The zwichenzug is necessary b/c 25.Bf4 Re1+ 26.Rxe1 Qxe1+ 27.Kh2 d5.

Did you have Shreddr analyze from my blunder..d5, or from 25.Bf4 itself? Because I still think the pendulum is better than 25.Bf4 directly: 25.Bf4 Re1+ 26.Rxe1 Qxe1+ 27.Kh2 Qf1 seems to work (eg 28.b3 d5 29.Bg6 Qa1 30.Qc7 Qf6 holds together doesn't it?).

I think I mis-analyzed 25..g6 too: 25..g6 26.Bf4 as in the gme, and if 26..Re1+ 27.Rxe1 Qxe1+ 28.Kh2 then 28..Qf1 now fails to 29.Bh6 (that bishop must be getting tired...).

Game 10

Played in Munich on September 14

<Translated from Tarrasch's book on the match (http://www.google.com/books?id=0CgC...) with additional commentary from Kasparov's <OMGP I>. Tarrasch's comments are in plain text; all other comments are in brackets. German that I was particularly unsure about is also in brackets, right after the translation. Corrections are welcome.>

1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. 0-0 Nxe4 5. d4 Be7 6. Qe2 Nd6 7. Bxc6 bxc6 8. dxe5 Nb7 9. Nc3 0-0 10. Re1 Nc5 11. Nd4 Ne6 12. Be3 Nxd4 13. Bxd4 13. c5 14. Be3 d5 15. exd6 Bxd6 16. Ne4

So far this game is a repetition of the eighth. <See Tarrasch vs Lasker, 1908, where the opening is discussed in detail.> The knight move is much stronger than the queen move to h5 played there. Now Black may not proceed with 16....Bxh2+ 17. Kxh2 Qh4+ 18. Kg1 Qxe4 19. Bxc5 Qxe2 20. Rxe2 Rd8 21. Re7 Bd7 22. Rd1, etc. 20....Ba6! (instead of 20....Rd8) 21. Re7 gives White the significantly better game, despite the opposite-colored bishops.

<OMGP I says this is equal after 21....Rfc8 22. Rae1 Bc4 23. Bxa7 h6 "(Vukic recommended 23....Bxa2 24. b3 Bb1, but after 25. Rxb1 Rxa7 26. Rd1 Black can resign).">

16....Bb7 17. Nxd6

17. Nxc5 would after the continuation 17....Bxc5 18. Bxc5 Qg5 cost White a piece. This constant threat on g2 is generally very irksome for White. <Diese stete Bedrohung des Punktes g2 ist uberhaupt fur Weiss hochst lastig.> While the exchange on d6 undoubles the black pawns, the dark squares will [otherwise] remain vulnerable to attack, so White must always seek to exchange one of the dangerous bishops <allein auf der schwarzen Feldern sind sie immer noch Angriffen zuganglich, und einen von den gefahrlichen Laufern muss Weiss immer zu beseitigen suchen>. However, the opposite-colored bishops, which readily occur in the Brazilian variation, give Black drawing chances in the event he loses a pawn <die uberhaupt in der brasilianischen Verteidigung leicht auftauchen, geben dem Schwarzen immer auch fur den Fall, dass er einen Bauern einbusst, Remischancen>. In any case, Black obtains a very satisfactory game in the continuation White chooses here.

17....cxd6 18. Rad1 Qf6 19. c4

Thus White makes the d-pawn permanently backward, so that it may later become a target. <Damit halt Weiss den d-Bauern dauernd ruckstandig, um ihn spater zum Angriffsobjekt zu machen.>

19....Re8

Now threatening, as in the eighth game, to attack the g2 square subsequently with ....Re6, ...Qe7 and ...Rg6.

20. Qg4

Here Black cannot take the b-pawn, as after 20....Qxb2 21. Rb1 Qc3 22. Rec1 he would be worse. White cannot take the bishop after 22....Qa3, because of ...Qxc1+ and Re1#, but there can follow instead 23. Bh6 g6 24. Qf4, and now both Qf6 and Rxb7 are threatened. The standing threat of Bh6 plays a big role in the subsequent course of the game.

The continuation 20....Re4 21. Qd7 Qxb2 22. f3! (not 22. Rb1 because of 22....Rxe3!) is disadvantageous for Black.

20....Bc6?

The decisive mistake. Black parries the threat of Qd7, but he could have done that with ...Re6, with a very satisfactory game. Bg5 after ...Rxe1+ and ...Qxb2 would be bad for White.

<After 20....Re6, OMGP I gives 21. Bf4 (21. Re2?! Rae8 22. Red2? Qe5! or 22. Qg3 Qg6!; 21. b3 Rae8 22. Qg3 h6 is equal) 21....Rae8 22. Rxe6 Rxe6 23. h4 (23. b3? Qe7!) 23....h6 or 21. Qf4 Qg6 22. f3 Rae8 23. Bf2 Rxe1+ 24. Rxe1 Rxe1+ 25. Bxe1 Qd3 26. Bc3 f6 "with equal chances in each case.">

21. Re2

A very simple yet strong move, after which Lasker spent no fewer than 56 minutes on his response. The difficulty for Black lies in the fact that his powerful development via ...Re6 is now impossible, because then Bg5 and the rook exchange would follow, with advantage for White. <Die Schwierigkeit fur Schwarz liegt namlich darin, dass seine kraftvolle Weiterenwicklung mittels Te6 jetzt untunlich ist, denn darauf folgt Lg5 nebst Turmtausch zum Vorteil von Weiss.> While his attacking move is thus stymied, he is threatened with an attack by four pieces on the queen pawn, which can only be defended by three pieces.

21....Re4(?)

21....Rab8 22. b3 d5 also came into consideration, but then White would win a pawn via 23. cxd5 Bxd5 24. Red2 Be6 25. Qa4.

<OMGP I: Not without reason do they say that mistakes come in pairs. 21....Re6? no longer worked because of 22. Bg5, but better, nevertheless, was Vukic's recommendation 21....h6!? 22. Red2 Rad8 23. Qg3 (23. Qf4 Qxf4 24. Bxf4 Re6 is equal) 23....Re6 24. h4 (24. Bxc5? Qf5 and ...Rg6) 24....Qe7 25. h5 Kh8, and Black's defenses hold: 26. Bxc5? Re1+ (<let us continue: 27. Kh2 Rxd1 28. Bd4 f6 29. Rxd1 Qe2 30. Rc1 Qxh5+ 31. Kg1 with equal chances -- G.K.>), or 26. Bf4 d5 27. cxd5 Bxd5 28. Rxd5? Re1+! (Neishtadt) However, after 28. Be3 White has a clear advantage, and so instead of 27....Bxd5? correct is 27....Re1+ 28. Rxe1 Qxe1+ 29. Kh2 Rxd5 30. Rxd5 Bxd5.>

22. Qg3 Qe6

A surprising move, which has many surprising twists in its wake. Expected was 22....Rxc4 with the continuation 23. Rxd6 Rd8! 24. Rxd8+ Qxd8, whereupon 25. h3! would give White the advantage because of the threats Bh6, Rd2, etc. After the text move ...Rg4 is the foremost threat; it would follow even after Qxd6, whereupon not only a catastrophe on g2 but also the loss of the exchange by ...Bf3 would be threatened. Satisfactory for White after ...Qe6 is the riposte 23. Rxd6 Qxc4 24. Red2, when 24....Rg4 is not favorable because of 25. Rd8+ Be8 26. Qxg4 and 27. Rxa8. <But after 24....Qxa2 25. h3 Re6 things are not so clear -- OMGP I.> The following move, which quite simply parries the major threat <Hauptdrohung> ...Rg4, is much stronger, however.

23. h3!

After 23....Qxc4, White would obtain an excellent game with 24. Bh6 g6 25. Rxe4 Qxe4 26. Rxd6, as the bishop on h6 is very strong and supports constant mating threats <lasst immer Mattangriffe zu>; already 27. Rxc6 Qxc6 28. Qe5 f6 29. Qe7 and mate is threatened.

23....Rd8 24. Red2 Re5 25. Bh6

Black cannot take the bishop because of Qxe5.

<Re the question raised in the kibitzing about whether 25. Bh6 or 25. Bf4 was better, Tarrasch doesn't comment, and neither does OMGP I, though it does give 25. Bh6 an exclamation point. But I am not sure the move deserves one. After 25. Bh6 g6(!), the best Shredder can find for White is 26. f4 Rh5 27. f5 Qxf5 28. Rxd6 Rxd6 29. Qxd6 Rxh6 30. Qxc6, and White is better but is a long way from winning. After 25. Bf4 Re1+ 26. Rxe1 Qxe1+ 27. Kh2, on the other hand, Shredder thinks White is more than a pawn ahead. But after 25. Bh6 Qg6(?) 26. Bf4 as in the game, Shredder immediately rates White as nearly two pawns ahead. So, unless Shredder misevaluated, 25. Bh6 probably worked out better than it deserved.>

25....Qg6(?) <As noted, OMGP I and Shredder agree that 25....g6 was better.> 26. Bf4

Now the rook obviously cannot be taken because of the mate threat on g2. Even so, Qxg6 would be bad, as the Black king would be given a way out. With the bishop move, on the other hand, White finally has achieved his goal, with all his pieces attacking the queen pawn, which must now fall, followed soon afterwards by the c-pawn, and then the game is decided in White's favor.

click for larger view

26....Re6 27. Bxd6 Qh5

It was better to move up <aufziehen> the h-pawn.

28. Qg4

Not, e.g., 28. Be5? because of 28....Qxd1+ 29. Rxd1 Rxd1+ 30. Kh2 Rg6 and Black wins.

28....Qxg4 29. hxg4 Re4 30. Bxc5

One can now see how important it was not to give the Black king a way out by 26. Qxg6.

30....Rxd2

Somewhat stronger was 30....Re1+ 31. Rxe1 Rxd2, when it is not so easy for White to demonstrate his superiority <worauf es fur Weiss nicht so leicht ist, sein Ubergewicht zu behaupten>. White would then play the strong 32. b4, and win after 32....Rxa2 33. b5 Bd7 34. Rd1 Be8 35. Rd8 Re2 36. Bxa7 with his two connected passed pawns.

31. Rxd2 h5 32. Rd6 Resigns.

Black gives up perhaps a little too early, since the different-colored bishops always give some chances for a draw, though the bishop must retreat to e8 (32....Bb7? 33. Rd7 Bc6 34. Rc7 and 35. Bxa7) and White remains, after 33. Rd8 a5 34. gxh5 with a three-pawn advantage and the superior position. Thus Lasker held it useless to prolong the game (which was adjourned in this position) after he saw my sealed move, 32. Rd6 <so dass es Lasker fur zwecklos hielt, die Partie, die in dieser Stellung abgebrochen wurde, wieder aufzunehmen, als er meinen aufgeschriebenen Zug Rd6 sah>.

click for larger view

Finally, a game which I played flawlessly throughout! The effect on the game of the backwards queen pawn is clear and culminates in the four-way attack on the pawn on the 26th move. <Die Konsequenz des Spiels auf dem ruckstandigen Damenbauer liegt klar zutage und kulminiert in dem vierfachen Angriff auf den Bauer im 26. Zuge.> Although one might think that the Brazilian Defense was refuted by this game, that is in no way correct; rather, Lasker's move 20....Bc6 was bad and must be replaced by ...Re6, as I explained in the comments to the game. Naturally the result of this game gave me new heart. <Naturlich erfullte mich der Ausgang dieser Partie mit neuem Mut.>

<After this game, the margin in the match was narrowed to 2:5, with three draws.>

My source is Tarrasch's book on the match. So who you gonna believe, the Praeceptor Germaniae or some Gotham tabloid? :-)

Of course, my original source for move orders was the book by Lasker scholar Kenneth Whyld. I would have to go back and look at Lasker's match book.

<I will venture to add that Lasker's tournament book for St. Petersburg 1909 is riddled with errors. Just saying.>