Annotated Game #137: A (mostly) clean win in the Caro-Kann

This seventh-round win may not be completely clean, but it sure looks better than the last few games and shows how effective the Caro-Kann can be as a counterattacking opening.  My opponent avoids theory early on, probably not having prepared anything against the Caro-Kann, and enters a harmless variation that lets Black equalize quickly.

White nevertheless plays quite aggressively as the middlegame phase is entered, signaling with 11. fxg3 that he wants to try for a kingside attack.  However, with both bishops already exchanged off and a solid structure for Black in place, this plan is over-optimistic.  Ignoring an interesting sacrificial theme on f6, which would have forced a draw with a perpetual check, White instead overextends his kingside pawns and essentially traps his own rook on the h-file.  I am then able to switch to operations on the c-file and break into White's back ranks, finishing off his king as it tries to run up the board.  Although as you can see in the notes it would have been better for me to execute some of my ideas a move earlier than I actually played them, I'm generally pleased with my performance.  There are a number of other tweaks that I found in analysis that should also help with future play of similar positions, including avoiding wasted moves and looking out for sacs on f6.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class C"] [Black "ChessAdmin"] [Result "0-1"] [ECO "B11"] [Annotator "ChessAdmin/Houdini/Komodo 8"] [PlyCount "68"] [EventDate "2012.07.??"] [EventType "swiss"] [EventRounds "9"] {B11: Caro-Kann: Two Knights Variation} 1. e4 c6 2. Nc3 d5 3. exd5 cxd5 4. Bb5+ Bd7 5. Bxd7+ Qxd7 6. d4 Nc6 7. Nf3 e6 8. Bf4 Bd6 9. Bg3 $146 Nf6 {an interesting alternative would be to have the knight head for f5 via e7, given the presence of the Bg3 and the chance to add pressure on d4.} 10. O-O Bxg3 { an example of how Class players prematurely release tension. Exchanging now does not offer any advantage to Black.} (10... Rc8 $5) 11. fxg3 {a weakening move that indicates my opponent is looking to play aggressively by opening the f-file for his rook.} a6 {Controls b5, but this is essentially a wasted move, as White is not threatening anything on the kingside. I could (and should) have castled immediately, but wanted to see more of my opponent's intentions first.} 12. Qd2 O-O {I correctly assess that Black is solid and should not worry too much about White's threatening gestures.} 13. Rae1 Rac8 14. Ne5 { an impatient move. White was evidently hoping for an exchange, which would be to his favor with a pawn established on e5 and no good retreat square for the Nf6. However, I do not oblige and simply retreat the queen to a better square.} Qd8 15. h3 {White returns my earlier favor of a wasted move (...a6) with one of his own.} (15. Nxc6 bxc6 16. Na4 Ne4 $15) 15... b5 (15... Nxe5 {now is good for Black, since the Nf6 has d7 for a retreat square.} 16. dxe5 Nd7 17. Rf2 Qb6 18. Na4 Qc7 $17 {Komodo 8 here considers Black to have by far the better position, up a pawn equivalent. White's pieces are scattered and uncoordinated, while Black will have strong pressure on the queenside and against the e5 pawn. }) 16. a3 {Consolidates b4, notes Houdini via the Fritz interface.} Qc7 { a passive move and it also overlooks a tactical shot on f6, now only protected by the g-pawn, which would give White a forced draw.} (16... Nxe5 17. Rxe5 Rc4 $17) 17. g4 (17. Rxf6 gxf6 18. Qh6 fxe5 19. Qg5+ Kh8 20. Qf6+ $11) 17... Nxe5 $15 {now I play the right move.} 18. Rxe5 Nd7 19. Rh5 {White continues to play agressively and over-optimistically.} Rfe8 (19... Nb6 $5 {threatening to move to c4 and also freeing up the queen on the seventh rank.} 20. g5 Qe7 $15 { lending support to the f-file.}) 20. Qf2 {a rather obvious threat. I parry it easily, but could have obtained an advantage instead.} Nf6 {this puts the knight back into the tactical line of fire, which White could have exploited with a similar sacrificial theme as in the previous variation, the removal of the g-pawn.} (20... f6 21. g5 f5 $17) 21. Rh4 (21. Rg5 Kh8 22. Rxg7 Kxg7 23. Qxf6+ Kg8 24. Qg5+ Kh8 {the best option Black has is to continue the repetition of moves.}) 21... Rf8 {a rather ineffectual move.} (21... h6 { this little prophylactic move would have been excellent, controlling g5 and shoring up Black's defense.} 22. Rxh6 Nxg4 23. hxg4 gxh6 24. Qf6 $15) 22. g5 $11 Ne8 23. Qf3 {another obvious threat, this time to head to h5.} g6 {if White had a dark-square bishop, Black could be in trouble. In this case, however, White's attack has nowhere left to go.} 24. Rh6 {White's optimistic desire to attack now goes too far. The rook visually looks more threatening on h6, perhaps the reason for my opponent's choice, but it in fact does nothing for him there and becomes a liability shortly.} (24. Rf4 Nd6 $15) 24... f5 ( 24... Nd6 {immediately is better.} 25. Qf6 Rfe8 $17) 25. g4 $2 Nd6 $19 26. Qe3 Ne4 27. Nxe4 $2 {an oversight. But White was lost anyway, states Houdini via the Frtiz interface.} (27. gxf5 exf5 28. Nxd5 $19 Qd8 29. Nf6+ Nxf6 30. gxf6 Qxf6 31. c3 Rce8 32. Qf4 {White has to maintain control over the c1-h6 diagonal to prevent a queen fork on g5, which allows Black to penetrate.} Re2 $19) 27... fxe4 {opening the f-file now makes all the difference for Black.} ( 27... dxe4 $6 28. c3 fxg4 29. Rf6 $16) 28. Rf6 (28. c3 Rxf1+ 29. Kxf1 Rf8+ 30. Kg2 Rf3 $19) 28... Qxc2 {White is now effectively helpless as Black penetrates the back ranks with the heavy pieces.} 29. Qf4 Rxf6 30. Qxf6 Qb1+ $19 31. Kf2 Rc2+ 32. Kg3 Qg1+ 33. Kf4 Qh2+ 34. Ke3 Qg3+ {missing the more elegant mate in one in favor of the mate in two.} (34... Qd2#) 0-1