23-08-2013, 11:56 PM
The Roaring Thirties (AVRO 1938)
The mid to late 30s was a particularly fascinating period in chess history, with the relatively quick emergence of at least five strong title contenders in Botvinnik, Keres, Fine, Reshevsky and Flohr to challenge world champions Alekhine, Euwe and Capablanca, who was on his comeback-trail.
Questions of supremacy are usually settled, officially at least, by championships but the war prevented this. As is well known AVRO 1938, one of the strongest tournaments ever held up to that time, was won by Keres and Fine who tied, Keres winning on tie break the right to challenge Alekhine, a challenge that was never made. Capablanca finished a disappointing seventh (table here, <!-- m --><a class="postlink" href="https://en.wikipedia.org/wiki/AVRO_1938_chess_tournament">https://en.wikipedia.org/wiki/AVRO_1938 ... tournament</a><!-- m -->).
This appeared to signal the sun setting on his brilliant career.
However in an interview given after the Buenos Aires Olympiad in 1939 (in which Capablanca scored and played very well), translated by Chess Historian Edward Winter, Capablanca made some interesting remarks:
“In the AVRO tournament I played under physical conditions that were absolutely abnormal. Although I am not up to date with chess literature, I played the openings well in all my games for the simple reason that I have judgment. But after the first three hours of play, I felt my head was splitting. It was impossible for me to think and coordinate ideas. Against Fine I had two won games; against Alekhine I should have won one game; and another one against Keres, thanks to an advantageous position which I built up conscientiously. But at the moment of transforming my advantage into victory, I found that my brain was not functioning and I then continued playing not with my head but with my hands. Despite the bitter cold of Holland in November, I immersed my congested head in icy water to try to clear it, although without any result ... I thus participated in the AVRO tournament playing like an automaton after the third hour, and it is therefore understandable how frequently I failed to win.”
<!-- m --><a class="postlink" href="http://www.chesshistory.com/winter/extra/capablanca11.html">http://www.chesshistory.com/winter/extr ... nca11.html</a><!-- m -->
This is the reason for this post. Capablanca’s high blood pressure at the time is of course well documented, as were the ‘absolutely abnormal’ playing conditions, the main complaint being that each round was played in a different Dutch town/city. Alekhine also stated that the travel and upheaval involved disadvantaged the older players. Capablanca also said that his high blood pressure was responding to treatment; however he succumbed to it in 1942.
Capablanca’s claim that due to his health and the conditions four wins vanished in the later stages of the game is intriguing, and if true could have a significant effect on any evaluation of the pre-war balance of chess strength. In the almost equally strong Nottingham tournament of 1936 (which contained all the AVRO players except Keres), Capablanca had tied for first with Botvinnik, and prior to that he had won a strong event at Moscow ahead of Botvinnik, Flohr and Lasker (who was still near the top in the mid 30s). In the interview Capablanca also said he could have won Nottingham outright but because of his ‘brain lapses’ he failed to remember his adjournment analysis in his last round against Bogolyubov.
So, what a find it was to discover that the AVRO games are available online
(Here <!-- m --><a class="postlink" href="http://www.chessgames.com/perl/chesscollection?cid=1006941">http://www.chessgames.com/perl/chesscol ... id=1006941</a><!-- m -->, or at the upgraded site <!-- m --><a class="postlink" href="http://www.chessgames.com/perl/chess.pl?tid=79229">http://www.chessgames.com/perl/chess.pl?tid=79229</a><!-- m -->), which allows his claims to be looked at.
1) Fine – Capablanca.
[Fine-Capablanca]
[pgn]1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 c5 5. Bd2 cxd4 6. Nb5 Bxd2 7. Qxd2 Nc6 8. Nf3 f6 9. Qf4 Nh6 10. Nd6 Kf8 11. Bb5 Nf7 12. Nxf7 Kxf7 13. Bxc6 bxc6 14. exf6 gxf6 15. Ne5 Kg7 16. Qg3 Kf8 17. Nxc6 Qd7 18. Nxd4 e5 19. Nb3 Qf5 20. Qd3 d4 21. O-O Rg8 22. f4 Bb7 23. Rf2 Be4 24. Qd2 Kf7 25. Re1 Rg4 26. Nc5 Bxg2 27. Rxg2 Rag8 28. Ree2 exf4 29. Nb7 Qd5 30. Rxg4 Rxg4 31. Rg2 Rxg2 32. Qxg2 f3 33. Qh3 Qg5 34. Qg3 Qc1 35. Kf2 Qe3 36. Kf1 Qe2 37. Kg1 Qd1 38. Kf2 Qxc2 39. Kxf3 Qc6 40. Ke2 Qxb7 41. b3 Qe4 42. Kd2 Qe5 43. Qh3 Qg5 44. Kd3[/pgn]
After an eventful middlegame and a hair-raising time scramble, they reached the following position after White’s 42nd move, Capablanca finding himself with an extra pawn:
[pos]8/p4k1p/5p2/8/3pq3/1P4Q1/P2K3P/8 b - - 2 42[/pos]
Past the time control, Blacks prospects appear very good after a simple move like …f5 or …h5 to improve his position. It’s probably not exactly ‘won’ as there is work to be done, but I think it would reasonable to describe it as ‘probably winning’. It’s not the precise evaluation that matters, but the next two moves. Capablanca played
42 … Qe5
This looks like a strange choice to me – the move doesn’t even attack the WQ. White could for example play Kd3 and if Black traded queens the winning chances would be with White. Black could instead continue with the tricky Qc5, but I doubt Capa had any such intention. Had he played the preparatory h5 first inducing h4 the move Qe5 would have packed real punch. The actual Qe5 just looks like a hasty oversight. In fact White replied
43 Qh3, gaining a bit of counterplay.
[pos]8/p4k1p/5p2/4q3/3p4/1P5Q/P2K3P/8 b - - 4 43[/pos]
After any sensible move Black still has good winning chances. However Capablanca then played
43 … Qg5+ ?
and after White’s obvious reply Kd3, a draw was agreed.
43 … Qg5+ I think is a really weak move with no upsides, other than being check. Just the sort of move you might make if you had a blinding headache and needed to go and put your head under the tap?
2) Capablanca-Fine.
[pgn]1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. cxd5 Qxd5 6. Nf3 c5 7. Bd2 Bxc3 8. Bxc3 Nc6 9. Rd1 O-O 10. e3 b6 11. a3 Bb7 12. dxc5 Qxc5 13. b4 Qh5 14. Bxf6 gxf6 15. Rd7 Rac8 16. Qb2 Rfd8 17. Rxb7 Ne5 18. Be2 Nxf3 19. Bxf3 Qe5 20. Qxe5 Rc1 21. Bd1 Rcxd1 22. Ke2 R1d2 23. Kf3 fxe5 24. Rxa7 e4 25. Kg3 Ra2 26. Ra6 Rdd2 27. Rf1 Rdb2 28. Rxb6 Rxa3 29. b5 Kg7 30. h4 Rab3 31. Kf4 Rxb5 32. Rxb5 Rxb5 33. g4 Rb4 34. Rc1 Rb2 35. Kg3 Kf6 36. Rc4 Ke5 37. Rc8 Kf6 38. Rg8 h6 39. g5 hxg5 40. Rxg5 Rb8 41. Kh3 e5 42. Rg1[/pgn]
The second game between Capablanca and Reuben Fine was, like their first dogfight, a terrific tussle, with the action not over even when they reached a drawish 4 vs 4 rook ending.
[pos]8/5pkp/4p3/8/1r2pKPP/4P3/5P2/5R2 w - - 1 34[/pos]
Capablanca played 34 Rc1. This is a slight inaccuracy; if White had played Rd1, the rook could have later gone to d4.
34 … Rb2 35 Kg3
[pos]8/5pkp/4p3/8/4p1PP/4P1K1/1r3P2/2R5 b - - 4 35[/pos]
35 … Kf6
A mistake, Black should have gone Rb4. With the WR on d1 W would then have the option of playing Rd4 (not forgetting to play g5+ first, or Black wins a rook with e5+!). Now Capablanca spots a chance.
38 Rc4 Ke5 37 Rc8! a good practical chance.
37 … Kf6 38 Rg8!
[pos]6R1/5p1p/4pk2/8/4p1PP/4P1K1/1r3P2/8 b - - 10 38[/pos]
White is after Black’s h-pawn! Or…
38 … h6?? At first sight this solves B’s problems, but B should have let the pawn go.
39 g5+ hxg5 (Ke7 doesn’t help)
[pos]6R1/5p2/4pk2/6p1/4p2P/4P1K1/1r3P2/8 w - - 0 40[/pos]
40 Rxg5?? White hopes to get his h-pawn moving asap, but it’s hard to believe that Capablanca would not see the immediate h5! after which the h-pawn can’t be stopped. Both players were short of time (if you look at the opening and middle game you'll see why) but even so… Strangely enough, this moment seems to have gone unnoticed, at least in the English-speaking press, for 13 years until surfacing in Gerald Abraham’s beginners book Teach Yourself Chess.
40 … Rb8 41 Kh3 (the moment has passed) e5 42 Rg1 and a draw was agreed.
3) Capablanca-Keres
[pgn]1. d4 e6 2. c4 Bb4 3. Nc3 c5 4. e3 Nf6 5. Ne2 cxd4 6. exd4 O-O 7. a3 Be7 8. Nf4 d5 9. cxd5 Nxd5 10. Nfxd5 exd5 11. Qb3 Nc6 12. Be3 Bf6 13. Rd1 Bg4 14. Be2 Bxe2 15. Kxe2 Re8 16. Kf1 Ne7 17. g3 Qd7 18. Kg2 Rad8 19. Qb5 Nf5 20. Qxd7 Rxd7 21. Rd3 h6 22. h4 Rc8 23. h5 b5 24. g4 Nxe3 25. fxe3 a5 26. b4 axb4 27. axb4 Be7 28. Rb1 Rc4 29. Nxb5 Rxb4 30. Rxb4 Bxb4 31. Kf3 g6 32. Rb3 Ba5 33. Ra3 Bd2 34. Ke2 Rb7 35. Nd6 Rb2 36. Ra8 Kh7 37. Nxf7 gxh5 38. Ne5 Bc1 39. Kd3 Rd2 40. Kc3 Rg2 41. gxh5 Bxe3 42. Ra7[/pgn]
They reached this position
[pos]R7/5N1k/7p/3p3p/3P2P1/4P3/1r1bK3/8 w - - 0 38[/pos]
Black has just played 37 ... gxh5. After the obvious recapture W has good winning chances; he has an extra pawn and the BK is exposed to checks and if the weak d-pawn should fall W would probably win. In addition Black’s pieces are a bit offside and the N is active. Black’s main chances stem from the reduced number of pawns and counterplay against the e-pawn.
38 Ne5? This does threaten to win a piece with Nd3, but B can get the B out of harm’s way with a discovered check.
38 … Bc1+ 39 Kd3 Rd2+ 40 Kc3 Rg2
h4! was more accurate, with the idea that if Nf3, h3! but W has a perpetual as the K can’t go to f8 due to Ng6+ winning the passed h-pawn - the one that White gave Black for free - but it doesn’t matter much:
41 gxh5 Bxe3 and after Ra7+ the players agreed to a draw.
4) Capablanca-Alekhine reached the fascinating position underneath, just after the time control:
[pgn]1. d4 Nf6 2. c4 e6 3. Nf3 b6 4. g3 Bb7 5. Bg2 Be7 6. O-O O-O 7. Qc2 Be4 8. Qc3 Nc6 9. Be3 d5 10. Nbd2 dxc4 11. Qxc4 Bd5 12. Qc3 h6 13. Rfd1 Rc8 14. Bf4 Qe8 15. a3 Na5 16. b4 Nb7 17. Ne1 Bxg2 18. Nxg2 c5 19. bxc5 bxc5 20. Qb2 Na5 21. dxc5 Rxc5 22. Rab1 Qc6 23. Ne1 Nd5 24. Be5 Nc3 25. Bxc3 Rxc3 26. Rdc1 Rc8 27. Rxc3 Qxc3 28. a4 Qxb2 29. Rxb2 Rc1 30. Rb1 Rc3 31. Rb8 Kh7 32. Rb5 Nc4 33. Nxc4 Rxc4 34. Rb7 Bf6 35. Nd3 Rxa4 36. Rxf7 a5 37. Nc5 Ra1 38. Kg2 a4 39. Ra7 a3 40. Nxe6 Bb2 41. Nf4 Bd4 42. Ra4 Bb2 43. e4 g5 44. Ra7 Kg8 45. Nd5 Bd4 46. Ra8 Kf7 47. Nb4 Rb1 48. Nc2 Bxf2[/pgn]
[pos]8/R5pk/4N2p/8/8/p5P1/1b2PPKP/r7 w - - 1 41[/pos]
Again, Capablanca is a pawn ahead. Even with the help of the bishop, Black is unable to push his a-pawn without tying up his rook. There are interesting ‘Kibitzer’ comments below the game on the website. White seems to have good prospects by pushing the pawns on the K-side to combine the possibility of queening a pawn with threats against the K. Without going into specific variations W can start with h4, f4 or e4 with different possibilities, in combination with bringing the K over to the Q-side. h4, bring the N to g6 and push the e pawn, possibly the f pawn as well. By my reckoning (with or without using a computer) White has good winning chances, probably more than 50-50 because of the tricks. Objectively my guess would be that Black probably can hang on to a draw if he plays accurately. However rather than push any of his pawns, Capablanca played
41 Nf4?
This move seems to significantly reduce White’s winning chances. Alekhine, as usual, is quick to seize the chance to get more active.
41…Bd4 42 Ra4 Bb2 43 e4 g5!
[pos]8/7k/7p/6p1/R3PN2/p5P1/1b3PKP/r7 w - g6 0 44[/pos]
44 Ra7+ This is a slight mistake, allowing the B to come to d4. After Nd5 W still has a little something.
…Kg8 45 Nd5 Bd4 46 Ra8+ Kf7 47 Nb4 Rb1 48 Nc2
[pos]R7/5k2/7p/6p1/3bP3/p5P1/2N2PKP/1r6 b - - 9 48[/pos]
Now Alekhine found Bxf2! and a draw was agreed. W can’t take the B or B will play Rb2 and use the a-pawn to regain the piece with a definitely, dead drawn rook ending.
According to the Kibitzer chat, even in the final position White can try to play for a win with 49. Nxa3 Rb2 50. Nc4 Rc2 51. Nd6+ Ke6 52. Nf5 when the initial suggestion Bc5+ seemed to lose. But after 52…g4! Black would be OK. White could safely play 51 Ra4 also. I find it a little surprising that Capablanca didn’t play a few more moves just to have a look. He won a similarly simplified ending against Alekhine in the world championship match with a weak extra pawn.
And lastly, here’s a ‘bonus blunder’. This is not one of the four games where Capablanca claimed to have had a win.
5) Capablanca-Reshevsky.
[pgn]1. d4 Nf6 2. c4 g6 3. Nc3 d5 4. Bf4 Bg7 5. e3 O-O 6. Rc1 c5 7. dxc5 Qa5 8. cxd5 Rd8 9. Qa4 Qxa4 10. Nxa4 Nxd5 11. Bb5 Nxf4 12. exf4 Be6 13. b3 Bd5 14. Nf3 Bxf3 15. gxf3 Nc6 16. Bxc6 bxc6 17. Ke2 Bh6 18. Rc4 e5 19. fxe5 Rd2 20. Kf1 Rxa2 21. Kg2 Be3 22. Rf1 Rb8 23. Kg3 Bd2 24. Nc3 Bxc3 25. Rxc3 Rb2 26. Rd1 R2xb3 27. Rxb3 Rxb3 28. Rd6 Rc3 29. Rxc6 a5 30. Kf4 Kf8 31. Rc7 a4 32. c6 a3 33. Kg3 a2 34. Ra7 Rxc6 35. Rxa2 Kg7 36. Ra7 Re6 37. f4 Rb6 38. h3 Rc6 39. f5 gxf5 40. Kf4 Kg6 41. Rd7 Rc4 42. Kg3 h5[/pgn]
[pos]8/R4pkp/4r1p1/4P3/5P2/6K1/5P1P/8 b - - 0 37[/pos]
Yet again Capablanca has managed to obtain an extra pawn, though the position is probably drawn (reminiscent of another rook ending discussed here recently, perhaps), but is it dead drawn? White needs to get f5 in. Interestingly, Max Euwe said he couldn’t say for sure it was drawn because of the f5 possibility. Anyway drawn or not, the main interest is in Capablanca’s next move.
39 f5??
Black captured gxf5, and after Kf4 simply defended the f5 pawn with Kg6. Draw agreed the next move.
To conclude, there’s enough evidence here (for me at least) to believe that Capablanca’s play towards the end of the tournament was probably affected to the level he claimed. They don’t show what should have happened, but suggest what could have happened. From the positions he reached he should have recorded two wins against Fine, and he also quickly blew very good chances to win against Alekhine and Keres, and blundered badly against Reshevsky, albeit in a drawish position. The tournament was very tight, the winning score being only +3 - a few more wins would have put Capablanca into the mix in the tournament, and maintained his run of top results in the strongest events.
This is only my view of the story of the late blunders, and of course can’t be the full story. Two wins against Fine would have been hard on Fine, who should have won the White game in the middlegame. Capablanca lost games too at AVRO, of course. When he said he played the openings well, he seems to have just been thinking about his White games - he looked generally unprepared to play against top flight players with Black, getting into inferior opening positions against Keres, Fine, Euwe and Alekhine. I think any speculative claim to overall supremacy (as opposed to just still being on a par with these players as late as 1938) would need to start by assuming that he had worked on his Black openings.
There were other might-have-beens. Fine was forced to play an adjourned game with Reshevsky at zero notice, then lost the game on time with both flags down, having battled to a drawn position. Alekhine omitted to simply force a pawn forward to queen in his game against Reshevsky, and also seems to have missed a win against Keres right at the end of the session (hat-tip to the Quality Chess book referred to in the link to the game). Check out Keres’ amazing escape against Euwe in Round 1. And the very famous Botvinnik-Capablanca game, if you happen not to have seen it.
If you’ve any interest in this old stuff at all, it’s really worth while checking out the link and reading the comments. Even if you only read the amusing aftermath of the Fine-Reshevsky game...
The mid to late 30s was a particularly fascinating period in chess history, with the relatively quick emergence of at least five strong title contenders in Botvinnik, Keres, Fine, Reshevsky and Flohr to challenge world champions Alekhine, Euwe and Capablanca, who was on his comeback-trail.
Questions of supremacy are usually settled, officially at least, by championships but the war prevented this. As is well known AVRO 1938, one of the strongest tournaments ever held up to that time, was won by Keres and Fine who tied, Keres winning on tie break the right to challenge Alekhine, a challenge that was never made. Capablanca finished a disappointing seventh (table here, <!-- m --><a class="postlink" href="https://en.wikipedia.org/wiki/AVRO_1938_chess_tournament">https://en.wikipedia.org/wiki/AVRO_1938 ... tournament</a><!-- m -->).
This appeared to signal the sun setting on his brilliant career.
However in an interview given after the Buenos Aires Olympiad in 1939 (in which Capablanca scored and played very well), translated by Chess Historian Edward Winter, Capablanca made some interesting remarks:
“In the AVRO tournament I played under physical conditions that were absolutely abnormal. Although I am not up to date with chess literature, I played the openings well in all my games for the simple reason that I have judgment. But after the first three hours of play, I felt my head was splitting. It was impossible for me to think and coordinate ideas. Against Fine I had two won games; against Alekhine I should have won one game; and another one against Keres, thanks to an advantageous position which I built up conscientiously. But at the moment of transforming my advantage into victory, I found that my brain was not functioning and I then continued playing not with my head but with my hands. Despite the bitter cold of Holland in November, I immersed my congested head in icy water to try to clear it, although without any result ... I thus participated in the AVRO tournament playing like an automaton after the third hour, and it is therefore understandable how frequently I failed to win.”
<!-- m --><a class="postlink" href="http://www.chesshistory.com/winter/extra/capablanca11.html">http://www.chesshistory.com/winter/extr ... nca11.html</a><!-- m -->
This is the reason for this post. Capablanca’s high blood pressure at the time is of course well documented, as were the ‘absolutely abnormal’ playing conditions, the main complaint being that each round was played in a different Dutch town/city. Alekhine also stated that the travel and upheaval involved disadvantaged the older players. Capablanca also said that his high blood pressure was responding to treatment; however he succumbed to it in 1942.
Capablanca’s claim that due to his health and the conditions four wins vanished in the later stages of the game is intriguing, and if true could have a significant effect on any evaluation of the pre-war balance of chess strength. In the almost equally strong Nottingham tournament of 1936 (which contained all the AVRO players except Keres), Capablanca had tied for first with Botvinnik, and prior to that he had won a strong event at Moscow ahead of Botvinnik, Flohr and Lasker (who was still near the top in the mid 30s). In the interview Capablanca also said he could have won Nottingham outright but because of his ‘brain lapses’ he failed to remember his adjournment analysis in his last round against Bogolyubov.
So, what a find it was to discover that the AVRO games are available online
(Here <!-- m --><a class="postlink" href="http://www.chessgames.com/perl/chesscollection?cid=1006941">http://www.chessgames.com/perl/chesscol ... id=1006941</a><!-- m -->, or at the upgraded site <!-- m --><a class="postlink" href="http://www.chessgames.com/perl/chess.pl?tid=79229">http://www.chessgames.com/perl/chess.pl?tid=79229</a><!-- m -->), which allows his claims to be looked at.
1) Fine – Capablanca.
[Fine-Capablanca]
[pgn]1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 c5 5. Bd2 cxd4 6. Nb5 Bxd2 7. Qxd2 Nc6 8. Nf3 f6 9. Qf4 Nh6 10. Nd6 Kf8 11. Bb5 Nf7 12. Nxf7 Kxf7 13. Bxc6 bxc6 14. exf6 gxf6 15. Ne5 Kg7 16. Qg3 Kf8 17. Nxc6 Qd7 18. Nxd4 e5 19. Nb3 Qf5 20. Qd3 d4 21. O-O Rg8 22. f4 Bb7 23. Rf2 Be4 24. Qd2 Kf7 25. Re1 Rg4 26. Nc5 Bxg2 27. Rxg2 Rag8 28. Ree2 exf4 29. Nb7 Qd5 30. Rxg4 Rxg4 31. Rg2 Rxg2 32. Qxg2 f3 33. Qh3 Qg5 34. Qg3 Qc1 35. Kf2 Qe3 36. Kf1 Qe2 37. Kg1 Qd1 38. Kf2 Qxc2 39. Kxf3 Qc6 40. Ke2 Qxb7 41. b3 Qe4 42. Kd2 Qe5 43. Qh3 Qg5 44. Kd3[/pgn]
After an eventful middlegame and a hair-raising time scramble, they reached the following position after White’s 42nd move, Capablanca finding himself with an extra pawn:
[pos]8/p4k1p/5p2/8/3pq3/1P4Q1/P2K3P/8 b - - 2 42[/pos]
Past the time control, Blacks prospects appear very good after a simple move like …f5 or …h5 to improve his position. It’s probably not exactly ‘won’ as there is work to be done, but I think it would reasonable to describe it as ‘probably winning’. It’s not the precise evaluation that matters, but the next two moves. Capablanca played
42 … Qe5
This looks like a strange choice to me – the move doesn’t even attack the WQ. White could for example play Kd3 and if Black traded queens the winning chances would be with White. Black could instead continue with the tricky Qc5, but I doubt Capa had any such intention. Had he played the preparatory h5 first inducing h4 the move Qe5 would have packed real punch. The actual Qe5 just looks like a hasty oversight. In fact White replied
43 Qh3, gaining a bit of counterplay.
[pos]8/p4k1p/5p2/4q3/3p4/1P5Q/P2K3P/8 b - - 4 43[/pos]
After any sensible move Black still has good winning chances. However Capablanca then played
43 … Qg5+ ?
and after White’s obvious reply Kd3, a draw was agreed.
43 … Qg5+ I think is a really weak move with no upsides, other than being check. Just the sort of move you might make if you had a blinding headache and needed to go and put your head under the tap?
2) Capablanca-Fine.
[pgn]1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 d5 5. cxd5 Qxd5 6. Nf3 c5 7. Bd2 Bxc3 8. Bxc3 Nc6 9. Rd1 O-O 10. e3 b6 11. a3 Bb7 12. dxc5 Qxc5 13. b4 Qh5 14. Bxf6 gxf6 15. Rd7 Rac8 16. Qb2 Rfd8 17. Rxb7 Ne5 18. Be2 Nxf3 19. Bxf3 Qe5 20. Qxe5 Rc1 21. Bd1 Rcxd1 22. Ke2 R1d2 23. Kf3 fxe5 24. Rxa7 e4 25. Kg3 Ra2 26. Ra6 Rdd2 27. Rf1 Rdb2 28. Rxb6 Rxa3 29. b5 Kg7 30. h4 Rab3 31. Kf4 Rxb5 32. Rxb5 Rxb5 33. g4 Rb4 34. Rc1 Rb2 35. Kg3 Kf6 36. Rc4 Ke5 37. Rc8 Kf6 38. Rg8 h6 39. g5 hxg5 40. Rxg5 Rb8 41. Kh3 e5 42. Rg1[/pgn]
The second game between Capablanca and Reuben Fine was, like their first dogfight, a terrific tussle, with the action not over even when they reached a drawish 4 vs 4 rook ending.
[pos]8/5pkp/4p3/8/1r2pKPP/4P3/5P2/5R2 w - - 1 34[/pos]
Capablanca played 34 Rc1. This is a slight inaccuracy; if White had played Rd1, the rook could have later gone to d4.
34 … Rb2 35 Kg3
[pos]8/5pkp/4p3/8/4p1PP/4P1K1/1r3P2/2R5 b - - 4 35[/pos]
35 … Kf6
A mistake, Black should have gone Rb4. With the WR on d1 W would then have the option of playing Rd4 (not forgetting to play g5+ first, or Black wins a rook with e5+!). Now Capablanca spots a chance.
38 Rc4 Ke5 37 Rc8! a good practical chance.
37 … Kf6 38 Rg8!
[pos]6R1/5p1p/4pk2/8/4p1PP/4P1K1/1r3P2/8 b - - 10 38[/pos]
White is after Black’s h-pawn! Or…
38 … h6?? At first sight this solves B’s problems, but B should have let the pawn go.
39 g5+ hxg5 (Ke7 doesn’t help)
[pos]6R1/5p2/4pk2/6p1/4p2P/4P1K1/1r3P2/8 w - - 0 40[/pos]
40 Rxg5?? White hopes to get his h-pawn moving asap, but it’s hard to believe that Capablanca would not see the immediate h5! after which the h-pawn can’t be stopped. Both players were short of time (if you look at the opening and middle game you'll see why) but even so… Strangely enough, this moment seems to have gone unnoticed, at least in the English-speaking press, for 13 years until surfacing in Gerald Abraham’s beginners book Teach Yourself Chess.
40 … Rb8 41 Kh3 (the moment has passed) e5 42 Rg1 and a draw was agreed.
3) Capablanca-Keres
[pgn]1. d4 e6 2. c4 Bb4 3. Nc3 c5 4. e3 Nf6 5. Ne2 cxd4 6. exd4 O-O 7. a3 Be7 8. Nf4 d5 9. cxd5 Nxd5 10. Nfxd5 exd5 11. Qb3 Nc6 12. Be3 Bf6 13. Rd1 Bg4 14. Be2 Bxe2 15. Kxe2 Re8 16. Kf1 Ne7 17. g3 Qd7 18. Kg2 Rad8 19. Qb5 Nf5 20. Qxd7 Rxd7 21. Rd3 h6 22. h4 Rc8 23. h5 b5 24. g4 Nxe3 25. fxe3 a5 26. b4 axb4 27. axb4 Be7 28. Rb1 Rc4 29. Nxb5 Rxb4 30. Rxb4 Bxb4 31. Kf3 g6 32. Rb3 Ba5 33. Ra3 Bd2 34. Ke2 Rb7 35. Nd6 Rb2 36. Ra8 Kh7 37. Nxf7 gxh5 38. Ne5 Bc1 39. Kd3 Rd2 40. Kc3 Rg2 41. gxh5 Bxe3 42. Ra7[/pgn]
They reached this position
[pos]R7/5N1k/7p/3p3p/3P2P1/4P3/1r1bK3/8 w - - 0 38[/pos]
Black has just played 37 ... gxh5. After the obvious recapture W has good winning chances; he has an extra pawn and the BK is exposed to checks and if the weak d-pawn should fall W would probably win. In addition Black’s pieces are a bit offside and the N is active. Black’s main chances stem from the reduced number of pawns and counterplay against the e-pawn.
38 Ne5? This does threaten to win a piece with Nd3, but B can get the B out of harm’s way with a discovered check.
38 … Bc1+ 39 Kd3 Rd2+ 40 Kc3 Rg2
h4! was more accurate, with the idea that if Nf3, h3! but W has a perpetual as the K can’t go to f8 due to Ng6+ winning the passed h-pawn - the one that White gave Black for free - but it doesn’t matter much:
41 gxh5 Bxe3 and after Ra7+ the players agreed to a draw.
4) Capablanca-Alekhine reached the fascinating position underneath, just after the time control:
[pgn]1. d4 Nf6 2. c4 e6 3. Nf3 b6 4. g3 Bb7 5. Bg2 Be7 6. O-O O-O 7. Qc2 Be4 8. Qc3 Nc6 9. Be3 d5 10. Nbd2 dxc4 11. Qxc4 Bd5 12. Qc3 h6 13. Rfd1 Rc8 14. Bf4 Qe8 15. a3 Na5 16. b4 Nb7 17. Ne1 Bxg2 18. Nxg2 c5 19. bxc5 bxc5 20. Qb2 Na5 21. dxc5 Rxc5 22. Rab1 Qc6 23. Ne1 Nd5 24. Be5 Nc3 25. Bxc3 Rxc3 26. Rdc1 Rc8 27. Rxc3 Qxc3 28. a4 Qxb2 29. Rxb2 Rc1 30. Rb1 Rc3 31. Rb8 Kh7 32. Rb5 Nc4 33. Nxc4 Rxc4 34. Rb7 Bf6 35. Nd3 Rxa4 36. Rxf7 a5 37. Nc5 Ra1 38. Kg2 a4 39. Ra7 a3 40. Nxe6 Bb2 41. Nf4 Bd4 42. Ra4 Bb2 43. e4 g5 44. Ra7 Kg8 45. Nd5 Bd4 46. Ra8 Kf7 47. Nb4 Rb1 48. Nc2 Bxf2[/pgn]
[pos]8/R5pk/4N2p/8/8/p5P1/1b2PPKP/r7 w - - 1 41[/pos]
Again, Capablanca is a pawn ahead. Even with the help of the bishop, Black is unable to push his a-pawn without tying up his rook. There are interesting ‘Kibitzer’ comments below the game on the website. White seems to have good prospects by pushing the pawns on the K-side to combine the possibility of queening a pawn with threats against the K. Without going into specific variations W can start with h4, f4 or e4 with different possibilities, in combination with bringing the K over to the Q-side. h4, bring the N to g6 and push the e pawn, possibly the f pawn as well. By my reckoning (with or without using a computer) White has good winning chances, probably more than 50-50 because of the tricks. Objectively my guess would be that Black probably can hang on to a draw if he plays accurately. However rather than push any of his pawns, Capablanca played
41 Nf4?
This move seems to significantly reduce White’s winning chances. Alekhine, as usual, is quick to seize the chance to get more active.
41…Bd4 42 Ra4 Bb2 43 e4 g5!
[pos]8/7k/7p/6p1/R3PN2/p5P1/1b3PKP/r7 w - g6 0 44[/pos]
44 Ra7+ This is a slight mistake, allowing the B to come to d4. After Nd5 W still has a little something.
…Kg8 45 Nd5 Bd4 46 Ra8+ Kf7 47 Nb4 Rb1 48 Nc2
[pos]R7/5k2/7p/6p1/3bP3/p5P1/2N2PKP/1r6 b - - 9 48[/pos]
Now Alekhine found Bxf2! and a draw was agreed. W can’t take the B or B will play Rb2 and use the a-pawn to regain the piece with a definitely, dead drawn rook ending.
According to the Kibitzer chat, even in the final position White can try to play for a win with 49. Nxa3 Rb2 50. Nc4 Rc2 51. Nd6+ Ke6 52. Nf5 when the initial suggestion Bc5+ seemed to lose. But after 52…g4! Black would be OK. White could safely play 51 Ra4 also. I find it a little surprising that Capablanca didn’t play a few more moves just to have a look. He won a similarly simplified ending against Alekhine in the world championship match with a weak extra pawn.
And lastly, here’s a ‘bonus blunder’. This is not one of the four games where Capablanca claimed to have had a win.
5) Capablanca-Reshevsky.
[pgn]1. d4 Nf6 2. c4 g6 3. Nc3 d5 4. Bf4 Bg7 5. e3 O-O 6. Rc1 c5 7. dxc5 Qa5 8. cxd5 Rd8 9. Qa4 Qxa4 10. Nxa4 Nxd5 11. Bb5 Nxf4 12. exf4 Be6 13. b3 Bd5 14. Nf3 Bxf3 15. gxf3 Nc6 16. Bxc6 bxc6 17. Ke2 Bh6 18. Rc4 e5 19. fxe5 Rd2 20. Kf1 Rxa2 21. Kg2 Be3 22. Rf1 Rb8 23. Kg3 Bd2 24. Nc3 Bxc3 25. Rxc3 Rb2 26. Rd1 R2xb3 27. Rxb3 Rxb3 28. Rd6 Rc3 29. Rxc6 a5 30. Kf4 Kf8 31. Rc7 a4 32. c6 a3 33. Kg3 a2 34. Ra7 Rxc6 35. Rxa2 Kg7 36. Ra7 Re6 37. f4 Rb6 38. h3 Rc6 39. f5 gxf5 40. Kf4 Kg6 41. Rd7 Rc4 42. Kg3 h5[/pgn]
[pos]8/R4pkp/4r1p1/4P3/5P2/6K1/5P1P/8 b - - 0 37[/pos]
Yet again Capablanca has managed to obtain an extra pawn, though the position is probably drawn (reminiscent of another rook ending discussed here recently, perhaps), but is it dead drawn? White needs to get f5 in. Interestingly, Max Euwe said he couldn’t say for sure it was drawn because of the f5 possibility. Anyway drawn or not, the main interest is in Capablanca’s next move.
39 f5??
Black captured gxf5, and after Kf4 simply defended the f5 pawn with Kg6. Draw agreed the next move.
To conclude, there’s enough evidence here (for me at least) to believe that Capablanca’s play towards the end of the tournament was probably affected to the level he claimed. They don’t show what should have happened, but suggest what could have happened. From the positions he reached he should have recorded two wins against Fine, and he also quickly blew very good chances to win against Alekhine and Keres, and blundered badly against Reshevsky, albeit in a drawish position. The tournament was very tight, the winning score being only +3 - a few more wins would have put Capablanca into the mix in the tournament, and maintained his run of top results in the strongest events.
This is only my view of the story of the late blunders, and of course can’t be the full story. Two wins against Fine would have been hard on Fine, who should have won the White game in the middlegame. Capablanca lost games too at AVRO, of course. When he said he played the openings well, he seems to have just been thinking about his White games - he looked generally unprepared to play against top flight players with Black, getting into inferior opening positions against Keres, Fine, Euwe and Alekhine. I think any speculative claim to overall supremacy (as opposed to just still being on a par with these players as late as 1938) would need to start by assuming that he had worked on his Black openings.
There were other might-have-beens. Fine was forced to play an adjourned game with Reshevsky at zero notice, then lost the game on time with both flags down, having battled to a drawn position. Alekhine omitted to simply force a pawn forward to queen in his game against Reshevsky, and also seems to have missed a win against Keres right at the end of the session (hat-tip to the Quality Chess book referred to in the link to the game). Check out Keres’ amazing escape against Euwe in Round 1. And the very famous Botvinnik-Capablanca game, if you happen not to have seen it.
If you’ve any interest in this old stuff at all, it’s really worth while checking out the link and reading the comments. Even if you only read the amusing aftermath of the Fine-Reshevsky game...