Akiba Rubinstein s Masterpiece As Savielly Tartakover wrote: Rubinstein is a rook ending of a chess game that was started by Gog a thousand years ago. Rubinstein Spielmann Notes by Dvoretsky, Levenfisch, Smyslov, Kmoch and Kasparov
Spielmann - Rubinstein [A positional disadvantage that occurs often is an abundance of "pawn islands." White has four islands against Black's two; this means that White has more vulnerable pawns that cannot protect each other. Therefore his position is inferior. ] 40...Tb8 a8! [ The first stage of Black's plan is to attack White's pawns so that the white rook will be chained to their protection.] [Chasing after material gain with 40...Tb8 b3? would have been a grave error, because after 41.Tc2 a2 Tb3 d3 42.a3 a4 Td3xd4 43.a4 a5 Td4 c4 44.a5 a6 the white rook is actively placed behind a passed pawn while the black rook must stand passively on a8. 44...Tc4 c8 45.a6 a7 Tc8 a8 46.Kh2 g3 ] 41.Tc2 c3 [The position of the rook on the 3rd rank has its own virtues, only Spielmann fails to exploit them in the further process. ] [41.Tc2 a2 Ta8 a4 42.Kh2 g3 Kf8 e7 (42...Ta4xd4? 43.a3 a4 ) 43.Kg3 f3 Ke7 e6 44.Kf3 e4 G. Levenfish and V. Smyslov also analyze 44...g6 g5!? This continuation is more dangerous, but their line shows that White maintains sufficient defensive resources: (44...d6 d5+ 45.Ke4 d3!? (Spielmann thinks that the rook stands even worse on a2 and explains this judgment with the line 45.Ke4 e3 Ke6 f5 However, the final position of this line is far from clear. And secondly, instead of 6 Ke3 White can play 6 Kd3!?.) 45...Ke6 f5 (45...Ke6 d6 46.Kd3 c3 Kd6 c6 47.Kc3 d3 Kc6 b5 48.Ta2 b2+) 46.Kd3 c3 Kf5 e4 47.Ta2 e2+ As we can see, the rook behind the pawn has some hidden potency although it is currently passive. It chains the black rook; for as soon as the black rook leaves a4 the white rook supports the advance of the a-pawn.) 45.Ta2 a1 f7 f6 46.Ta1 a2 f6 f5+ 47.Ke4 d3 Ke6 d5 48.Kd3 c3 Ta4 c4+ 49.Kc3 b3 Tc4xd4 50.a3 a4 Td4 d3+ (50...Kd5 c6 51.a4 a5 Kc6 b7 52.a5 a6+ Kb7 a7 53.Ta2 a5 Td4 f4 54.Ta5 d5 Tf4xf2 55.Td5xd6 Tf2 f3+ 56.Kb3 c4 Tf3xh3 57.Td6 g6 g5 g4 58.Kc4 b5 Th3 b3+ 59.Kb5 a5=) 51.Kb3 b4 Td3xh3 52.a4 a5 Th3 h8 53.a5 a6 Kd5 e4 54.a6 a7 Th8 a8 55.Kb4 b5 Ke4 f3 56.Kb5 b6=]
41...Ta8 a4 42.Tc3 d3 Kf8 e7 [ The second stage: the king goes to the center.] 43.Kh2 g3 [43.d4 d5 is met with 43...g6 g5! (43...Ke7 f6 44.Td3 f3+; 43...f7 f5 44.Td3 e3+ e6) 44.Kh2 g2 Ke7 f6 45.Td3 f3+ Kf6 g6 d4 46.Tf3 d3 f7 f6! f5; VariantDvoretsky 43.Kh2 g2!?; Variant Rybka 3.0 43.Td3 e3+; Variant Deep Fritz 10 43.h3 h4] 43...Ke7 e6 9+-+-+pzp-0 9-+-zpk+p+0 9+-+-+-+-0 9r+-zP-+-+0 9zP-+R+-mKP0 9-+-+-zP-+0 9+-+-+-+-0 44.Kg3 f3? [ In my opinion this is an obvious positional error that was somehow left unnoticed by the annotators. Letting the black king pass to d5, White condemns himself to a passive defense that, as we know, forebodes gloom in rook-and-pawn endings.] [He could get excellent chances for a draw by playing 44.Td3 e3+! Ke6 d7 (44...Ke6 d5 45.Te3 e7 Ta4xa3+ 46.Kg3 g2) and now either 45.Te3 d3 (or 45.Te3 f3!? f7 f6 46.d4 d5 Ta4 d4 47.Tf3 b3 Td4xd5 48.Tb3 b7+ Kd7 e6 49.Tb7xg7 g6 g5 50.Tg7 a7) 45...Kd7 c6 46.Td3 c3+ Kc6 d5 47.Tc3 c7 Ta4xa3+ 48.Kg3 g2 Kd5 e6 49.d4 d5+ Ke6 f6 50.Tc7 d7 Ta3 a6 51.h3 h4] 44...Ke6 d5 45.Kf3 e2?! [ Another inaccuracy.]
[A good idea was to restrain Black's pawns on the kingside by means of ¹45.h3 h4! It's worth mentioning that here, as well as later on, White is not afraid of 45...Ta4xd4 because he has a distant passed pawn in the pawn endgame after 46.Kf3 e3] 45...g6 g5! [ The third stage of the plan: it is important to improve the pawn structure on the kingside.] [45...Ta4xd4? 46.Ke2 e3] 46.Td3 b3 9+-+-+pzp-0 9-+-zp-+-+0 9+-+k+-zp-0 9r+-zP-+-+0 9zPR+-+-+P0 9-+-+KzP-+0 9+-+-+-+-0 46...f7 f6! [46...Ta4xd4? can be met either with 47.Tb3 b5+ (or 47.Tb3 d3!?=) ; In case of 46...Kd5xd4?! 47.Tb3 b7 White has considerably reduced the number of pawns. (47.Tb3 g3!? Ta4 a5 48.a3 a4 Dvoretsky) Although G. Kasparov claims the ending after 47...f7 f6! (47...Ta4xa3? 48.Tb7xf7 Ta3xh3 49.Tf7xg7³) 48.Tb7xg7 Ta4xa3 49.h3 h4! g5xh4 50.Tg7 g4+ Kd4 c3 51.Tg4xh4 Ta3 a2+ to be won for Black, it is not clear whether his judgment is correct. In addition, Black should take 8 Rg3!? Ra5 9 a4 into account. 52.Ke2 e3 d6 d5 53.Th4 h3 d5 d4+ 54.Ke3 e4+ Kc3 c4 + Rubinstein's move is safer.]
47.Ke2 e3 9-+-zp-zp-+0 9+-+k+-zp-0 9r+-zP-+-+0 9zPR+-mK-+P0 9-+-+-zP-+0 9+-+-+-+-0 [47.Tb3 b7 can be met now with 47...Ta4xa3 48.Tb7xg7 Ta3xh3 (48...Kd5xd4µ brings us to the above-mentioned line.) 49.Tg7 g6 (49.Tg7 f7 Kd5 e6! 50.Tf7 f8 f6 f5 51.Tf8 e8+ Ke6 d5 52.Te8 f8 Kd5 e4 g4, f3 +) 49...Kd5 e6 50.Tg6 g8 Th3 h4 51.Ke2 e3 Ke6 d5 This position, as Kasparov has proven, is winning, and here I agree with him. One who fights for a win should avoid pawn exchanges; in this line, an extra pawn pair remains on the board compared with the 7...Kxd4 line. 52.Tg8 g6 (52.Tg8 f8 Th4 e4+ 53.Ke3 f3 Te4 e6 54.Kf3 g4 Kd5xd4 55.Kg4 f5 Te6 e2 +) 52...Th4 e4+ 53.Ke3 f3 Te4 e6 (53...g5 g4+ 54.Kf3 g3 f6 f5 55.Tg6 f6 Te4xd4 56.Tf6xf5+ Kd5 c4µ e4; 53...Te4 f4+ 54.Kf3 e3 Kd5 c4 55.f2 f3 Kc4 c3 56.Tg6 h6 d6 d5 57.Th6 g6 Tf4xd4 58.Tg6xf6 Td4 d3+ 59.Ke3 f2 Td3 d2+ 60.Kf2 e3 d5 d4+ 61.Ke3 e4 Td2 e2+ 62.Ke4 f5 Te2 g2 63.Kf5 e4 d4 d3 64.Tf6 c6+ Kc3 d2 65.Tc6 d6 Tg2 e2+ 66.Ke4 f5 Te2 e3 +) 54.Kf3 g4 Kd5xd4 55.Kg4 f5 Te6 e2! 56.Kf5xf6 Te2xf2+ 57.Kf6 e6 Tf2 e2+ 58.Ke6xd6 Te2 e5 + Kasparov] 47...Kd5 c4 48.Tb3 d3 [48.Tb3 b7 Ta4xa3+ 49.Ke3 e4 d6 d5+ 50.Ke4 f5 Ta3xh3 51.Tb7xg7 Th3 f3+ + is hopeless.] 48...d6 d5 [ Black has improved his pawn structure and optimally placed his king. Now it is time for the rook. It has completed its mission on a4 and may find a new application for its talents.]
49.Ke3 d2 Ta4 a8 50.Kd2 c2 [50.Kd2 e2? Ta8 b8 (50...Ta8 e8+) 51.Ke2 d2 Tb8 b2+ 52.Kd2 e3 Tb2xf2 Dvoretsky (Or 52...Tb2 a2 Spielmann) ] 50...Ta8 a7! 51.Kc2 d2 Ta7 e7! [ ] 52.Td3 c3+! [ The last chance to display activity.] [A continued passive policy would have led to an inglorious demise: 52.Kd2 c2 Te7 e2+ 53.Td3 d2 Te2xd2+ 54.Kc2xd2 Kc4 b3! + Spielmann; Or 52.Td3 e3 Te7xe3! (52...Te7 b7! 53.Te3 d3 Tb7 b2+ 54.Kd2 e3 Tb2 b3 + Spielmann (54...Tb2 a2 +) ) 53.f2xe3 (53.Kd2xe3 Kc4 b3) 53...f6 f5! 54.Kd2 c2 (54.Kd2 e2 Kc4 b3 55.Ke2 f3 Kb3xa3 56.Kf3 g3 Ka3 b4 57.h3 h4) 54...g7 g6 + 55.Kc2 d2 (55.Kc2 b2 g5 g4) 55...Kc4 b3 + Dvoretsky] 52...Kc4xd4 53.a3 a4! Te7 a7 54.Tc3 a3 Ta7 a5! [The pawn must be stopped as soon as possible. Black intends to approach it with his king: either simply for winning it or for blocking it and releasing the rook from its passive position.] 55.Ta3 a1 Kd4 c4 9-+-+-zp-+0 9tr-+p+-zp-0 9P+k+-+-+0 9+-+-+-+P0 9-+-mK-zP-+0 9tR-+-+-+-0
56.Kd2 e3?! [56.Ta1 c1+! Kc4 b4 57.Tc1 b1+! Kb4xa4 58.Kd2 d3 deserved attention: (58.Tb1 b7? Ta5 b5! is not so good.) Levenfish and Smyslov analyzed 58...Ta5 c5! (The position of the black king on the edge could give some practical chances. For example, after 58...Ta5 b5? 59.Tb1 a1+ Ka4 b4 60.Kd3 d4± it was Black who had to fight for a draw.) 59.Kd3 d4 (59.Tb1 b7? Tc5 c4 60.Tb7xg7 Ka4 b5 +) 59...Tc5 c2 60.Tb1 b7 (60.Kd4xd5 Tc2xf2 61.Kd5 e6 Tf2 h2 62.Ke6 f7 Th2xh3 +) 60...Tc2xf2 61.Tb7xg7 in their opinion, White maintains chances for a draw. Kasparov extended this line with 61...Tf2 d2+! 62.Kd4 c5 Ka4 b3 63.Tg7 g6 (63.Tg7 d7 Td2 d3 (63...Kb3 c3 64.Td7xd5 Td2 f2 65.Kc5 d6; 63...Td2 c2+ 64.Kc5 d4) 64.Td7xd5 Td3xh3 65.Td5 f5 Kb3 c3! (65...Th3 h6 66.Kc5 d4) 66.Tf5xf6 g5 g4 + Dvoretsky) 63...Kb3 c3 64.Tg6xf6 d5 d4 and evaluated the final position as winning. I think he is incorrect here: White saves himself by 65.Tf6 a6! Kc3 d3 a) 65...Td2 d3 66.Ta6 a4=; b) 65...d4 d3 66.Ta6 a3+ Kc3 b2 67.Kc5 b4 Td2 d1 68.Ta3 b3+ (68.Ta3 c3!=) 68...Kb2 c2 69.Kb4 c4 d3 d2 70.Tb3 c3+ Kc2 b2 71.Tc3 b3+ Kb2 a2 72.Kc4 c3=; c) 65...Td2 d1 66.Ta6 a3+ Kc3 b2 67.Kc5 b4 Td1 e1 (67...d4 d3 68.Ta3 c3=) 68.Ta3 d3 Te1 e3 69.Td3xd4 Te3xh3 70.Td4 d2+ Kb2 c1 71.Td2 g2=; 66.Kc5 d5 Kd3 e3 67.Ta6 e6+ Ke3 f4 68.Te6 f6+ Kf4 g3 69.Tf6 g6= It seems that, in spite of previous errors, Spielmann's position remained tenable. (Dvoretsky)]
56...d5 d4+ 57.Ke3 d2 9-+-+-zp-+0 9tr-+-+-zp-0 9P+kzp-+-+0 9+-+-+-+P0 9-+-mK-zP-+0 9tR-+-+-+-0 57...Ta5 f5! [ Black's precise 15th move tells: the rook may leave the blockade position.] 58.Kd2 e1 [If 58.a4 a5 then 58...Tf5xf2+ 59.Kd2 e1 Tf2 b2! (Rather than 59...Tf2 h2? 60.Ta1 a4+ Kc4 b5 61.a5 a6!) 60.a5 a6 Tb2 b8 61.a6 a7 Tb8 a8 62.Ke1 d2 Kc4 c5 63.Kd2 d3 Kc5 b6 64.Kd3xd4 (64.Ta1 b1+! Kb6xa7 65.Kd3xd4 Ta8 b8 66.Tb1 a1+ Ka7 b6 67.Kd4 d5 Tb8 e8 68.Ta1 a3 is more stubborn, and Black's win is still not a simple matter. (K. Müller)) 64...Ta8xa7 + Spielmann] 58...Kc4 b4! [ A typical reassignment of pieces: the king will block the past pawn while the rook will attack White's weak pawns.]
59.Ke1 e2 Kb4 a5! 9-+-+-zp-+0 9mk-+-+rzp-0 9P+-zp-+-+0 9+-+-+-+P0 9-+-+KzP-+0 9tR-+-+-+-0 60.Ta1 a3 [After 60.Ta1 b1 Ka5xa4 61.Tb1 b7 Kasparov suggests 61...d4 d3+ (61...g7 g6! + is much stronger. (Dvoretsky)) 62.Ke2xd3 Tf5xf2 63.Tb7xg7 Tf2 f3+ 64.Kd3 e4 (64.Kd3 c4! Ka4 a3 (64...Tf3 f4+ 65.Kc4 d5 Ka4 b3 66.Kd5 e6=) 65.Tg7 g6 Ka3 b2 66.Kc4 d5 Kb2 c3 67.Kd5 e6=) 64...Tf3xh3 + This line is erroneous: White holds by means of 64 Kc4!] 60...Tf5 f4 61.Ta3 a2 [61.Ke2 f1 Tf4 h4 62.Kf1 g2 Ka5 b4! 63.Ta3 a1 d4 d3 64.a4 a5 d3 d2 65.a5 a6 Th4 h8 66.a6 a7 (66.Kg2 f1 Kb4 c3) 66...Th8 a8 67.Kg2 f3 Ta8xa7 + Spielmann] 61...Tf4 h4 62.Ke2 d3 [62.Ta2 a3 Ka5 b4] 62...Th4xh3+ 63.Kd3xd4 Th3 h4+ 64.Kd4 d3 [64.Kd4 e3 Th4xa4 65.Ta2 d2 Ka5 b6! and if 66.Td2 d7 then 66...Ta4 a7] 64...Th4xa4 65.Ta2 e2 [ e7] 65...Ta4 f4! [65...Ka5 b6? is wrong 66.Te2 e6+! e7]
66.Kd3 e3 Ka5 b6 67.Te2 c2 [ c8 g8] 9-mk-+-zp-+0 9-+-+-tr-+0 9+-+-mK-+-0 9-+R+-zP-+0 9+-+-+-+-0 67...Kb6 b7! [ Accurate to the last! Black prevents the maneuver Rc8 g8 and prepares to cross the c-file with his king after Ra4 a6 c6.] 68.Tc2 c1 Tf4 a4 69.Tc1 h1 Kb7 c6 70.Th1 h7 Ta4 a7 71.Ke3 e4 Kc6 d6 72.Ke4 f5 g7 g6+! 73.Kf5xg6 Ta7xh7 74.Kg6xh7 Kd6 e5 75.Kh7 g6 g5 g4 0 1