Endgame riddle solved: Najdorf could have drawn against Tal in Bled 1961

by Karsten Müller
4/15/2022 – On 3 April Karsten Müller invited ChessBase readers to analyse the bishop vs knight endgame from Tal vs Najdorf, Bled 1961. The endgame was tricky - Tal won the game but a close look revealed that Najdorf missed a couple of chances to save the draw and that both sides made mistakes.
Rules of thumb are the key to everything when you are having to set the correct course in a complex endgame. In this final DVD of his series on the endgame, our endgame specialist introduces you to the most important of these rules of thumb.

Endgame analysis of M. Tal vs M. Najdorf, Bled 1961

Conclusions by Zoran Petronijevic

  • 1. Our initial position is winning for White.
  • 2. 27.Qb3? allows Black to hold. After the stronger 27.Rd3 White has a decisive advantage.
  • 3. 30…Ke7? is a mistake. After the stronger 30…Ke8 White only has a symbolic advantage.
  • 4. 31…axb5? is a serious mistake after which Black is lost. Better was 31…a5 after which Black has good chances to hold.
  • 5. Tal’s straightforward move 33.Rb6? is a mistake that throws away his advantage. Better was the strategic move 33.g3 or first 33.Bd5 with good winning chances.
  • 6. 34…Rf4?! is dubious and the first sign that Black will make a decisive mistake – on his next move. After the better alternatives 34…Rf8 or 34…Re7 Black has good chances to hold.
  • 7. 35…Ra4?! is also dubious. After 35…Rf8 Black should be able to hold.
  • 8. After 36…Ra1? Black is lost. 36…Ra5 with good drawing chances was better; but to save this position Black must find a lot of good moves, which in practical game is difficult.
  • 9. After 37.Kg2 White has a winning position.
  • 10. After the adjournment Tal made an important mistake: 43.Ra8? spoils the win. Better was 43.h4.
  • 11. 44…Nb7? is the final error in this game. Better was 44…Kd8 after which Black has good chances to save the game.
  • 12. 47.Ra8?! is dubious, but Black is lost anyway.

As usual, Charles Sullivan and Zoran Petronijevic provided the main part of the analysis:

 
Tal, Mikhail1–0Najdorf, Miguel
Bled
03.10.1961[C.Sullivan & Z.Petronijevic]
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.Bg5 Be6 9.0-0 0-0 10.Bxf6 Bxf6 11.Qd3 Nc6 12.Nd5 Bg5 13.Rfd1 Kh8 14.c3 f5 15.Bf3 Bxd5 16.Qxd5 fxe4 17.Qxe4 Qe7 18.Qd5 Rf6 19.Nd2 Bxd2 20.Rxd2 Qc7 21.Re1 Raf8 22.Re3 g6 23.Be4 Kg7 24.Rf3 Rxf3 25.Bxf3 Rf6 26.Be4 Qf7 27.Qb3? Hans Kmoch in Chess Review (November 1961) awarded a "!" for this move, but there was no need to allow doubled pawns. If White is patient, Black will soon run out of good moves: 27.Rd3! Kf8 28.Qxf7+ Kxf7 29.Bd5+ Ke7 30.Rh3 h5 31.Rg3 a5 Trying to avoid the further weakness of ...a5 does not help: 31...Ke8 32.h3 Kf8 33.Kf1 Ke8 34.Ke1 Kf8 35.a4 Ke7 36.Ke2 and now it must be played: a5 37.Rf3! Rf5 38.h4 Nd8 39.Ke3 b6 40.Rxf5 gxf5 41.Bf3+- 32.Be4 Kf7 33.Kf1 h4 34.Bd5+ Kf8 35.Rh3 g5 36.Ke2 Ne7 36...Nd8 37.Rf3! Ke7 38.Rxf6 Kxf6 39.Kd3 b6 40.Kc4+- 37.Bxb7 Ng6 38.Re3 Rf7 39.Bc6 Nf4+ 40.Kd1 Rc7 41.Bf3 Ke7 42.g3 Ne6 43.Re4 Rc8 44.Ra4 Rb8 45.b3 Rf8 46.Ke2 Rc8 47.Kd2 Rf8 48.Ke3 Rc8 49.c4 Rc5 50.Bd5 Nd4 51.Kd2 h3 52.Kc3 Nb5+ 53.Kb2 Nc7 54.Be4 Ne8 55.Bf5 Nf6 56.Kc3 Nd5+ 57.Kd3 Nb6 58.Ra3 d5 59.cxd5 Nxd5 60.Ra4 Rc3+ 61.Kd2+- (if Rf3 62.Be4 wins material). 27...Qxb3 28.axb3 Nd8 Kmoch notes that 28...a5 is weak and suggests 29.Rd5 (with the threat of Rb5). ZP: White is clearly better, but Black can put up some resistence after Rf7 ZP: 28...a5 is simply a mistake. After 29.Kf1 White is sim ply winning: his plan is to drive his king to b5 -- Ke2-d3-c4-b5, after which Black is lost. 29.b4 Tal hopes to turn his doubled pawns into a weapon. Kf7 30.Rd5 Ke8 ZP: ? ZP In my opinion the game move is a mistake. Stronger is 30...Ke7 31.Ra5 31.b5 doesn't work here due to Ne6 Pd6 is protected. White only has slight edge. 32.bxa6 bxa6 33.Ra5 Nc7 31...Kd7 32.b5 axb5 33.Rxb5 Kc7 34.g3 b6 35.Rb4 Nb7 36.f3 Nc5 37.Bd5 h6 38.Kf1 g5 39.Ke2 Rg6 White has slight edge, but Black should hold. 31.b5 Nothing special is gained by 31.Ra5 Kd7 32.b5 axb5 33.Rxb5 Kc7 ZP Transposes to the line after 30...Ke7. 31...axb5? Kmoch thought that Black could not save the game with 31...a5! because of 32.b6 Nc6 33.Bc2 but Black can hold despite losing a pawn: ZP Stronger is 33.h4 although after a proper play, Black should hold: Kd7 34.f3 Rf8 35.Kh2 a4 36.Kh3 Ra8 37.Rb5 Nd8 38.Kg4 Ke6 39.c4 a3 40.bxa3 Rxa3 41.Bd5+ Ke7 33...Ne7 34.Rxa5 d5 35.Bb3 35.Ba4+ Kf7 36.Rb5 Ke6 drawish 35...Rxb6 36.Bxd5 Nxd5 37.Rxd5 Rxb2 38.Rxe5+ Kf7 drawish ZP It is an easy draw. 32.Rxb5 Rf7 33.Rb6? White chooses the wrong "active" move. 33.Bd5! (Or 33.g3, when Black has nothing better than 33...Rc7 34. Bd5!, transposing to this main variation) Rc7 34.g3! A mistake would be 34.h4 Rc5! 35.Rxc5 dxc5 36.Kh2 b5 37.Kh3 Ke7 38.Kg4 Kf6 39.h5 c4 40.Kh4 ZP Interesting is 40.hxg6 hxg6 41.f4 Ne6 42.fxe5+ Kxe5 43.Bxe6 Kxe6 44.Kg5 Ke5 45.g4 Ke4 46.Kxg6 Kd3 47.Kf7 Kc2 48.g5 b4 49.cxb4 Kxb2 50.b5 c3 51.b6 c2 52.b7 Ka2 53.b8Q c1Q the position is even according to TBs. In real tournament game White has good practical chances. 40...g5+ 41.Kg3 Kf5 42.h6 e4 43.Kh2 Ne6 44.Bc6 Nf4 45.Bxb5 Nd3 46.Bxc4 Nxb2 47.Bb3 Kg6 48.Bc2 Kxh6 49.Kg3 Kg7 50.Bxe4 Na4 51.c4 h6 52.Kf3 Kf6 53.Bc2 Nc5 and Black, a pawn down, can still hold the position. 34...Kd7 35.Kg2 Rc8 Black can try to keep material equality, but 35...Kc8 36.Kh3 Re7 37.Kg4 Re8 38.Be4 Re6 39.c4 Kd7 40.Rb6 a more adventurous win is 40.Bxb7 Rf6 41.f3 Nc6 42.c5 e4 43.fxe4 Ne5+ 44.Kg5! Rf2 45.Kh6 Rxh2+ 46.Kg7 Nf3 47.b4 Nd4 48.Ra5 Ne6+ 49.Kg8! dxc5 50.bxc5 Rc2 51.Kxh7 Nxc5 52.Ba8 g5 53.Kg6 g4 54.Kf5+- ZP Probably faster win is after 40.h4 h5+ 41.Kh3 Rf6 42.Kg2 Kc7 43.Rb3 Nc6 44.Bd5 Rf5 45.Rf3 Nd4 46.Rxf5 gxf5 47.Bf7 Kd7 48.Bxh5 Ke6 49.Kf1 Nb3 50.Bd1 Nc5 51.Ke2 Kf6 52.f3 Ne6 53.Kf2 Kg7 54.Bc2 Nd4 55.Bd3 b6 56.Ke3 Kh6 57.g4 f4+ 58.Ke4+- 40...Re8 41.h4 Rf8 42.f3 Kc7 43.Rb5 b6 44.Rb3 Ne6 45.Ra3 h5+ 46.Kh3 Rf6 47.Ra7+ Kd8 48.Ra8+ Ke7 49.Rg8 Kf7 50.Rb8+- and the creation of a passed pawn on the queenside will doom Black. 36.Bxb7 Rb8 37.Ba6 Ra8 38.Ra5 Nc6 39.Ra2 Rb8 40.Bc4 Ne7 41.Ra7+ Kd8 42.b4 d5 43.Bb3 Rb6 44.Ba4 Rd6 45.Kf3 h6 46.Ke2 Nf5 47.Ra5 ZP Also strong seems 47.f3 Ne7 48.Bb5 Re6 49.Ra8+ Kc7 50.Re8 Kd6 51.Kd3 Kc7 52.c4 dxc4+ 53.Kc2 Kd6 54.Rd8+ Kc7 55.Rd7+ Kc8 56.Kc3 Nf5 57.Rf7 Rd6 58.Bxc4 h5 59.b5 h4 60.gxh4 Ne3 61.Bb3 Rd4 62.Re7 Nd5+ 63.Bxd5 Rxd5 64.Kc4 Rd4+ 65.Kc5 Rxh4 66.Kc6 Rc4+ 67.Kb6 Rc3 68.Rxe5 Rxf3 69.Re8+ Kd7 70.Rg8+- 47...Ne7 48.Kd3 Nf5 49.Bb3 Ne7 50.h4 Kc7 51.Ra7+ Kd8 52.Ra8+ Kd7 53.Ba4+ Kc7 54.Ra7+ Kd8 55.Ke2 Nf5 56.Ra8+ Kc7 57.Re8 e4 58.h5 Rf6 59.Bb3 Kd7 60.Re5 Ne7 61.hxg6 Rxg6 62.c4 dxc4 63.Bxc4 Kd6 64.Rxe4+- 33...Kd7! White must have overlooked this move -- Black improves the position of his king. 34.Bd5 34.Bxb7?? Kc7-+ was pointed out by Kmoch. 34...Rf4 ZP Interesting is also 34...Rf8 35.Rb4 Kc7 36.Rh4 h5 37.Rc4+ Nc6 38.h4 Rd8 39.b3 Kb6 40.Ra4 Rb8 41.g3 Ne7 42.Bg2 Kc5 43.Rc4+ Kb6 44.Rb4+ Kc5 45.Rxb7 Rxb7 46.Bxb7 d5 Black's King is so strong that the position is close to even. 47.Ba6 d4 48.cxd4+ exd4 49.Kf1 Kb4 50.Ke2 Kc3 51.b4 Kxb4 52.Kd3 Kc5 53.Ke4 Kd6 54.Bc4 d3 55.Bxd3 Ke6= 35.g3 Ra4?! Black's sturdiest defense is 35...Rf8 36.Rb4 Kc7 37.Rh4 h5 38.Rc4+ Nc6 and Black can hold; for example: 39.h4 Rd8 40.Bf7 ZP Slightly preferable is 40.Kg2 although doesn't give more than a slight edge. 40...b5! 41.Re4 d5 42.Re1 Rd6= 36.Bxb7 Is it possible for Black to save this position? Ra1+? At first, the machines think that White has a win after 36...Ra5 37.Bf3 d5 38.Kg2 Kc7 39.Rb4 Ne6 40.Rh4 Nf8 41.b4 Rb5 42.c4 Rxb4 43.Bxd5 Kd6 44.Rh6! Black's salvation is Rb3‼ which takes away the f3-square from White's king: 44...Rb2 45.h4! e4 45...Ke7 46.h5! Kf6 47.hxg6 hxg6 48.Rh8 Ne6 49.Kf3 Kg7 50.Ra8 Ng5+ 51.Ke3 Rb3+ 52.Ke2 Rb2+ 53.Kd3 Rxf2 54.Rg8+ Kh6 55.Re8 Nh7 56.Rxe5+- 46.Bxe4 Ke5 47.Bd5 Kf5 48.Kf3! Rb3+ 49.Ke2 Rc3 50.f3 Rb3 51.g4+ Kf4 52.Kd2 Ra3 53.Kc2 Re3 54.Be4 Ra3 55.Kb2 Ra7 56.Kc3 Rc7 57.Kd4 Rd7+ 58.Kc5 Rc7+ 59.Kd5 Rd7+ 60.Kc6+- 45.h4 looks like a winner, but e4‼ 46.Bxe4 Ke5 47.Bd5 Kf5 48.Kf1 48.Bg8 Kf6! 48...Kf6 49.g4 Kg7 50.g5 Nd7 51.h5 Nf6! 52.hxg6 hxg6= . The ending might be 53.Bc6 Rb6 54.Bg2 Rb1+ 55.Ke2 Rb2+ 56.Kf3 Nh7 57.Rh1 Nxg5+ 58.Ke3 Ne6= 37.Kg2 Kc7 Too late is 37...Ra5 38.Bf3 Kc7 39.Rb4 d5 40.Rh4 h5 41.b4 Rb5 42.c4 dxc4 43.Rxc4+ Kd6 44.h4 Ne6 45.Bc6 Rb6 46.Be8 g5 47.hxg5 Nxg5 48.Bxh5+- 38.Ra6 Rb1 39.Bd5 Rxb2 40.Ra7+ Or White could immediately play the winning 40.h4 Nb7 41.Ra8 e4 42.Kf1 e3 43.fxe3 Nc5 44.Ra7+ Kd8 45.Rxh7 Nd3 46.Be4 Ne5 47.Rh6 Ke7 48.Bxg6+- 40...Nb7 According to the Associated Press, this game was adjourned after 40 moves; Tal "was confident he would win," reported the AP. The next day, the game was resumed. 41.Kf3 Kmoch: "White had to surrender his Pawn plus. But the indirect exchange of Knight Pawns has greatly increased the activity of his pieces. Black is in a bind which he cannot bear for long." The best move is 41.h4! Kb8 42.Ra4 Nc5 43.Ra8+ Kc7 44.Ra7+ Kd8 44...Nb7 45.Kf3 Kb8 46.Ra6 Kc7 47.c4 h5 48.Be4 Rb3+ 49.Ke2 Rb4 50.Ra7 Kb8 51.Ra2 Rxc4 52.Bxg6 Rc8 53.Bxh5 Nc5 54.Bf7+- 45.Rxh7 e4 46.Kf1 Rc2 47.Rg7 Rc1+ 48.Ke2 Rc2+ 49.Ke1 e3 50.fxe3 Rxc3 51.Ke2 Rc2+ 52.Kf3 Nd3 53.Bf7 Ne5+ 54.Ke4 Rc7 55.Rg8+ Ke7 56.Bxg6+- 41...Kb8 42.Ra6 Kc7 43.Ra8? If the Associated Press is correct that the 41st move was sealed for adjournament, it is difficult to understand the wobbly play by White on this move and by Black at the next. White misses his third chance to play the winning 43.h4! Nc5 44.Ra7+ Kd8 45.Rxh7 Rc2 46.c4 e4+ 47.Ke3 Rc3+ 48.Kd2 Rb3 49.Rf7 e3+ 50.fxe3 Rb2+ 51.Ke1 Nd3+ 52.Kf1 Ne5 53.Rf6 Kc7 54.Re6 Rd2 55.Ke1 Rc2 56.Kd1 Rc3 57.Kd2 Ra3 58.Ke2 Ra2+ 59.Kd1 Ra3 60.Kd2 Kd7 61.Ke2 Ra2+ 62.Kd1 Ra3 63.e4 Rc3 64.g4! Nxg4 65.Kd2 Rh3 66.Rxg6+- and if Rxh4 67.Be6+ Ke7 68.Bxg4 wins a piece. 43...Nc5 44.Ra7+ Nb7? Perhaps Najdorf hoped for a draw by repetition? Removing the knight from its excellent square relinquishes the draw that was available after 44...Kd8! 45.Rxh7 Rc2 46.c4 e4+! 47.Ke3 47.Bxe4 Rc3+! 48.Kf4 Ne6+ 49.Kg4 Rxc4= 47...Rc3+ 48.Kd2 Rb3 49.Rh4 49.g4 Rb2+ 50.Ke3 Rb3+ 51.Ke2 Rb2+ 52.Kf1 Nd3 53.Bxe4 Nxf2! 54.Bxg6 Nxg4 55.h3 Ne3+ 56.Ke1 Nxc4= 49...g5= and Black's aggressive pieces hold the draw. If 50.Rg4 Rb2+ White cannot retreat to his kingside: 51.Ke1?? Nd3+ 52.Kf1 Rxf2+ 53.Kg1 e3 54.Re4 e2-+ 45.h4! Other moves that win for sure are c4, Kg4, and Ke4. Simplifying to a pawn ending wins, if White plays precisely: 45.Ke4 Re2+ 46.Kd3 Rb2 47.Kc4 Kc8 48.Rxb7 Rxb7 49.Bxb7+ Kxb7 50.Kd5 Kc7 51.c4 Kd7 52.h4! h6 53.g4! 53.f3? h5!= 53...g5 54.h5! 54.hxg5?? hxg5= 54...Ke7 55.Kc6 Ke6 56.f3 Ke7 57.Kc7 Ke6 58.Kd8 d5 59.c5! d4 60.c6 d3 61.c7 d2 62.c8Q++- 45...Kb8 46.Ra6 Kc7 47.Ra8 47.c4 can be played first, also leading to a win. 47...Rb5 Or 47...Nc5 48.Ra7+ This position could have been reached had White played 43.h4 Nc5 44.Ra7+. See the note at White's 43rd move. 48.c4 Rb3+ A slightly different finish follows 48...Rb2 49.Rh8 h5 50.Rh7+ Kc8 51.Be4 Na5 52.Bxg6 Nxc4 53.Bxh5 d5 54.g4 e4+ 55.Kg3 Nd6 56.Bf7 Rb3+ 57.Kg2 Rb2 58.Be6+ Kb8 59.Kg3 d4 60.Rd7 e3 61.fxe3 Ne4+ 62.Kf3 Nc5 63.Rd8+ Kc7 64.Rc8+ Kd6 65.Bf5 d3 66.Bxd3+- 49.Kg4 49.Kg4 Rb2 50.Ra7 Kb8 51.Rxb7+ Rxb7 52.Bxb7 Kxb7 53.Kg5 Kc6 54.Kh6 Kc5 55.Kxh7 Kxc4 56.Kxg6 d5 57.h5 d4 58.h6 d3 59.h7 d2 60.h8Q d1Q 61.Qxe5 "and White wins surely because his Pawns are safe" ?- Kmoch. 1–0

 

Wolfram Schoen explains the endgame with more text and less variations:

 
Tal, Mihail1–0Najdorf, Miguel
Bled
1961[Schön,Wolfram]
1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be2 e5 7.Nb3 Be7 8.Bg5 Be6 9.0-0 0-0 10.Bxf6 Bxf6 11.Qd3 Nc6 12.Nd5 Bg5 13.Rfd1 Kh8 14.c3 f5 15.Bf3 Bxd5 16.Qxd5 fxe4 17.Qxe4 Qe7 18.Qd5 Rf6 19.Nd2 Bxd2 20.Rxd2 Qc7 21.Re1 Raf8 22.Re3 g6 23.Be4 Kg7 24.Rf3 Rxf3 25.Bxf3 Rf6 26.Be4 Qf7 27.Qb3 Qxb3 28.axb3
Rook + bishop vs rook + knight, six pawns each. Looking at the pawn structure shows that White holds a double pawn. But it is a good one, as a2xb3 has been an innerswap, taking towards the centre. Black has kept both his central pawns, White none. On the other hand, Black has three pawn groups, while White has only two. Most notably is the black d6. Backward, under the fire of the white rook on d2 and with no prospect of getting side by side with his neighbour on e5 by an advance with d6-d5. For now. Piece configuration: The white rook attacks, the black rook defends. The white bishop occupies a central square on e4. Thus he has central control on d5 and exerts pressure on both sides simultaneously (c6,b7 as well as g6,h7). Contrary to this the black knight isn't bad, but he has no perspective for an outpost at all. Overall White has an advantage due to his superior minor piece and Black's weak d6-pawn. 28...Nd8?! Simply not good. At least because of the principle "knight on the rim". I suspect Najdorf wanted to eliminate one of White's options thereby. However after the text move the knight is tied to the b7-pawn, until some other piece comes to help. 28...Kf8 This or 28...Kf7 would have been the most direct way to deal with Black's problems. Once the black king is on c7, he would protect d6 and b7 simultaneously, thus the Nc6 and the Rf6 would be flexible again. Therefore White's advantage in the diagram position isn't of a permanent nature, but just temporarily or of a dynamic nature. He will try to make something out of the time Black is busy with his king manoeuvre. 29.b4 Ke7 30.Rd3 30.Bxc6?! Maybe Najdorf has been afraid of this exchange, giving Black an isolated a6-pawn. bxc6 31.Rd1 However, it is a rook ending, where often a deficit of a pawn can be compensated by some activity. c5!? A drastic defense. Not mandatory, but sufficent. 32.bxc5 dxc5 33.Rd5 Rd6 34.Rxe5+ Kf6 35.f4 Rd2 36.Rxc5 Rxb2= 30...Kd7 31.Rh3 The h7 is White's new attractive target after the black king has left this area. Rf7 32.Bd5 Rg7 33.Kf1 Black has new problems because of his inactive Rg7 and the prospect of a white rook infiltration by Rh3-f3-f8. Yet, his position is defensible and he should be able to make some modest progress over the time to gradually equalize, e.g. Kc7 34.Rf3 Ne7 35.Bb3 Kc6 36.Rf8 Nd5 37.g3 Nb6 38.Ke2 Re7 39.h4 Kc7 40.h5 Nd7 41.Rf3 Kc6 42.Ba4+ Kc7 43.Bc2 gxh5 44.Be4 b6 45.Rf5 h4 46.gxh4 a5 29.b4 In case Black had any hopes for a set-up with b7-b6 and a6-a5, these have now become an illusion. Kf7 30.Rd5 Direct play by Tal. The idea is to increase the pressure on the b7 by means of the rook, via a pawn exchange on b5. 30.b5 Same idea, but this move order backfires. axb5 31.Rd5 b4! 32.c4 Ne6! 33.Rb5 Nc5 34.Bd5+ Ke7 35.Rxb4 Rf8= 30.Rd3 As an alternative, Whilte could try to go for the h7 here as well. Like in the other side line, he can keep the initiative for some time, but in the end it will fizzle out. Ke7 31.Rh3 Rf7 32.Bd5 Rg7 33.Re3 Kd7 34.g3 b6 35.Re4 g5 36.h4 h6 37.hxg5 hxg5 38.Bc4 b5 39.Bb3 Nc6 40.Re1 Kc7 41.Kg2 a5 42.bxa5 Nxa5 43.Bd5 Nc6 44.Kf3 b4 30...Ke8?! 30...Ke7! On his way over to c7 the black monarch should take the chance to cover his d6-pawn at the same time. 31.Ra5 31.b5 Ne6! The main difference to Najdorf's move 30. 32.bxa6 bxa6 33.Ra5 Nc7 White can't capitalize on the weak a6, for example 34.Bb7 Rf8 35.b4 Nb5 36.c4 Nc7 37.Bxa6 Rf4 38.g3 Rd4= 31...Kd7 32.b5 axb5 33.Rxb5 Kc7 White has created the pawn configuration he wanted. But as he had to spend an extra tempo with his rook, the black king has made it in time to c7. So Black can use rook and knight for the better next. Small advantage for White for a long time, that's all. 34.g3 Nc6 35.Kg2 Rf8 36.c4 Ra8 37.c5 dxc5 38.Rxc5 Ra4 Forcing a harmless rook ending, although a pawn down. 38...Kb6 Alternatively Black can go for a pure minor piece ending. 39.Rd5 Ra4 40.f3 Rd4 41.Rxd4 exd4 42.Kf2 39.Bxc6 bxc6 40.Rxe5 Kd6 41.Re3 Rb4 42.b3 g5 43.Kf3 h5 31.b5 axb5? The natural reaction of the defender. Exchanging pawns, especially if they have crossed the middle-line. In this case, however, it is actually already the losing move. 31...a5! Because of his previous inaccuracies, Najdorf had to find a non-standard defence here. 32.b6 32.Rd1 b6 33.b4 axb4 34.cxb4 Ke7 35.Ra1 d5! 36.Bxd5 Rd6 Black has neutralised White's idea of a rook infiltration via a7. 32...Nc6 Despite the black weaknesses on a5, b7 and d6 there seems to be no way for White to get a serious advantage. In some lines his b6 becomes the source of counterplay. 33.Rb5 33.Bc2 Ne7! 34.Rxa5 d5 35.Rb5 Kd7 36.Ba4 Kd6 37.g3 Rf8 38.Ra5 Nc6= 33...Kd7 34.f3 Rf8 35.Kf2 Ra8 The black rook supports his a-pawn for a file opening, while White attempts to attack on the kingside with the black h7-pawn being the weakest link. 36.Kg3 a4 37.Kg4 Ke7 38.Kg5 Nd8! This time for a good reason. 39.h4 Nf7+ 40.Kg4 Nd8 41.c4 41.h5 gxh5+ 42.Kxh5 a3 43.bxa3 Rxa3 44.c4 Ra2= 41...a3 42.bxa3 Ra4 43.Bd5 Rxa3 Due to Black's improved rook, the worst is over for him. 32.Rxb5 Rf7 Compared with the starting position White has made solid progress, as Black now has two pawn weaknesses on b7 and d6. Moreover, the black rook is now tied to the b7-pawn. Unlike to the 30...Ke7-line, where the black king could take over this job. Nevertheless, Black has two plans to relieve the pressure. Namely Rf7-c7-c5 and Rf7-c7, Ke8-d7-c8. 33.Rb6? With the idea to cause an even stronger bind, by forcing the black rook to defend d6 as well. But this is too optimistic and likely Tal simply has overlooked Najdorf's little defensive trick. 33.g3 In fact, White can already claim a decisive advantage. The general approach is this: Use the time Black needs for his plans for good. In case Black goes for the exchange of rooks, ensure the minor piece ending is winning and in case Black goes for the set-up with Kc8, make progress on the kingside. In short, the white king has to be activated quickly to perform both tasks. Rc7 34.Kg2 Kd7 34...Rc5 35.Rxc5 dxc5 36.Kf3 Ke7 37.Bd5 Bishop corrals knight. Kd6 38.Ke4 Central king, just in time. Ne6 38...b5 39.g4! h5 40.gxh5 gxh5 41.b3+- 39.Bxb7 c4 40.Ke3 Nc5 41.Be4 Na4 41...Ke6 42.Bc2 Nd7 43.f4 Nf6 44.Kf3 Kd5 45.Ba4+- 42.Bc2 Nxb2 From one corral to another. 43.h4 h6 44.Kd2 g5 45.f3! gxh4 46.gxh4 Ke6 47.Kc1 Nd3+ 48.Bxd3 cxd3 49.Kd2 Kf5 50.Kxd3 Kf4 51.Ke2 e4 52.fxe4 Kxe4 53.h5+- 35.h4 Kc8 35...Rc5 36.Rxc5 dxc5 37.h5 gxh5 38.Kh3+- 35...Ne6 36.Bxb7 Rc5 37.Ba6 Rc7 38.Ra5 Nc5 39.Bc4 Ke7 40.Kf3 Nd7 41.Ba2 Rb7 42.b4 Rc7 43.Ra3+- 36.Bd5! Preventing Nd8-e6, as well as Rc7-f7. In the next stage the white king wants to profit from Black's holes on the dark squares by marching in on h6. For quite some time Black can try different set-ups, occasionally switching to an active defense for the price of a pawn or two. Re7 36...Nc6 37.Kf3 Ne7 38.Be6+ Kd8 39.Kg4 h6 40.h5+- 37.Kf3 Re8 37...Kc7 38.Kg4 Nc6 39.Kg5 e4 40.Kf4 e3 41.fxe3 b6 42.e4 h6 43.Rb3 Ne5 44.Ra3+- 38.Kg4 h6 Holes plugged, but now the g6-pawn is a new black weakness. 38...Rf8 39.f3 Rf5 40.Be4 h5+ 41.Kh3 Rf6 42.c4 Kc7 43.Rb3 Nc6 44.g4 hxg4+ 45.fxg4 Rf1 46.h5 In this line White has converted his kingside superiority into a distant passed pawn, with desicive effect. gxh5 47.gxh5 Rf4 48.Bd5 e4 49.Re3 Rf5 50.Kg4 Rf2 51.Rxe4 Ne5+ 52.Kg3+- 39.Be4 Re6 40.c4 Nc6 41.c5 dxc5 42.Rxc5 Rf6 43.f3 Kd7 44.h5 gxh5+ 45.Kxh5 White has exchanged two pairs of pawns, but meanwhile all three remaining black pawns are weak and spread across an open board. One example of a possible finish, displaying noticeably the strength of a bishop in positions of this kind: Kd6 46.Rb5 Kc7 47.g4 Re6 48.Rc5 Kb6 49.Rd5 Rf6 50.b4 Kc7 51.Rb5 b6 52.Rd5 Nxb4 53.Rxe5 Na6 54.Rf5 Re6 55.Rf7+ Kd6 56.Bf5 Re3 57.Rf6+ Kc7 58.Bh7 Nc5 59.Kxh6 b5 60.g5 Re1 61.g6 Rh1+ 62.Kg7 b4 63.Rf7+ Kd6 64.Kh8 Ne6 65.Rf6 Ke5 66.Rxe6+ Kxe6 67.g7 Rg1 68.f4! b3 69.g8Q+ Rxg8+ 70.Kxg8+- 33...Kd7! Najdorf seizes the chance to correct his set-up and transfers his king towards c7 eventually. 33...Rd7? 34.Bd5 Ke7 35.g3+- An even better version of a bind than after 33.g3. 34.Bd5 I like that move. By now Tal must have realised his mistake in the previous move, resulting in the loss of most of his advantage. Yet, he immediately poses a new practical problem for Najdorf, who is forced to make a decision for his rook. Rf4?! Najdorf opts for an active use of his rook. Natural, but it is indeed going in the wrong direction. 34...Re7 In my mind the rook should completely stay at home, prophylactically protecting h7. E.g. 35.g3 Kc7 36.Rb3 Nc6 37.Kg2 Re8 38.c4 Na5 39.Rf3 Re7 40.Rf6 Kb6! Now is the chance for activity. 41.b4 Nc6 42.Rxd6 Kc7 43.Re6 Rxe6 44.Bxe6 Nxb4 45.Kf3 Kd6 46.Bg8 h6 47.Bh7 g5 48.Be4 b6 49.Kg4 Kc5= Based on counterplay with the passed b-pawn to come. 35.g3 Ra4?! 35...Rf8! The correct move, but almost impossible to play in practice. As a valid alternative to 34...Re7, this rook could have gone to f8 one move before, without White getting the useful move g2-g3 for nothing. On the other hand the insufficiency of the text move is everything else but obvious. 36.Rb4 36.c4 Kc7 37.Rb3 Nc6 38.Kg2 Rb8!? A non-forced tactical solution. 39.Rf3 b5 40.Rf7+ Kb6 41.cxb5 Kxb5 42.Rxh7 Kc5 43.Bxc6 Kxc6 44.Rg7 Rxb2 45.Rxg6 Kd7! 46.h4 e4! 47.Kf1 d5! 48.h5 Rb1+! 49.Kg2 Rb2 50.Kf1 Rb1+ 51.Ke2 d4! 52.Rg4 d3+!= 36...Kc7 37.Rh4 h5 Now Black has a more vulnerable weakness on g6 than on h7 before. This is the reason why I prefer 34...Re7. 38.Rc4+ Nc6 39.h4 Fixing g6 and giving White some advantage. But he can't keep the control and make progress at the same time. So sooner or later Black will find a way to break free. One examplary line is Rd8 40.b3 Kb6 41.Kg2 Kc7 42.Kf1 Kb6 43.Ke2 Kc7 44.Ke3 Ra8 45.Ra4 Rb8 46.Be4 Ne7 47.Ra5 Rg8 48.f4 Kb6 49.b4 exf4+ 50.gxf4 Rc8 51.Kd3 Rf8 52.c4 d5 53.Rb5+ Kc7 54.Bxd5 b6! 55.Ke4 Nf5 56.c5 bxc5 57.Rxc5+ Kb6= 36.Bxb7 Ra1+?! Gaining the b2-pawn and thus restoring material equality. Unfortunately for Najdorf, again the most natural is not the best and White gets a winning position after all. 36...Ra5! In comparison this offers Black way better chances. The idea is totally different from the text move. First point, the Ra5 is a knight's move apart from b7. That is, Kd7-c7 is a threat now. 37.Bf3 d5 Second point, suddenly Black can mobilise his central pawns. 38.Kg2 Kc7 39.Rb4
This position is highly interesting, very difficult and borderline. Meaning the evaluation is between a white win and a draw. White holds a sound extra pawn and the better minor piece, but Black has the central domination. In the summary of the analysis, I think the following points stand out: a) Black doesn't have a clear drawing concept. His job is to place his pieces optimally and wait for White's commitments in order to have an appropriate reaction at hand. b) White's general plan is to challenge Black's central superiority. His main idea is to exchange the black central pawns vs his c- and f-pawns. With more open space he can try to disorganise Black's pieces, especially with his long-ranged bishop vs the more short-ranged black knight. c) Black's biggest asset is the notorious combination of the white bishop and the wrong-coloured h-pawn. So another of White's ideas is to exchange this pawn, for example by h2-h4-h5xg6. d) In case of the exchange of rooks, the bishop vs knight ending is winning for White, if his passed b-pawn makes it to b5. Otherwise it is drawn. All in all, I would say that Black can just about hold on. With some uncertainty, because there is still a lot of play left for both sides. The lines to come are by no means complete or exhaustive, but primarily showing the typical ideas. 39...Ne6 39...e4 Seems to be a touch weaker. 40.Bd1 Nc6 41.Ra4 Rb5 42.Ra2! d4 43.cxd4 Rb4 44.b3 Rxd4 45.Be2 Kb6 46.Kf1 Ne5 47.Ra4 Kc5 48.Rxd4 Kxd4 As White's pawn will come to b5, this minor piece ending is lost for Black. But just after a long fight. So the latter might choose one of the several alternatives in the recent moves. Whether this brings him salvation is another question. 49.b4! Nd7 50.b5 Kc5 51.Ke1 Nb6 52.Kd2 Nd5 53.f3! A key move, according to the outlined plan. exf3 54.Bxf3 Ne7 55.Be2 Nd5 56.Kd3 h6 57.Bf1 Nf6 58.Ke3 g5 59.h3 Nh5 60.Kf2! Nf6 61.Kf3 Zugzwang Kd5 62.Ke3 Nh5 63.Kf2! Kc5 64.Be2 Nf6 65.Ke3 Zugzwang Nd5+ 66.Ke4 Nc3+ 67.Kf3 h5 68.Bd3 Kd5 69.b6 Kc6 70.g4! hxg4+ 71.Kxg4 Nd1 72.Kxg5 Nf2 73.Bb5+!+- 40.h4 Keeping many options open for the moment. 40.Rh4 Tempting, but too forcing. The refutation is one of these study-like defences by Charles Sullivan. The base idea is that the white rook will come into a kind of golden prison on h6. The only way to liberate him results into too many weak white pawns. Nf8! 41.b4 Rb5 42.c4 Rxb4 43.Bxd5 Kd6 44.Rh6 Rb3! 45.h4 e4‼ 46.Bxe4 Ke5 47.Bd5 Kf5! 48.Kf1 Kf6 49.g4 Kg7 50.g5 Nd7 51.h5 Nf6! 52.hxg6 hxg6 53.Bc6 Rb6 54.Bf3 Nh7= 40.c4 Another critical test for Black. e4 41.Bg4 Nd4 42.cxd5 Rxd5 White has exchanged the first of the black central pawns. 43.Rc4+ After this Black can force a draw, more or less. Instead White could try some more restrained alternatives. Kd6 44.Bc8 Ke5! For this central cluster Black is willing to sacrifice another pawn. 45.Rc7 Nf5 46.Rxh7 Nd6 47.Bg4 Rd2 48.Re7+ Kd4 49.b3 Rb2 50.h4 Rxb3 51.Re6 Nc4 52.Rxg6 Rb2 53.Rf6 e3 Not forced, but pretty. 54.Rf4+ Kd5! 55.Bf3+ Kd6! 56.Bh5 Kc5! With zugzwang for White. To be more precise: It is a strong mutual zugzwang. 57.Rf8 Kd6! 58.Kf1 exf2! 59.Rxf2 Ne3+! 60.Kg1 Rxf2 61.Kxf2 Nf5!= 40...Kc6 40...e4 41.Be2 Ra2 42.f3 Nc5 43.fxe4 Nxe4 So this is a sample line, where White exchanges the black e-pawn first. 44.c4 Too early, but still very dangerous for Black. d4 45.Kf3 Nd2+ 46.Kf4 Ra8 47.Bd3 Rf8+ 48.Ke5 Rf3 49.Kxd4 Rxg3 50.c5 g5 51.Rb6 gxh4 52.Bf5 h3 53.Rh6 h2! 54.Rxh7+ Kc6 55.Bd7+ Kb7 56.Bb5+ Kb8 57.Bc6 Rb3! 58.Rxh2 Rxb2 59.Kd5 Kc8!= No mate, no chance to exploit the pin, no win for White. 41.Rb8 Ra2 The more active choice for the black rook. 41...Ra7 42.h5 gxh5 43.Bxh5 A variation, where White could get rid of his notorious h-pawn. e4 44.Kf1 Ra1+ 45.Ke2 Rb1 46.Ke3 Re1+ 47.Be2 Nc5 48.b3 Kc7 49.Rb5 Kc6 50.c4 d4+ 51.Kxd4 Nxb3+ 52.Rxb3 Rxe2 As before. Because this rook ending is narrowly lost for Black, he should look for other ways in the previous moves. 53.c5 Rxf2 54.Rb6+ Kc7 55.Rh6 Rf3 56.Rxh7+ Kc6 57.Rh6+! Kc7 58.Rg6! e3 59.Kd3! Rf5 60.Kxe3! Rxc5 61.Kf4 Kd7 62.Rg5! Rc4+ 63.Kf5! Ke7 64.Kg6!+- 42.c4 e4 43.cxd5+ Kxd5 44.Be2 Ra7 45.b3 45.f3 Alternatively White can go for the exchange of the last black central pawn. exf3+ 46.Bxf3+ Ke5 47.Rb5+ 47.Rb7 Rxb7 48.Bxb7 The pure minor piece ending, with White's distant passed b-pawn on his starting square. Black can hold, but it is no walk in the park. Nc5 49.Bc6 Nd3! 50.b3 Kd4 51.Bb5 Ne5! 52.Kf2 Kc3 53.Ke3 Kxb3 54.Kf4 Nf7! 55.Bd7 Kc3 56.Be6 Nd8 57.Bg8 h6 58.Ke5 Nc6+ 59.Kf6 h5! 60.Bh7 Nd4! 61.Kxg6 Ne2= 47...Kd4 48.Rb4+ Ke5 49.Re4+ Kd6 50.Kf2 Ra1 White probably can't get any further from here. E.g. 51.Be2 Rb1 52.Rb4 Kc5 53.Rb7 Kd4 54.Rb6 Kd5 55.Bf3+ Ke5 56.Rb5+ Kd6 57.Rd5+ Ke7 58.Rd2 Rc1 59.Ke3 Rg1 60.Rg2 Ra1 61.Rc2 Rg1 62.Kf2 Ra1 63.Be2 Kd6 64.Rc4 Rb1 65.Rb4= With the same position as on move 52. 45...Kd4 46.Rb4+ Ke5 47.Bc4 Nd4 48.Ra4 Rb7 49.b4 White has made some progress with his passed b-pawn. Nc2 As Charles Sullivan discovered, it is best for Black to keep controlling f3 with the knight for some more moves. 49...Nf5?! 50.b5 Nd6 Black has given up a bit of his central piece configuration. I can't say whether this is decisive already. 51.Be2 Rc7 52.Ra2 Rb7 53.Rb2 Kd4 54.Rb3 Rb8 55.f3! A small progress. Rb6 56.Bf1 Kc5 57.Rb1 Kd5 58.Ra1 Rb8 59.fxe4+ Nxe4 60.Bd3 Nf6 61.Ra7 Kd4 62.Be2 Ke3 63.Re7+ Kd4 64.Rf7 Ke5 65.Ra7 Rb6 66.Kf3 Rb8 67.Re7+ Kd6 68.Rf7 Ke5 69.Bf1 Rb6 70.Re7+ Kd4 71.Ra7 Rb8 72.Be2 Rc8 73.Ra4+ Ke5 74.Ra3 Rf8 75.g4 h6 76.Kg2 Rb8 77.Rb3 Rb6 78.Bf3 Kd6 79.Kg3 Nd7 80.Kf4 Rb8 81.Rd3+ Kc7 82.Rd5 Rf8+ 83.Kg3 Rf6 84.Be4 Nf8 85.Rc5+ Kb6 86.Rc6+ Rxc6 87.Bxc6 Another version of bishop vs knight, with the white pawn on b5. Again White can demonstrate a win hereupon. Nh7 88.Be4 Nf8 89.Bd3 g5 90.h5 Nd7 91.Kf2 Nf6 92.Be2 Nd5 93.Kf3 Nf6 94.Ke3 Kc5 95.Bf3! Kxb5 96.Kd4 Nd7 97.Kd5 Kb6 98.Kd6 Nc5 99.Ke7 Kc7 100.Kf6 Kd6 101.Kg6 Ke5 102.Kxh6 Kf6 103.Kh7+- 50.b5 Ne1+ 51.Kf1 Nf3 52.Ke2 Nd4+ 53.Kd2 Rb8 54.Ra5 Nf5 55.Ra7 Nd6 56.Be2 Rh8 Sometimes cowering in the corner is best. Although Black's piece placement isn't optimal anymore, White gets no chance to enforce the exchange of the last black central pawn by f2-f3. 57.Ra3 Rf8 58.Ke1 Kd5 59.Ra7 Rh8 60.Kd1 Ke5 61.Kd2 Kd4 62.Rd7 Ke5 63.Ra7 Kd4 64.Ra4+ Kd5 65.Rb4 As more manoeuvring doesn't promise anything, White has to try someting sooner or later. Rf8 66.b6 Kc5 67.Rb1 Rb8 68.Ke3 Rxb6 69.Rxb6 Kxb6 70.Kf4 Kc6 71.Kg5 Kd5 72.Kh6 e3! 73.fxe3 Nf5+ 74.Kxh7 Ke4! 75.Kxg6 Nxg3= 37.Kg2 Kc7 37...Ra5?! After White's king has abandoned the first rank, this idea runs into a tactical refutation. 38.Bf3 d5?! 39.b4+- 38.Ra6 Rb1 39.Bd5! To form a powerful duo with the isolated white pawn on c4 in the future. Rxb2 The new problems of Black are the opened 7th rank and thus the mega-weakling on h7, as well as the placement of his knight. Who is still on the rim, costing time to improve. 40.Ra7+ 40.h4 Nb7 41.Ra8 This is a shortcut to a position to be discussed a little later, in the side line 41.h4 after 43.Ra8. 40...Nb7 41.Kf3 Very often in similar situations White can abuse such a pin of a piece by making some preparatory moves, exchanging twice on b7 and scoring the full point via an easy pawn ending. So improving the king is the most natural. Alas, here it isn't that simple. 41.h4 Alternatively, White can opt for a set-up where his king stays put on g2 for some time. Kb8 42.Ra6 Kc7 43.Ra8 Nc5 43...e4 44.Rh8 Nc5 45.Rxh7+ Kd8 comes to the same 44.Ra7+ Kd8! A picture we will see several times again in slightly different versions. Once Black's knight has improved to c5, his best chance is to give up the h7-pawn. 45.Rxh7 e4 46.Kf1 For his pawn deficit Black has active play with his pieces and he can cause real trouble. However, in the end White will emerge victorious, even if it takes its time. For example Rc2 46...e3 47.fxe3 Nd3 48.Be4 Ne5 49.Rh6 Ke7 50.Bxg6 Ng4 51.Rh8 Nxe3+ 51...Kf6 52.Kg1 Ne5 53.Bh5 Rc2 54.Rf8+ Ke6 55.e4+- 52.Ke1 Kf6 53.Bd3 Ng2+ 54.Kd1 Ne3+ 55.Kc1 Rh2 56.Rf8+ Ke7 57.Rf3 Ng4 58.Bc4 Ne5 59.Re3 Kf8 60.Bd5+- 47.Rg7 Rc1+ 48.Ke2 Rc2+ 49.Ke1 e3 50.fxe3 Rxc3 51.Ke2 Rc2+ 52.Kf1 Rc1+ 53.Kg2 Rc2+ 54.Kh3 Nd3 55.Bf7 Nf2+ 56.Kg2 Ng4+ 57.Kf3 Ne5+ 58.Ke4 Rg2 59.Bxg6 Rxg3 60.h5 Rh3 61.Kd5 Rxe3 62.Kxd6 White has just one pawn left, but the black knight is too short-ranged to deal with the passed pawn and support his king simultaneously. Nc4+ 63.Kc5 Ne5 64.Bf5 Ra3 65.Kd5 Ra5+ 66.Ke6 Nc6 67.Bg4 Ra6 68.Rg5 Kc7 69.Rc5 Kb6 70.Rd5+- 41...Kb8 42.Ra6 As the pawn ending offers nothing for now, this rook has to retreat. 42.Ra2? Tempting, but inappropriate in the end. White's basic point of the rook exchange is to run over Black's weak kingside. However, the later can count on an instructive resource. Rxa2 43.Bxa2 Kc7 44.Bg8 h6 45.Bh7 Kd7! 45...g5? 46.Kg4 Kd7 47.Kh5 Ke7 48.Kxh6 Kf6 49.c4 Nc5 50.f3+- 46.Bxg6 Ke6 Black will in fact build up a blockade on the dark squares. 47.c4 47.Kg4 Kf6 48.Kh5 Kg7 and White's king has to return sooner or later. 47...Nc5 48.Be4 Kf6 49.Ke3 Nb3 50.Bd5 Nc5 White can try some set-ups here, but going with the king the long way around to c6 seems the most dangerous. 51.f3 Kf5 52.Kd2 Nb3+ 53.Kc3 Nc5 54.Kb4 Kf6 55.Kb5 Nd3 56.Kc6 Ke7 57.h4 Nb4+ 58.Kc7 Nd3 59.g4 Nf4! Black needs to be alert. His knight has to guard the white kingside pawns. 60.Kc6 Nd3 61.Be4 Nf4 62.Kc7 62.g5 hxg5 63.hxg5 Ne6 64.g6 Ng7= 62...Ne6+ 63.Kb6 Nf4 64.Kc6 Ne6 65.g5!? After a king's triangle to hand the move to Black, White can try a breakthrough as his last chance. hxg5 66.h5 Nd8+! 67.Kd5 Nf7! 68.Bf5 Nh6 69.Be6 e4 70.fxe4 g4 71.e5 dxe5 72.Kxe5 g3 73.Bh3 Nf7+! 74.Kf4 Kf6= 42...Kc7 42...Nc5 Not as unreasonable as it might look at first sight. A lot of pawns from both sides will leave the board. 43.Rxd6 Nd3 44.Bg8 Nxf2 45.Bxh7 Nh3 46.Bxg6 Rxh2 47.Rd5 Ng1+ 48.Kg4 Ne2 49.Rb5+ Ka7 50.c4 Rg2 51.Rb3 Nd4 52.Re3+- Finally winning a second pawn, while the black pieces run into the void. 43.Ra8? Missing the chance for a necessary second set-up move, following Kg2-f3, before running over the black kingside. Next, the black knight will hop to c5 anyway, with the intention to let the h7 go. Thereupon White will reply 44.Ra7+, no matter if the rook starts on a6 or a8. 43.h4! Best, as it secures the white h-pawn. Nc5 43...h6 44.Ra7 Kb8 45.Ra4! Rd2 46.c4 Rd3+ 47.Kg2 Rb3 48.Ra6 Kc7 49.c5 Rd3 50.Be4 Rd4 51.cxd6+ Nxd6 52.Bxg6 e4 53.Ra8+- 44.Ra7+ Kd8 45.Rxh7 The entire play is very similar to the line with 41.h4. Black annoys the white king and tries to reduce the white pawn mass. Rc2 46.c4 e4+ 47.Ke3 Rc3+ 48.Kd2 Rb3 49.g4!? Maybe not very elegant, but a very forcing solution. 49.Rf7 The more controlled way. Now Black can and has to invest a second pawn, to keep his initiative alive. However, White will prevail, even if it takes a lot of patience. E.g. e3+ 50.fxe3 Rb2+ 51.Ke1 Nd3+ 52.Kf1 Ne5 53.Rf6 Kd7 54.Re6 Rc2 55.Ke1 Kc7 56.Kd1 Rc3 57.Kd2 Ra3 58.Ke2 Ra2+ 59.Kd1 Ra3 60.Kd2 Kd7 61.Ke2 Ra2+ 62.Kd1 Ra3 63.e4! Rc3 64.g4! Kc7 64...Nxg4 65.Kd2 Rh3 66.Rxg6 Rh2+ 67.Kc3 Ne5 68.Rh6+- 65.h5 gxh5 66.gxh5 Kb6 67.h6 Rh3 68.Ke2 Kc5 69.Kf2 Ng4+ 70.Kg2 Rh5 70...Rxh6 71.Rxh6 Nxh6 72.Kg3+- Knight in prison. 71.Kg3 Nxh6 72.Rg6! Rh1 73.Kg2 Rh4 74.Kf3 Rh1 75.Ke3 Rh4 76.Be6! A new corral for the knight. Rh1 77.Rg5+ Kc6 78.Kf2 Rh2+ 79.Kg3 Rh1 80.Kg2 Rh4 81.Kf3 Kb6 82.Ke3 Rh1 83.Rg7 Rh4 83...Kc5 84.Kf4 Re1 85.Rg5+ Kb6 86.Rg6 Rh1 87.Kg5+- 84.Kd4 Rh1 85.Bd5 After a lot of manoeuvring Black's coordination isn't as good as before. For example his king has been driven away from the centre. Rd1+ 86.Kc3 Rc1+ 87.Kd3 Rd1+ 88.Ke2 Rh1 89.Kf2 Rh3 89...Kc5 90.e5! 90.Kg2 Rh5 91.Kg3 Kc5 92.Kf4 Kd4 93.Rd7 Kc5 94.Rc7+ Kd4 95.Rc6+- As the d6-pawn is lost. 49...Rb2+ 50.Ke3 Rb3+ 51.Ke2 Rb2+ 52.Kf1 Nd3 53.Rf7 Ne5 54.Rf6 Nxg4 55.Rxd6+ Kc7 56.Rc6+! Kd8 57.Rxg6 Rxf2+ 58.Ke1 Rf4 59.h5 Ne5 59...Nf6 60.h6 Ke7 61.h7 Rh4 62.Bg8 Rh2 63.Kf1 e3 64.Ke1 Nd7 65.Rg2 Rh1+ 66.Ke2+- 60.Rd6+ Kc7 61.Re6 Rf5 62.h6 Rh5 63.Bxe4 Kd7 64.Ra6 Nxc4 Tablebase land, confirming a white win based solely on tactics. 65.h7! Nd6 66.Bd5! Ke7 67.Ra8! Nf7 68.Ra7+! Kd6 69.Rxf7! Kxd5 70.Rf5+!+- White walked on a thin line for a long time. 43...Nc5! 44.Ra7+ Nb7? A fairly big error by Najdorf this time, I am sorry to say. This passive retreat loses clearly, as now White gets the opportunity for even two more set-up moves. 44...Kd8! Even without knowing the details to come, Black should be very willing to keep his finally useful knight on c5 and invest his weakest pawn on h7. Indeed, in this version Black gets enough counterplay to keep the balance in the end, since the white pawn has remained on h2. 45.Rxh7 Rc2 46.c4 e4+ 47.Ke3 Rc3+ 48.Kd2 Ra3! 49.Rf7 49.Rh4 g5 50.Rh8+ Ke7 51.h4 gxh4 52.Rxh4 e3+ 53.fxe3 Ra2+ 54.Kd1 Nd3 49.h3 Ra2+ 50.Ke3 Ra3+ 51.Ke2 Ra2+ 52.Kf1 Ra1+ 53.Kg2 Ra2 54.Rh8+ Ke7 55.Ra8 Re2 49...e3+! 50.fxe3 Ra2+ 51.Kc3 51.Ke1 Rxh2= Demonstrating the difference to the line following 43.h4. 51...Ra3+ 52.Kb4 Rxe3 53.Rg7 Nd3+ 54.Kc3 Ne5+ 55.Kd4 Re1 Not many pawns left on the board and Black's knight on e5 is eventually something to be proud of. 45.h4! This little improvement for the h-pawn is so useful, as we have seen. 45.Ke4 Thanks to Najdorf's latest mistake it is even possible to prepare a winning transition. Although the resulting pawn ending needs exact handling for some time, so Tal wisely refrains from it. Re2+ 46.Kd3 Rb2 47.Kc4 Kb8 48.Rxb7+ Rxb7 49.Bxb7 Kxb7 50.Kd5 Kc7 51.c4 Kd7 52.h4! h5 53.f3! Ke7 54.c5! dxc5 55.Kxe5! Kf7 56.Kd5! Kf6 57.Kxc5 Ke5 58.Kc4!+- 45...Kb8 46.Ra6 Kc7 47.Ra8?! Tal commits the same kind of mistake as in move 43. 47.c4! Much better to use the time for a next set-up move. h5 47...Nc5 48.Ra7+ Kd8 49.Rxh7 Rb3+ 50.Kg4 e4 51.Kg5+- 48.Be4 Nc5 49.Ra7+ Kb6 50.Ra8 Kc7 50...Rb3+ 51.Kg2+- 51.Bxg6 e4+ 52.Bxe4 Nxe4 53.Kxe4 Rxf2 54.Ra5+- With a fairly easy win in all the lines. 47...Rb5?! I guess by now Najdorf has lost all hopes of salvation already. His move poses no difficulties for Tal anymore. 47...Nc5! 48.Ra7+ Kd8 49.Rxh7+- But White has to work very hard and very long for his win. See 43.h4 Nc5 44.Ra7+ Kd8 45.Rxh7 for all the details of this position. 48.c4! A fairly obvious move. However, at this point Tal demonstrates the optimal pawn set-up, albeit with some delay. Rb3+ 48...Rb2 49.Rg8+- 49.Kg4 With the white king on his way to the black kingside pawns and a lost pawn ending via Ra8-a7 looming, Najdorf resigns. 1–0

Links

Karsten Müller is considered to be one of the greatest endgame experts in the world. His books on the endgame - among them "Fundamentals of Chess Endings", co-authored with Frank Lamprecht, that helped to improve Magnus Carlsen's endgame knowledge - and his endgame columns for the ChessCafe website and the ChessBase Magazine helped to establish and to confirm this reputation. Karsten's Fritztrainer DVDs on the endgame are bestsellers. The mathematician with a PhD lives in Hamburg, and for more than 25 years he has been scoring points for the Hamburger Schachklub (HSK) in the Bundesliga.
We use cookies and comparable technologies to provide certain functions, to improve the user experience and to offer interest-oriented content. Depending on their intended use, analysis cookies and marketing cookies may be used in addition to technically required cookies. Here you can make detailed settings or revoke your consent (if necessary partially) with effect for the future. Further information can be found in our data protection declaration.