In your A line I'd agree that White has more than enough compensation, but 14...Qxc3 doesn't look a good idea to me. With BOC Black should be looking to keep his position solid and trade down to the drawn ending. An extra pawn here or there will make no difference.

As such I see two possible plans here - the solid 14...Re8, intending Re6, and the one I like - saccing a pawn back by 14...d5 15.cxd6 f4 followed by blockading the pawn with Bd7. The pawn c3 is still hanging and f3 is also threatened. Might I suggest that this looks rather drawish?

Can Black just castle back, after Nxd4/e5?

I would rather have White. Black has to rely on his pawns and as soon as they advance White's queen will invade and pick up a few here and there. But if I am wrong, as so often, there is also 12.Rb1, see the initial post.

drkodos wrote on 12/05/07 at 20:58:17:


Well the fun only starts of course after accepting the sacrifice. 8.., Qc7 is possible but there have been changed some things.
1) The black queen committed herself to a square already. Which makes ideas with Qa5, Qf6 (once the knight has moved), ... impossible
2) After 8.., Qc7 9. Nc2 becomes possible and stronger than 8. Nc2 because white doesn't need to worry anymore about taking on c3 , followed up with d5 which equalises on the spot
3) After 8..., Qc7 9. Be3 is now stonger than immediately because after the standard reply 9.., Ne5 follows 10. Nf3, Nc4: 11. Bc4: , Qc4: 12. Na4! with strong threats
4) Also after 8.., Qc7 9. Nc6:, dc6: 10. e5 is still interesting although I must admit against a strong defense the pawnsacrifice looks a bit light.
5) After 8 .., Qc7 9. Bc2 Black has already committed to the Qc7 line which in comparison with 8..., Ne5 is regarded by theory as less good.
6) 9. Nf3 is more interesting than 8. Nf3 because the queen already committed to c7.
7) White can still choose to return to the mainline with 9. Nde2

Conclusion even if black doesn't accept the challenge then white has the benefit to bring black at least on unfamiliar ground very early in the game. I don't claim any advantage for white but lots of new unexplored possibilities which is exactly what a tournament player looks for.

By analysing one of my recent clubchampionshipgames I bumped on a spectacular novelty very early in the paulsen/kan system. I was surprised that in the almost 1000 games I found with the position, this move was never played. I must admit that it looks not logical at all.

1. e4, c5 2. Nf3, e6 3. d4, cd4: 4. Nd4:, a6 5. c4, Nf6 6. Nc3, Bb4 7. Bd3, Nc6 8. 0-0 N ??!!! Yes the knight on d4 hangs but things aren't so simple :

8 .., Nd4: (Of course black can refuse but then it will certainly not refute whites last move. On top refusing will mean that white gets extra time to find the optimal square for the knight.)
9. e5!, Nc6 (Giving back the piece is the best policy because after 9.., Ng8 follows 10. Qg4 which gives white a huge advantage)
10. ef6: and now black must choose between A. 10.., Qf6: or B. 10. .., Bc3:

A. 10.., Qf6: 11. Ne4!, Qe5 (11... , Qe7 is worse due to 12. a3.) 12. a3, Be7 13. Re1, f5 14. Nc3, Qf6 15. Bf4, 0-0 16. Bf1 and white has excellent compensation for the pawn

B. 10.., Bc3: 11. bc3:, Qf6 12. Rb1!, 0-0 13. Ba3 (13. Qc2 is also interesting see 13 ..., g6 14. Ba3, Re8 15. c5, Ne5 16. Be2, Nc6 17. Rfe1, Qe7 18. Qb3) , Rd8 14. Bd6, Qc3: 15. Rb3, Qf6 16. Qb1, g6 17. c5, Nd4 18. Qb2, Qg7 19. Rb6 with a very passive position for black and lots of compensation in both lines for white

Let me know if somebody is willing and able to test this in practice.