4. THE OTHER HALF OF THE BARGAIN Band 4: Sacrifices. Sacrifices can be classified in many ways - one is to distinguish between positive and negative sacrifices. Most sacrifices are positive, in that one lets the opponent capture something - a negative sacrifice is where one refrains from capturing something that could have been captured. A subtype that I especially like is: the exchange with only one exchanger. An obvious, almost trivial, but very clear example is Mednis - Zapata, Amsterdam 1986 (See Diagram) White: Kg1, Qa4, Ra1, Ra7, Bh1, Nc5, Nf3; pawns - c4, d3, e2, f2, g3, h2 Black: Kh8, Qe8, Rb8, Rg7, Bh3, Nd6, Ng4; pawns - c6, d5, e4, f4, g5, h6 After 32.Rxg7, Black did of course not recapture, but played 32...e3 which is a double negative sacrifice: apart from not recapturing on g7, he didn't take the Nf3. After 33.Bg2 exf2+ 34.Kh1 Bxg2+ 35.Kxg2 Qxe2 36.Rh7+ Kxh7 White resigned, because 37.Rf1 is met by Ne3+ Although the difference isn't always clear, this negative sacrifice is not a Zwischenzug, which is normally meant to postpone a capture. In a 'non-recapture', one side does not go along with the illusion that recapturing is obligatory. One of the first cases that struck me was the following. (See Diagram) White: Kg2, Qf3; pawns - b4, h2 Black: Ka7, Bb7; pawns - a5, b6 White to play; Ghitescu - X. Fischer, Romanian championship 1961. (The X stands for 'not Bobby'.) White resigned in view of 1.bxa5 bxa5 2.Kf2 Bxf3 3.Kxf3 a4 and Black queens first. He hadn't realized that recapturing is just an option; after 3.Ke3 there will be no queening, and the game is a draw. The most famous case of not recapturing is this: (See Diagram) White: Kf2, Ne3; pawns - a3, b4, d4, f3, g3 Black: Kd6, Bb1, pawns - a6, b5, d5, f6, h5, h4 In Karpov - Kasparov, 9th game of the first match, Moscow 1984, Black had just played 46...gxh4, expecting White to fulfill his half of the bargain with 47.gxh4. Black would then be safe, because without the h4-square, the white King has no way to enter the black fortress. 47.Ng2!! One exclamation mark is for the fact that this happened shortly after the adjournment; Kasparov's whole team had apparently suffered of materialistic conditioning, but Karpov had seen it. Now White will get the all-important square h4 - and the pawn, too. After 47...hxg3+ (Or 47..h3 48.Nf4 and both h-pawns will soon be lost.) 48.Kxg3 Ke6 (48...Bg6 49.Nf4 Bf7 50.Kh4 etc.) 49.Nf4+ Kf5 50.Nxh5 the King was able to enter via the kingside. Kasparov defended heroically, but had to resign on move 70. Karpov's reason for not recapturing was to keep a square unblocked - in the next example, the blocking idea works the other way around. (See Diagram) White: Kh3, Rg8; pawns - h4 Black: Kf6, Ne4, Ng3; pawns - f5, f3 White to play; Olij - Smit, Dieren 1979. 1.Rxg3 Other moves lose as well, but it's surprising that this one does: 1...Nd2! Leaving g3 blocked - there is nothing against f2 and promotion now. White resigned. The following endgame study combines the blocking and unblocking ideas in a beautiful way - first, a piece is lured to a blocking square, then that piece is left standing there to maintain the block. (See Diagram) White: Kf5, Ng2; pawns - a6, b5, e4 Black: Kh1, Ba1; pawns - a7, b6, h2 White to play and win; F. Bondarenko & M. Liburkin, 2nd Prize Shakhmaty v SSSR 1950. First, White must stop Black's pawn. 1.Nh4! We'll see later why 1.Ne1 doesn't work. 1...Kg1 2.Nf3+ Kg2 3.Nxh2 Kxh2 4.e5 Why bother about that pawn at all, one might think - why not head straight for the goodies on a7 and b6? But after 4.Ke6 Kg3 5.Kd7 Kf4 6.Kc8 Ke5! 7.Kb7 Kd6 8.Kxa7 Kc7 Black is in time. 4...Bxe5 Or 4...Bc3 5.e6 Bb4 6.Ke5 Kg3 7.Kd5 Kf4 8.Kc6 Ke5 9.Kb7 Kd6 10.e7! and wins 5.Ke6!! The marvellous point: capturing, which one would do almost automatically here, is not obligatory in chess. White leaves a whole bishop on the board, just to obstruct Black's King's path; after 5.Kxe5? Kg3 Black would again reach c7 in time. 5...Kg3 6.Kd7 Kf4 7.Kc8 White could still go wrong with 7.Kc6? Bb8! and Black has won the extra move he needs for his King: 8.Kb7 Ke5 9.Kxb8 Kd6 10.Kxa7 Kc7 draw. Now, after 7...Kf5 8.Kb7, White wins. After 1.Ne1? Bc3! Black has won a decisive tempo for his Bishop; 2.Nf3 Kg2 3.Nxh2 Kxh2 4.e5 (After 4.Ke6, Black reaches c7 in time.) Now, because the Bishop is already on c3, Black does not need to block his King's path with Bxe5. Instead, he plays 4...Kg3! and whatever White does, Black is in time, e.g. 5.Ke6 Kf4 6.Kd6 Bxe5+ 7.Kd7 Bb8 8.Kc8 Ke5 9.Kxb8 Kd6 10.Kxa7 Kc7 or 5.e6 Bb4 6.Ke5 Bf8 7.Kd5 Kf4 8.Kc6 Ke5 etc. Another obvious motivation for not obliging to recapture can be that the capturing piece would lose control over a vital square - sometimes the square which it would capture to. (See Diagram) White: Kg1, Ra4, Be1; pawns - g2 Black: Kc7, Rf2, Bf4; pawns - e3, g3 Black to play; Wojciehowski - Sandler, Riga 1982. After 1...e2! White probably saw that 2.Rxf4 Rxf4 3.Bxg3 Kc6 loses, but he may have had hopes, perhaps even of winning, for 2.Bxf2 However, Black kept control over f2 and played 2...Be3, and White resigned. Same thing in a game Pohl - Wolff, Pasewalk 1978. (See Diagram) White: Kd1, Rc3; pawns - b5, g4, h3 Black: Kd4, Nc1; pawns - d2, e3, g5 What White was thinking when he played 1.Rxc1? is hard to say, because 1.Rc8 (1...Nd3 2.Rd8+ Ke4 3.b6) is probably winning, and with the text move he could not have hoped for more than a draw; 1...dxc1Q+ 2.Kxc1 Kc5 etc. However, Black played 1...Kd3 and White resigned immediately. In the following interesting game, which does not seem to have found its way to the big databases, the 'non-capturing' is again meant to keep control of certain squares - even at the cost of a move. Voronov - Estrin, Leningrad 1971 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.O-O Be7 6.Re1 b5 7.Bb3 O-O 8.h3 d6 9.c3 Na5 10.Bc2 c5 11.d4 Qc7 12.Nbd2 Nc6 13.dxc5 dxc5 14.Nf1 Be6 15.Ne3 Rad8 16.Qe2 g6 17.Ng5 Nh5 18.Nxe6 fxe6 19.g3 c4 20.Ng4 Bc5 21.Kg2 Nd4 A standard sacrifice. 22.cxd4 exd4 23.Bg5 d3?! Black does not want to lose any time and makes it a speculative adventure. 24.Bxd8 (See Diagram) 24...Qb8 He keeps aiming at f2 and g3, even if he's a piece and a Rook down now. 25.Nh6+ According to Osnos, White could have refuted the whole thing with 25.Bxd3 Bxf2 and now, instead of 26.Nxf2 Qxg3+ when Black has a winning attack, 26.Bh4! Bxg3 27.Rf1 In the game, there followed 25...Kg7 26.Qd2 Rxf2+ 27.Qxf2 Bxf2 28.Bf6+ (28.Kxf2 Qxg3+ 29.Kf1 Qf4+ 30.Kg2 Qxh6 and Black wins) 28...Nxf6 29.Kxf2 Qf8 and White resigned Most of the time however, not recapturing is about time. (See Diagram) White: Ke5, Bh8; pawns - g2, h5 Black: Kf7, Nf5; pawns - b3, g6, h7 In Seirawan - P. Nikolic, Sarajevo 1987, Black may have been winning anyway, but when White played 39.hxg6+? the surprising Kg8! gave the Pb2 an immediate free walk. (See Diagram) White: Kh1, Rf1; pawns - a2, c6, d6, h2, h6 Black: Kh7, Rf8; pawns - b6, d3, e4, f2 Gligoric - Stein, Lvov 1962. After 48.Rxf2, Black played the incredible 48...e3! Instead of capturing a Rook, he sacrifices one. True, the speed of the pawns is all that matters, but this was a demonstration; 48...Rxf2 would have won as well, e.g.: 49.d7 d2 50.d8Q Rf1+ 51.Kg2 d1Q, and White has no perpetual. 49.Rxf8 d2 50.c7 d1Q+ and Black won nicely: 51.Kg2 Qg4+ 52.Kf1 Qc4+ 53.Kg2 e2 54.Kf2 Qe6 55.Ke1 Qxd6 56.Rh8+ Kg6 57.Rg8+ Kh5 and White resigned. A very funny example happened in the following game. Kotkov - Vistanetskis, Vilnius 1961: 1.e4 e5 2.Nf3 Nf6 3.d4 exd4 4.e5 Ne4 5.Qxd4 d5 6.exd6 Nxd6 7.Bd3 Nc6 8.Qf4 g6 9.Bd2 Qe7+ 10.Be2 Ne4 11.Nc3 Nxd2 (See Diagram) 12.0-0-0 To avoid the inconvenience of Qxd2 or Nxd2. It's not a Zwischenzug, because 12...Nxf3 was now possible; after 13.Nd5 Qg5 14.Nxc7+ Ke7 White had then better take the perpetual with 15.Nd5+ 12...Be6 13.Nd4 The Nd2 is not jumping anywhere. 13...Nxd4 14.Qxd4 Nb3+ 15.axb3 Bh6+ 16.Kb1 0-0 17.Qe4 c6 18.Bc4 Bg7 and a draw was agreed. Sometimes, recapturing seems to be a matter of belief in authority. (See Diagram) White: Kg1, Rb6, Rf1, Bf4; pawns - c3, c4, d5, g3, h6 Black: Kd7, Re4, Rg6, Bh7; pawns - c5, d6, f7, f3, g4 Black to play; Cooper - Petrosian, Buenos Aires 1978. The former world champion played 36...Rxf4, and the humble 2315 Welshman believed him with 37.gxf4 With 37.Rb7+ he could have mated Black in a few moves; 37...Kc8 (37...Ke8 or Kd8 38.Ra1) 38.R1b1 Kd8 39.Ra7 Kc8 40.Ra8+ Kc7 41.R8b8 and mate) In the game, he lost after 37...Rxh6 38.Rb7+ Ke8 39.Ra1 Kf8 The Ph6 is gone! 40.Raa7 Rf6 41.Kf2 Rxf4 42.Ra1 Rxc4 43.Kg3 Be4 44.Rb8+ Ke7 45.Kxg4 Bxd5+ 46.Kg3 Rxc3 47.Ra7+ Kf6 48.Rd8 Rc2 49.Rxd6+ Ke5 and White resigned. (See Diagram) White: Kg1, Qf3, Ra1, Rf2, Bb5, Nc3; pawns - a2, b2, d5, e3, g2, g3, h3 Black: Kf8, Qe7, Ra8, Rh8, Nb8, Bg7; pawns - a7, b7, c5, d6, f7, g5, h6 When in Dzyuban - Yurtaev, Zelinograd 1978, Black played 16...Bxc3, White didn't lose time over that Bishop; after 17.bxc3 Kg7, followed by Rf8, Black could have put up a defense. 17.Raf1! Kg8 (Rh7 18.Qf5) 18.Qxf7+ Qxf7 19.Rxf7 Bg7 20.Rxb7 and the end was sadistic: 20...h5 21.Rff7 Bf8 22.Rfc7 h4 23.g4 Rh6 24.Bd3 Rf6 25.Bh7+ Kh8 26.Rc8 a5 27.Bf5 a4 28.a3 No need to take the Nb8 - this is a permanent negative sacrifice. 28...Kg8 29.Be6+ Kh8 30.g3 Against a big man with big fists, I'd just play Re8. 30...hxg3 31.Kg2 c4 32.Kxg3 and Black resigned. Bruckner - S. Rubinstein, Vienna 1936 1.Nf3 d5 2.c4 dxc4 3.Qa4+ Nc6 4.Qxc4 e5 5.d3 Be7 6.g3 Nf6 7.Bg2 O-O 8.O-O Nd7 9.Qc2 Nc5 10.Be3 Ne6 11.Rc1 f5 12.Qb3 f4 13.Rxc6 bxc6 14.Nxe5 fxe3 15.Nxc6 exf2+ 16.Kf1 Qe8 17.Nxe7+ (See Diagram) After 17...Qxe7 18.Bxa8 Kh8, White's combination would still have been dubious, but 17...Kh8! was immediately killing. Wasting no time, all Black wants is to unpin his own Knight, which gives him a winning attack. 18.Qb4 (18.Bxa8 Nd4, followed by Bh3+ etc.) 18...c5 19.Qh4 Rb8 20.Be4 Ng5! 21.Ng6+ Qxg6 22.Bxg6 Bh3+ 23.Qxh3 Nxh3 24.Be4 Rxb2 and Black won quickly. (See Diagram) White: Ke1, Qc2, Ra1, Rh3, Bd2, Bd3; pawns - a2, b2, d4, e5, f2, g2, h4 Black: Kg8, Qd8, Ra8, Rf8, Be7, Bf3, Nd7; pawns - a5, b4, c5, e6, f7, g7, h6 This is Portisch - Johannessen, Havana 1966. Black just exchanged his Bishop for a Knight on f3 - or so he thought. Instead of racapturing, Portisch sacrificed a piece himself: 18.Bxh6 Ignoring the Bf3, he threatens Rg3. It's deep and brilliant, but not entirely clear - and possibly unsound, as computer-aided analysis shows. 18...Bxh4 If 18...gxh6, then 19.Rg3+ Kh8 20.Qd2 Bg5 21.hxg5 and wins. Or 18...Bg4 19.Rg3 f5 (gxh6 20.Rxg4+ Kh8 21.Qd2) 20.Rxg4 fxg4 21.Bh7+ Kh8 22.Bxg7+ Kxg7 23.Qg6+ Kh8 24.Bg8 Rf7 25.Bxf7 Qf8 and White gets at least four pawns for the Knight, and must be winning. The text move was best. 19.gxf3 gxh6 20.Ke2 Nxe5(!) 20...Bg5 21.f4 Bxf4 22.Rg1+ Kh8 22.Qd2! and mates. Or 20...cxd4 21.Rg1+ Bg5 22.Rxh6 Nxe5 23.Qc5 Nxf3 24.Kxf3 f6 25.Ke2 Kf7 26.Rxg5 fxg5 27.Qe5 and wins. 21.dxe5 Qd4 22.Rxh4 Qxh4 23.Rg1+ Kh8 24.Qc1 f6 This loses immediately (25.Rg6 and Black resigned) but 24...c4 might even have been winning: 25.Be4 (what else?) Rad8 26.Rh1? Rd2+! and Black wins. The following endgame study features a case of mutual non- (re)capturing. (See Diagram) White: Kh2, Rg4, Nd7; pawns - d4 Black: Kh8, Rh7, Nh3; pawns - d5, e6, f6 White to play and draw, L. Lindner, 2nd/3rd Prize Hungarian Chess Federation, 1976. 1.Rg8+! And not the obvious 1.Nxf6 Nf2+ 2.Nxh7 Nxg4+ 3.Kg3 Kxh7 4.Kxg4 Kg6 and Black has the opposition and wins. 1...Kxg8 2.Nxf6+ Kh8! 3.Nxh7 Kg7! Instead of recapturing, he plays the tempo move he wouldn't have had after 2...Kg7 The reason is that after 3...Kxh7 4.Kxh3, White has the opposition, and draws. But if now 4.Kxh3 Kxh7! and the roles are reversed, and Black wins. 4.Kg2! White also refrains from capturing. But now Black has nothing better than 4...Nf4+ 5.Kf3 Kxh7 6.Kxf4 Kg6 7.Kg4 (a last subtlety; 7.Ke5 Kf7 loses) 7...Kf6 8.Kf4 and draws (Ke7 9.Ke3 etc.) The following could be seen as a Zwischenzug, but for the non- recapture of a Queen, I can be a little less strict. Vladimirov - Epishin, Tashkent 1987 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e6 6.g4 h6 7.h4 Be7 8.g5 hxg5 9.Bxg5 Nc6 10.Qd2 a6 11.O-O-O Bd7 12.f4 Nxd4 13.Qxd4 Bc6 14.Rg1 Qa5 15.Bh3 Qc5 16.Qd3 b5 17.f5 b4 18.Be3 Qa5 19.Qc4 Bb7 20.fxe6 bxc3 21.Rxg7 d5 22.Qb3 cxb2+ 23.Kb1 Qb5 24.exf7+ Kf8 25.Rdg1 Qxb3 (See Diagram) 26.Bh6 A wonderful move. The immediate threat is Rh7 mate, and on Rxh6, there follows Rg8+, also with mate next move. Black can do nothing with his temporary extra Queen (and Knight.) 26...Ng4 27.Rh7+ Nxh6 28.Rxh8+ Kxf7 29.Rh7+ and Black resigned; after Ke8 30.axb3 he will lose decisive material. The final, well known word is to the Magician. Tal - Feuerstein, simul, Stuttgart 1958 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 g6 6.Be3 Bg7 7.f3 Nc6 8.Qd2 Bd7 9.O-O-O Qa5 10.Kb1 Rc8 11.g4 h6 12.h4 a6 13.Be2 Ne5 14.g5 hxg5 15.hxg5 Rxh1 (See Diagram) 16.gxf6! Rxd1+ Even if it is the best move, Rh7 is 'no move'. 17.Nxd1 Qxd2 18.fxg7! The theme doubled. 18...Kd8 19.g8Q+ Kc7 20.Qxc8+ and Black resigned.