17 OCTOBER 1992, Page 44

CHESS

Two of the best

Raymond Keene

Fischer has taken a commanding lead in Belgrade. Game 16 saw an idea which he must have kept secret for a quarter of a century, while game 17 exhibited a filigree display of endgame artistry in the grand tradition of the great endgame virtuosi Rubinstein and Capablanca.

Spassky — Fischer: 'World Chess Cham- pionship', Sveti Stefan, Game 16; Benoni.

1 d4 Nf6 2 c4 c5 Fischer revives happy memories of his first ever win against Spassky, the Benoni in game 3 of the 1972 encounter. 3 d5 d6 4 Nc3 g6 5 e4 Bg7 6 Bg5 This is risky, though considered sound. Safe and solid is 6 Nf3 followed by Be2. 6. . . h6 7 Bh4 g5 Choosing the sharpest reaction. 8 Bg3 Qa5 9 Bd3 Accept- ing the challenge. The alternative is 9 0d2, after which, however. Black can play to achieve . . . b5 or try 9 . . . Nh5 to trade White's queen's bishop. 9. . . Nxe4! Only apparently a sacrifice, since Black can recoup the piece. Still, it is a dramatic move, but one considered doubtful by theory for over a quarter of a century. 10 Bxe4 Bxc3+ 11 bxc3 Qxc3+ 12 Kfl 15 13 Rd l 13 Ne2 Qf6 14 Bc2 f4 15 h4 Rf8! 16 hxg5 hxg5 is better for Black, Stein — Geller, Moscow 1966 (0-1, 23). In his subsequent analysis Geller gave 13 Rd l as the probable refutation of Black's play.

Position after 14. . . g4!!

13 . . . Qf6 14 h4 g4!! (Diagram) A vital theoretical novelty which rejuvenates Black's prospects. Geller cited only 14 . . . fxe4 15 0h5+, claiming a White advantage. Fischer's new move overturns this ancient and hallowed verdict. 15 Bd3 14 I now see no good move for White. If 16 Bh2 g3 17 fxg3 fxg3 is discovered check. Or 17 0h5+ Kd8 18 Bxg3 fxg3 19 NO Rg8! with the terrible threat of. . . Bg4. 16 Ne2 fxg3 17 Nxg3 RN 18 Rc2 18 Ne4 may be a slight improvement, but it is not convincing. 18 . . . Nd7! Giving back the pawn for swift mobilisa- tion, and the chance of cementing the knight on the dominating e5 square. 19 Qxg4 Ne5 20 Qe4 Bd7 21 Kgl 0-0-0 Although in the given cir- cumstance this is a very logical move, it is highly unusual for Black to castle queenside in a Benoni formation. 22 Bfl Rg8 2314 Nxc4 24 Nh5 (#7 25 Qxc4 Qxh5 26 Rb2 Rg3 27 Be2 Qf7 28 813 Rdg8 29 Qb3 b6 30 Qe3 Qf6 31 Re2 Bb5 32 Rd2 32 Qxe7 Qxe7 33 Rxe7 Itxf3 wins a piece for Black. 32. . . e5! An inspired way to terminate. 33 dxe6 Bc6 34 Kfl Bxf3 White resigns If now 35 gxf3 Oaf + 36 Qel Rgl + 37 Rgl Rxgl+ 38 Kxgl Qxel + wins the house.

Fischer — Spassky: Game 17: Sicilian Defence.

1 e4 c5 2 Nc3 Nc6 3 Nge2 e6 4 g3 d5 5 exd5 exd5 6 Bg2 d4 Fischer, it transpires, is no stranger to this variation. Fischer — Bertok, Zagreb 1970, had continued 6. . . Nf6 7 d4 cxd4 8 Nxd4 Bg4 9 Qd3 with pressure against Black's isolated queen's pawn. 7 Nd5 Nf6 8 Nef4 Nxd5 9 Nxd5 Bd6 10 0-0 0-0 11 d3 Be6 12 Nf4 Apparently an innovation, 12 Qh5 having previously been played. 12 . . . Bf5 13 h3 Preparing, in due course, to expand by means of g4. 13 . . . Rb8 Sources in Belgrade criticised this move, sug- gesting instead 13 . . . 0c7. Another possible move is 13 . . . Qd7. But if Spassky' 13th move is an error, it is a very subtle one. 14 Bd2 Re8 15 Rel Rxel+ 16 Qxel Qd7 17 g4 Re8 18 Qdl Bxf4 19 Bxf4 Be6 20 Qf3 Thematically, heightening his pressure along the hl-a8 diagonal. In reply, Spassky implements a counter-attack against White's queenside pawns. 20 . . . NM 21 Qxb7

Position after 24 Be4 Nxc2 22 Rd l Qxb7 23 Bxb7 NM 24 1e4 (Diagram) This is the moment of crisis. Spass- ky's next move is plausible, but he has over- looked or underestimated the force of Whites 25th. The most combative line is 24 . . . c4 25 dxc4 Bxg4 (if 25 . . . Nxa2 26 Rdl Bxc4 27 Rxd4, with two raking bishops on an open board). After 25. . . Bxg4 White cannot play 26 Bxh7+ Koch7 27 hxg4 on account of . . . Nd3 winning for Black. Furthermore, 25. . . Bxg4 26 Rd l Bf51 27 Bd2 Ftxe4 28 Ftxe4 Bxe4 29 Bxb4 leads only to a draw. Therefore, if White wants to play for a win, he must try an immensely subtle pawn sacrifice, and I am sure this is what Fischer had in mind. It goes like this: 24 . . • e4 25 dxc4 Bxg4 26 Bbl! Bxh3 27 a3 Nc6 28 b4 f6 29 b5 Ne5 30 Bxe5 fxe5 31 Be4. In this position White's golden horde of queenside pawns are very hard to stop. 24. . . Bx.a2 Maybe Spassky expected the immediate 25 Rxc5 in response, but then 25 . . . Bbl! would have solved all of Black's problems. 25 Bd2 Bd5 26 Bxd5 Nxd5 21 11xc5 Nb6 A most instructive endgame has arisen. Fischer holds the advantage, since !us rook and bishop are respectively more active than Black's rook and knight. Additionally, the black pawn on d4 is weak. Nevertheless, it is a pleasure over the next few moves to observe the efficiency with which Fischer capitalises on his plusses. Spassky also defends like a demon, and if anyone needs a reminder how difficult it can be to cash in on a won endgame like this, I need only refer them to game 6 of this match, one of those that got away. 28 Kfl 1629 Ra5 Re7 30 Bb4, Rd7 31 Bc5 Kf7 32 Ke2 g5 33 Kf3 Kg6 34 Ke410 35 Bxd4 Re7+ 36 Kf3 h4 37 Bc5 Re! Spassky finds his best chance — fix White's h3 pawn MS! future target and activate the rook. 38 Rxa7 NdS 39 Bf8 Re8 40 Bd6 Re6 41 Rd7 Nb6 42 Rd8 Nd5 43 b4 Re! 44 b5 Rbl 45 Rb8 Rb3 46 Ke4 Nc3+ 47 Kd4 Nxb5+ 48 Kc4 Rc3+ 49 ICxb5 Rxd3 Spass!‘Y only has to eliminate or exchange the immobile and seemingly defenceless trio of white pawns on h3, 12 and g4 and it will be a draw. .F11-5 exertions have been heroic, but now an inspired king march by Fischer terminates his resistance, 50 Kc6 Rxh3 51 Kd5 Rf3 52 Ke6 Rxf2 53 Rg8+ ICh7 54 Kf7 Ra2 If 54 . . . f555 Rg7+ Kh656 Rg6+ Kh7 57 Rf6 threatening both gxf5 and Bf8. 55 Rg7+ Kh6 56 Bf8 Ra7+ 57 KA* R56+ 58 1Cf7 Black resigns A titanic struggle and One of Fischer's classic endgames. If 58 . . . Ka4 59 Rg6+ Kh7 60 Rh6 mate.

Running score:

Fischer 1 1/21/2 0 0 1/2

1 1 11/2101/21/21/2111/2 11

Spassky 0½½1 1 1/2 0 0

0 1/2 0 1 1/21/21/2 0 0 1/2 7