54. c4 seems more obvious, with lines like 54. c4 Ra2+ 55. Kf3 a4 56. Ra6 a3 57. Kf4 Rf2+ 58. Ke5 Rf3 59. h5 Rxg3 60. Ra7+ Kg8 61. Kxf5 looking good for White.

But maybe Black can try 54. c4 Ra3 cutting off White's king. I haven't got a line for White against that yet.

If, as in this score, 68 Rg6+ was played, then as <boomie> and <nightsend> say 69...Ra4 loses, and White would also have 69 Rg5 ! which may win. Black would probably be better to play 68 Rg6+ Kf7 when there is no clear way for White to make progress.

So Capa probably did mean 54. c4, but I only found lines like 54. c4 Ra3 55. Ra6 Kg7 56. h5 a4 57. Kf2 Rc3 58. Rxa4 Kh6 59. Ra5 Kg5 60. Rc5 Rc2+ 61. Ke3 Rc3+ 62. Kd4 Rxg3, and it's not clear to me that White necessarily should win against Black's counterplay with the f-pawn. This line is long and so possibly full of improvements, so I didn't take it further.

62.Kd4!? Rxg3 63.h6 Rg4+! 64.Kd3! Rg3+ 65.Ke2 Kxh6 66.Rxf5 Kg6=

White can't get out of check without getting into a drawn position. So ,amazingly, Schlechter did not miss a win in this game. Possibly, he even saw the line above and Lasker had to find 54...Re4!! in order to escape.

Here is the entire analysis on that move:

56... Rc4, 58... Kf6, 60... Kf6, same position.

What's more, later there's a similar notes saying "a MODERATE advantage ( ). To the best of my knowledge, this symbol denotes a CLEAR advantage, not 'moderate'.

I would suggest to clean old games, which are annotated by the greats of the past, from these meaningless comments. If computer analysis is to be included, do it wisely and with respect - giving variations unseen by the annotator, explaining things in words, something that will match the high and instructive standards of the commentator from the past.

Just a suggestion... I know that this website is created and maintained by very few and very hard working people.

Much love

I agree. I'm currently mapping this game with an engine, and although they're not great at endgames it's obvious these "moderate advantage" etc comments are both unnecessary and often wrong.

<Game moves: 1.e4 e5 2.Nf3 Nc6 3.Bc5 Nf6 4.0-0 d6 5.d4 Bd7 6.Re1 exd4 7.Nxd4 Be7 8.Nc3 0-0 9.Bxc6 bxc6 10.Bg5 Re8 11.Qf3 h6>

This position and some following moves are known to theory. Engine mapping commenced here. Evaluation after 16 ply of <11…h6> was <=0.15>.

<Game moves: 12.Bh4 Nh7 13.Bxe7 Qxe7 14.Rad1 Nf8 15.h3 Ng6 16.Qg3 Qg5 17.Qxg5 hxg5 18.f3 f6 19.Kf2 Kf7 20.Nde2 a5 21.b3 Reb8 22.Nc1 Be6 23.Nd3 c5 24.Nb2 Ne5 25.Nd5 Rb7 26.Re3 Nc6 27.Rc3>

Capablanca in his notes to this game thought this was a poor move. Until this move, the evaluation of the game position had been sitting close to 0 since the opening. The engine recorded a minor shift to <-0.18> on the forward slide and <-0.14> on the return slide (both 17 ply). Does this minute shift signal Capablanca’s sensitivity to the endgame?

<Game moves: 27…g6 28. a4 f5 29.Ne3 Re8 30.Nec4 Ra7 31.Re1 Bxc4 32.Nxc4 Kf6 33.Ne3 Ne5>

The engine evaluation of this move on the forward slide assessed an evaluation shift from <0.17> to <0.94>, a jump of <0.77>, just below the threshold of <0.80> for a bad move. However, the return slide reduced that shift to <0.61>, a jump of <0.44>, well inside the threshold of a bad move.

<Game moves: 34.exf5 gxf4 35.g3 Rh8 36.f4 gxf4 37.Nd5+ Kf7 38.Nxf4 Rb7 39.Kg2 c4 40.bxc4 Rb4 41.c5 Rxa4 42.cxd6 cxd6 43.Rc7+ Kf6 44.Nd5+ Kg5 45.h4+ Kh6 46.Ne7 Rf8 47.Rd1 Rf7 48.Rxd6+ Kh7 49.Re6>

Engine evaluation had steadily climbed to <1.47+>- at <45…Kh6> reaching a peak of <1.69 > at <48…Kh7>., before subsiding to <1.18=/-> at <49.Re6>. This was confirmed on the return slide when the peak evaluation reached was <1.32 > for <43…Kf6>.

<ANALYSIS>

<Note> The fluctuations generated in the relatively low (16 minimum) ply forward slide were smoothed out in the equivalent return slide. The corrected evaluations extracted from the return slide are used in this analysis, as they are considered more reliable than the raw evaluations generated on the initial forward slide.

<Evaluation range>: Between < 0.33> of <29.Ne3> - representing a microscopic advantage for Black - and < 1.32> in respect of <43…Kf6>. Although the latter technically represents close to a winning advantage ( 1.41>, it probably also represents an inflated endgame evaluation, as the game was clearly drawn. No winning variation has been demonstrated in any notes or analyses, and none appeared in the engine evaluation.

Black’s saving move was <56…Rc4>, according to Capablanca, and initially attracted a <1.88 > evaluation from the engine. This quickly subsided to <1.16 > on White’s response, which the engine considered to be White’s best move. The return slide evaluated both ply at <1.14 > graphically demonstrating the smoothing function it performed throughout the game.

The largest evaluation shift for White was <0.15> between <11…h6> (<=0.15>) and <12.Bh4> (<=0.00>). This represents a virtually flawless game by White.

The largest evaluation shift for Black was <0.39> between <33.Ne3> and <33…Ne5>, well outside the range of an evaluation shift representing a bad move. As the game was drawn, this could also be said to represent a flawless game by Black.

The engine evaluation of the final position was <0.94 >, in a dead drawn position, representing a minor endgame evaluation inflation by the engine. If the whole amount is treated as inflation and subtracted from the most favorable evaluation that occurred for White during the course of the endgame, then the maximum evaluation for White is reduced from < 1.32> at <43…Kf6> to < 0.36>, consistent with a perfectly played draw by both players who drew on close to all the resources available in the positions they generated.

<End note>: There was no sign of the win Capablanca thought could be claimed with <54.c3>. However it was noted that <c3> was a frequent motif in some variations that provided temporary cover from Black’s rook for the White king on the third rank.

<GAME RATING>: This game is rated at 0, representing no bad moves or blunders by either player.

Within the methods adopted for this quantitative analysis, this was a flawless game.