fling wrote on 02/03/12 at 11:59:06:

Kramnik-Carlsen was a pure reversed Sokolsky, 1.Nf3 b5 2.e4 Bb7 3.Lxb5 Bxe4 - that is much worse compared to 1.Nf3 Nf6 2.g3 b5.

Keano wrote on 02/02/12 at 16:55:21:

Yes, Keano! I think this is the best way to meet 2...b5. Some 20 years ago I played almost exclusively 1.Nf3 as a first player. And I myself tried 2...b5 against 1.Nf3 Nf6 2.g3 several times as a second player. With white I tried almost every possible setup against 2...b5 and found out that this 5.b3-setup was most annoying for black. 

Larsen_fan wrote on 04/05/11 at 17:02:11:


The only game with this exchange sacrifice I know of is Lago - Gaprindashvili, 1999. White made seemingly plausible moves and ended in a bad position. 10.d3 is a mistake but 10...Qc8 can easily be missed. The exchange sacrifice is highly speculative and possibly not quite correct. It's unlikely that Black succeeds in capturing the Ba8. 

Hickl apparently expected the pawn sacrifice 4...d5. I think he was surprised by 4...c6. At least that's the only explanation I have for his opening play in his game vs. Aronian.  

The oldest 3...e6 games with decent players I could find are Csom - Romanishin, 1976, and Szymczak - Prandstetter, 1979. In both cases White didn't dare to take the offer. Cautious white players can simply decline the offer and play their chess.  


[Date "1999.??.??"]
[White "Lago, Victor Miguel"]
[Black "Gaprindashvili, Valerian"]
[Result "0-1"]
[WhiteElo "2340"]
[BlackElo "2390"]

1. Nf3 Nf6 2. g3 e6 3. Bg2 b5 4. Nd4 c6 5. Nxb5 cxb5 6. Bxa8 d5 7. a4 b4 8. c4 Ba6 
9. cxd5 exd5 10. d3 Qc8 11. e4 dxe4 12. O-O Bxd3 13. Re1 Be7 14. Nd2 Nc6 15. Bxc6+ Qxc6 
16. Nf1 O-O 17. Be3 Qd7 18. Kg2 Rd8 19. Rc1 Nd5 20. Qb3 Qe6 21. Nd2 h5 22. Bxa7 h4 
23. Kg1 hxg3 24. fxg3 Bg5 25. Rcd1 Qg4 26. Nf1 Nf4 27. Be3 Nh3+ 28. Kg2 Rd6 29. Nd2 Be2 
30. Bxg5 Nxg5 31. Qe3 Bxd1 32. Nc4 Bf3+ 33. Kg1 Rf6 34. Ne5 Nh3+ 35. Kf1 Bh1+ 0-1


[Event "?"]
[Site "Naleczow"]
[Date "1979.??.??"]
[White "Szymczak,Zbigniew "]
[Black "Prandstetter,Eduard "]
[Result "1/2-1/2"]
[ECO "A04"]
[Round "?"]

1. Nf3 b5 2. g3 Nf6 3. Bg2 e6 4. O-O Be7 5. c4 O-O 6. b3 c5 7. Bb2 Nc6 
8. e3 bxc4 9. bxc4 Qb6 10. Qb3 Na5 11. Qxb6 axb6 1/2-1/2

[Event "?"]
[Site "Erevan"]
[Date "1976.??.??"]
[White "Csom "]
[Black "Romanishin,Oleg "]
[Result "1/2-1/2"]
[ECO "A05"]
[Round "?"]

1. Nf3 Nf6 2. g3 b5 3. Bg2 e6 4. O-O c5 5. d3 Bb7 6. e4 d6 7. Nbd2 Nc6 
8. h3 Be7 9. Nh2 Qb6 10. Kh1 d5 11. a4 dxe4 12. axb5 Qxb5 13. dxe4 Rd8 14. Re1 c4 
15. f4 e5 16. c3 O-O 17. Qa4 Qxa4 18. Rxa4 Nh5 19. Ndf1 exf4 20. Bxf4 Bh4 21. Rea1 Rd3 
22. Nf3 Nxf4 23. gxf4 Bf2 24. Rxc4 Rfd8 25. Kh2 Bb6 26. Rca4 f6 27. e5 fxe5 28. Nxe5 Nxe5 
29. Bxb7 Nf3+ 30. Kh1 Nh4 31. Bg2 R8d6 32. f5 Nxf5 33. Rf4 Rf6 34. Rf3 Rxf3 35. Bxf3 g6 
36. Bd5+ Kg7 37. b4 h5 38. Rd1 Kh6 39. c4 Bd4 40. c5 Bf2 41. Ra1 1/2-1/2

Don't engines still have non trivial problems with trapped pieces? It doesn't seem that the bishop on a8 is going to get out alive so more or less material equality.

If it isn't then walk the computer down the lines a bit and the evaluation should pick up when it notices this. Or, in the way of computers, it might find some contrived way to rescue it

Do have to say that e6 does seem an odd move and the whole e6/c6 etc set up rather artificial.

Quote:


4. Na3 isn't a theoretical threat - white often uses it a couple moves later anyway.

Even black's most blatantly obvious defense is sufficient for him - 4...b4 5. Nc4 a5 - that's not a good sign for white's fourth move having punch to it.

Larsen_fan wrote on 04/03/11 at 22:08:03:


3...Bb7 is normal. No idea why black would play 3...e6 - even if he gets compensation (B+N versus R+2 pawns) it seems unnecessary.

I think that the Spasskij variation is great! 
Black can play to win
The battle is bloody!