Note that up until the sacrifice he is playing with a bishop + knight against a bishop pair. He seemed able to make that matchup work.

Bogoljubov vs Lasker, 1924

Yates vs Lasker, 1924

Thanks for this, very pretty! Couldn't Lasker have also played 41....Ng2 in the game continuation instead of 41....c3+?

Great as the final combination (and its alternatives are), the commentary on the balance of the game over the course of the past 100 years has often been superficial if not downright wrong. The theme of all this bad commentary is that Lasker played excellent chess throughout while Lee played like a duffer. This is unfair and inaccurate. Lee played quite well for most of the game, and it is no insult to Lasker to point out that--like most humans--his play was not always flawless.

In this post, I will focus on the opening.

Lee played the old Steinitz 5. d3 line in the Ruy Lopez. Reinfeld-Fine in their book on Lasker's best games call Lee's play in moves 5 to 7 "so timorous that Lasker [was] soon enabled to seize the initiative without difficulty."

What nonsense! Quite apart from Steinitz' success with the 5. d3 line, in 1965 Bobby Fischer played 5. d3 against Smyslov in the 1965 Havana tournament (as well in other tournaments), and proved that it can be a formidable weapon.

Perhaps the most aggressive response to 5. d3 is 5...Bc5 as played by Paul Morphy against Andersen in Games 2 and 4 of their 1858 match. Another possibility suggested by Fischer was 5...b5. Lasker's actual response here, 5...d6, was arguably not as good as 5...Bc5 or 5...b5, but was a normal move and the very one played against Fischer by Smyslov in 1965.

Lee's sixth move, 6. c3, was arguably not as good as 6. c4 recommended by MOC-13 (and my computer likes 6. h3--which is at least as "timorous" as Lee's 6. c3). With regard to 6. c3 itself, suffice it to note that Fischer played this very move in the Smyslov game. Indeed, 6. c3 is a very normal move by White in the Ruy Lopez. Thus, through 6. c3, the game we are witnessing is in effect Fischer vs. Lasker. Would anyone accuse Fischer of being "timorous" here?

After Lasker's 6...b5 (which seems better than Smyslov's 6...Be7 or MCO-13's 6...g6) Lee retreated his Bishop with 7. Bc2. Both Hoffer in the Tournament Book and Reinfeld-Fine condemn this move, claiming 7. Bb3 was better. According to Hoffer, Lee should have played 7. Bb3 and then after 7...Be6 retreated with 8. Bc2. In other words, Lee should have lost time.

What nonsense!

After Lasker's 7...g6, Lee played 8. a4. Soltis in his book on Lasker condemns this move by Lee as not fitting the Steinitz 5. d3 formula. While 8. d4 might arguably have been better, there is nothing wrong that I can see with Lee's move.

As is obvious, the commentators are programmed to demean Lee's play and praise Lasker's every move.

Lasker's 8...Bb7 earns an exclamation point from Marco in his commentary on the game (8...Bd7 looks better to me).

Reinfeld-Fine state that Lasker was "already superior" after 10...d5, and heap praise on Lasker's 14...Na5. Soltis gives this move an exclamation point.

Given the above commentary in which Lee's play is repeatedly denounced while Lasker was praised to the heavens, one might think that Lasker had the game well in hand after 14...Na5 or at least had a significant advantage. In fact, here was the position after 14...Na5:

click for larger view

Is this position so much better for Lasker? Hardly. Fritz says that Lee had the better game as of this point. In fairness to Soltis, he does state that the claim that Lasker was better by move 10 was "too generous." But even he loves Lasker's 14...Na5. In fact, 14...b4 was much better.

The reality is more complicated than the lazy decision to praise Lasker's every breath and trash Lee's every effort. Rather, this was a toughly conducted well-played (even if not perfect) game by both players at least until Lee's mistake on move 30.

Lasker's play here and throughout the London 1899 tournament was at such a high level that there is no need to exaggerate his virtues or to demean Lee's excellent effort.

Such was the theme for many an annotated game until the rise of computers, which mercilessly bare even the greatest players' errors of omission and commission during their analyses.

<....Lee played the old Steinitz 5. d3 line in the Ruy Lopez. Reinfeld-Fine in their book on Lasker's best games call Lee's play in moves 5 to 7 "so timorous that Lasker [was] soon enabled to seize the initiative without difficulty."....>

Reinfeld-Fine and Soltis praise Lasker's 16...Qc7 as preparing for the advance of the c-pawn and anticipating 17. c4. In fact, 16...b4 was better.

Lee's 17. Qe1 looks awkward, and was probably not as good as 17. axb5. But Lee's move had its points (e.g., he threatened 18. c4) and was not all that bad. Whatever one thinks of 17. Qe1, it was far better than 17. Qf1 as recommended by the Tournament Book and by Reinfeld-Fine. As Soltis correctly notes, 17. Qf1 would have been even more "artificial" than Lee's actual 17. Qe1.

After 17. Qe1, the Tournament Book says that Lasker's 17...c4 wins a pawn. While I agree that Lasker's 17...c4 was excellent (Marco, Reinfeld-Fine, and Soltis all give it an !), it at most would have yielded a very small advantage to Lasker had Lee played 18. axb5, a move only Soltis seems to have noticed.

Lee's real trouble, as Soltis has noted, began with his 18. d4? Lasker now had a major advantage after 18...Nxe4 (18...Nd7 may have been even better), but after 19. NxN Lasker--incredible to relate--erred with the wooden 19...dxN. In fact, as only Soltis has pointed out, Lasker should have played the problem-like 19...exd4!! Amazing that none of the other commentators noticed this powerful move or the fact that Lasker had forfeited most of his advantage with his 19...dxN.

After Lasker's 21...Qxe5, Lasker was a pawn up, and the commentators proclaim the game over. The Tournament Book states that Lasker now had the "game in hand." Reinfeld-Fine state that "The rest requires only technical skill." Soltis' evaluation is similar.

In fact, though down a pawn, Lee had the two Bishops and all sorts of tactical chances. Here was the position after 21...Qxe5:

click for larger view

To note just one possibility, Lee could here have played 22. axb4 followed by 23. Bh6, or perhaps 22. Bh6 immediately. Does anyone really think that Janowski or Pillsbury (or Lasker) would have felt despondent with this position as White.

I agree with Soltis that after 22. Be3 Nc6 Lee's 23. b3 was ill-advised (23. Rd7 would have given him counterplay if not equality). But Lasker's 24...Nc6 was a mistake (24...Nb3!) and Lee's powerful 25. Rd7, though ridiculed by Reinfeld-Fine, gave him almost equal chances.

Lee's 26. Rdd1 was inferior to 26. Rad1, but he was still very much in the game and fighting hard to create attacking possibilities.

Only with 30. h4? did Lee finally reach a lost position. Had he played 30. Rd1, a move apparently only spotted by Soltis, Lee's position would still have been playable.

Lee then went downhill fast. His 33. Ra5 was bad (33. Rb6 was the only real chance) and his 36. Bd4? (36. BxN) opened the door to Lasker's fireworks.

14...b4 would have nixed these chances for Lee. I see no way for Lee to achieve more than equality after 14...b4. Best for White seems to be 15. exd5, but after 15...Qxd5 Lasker has at least equal chances. This would not have been possible for Lasker had Lee responded properly to 14...Na5.

Lee was already lost before his weak 36. Bd4? (36. BxN was "best") and 37. Qe3? (37. Bd1 was "best" but loses decisively to 37...Qh3). But Lasker didn't just win this won game--there were many roads to Rome at this point--he did it with sublime artistry.

Shortly after the 1899 London tournament, Tarrasch pointed out an arguably faster (and also brilliant) winning combination beginning with 37...Qg4+. an argument then developed: which line was better, Lasker's 37...RxB! or Tarrasch's 37...Qg4+.

Soltis casts his vote for Lasker's line, stating that it was more "elegant" than Tarrasch's. It is hard to argue with this, since 46...Bc8!!--a move Lasker says he had foreseen when playing 37...RxB, is one of the most gorgeous things I have ever seen on a chessboard.

Incredibly, Anderssen99 and keypusher have found improvements in the Tarrach and Lasker combinations.

Tarrasch's line begins with 37...Qg4+ 38. Kf1 Qg2+ 39. Ke1 Qg1+ 40. Kd2 Ng2 41. Qe2 e3+.

Anderssen99 has found an improvement no one else seems to have noticed in over a century...41...RxB+

This is beautiful. I love it. Bravo Anderseen99!

In Lasker's 37...RxB line, keypusher has found an even more devastating line than the 41...c3+ Lasker actually played. As only keypusher seems to have discovered, 41...Ng2! is utterly crushing.

Wow! Dynamite! Bravo keypusher!

These improvements, however, in no way detract from Lasker's transcendent play in closing out this game.

I agree entirely. I kept wondering: who's this guy who's playing so well that he's got Lasker playing positional micro-moves like 17...Qc7 and 30. Bc6. Frankly, maybe this shows a lack of sophistication on my part but I quickly get bored with positional games and was greatly relieved when Lasker finally came out of the closet as the great tactician we know and love with 37. Rxd4.

< after 19. NxN Lasker--incredible to relate--erred with the wooden 19...dxN. In fact, as only Soltis has pointed out, Lasker should have played the problem-like 19...exd4!! Amazing that none of the other commentators noticed this powerful move or the fact that Lasker had forfeited most of his advantage with his 19...dxN.>

If instead Lasker had played 20...Nc6 as you suggest, then 21. axb5 does look best for White, but 21...axb5 would then be a blunder, since 22. RxR is crushing (as you note, 22...BxR runs into 23. Nf6+, and 22...RxR still leaves White winning after 23. Nc3).

But Lasker didn't need to go in for any of this. After 21. axb5 in your line, Lasker would undoubtedly have played 21...dxe4! Now, if 22. bxN then 22...exN. 23. cxB RxQ+ 24. RxR Qxb7.

I think these lines confirm our agreed conclusion concerning the superiority of 19...dxe4. Whether Lasker had played 20...Nc6 (your move) or 20...Nb3 (my suggestion) the result appears to be the same: a win for Black.