Annotated Game #257: The dangerous Caro-Kann Classical

This next game, the third played, was the best of the tournament for me, in terms of the level of play and it being the most interesting. Over the years I've had excellent results with the Caro-Kann Classical defense, as it contains much more dynamic possibilities than its rather staid reputation. So I was happy when my opponent went into the variation and left theory around move 10 with a sub-par move.

As is usually the case, this theory deviation was hardly losing, but it did give me an easier game and essentially allowed Black to equalize immediately. My opponent adopted a very aggressive-looking plan of advancing his g- and h-pawns against my king, following queenside castling, but the thematic 16...Ng4 effectively blocked that possibility. To his credit, he found the tricky follow-up 17. Ng5, a sacrifice that complicated things considerably and made me have to be careful...occasionally too careful, as was determined in analysis. My forward-deployed knight, however, continued to serve a blocking function on h2 and was the determining factor in the win.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class B"] [Black "ChessAdmin"] [Result "0-1"] [ECO "B19"] [Annotator "ChessAdmin/Dragon 2.5.1 by Komodo"] [PlyCount "70"] 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 {the Classical Caro-Kann} 5. Ng3 Bg6 6. h4 h6 7. Nf3 Nf6 8. Bd3 Bxd3 9. Qxd3 e6 10. c3 {first deviation from theory by my opponent. This is not a usual feature of the Classical variation, since the extra reinforcement of d4 is not really needed.} Bd6 {my first real think of the game. Other standard development options would be ...Be7 or ...Nbd7. In my experience, it's useful to get the king castled more quickly, to avoid White developing tactical threats on the e-file, which means a bishop move. On d6, it opposes White developing his bishop to f4.} 11. Bd2 O-O 12. Ne4 { this forces the knight exchange, otherwise the Bishop becomes awkwardly placed after retreating to c7.} Nxe4 13. Qxe4 Nd7 14. O-O-O {this surprised me a little. By doing this, White is signaling his intention to play aggressively on the kingside. However, it also will give Black's queenside counterplay more bite.} Nf6 15. Qd3 Qc7 (15... Qa5 $5 {is a more active placement of the queen, adding pressure to the a-file while also influencing the 5th rank.}) 16. Rdg1 { an obviously aggressive move, with White planning to push the g- and h-pawns for a direct attack on Black's king position. The next move is a thematic response for Black, physically blocking the pawn advance.} Ng4 {White has no way of removing the knight, but finds the continuation that allows him to attempt to keep pressing an attack. I had seen the possibility of his next move, but had somewhat discounted it.} 17. Ng5 {threatening mate on h7. While it does not in fact lose a piece, Black should still be at least a little better.} (17. Qe2 $5 {is the engine's preference, going for a positional game that keeps the balance, although it gives Black a small edge.}) 17... hxg5 ( 17... f5 $5 {is what Komodo Dragon 2 prefers, despite the fork on e6.} 18. Nxe6 Nxf2 19. Nxc7 Nxd3+ {I had calculated this far and seen no benefit to Black after Kc2, but after} 20. Kc2 Nb4+ 21. Kb3 Bxc7 22. Kxb4 {White's exposed king position and Black's ability to put a rook on the e-file should provide an edge.}) 18. hxg5 f5 {forced.} (18... g6 $2 19. Rh4) (18... Rfe8 $2 19. Qh3) 19. Qe2 Rae8 $11 {with best play the Ng4 is trapped, which I correctly assessed at the time, so I did not worry about it too much. However, my opponent appears to have underestimated the difficulty of rounding it up without incurring further losses, specifically missing the fact that it could go to h2.} 20. f3 ( 20. Rh3 $5 {would make ...Nh2 pointless.}) 20... Nh2 {this effectively blocks the h-file and White's attacking ideas, at least for the moment. White should immediately resolve this, even at the cost of the g2 pawn, which would restore his dynamic chances on the kingside.} 21. Re1 $2 (21. g3 Bxg3 22. Qg2 Bf4 { was what I was expecting, which does not give White enough time to develop an attack.} 23. g6 Bxd2+ 24. Kxd2 Qf4+ 25. Ke2 Qxf3+ 26. Qxf3 Nxf3 27. Kxf3 Rf6 $11 {despite being a pawn down, White is more active in the double rook endgame, which counts for a lot.}) 21... g6 $6 (21... Bg3 $1 {I had looked at this as a candidate move, but did not find the most effective follow-up, so eventually went with a more defensive move.} 22. Rd1 {the point is that now White cannot get to the Nh2, so Black should be able to keep the material and have a winning advantage. For example} e5 $19 {and now Black's centrally placed rooks also get into the game.}) 22. Qc4 b5 $17 {by this point I have a comfortable advantage, even if I have to keep a watch on e6. With the pawn advance, my queenside counterplay is also now coming into focus.} (22... Bg3 { is still possible.} 23. Re3 Kf7 $19) 23. Qb3 Kf7 24. Re2 a5 {with the obvious threat of ...a4, but my opponent misses his chance to block it.} (24... Rh8 $5 {is the engine's preference, which activates the rook on the h-file and overprotects the Nh2, freeing up the Bd6 and Qc7.}) 25. g4 $2 {this accomplishes nothing and lets me roll on the queenside.} a4 $19 26. Qc2 Kg8 { a prophylactic move designed to avoid any possibility of a subsequent White attack down the h-file, for example threatening to place a rook on h7.} 27. gxf5 exf5 {now my Re8 springs to life and White collapses.} 28. Rf2 $2 { this makes it easier for me, with the next obvious follow-up move, but White was already lost.} Bg3 {forcing the next move and allowing the Nh2 to extricate itself.} 29. Rg2 Nxf3 30. Qd3 Nxd2 31. Rxg3 Ne4 32. Rg2 (32. Rgh3 Qf4+ 33. Kc2 Nxg5 34. Rh8+ Kf7 $19) 32... Qf4+ 33. Kb1 Nf2 {the fork is the final nail in the coffin.} 34. Rxf2 Qxf2 35. Qh3 {threatening mate on h7, but one tempo too short.} Re1+ 0-1