf5 18.Qd3 h4? 19.Bg4 Nd6 20.N1d2+- f5 21.a3 Qb6 22.c5



22...Qb5 23.Qc3 fxg4 24.a4 h3 25.axb5 hxg2+ 26.Kxg2 Rh3 27.Qf6 Nf5 28.c6 Bc8 29.dxe6 fxe6 30.Rfe1 Be7 31.Rxe6



1-0