Capablanca-Yates (Hastings, 1930) revisited Over ten years ago, I shared on this forum my analyses of Capablanca’s two classic rook endings with
e, f, and g pawns versus f and h pawns: against Yates (1930)
https://www.chesspub.com/cgi-bin/chess/YaBB.pl?num=1336829888/all and Duras (1913)
https://www.chesspub.com/cgi-bin/chess/YaBB.pl?num=1357755353/all . These analyses demonstrated that much of the traditional theory was flawed in several respects. In correspondence with Mark Dvoretsky, the fundamental results of these analyses were reflected in the 2. Russian edition of
Dvoretsky’s Endgame Manual (Moscow 2016, p. 245-246) and in the 5th English edition (Russell Enterprises, 2020, p. 201-202). However, several recent books dedicated to endgames, such as those by
E. Grivas (Your Jungle Guide to Rook Endings, Thinkers Publishing 2020) and
М. Sherechevsky (С молодьежю в эндшпиль (With the youth – into the Еndgame), vol. 2, Moscow 2021) continue to perpetuate the errors of old theory. Remarkably, in S.
Shankland's excellent work
Theoretical Rook Endings (Quality Chess UK Ltd, 2023, p. 184) there is no mention of the mistakes in the Capablanca-Yates game (63.Kf2? Ka3?) that were already known in the previous theory. Recently,
B. Gelfand also revisited the Capablanca-Yates game (
Decision Making in Major Piece Endings (Quality Chess UK Ltd, 2020). Although his analyses of the line 61...Ra4 (instead of 61...Re3 played in the game) do not contain errors, there are some contentious issues regarding the methods to achieve victory.
In this thread, I'll demonstrate that:
1. Gelfand's suggestion of 67...Rd4 in D. 3 is not an improvement compared to previous analyses, and 79.Rd7 (D. 9) makes the win very difficult. The easiest winning path for White remains the old plan proposed by Kopayev (Part 1).
2. Contrary to the prevailing opinion until now (Kopayev, Dvoretsky, Gelfand), 70.f5 in D. 2 after 67…Rc6 68.Ke4 Rc4+ 69.Kf3 Rc6 also leads to a win, albeit in a much more complex manner (Part 2).
Capablanca-Yates (Hastings, 1930) D. 1 
61.Rb6?! 61.Rd6 (Kopayev) is more precise, but, as established since 2012, 61.Rb6 does not spoil anything. Nevertheless, Grivas and Shereshevsky give this move a question mark and believe that the position is drawn. The disadvantage of this move is that the White rook cannot protect its king from side checks. Black gets the opportunity to push the White king to the unfavourable square f5.
In the game after 61.Rb6 Yates played 61…Re3 (II) which is weaker than 61…Ra4 (I).
I) 61…Ra4 62.Kf3
Kopayev warns against the line 62.Kg3(?!) Ra3+ 63.Kh4 Ra4 64.f5(?) Ra5=. Dvoretsky and Gelfand repeat this line without any comments. 62.Kg3 is in fact a useless move, but it does not throw away the win. After 62… Ra3+ 63.Kh4 Ra4 White has to backtrack by 64.Kg3. 64.f5? is of course a decisive error and gives away the win.
62...Ra3+ 63.Ke4 Ra4+ 64.Kf5 Rc4 D. 2 Now “in order to achieve victory White should a) displace the Black rook from the 4th rank, b) return the king to the centre, and c) prepare the advance of the f-piece” (Kopayev,
Shakhmatny Bulletin (‘Chess Bulletin’), n. 8-1956, p. 229).
65.Rb7 Kg8 66.Rb3 Kg7 67.Re3 D. 3 
In 1956 Kopayev analysed this position and come to the conclusion that White wins after 66…Rb4 67. Re4. 2 years later in the first edition of
Shakhmatnye okontchaniya (‘Chess Endings. Knight against Bishop, Rook endings’. General Editor Yu. Averbakh, Moscow, 1958, p. 333) he changed his mind and thought that by 67…Rc6 Black can save the game. In 2012 Dvoretsky and I found independently that after 68.Ke4 Rc4+ 69.Kf3 Rc6 White wins by 70.Ra3 f6 71.Ra7+ Kf8 72.ef6! (D. 7 of the thread Capablanca-Yates (Hastings, 1930).
In D. 3 instead of 67…Rc6 Gelfand suggests 67...Rd4 and gives this move an exclamation mark.
67...Rd4 According to Gelfand this move “is the toughest defence and is exceedingly difficult to refute in practice” (p. 113 of the Russian edition). Actually 67...Rd4 has no particular merits: in a couple of moves it leads necessarily to the positions which were analysed by Kopayev.
68.Re1 Gelfand gives an exclamation mark also to this move and says: "A waiting move, preparing for Black to leave the ideal square on d4". But why d4 is an ideal square? As in the case of Kopayev’s old move 67...Rb4, White can immediately chase the rook from the 4th rank with 68.Re4 and bring his king back to the centre. After 68.Re4 Rd1 69.Ra4 Rb1 70.Ke4 we transpose to the position which arose in the game Duras-Capablanca (1913) (rev. col.) after 61.Rb8 (=61...Rb1). See D. 10 in the thread
Capablanca-Yates (Hastings, 1930).
68...Ra4 69.Re4 Ra1 70.Rd4 Ra3 71.Rd7 Ra5 72.Ke4 Ra4+ 73.Rd4 Ra1 We have again the transposition to the same position of Duras-Capablanca game after 61.Rb8 (=61...Rb1) of the previous note (with the irrelevant difference that there the Black’s rook was on b1 instead of a1).
D. 4 Now, when the Black rook is displaced from the 4th rank and the White king returned to the centre, White can carefully prepare the advancement of the f-pawn. As I proved in 2013 (see the thread
Duras-Capablanca (1913) revisited, D. 1.-1) the immediate 3.f5 (played by Capablanca) is also winning for White but the win is very difficult. The right plan is according to Kopayev “to push the f-pawn to the 5th rank only after transferring the rook to the eighth rank, not earlier. In that way the Black’s king which is restricted in his mobility not only by White pieces but also by his own pawns is shut out of the game”.
Gelfand considers here immediate transfer of the rook to the 8th rank - 74.Rd8 (II) which is winning, but 74.Rb4 (I), suggested by Kopayev, makes the win easier. It is important to emphasize that after White has managed to bring back his king to the centre, keeping the rook on the d-file does not have much sense. It is better to place the rook far away from the Black king (on a-, b- or c- files). We will see the difference in the Gelfand's line 74.Rd8 after 81…Kf8 (D. 10) and Kopayev's line (D. 8).
A. 74.Rb4 I am following now the analysis of Kopayev in
Shakhmatny Bulletin (‘Chess Bulletin’), n. 9-1956, p. 266-267, of the game Duras-Capablanca (1913). Unlike Kopayev, here and in the thread
Duras-Capablanca (1913) revisited I have reversed the colours (in the game, Capablanca had the three pawns). In the actual game the stronger side’s rook was on d5 and after 61.Rb8 (=Rb1 in my analysis) Kopayev suggested 61…Ra5 (=74.Rb4 in our analysis). The fact that the rook of the stronger side in Kopayev's analysis was on the a-file, and the rook of the weaker side on the b-file, is of no significance.
74...Rg1 75.Kf3 Ra1 “Keeping the possibility of both rear and flank checks”.
75.Kf3 Ra1 76.Rb8 Rf1+ “Flank checks would only speed up the outcome, since after the transfer of the White king to h4 it is impossible to prevent the advancement of the f-pawn”.
77.Ke3 Re1+ 77...Rg1 78.f5 Rxg4 79.f6+! Kh7 80.e6+–(Kopayev).
78.Kf2 Re4 79.Kf3 Re1
D. 5 80.Rb7 “Black is in zugzwang. It should be noted that the move Rb7 became possible only after the White king had escaped flank checks”.
After 80.Rb7 Kopayev analysed 80…Kf8 (a), 80...Rf1+ (b) and 80…Rb1.
80…Rb1 offers little resistance after
81.e6 Kf6 82.exf7 Kg7 83.Ra6 Rb8 84.Kg3 Rf8 851.Kh4 Rxf7 86.f5 Rb7 87.Kh5+–.
a) 80...Kf8 “This move will allow White to win a decisive tempo in order to transfer the rook to the eighth rank”.
81.Rb4 “Threatening to dislodge the Black rook from the e-file by means of Kf2. The importance of this manoeuvre is evident from the following example variation”.
81...Kg7 81...Rb1 82.Ra8+ Ke7 (82...Kg7 83.f5+–) 83.Ra6 Rh1 84.Ke4 h5 85.g5 h4 86.Rh6 h3 87.Kf3 h2 88.Kg2 +– (Kopayev).
82.Kf2 Ra1 D. 6 83.Rb8 “The correct continuation, which gives White the opportunity to gain a decisive advantage, is to immediately move the rook to the 8th rank” (Kopayev,
Shakhmatny Bulletin (‘Chess Bulletin’), n. 8-1956, p. 229).
Kopaev rightly points out that 83.Rb7 is weaker due to the flank checks. However, Kopayev believed that after 83...Rb2+ White had nothing better to do than to allow his king to be driven to f5 with subsequent implementation of his plan, indicated in the note to the D. 2. In fact, White’s best option after
83.Rb7 Rb2+ is to send his king to h4.
84.Kg3 Rb3+ 85.Kh4 Re3 86.Ra7 Re1 87.Kg3 Re3+ 88.Kf2 Re4 89.Kf3 Re1 90.Rb7+–. We have now transposed to the D. 5.
83…h5 Black has nothing better. 83...Rb2+ 84.Kg3 Rb3+ 85.Kh4 is hopeless for Black.
84.g5 h4 85.Kf3 h3 86.Ra2+–.
“The win was achieved by subtle manoeuvres on the squares a1, a2, a3, a5.
These analyses show that the White rook plays a key role in achieving victory: it limits the defense options of Black and contributes to the advancement of its pawns”.
The key advantage of Kopayev's winning strategy lies in preventing Black from activating the king through Kf8-e7. By this means, White effectively avoids the emergence of the challenging f6 pawn setup. It is more than strange that this analysis by Kopayev from 1956 cannot be found in any endgame textbook, including the 2nd edition of
Shakmatye okonchaniya from 1984.
D. 7 Kopayev considers here only 81…Re1+ (b1), but it is worth looking at the waiting move 81…Ra1 (b2) closely, because it shows the important difference of keeping the rook far away from the Black’s king in comparison to the line of Gelfand (see below D. 10).
b1) 81…Re1+ 82.Kf5 Kopayev analyses only
82.Kd5(?!) Rg1(?!) 83.f5 Rxg4 84.e6, but instead of the weak 82…Rg1 Black can offer stiffer resistance by 82...Kf8. White can secure victory only by employing the f6 setup. I’ll provide for details in Part 2 of this thread.
82...Re3 83.Re7 Ra3 84.e6 Ra5+ 85.Ke4 Kf6 86.Rxf7+ Ke6 87.Rh7+–.
b2) 81...Ra1 82.f5 Kf8 D. 8 83.Rb8+ Ke7 84.f6+ With the rook far away from the Black’s king, White displaces the enemy king from the 8th rank and takes the f7-pawn by Rf8.
84…Kd7 85.Rf8 Re1+ 86.Kd5 Rd1+ 87.Kc4 Rc1+ 88.Kd3 Ke6 89.Re8+ Kd7 90.Re7+ Kd8 91.Rxf7 Ke8 92.Ra7+–.
Pages: 1
Capablanca-Yates (Hastings, 1930) revisited (Read 124 times)
