click for larger view

Yes I think it's still unclear. I tried <44...Rd2 45. Rxf3 Ne5 46.Re3 Ra5 47.Rb1 Rxd6 48.Kf2>

click for larger view

Black is a up, but all pawns are on the side, no passer. I think white shall hold this.

30...Ng7 looks like an interesting alternative of similar strength. But I wouldn't move 31.Ra1. Instead maybe 31.Be4

But I think Black can improve earlier. 50...f3?! looks premature. Pushing the pawn to f3 makes it overextended at this point. Probably Nimzo wanted to keep White's bishop immobilized, but after <50...Rxa2> White can't do much.

For example, 50...Rxa2 51. Be2 Kg5 52. Bf3 Nd7 53. Bg2 Ne5 54. c5 f3 55. Bf1 Kf4 looks like a better version of Black's attack.

My line was only an example. The point is after White plays the massively weakening 29. g4?!, we expect Black to get more than just a slightly better ending.

Robert Eugene Byrne says: <Kamsky used Aron Nimzovich's 4 . . . Ba6, instead of the standard fianchetto with 4 . . . Bb7, to make White decide how to defend the c4 pawn. Thus, 5 Qa4 may leave the queen misplaced later; 5 Qc2 invites 5 . . . c5, when 6 d5?! leads to a dubious gambit after 6 . . . ed 7 cd Bb7 8 e4 Qe7 9 Nc3 Nd5, and 5 Nd2 diverts the queen knight from its most effective post, c3. The move chosen by Karpov, 5 b3, is solid.>

Source: http://query.nytimes.com/gst/fullpa...

<4.g3 Ba6> White has telegraphed his intent to play Bg2; hence his B can't stay to defend c4, and <5.b3> he returns the tempo. Long-term, this tends to give White an immobile Q-side pawn structure. Also, it prevents Qb3 lines.

Now <5..Bb4+ 6.Bd2> Black forces the B off its strongest diagonal. 6.Nd2 Ne4 is the canonical Nimzo threat. The dropback 6..Be7 is also playable, and is called the "Intermezzo Check" line. Black hasn't lost a tempo because White's B is really poor on d2, and inevitably moves again.

<7..Bb7> is fine. This B did its dawn job (jog? :) by forcing White to commit. Now it's biting on granite at c4, so there's no point in staying, and it shifts to its day job.

<8..c5> Black doesn't fear a "bad" pawn structure: c5 isn't weak, and half-open b is more useful for his heavy pieces. a7's weakness doesn't matter, as he already plans the minority attack <15..a5 19..a4>. Then White's a2 turns out to be the real weakness.

Yet another interesting point is 18..Qb7. 18..a4 19. bxa4 Qd7 is worth considering.

Instead of 19.Qd3, maybe f3, reinforcing the center, is better. Over-protection vs. Nimzowitsch!

I agree that 29.g4 looks ugly. If White was so desperate to gain counter-attacking chances, 29.f4 would be the move. Then, put a rook on f2 and push with f5!

Another opportunity to put a harder fight was at move 48. I don't really understand the Rb2-e2 dance. The straight-forward 48.Re2 Ne4 49. Bg2 might be suggested.

The final mistake, it seems, is 54.h4+? wasting a vital tempo. 54.Re6 Kf4 doesn't work just as well for Black: 55.d6 Ra1 56.Rxh6 and I'd argue White isn't lost at all here.

<catfriend: The poor Bishop's doomed anyway.>

I think White wrote off the B as lost as early as <50.Re8>, abandoning the a2-pawn for active Rook checks from behind, and a thin hope that <53.Rxd6> might mobilize his pawns.