Instead, White is handed an excellent attacking opportunity, which he takes after completing his development, gaining a clear advantage by move 12. Black neglects his defense of the evil e-file and should have been punished for it on move 15, where the engines show White winning a piece. However, White loses his nerve and goes for two piece exchanges. The exchanges allow White to wreck Black's kingside pawn structure, but the disappearance of the attacking pieces and Black's extra pawn mean that the position is level. White makes some additional demonstrations on the kingside, but his decision to again exchange an attacking piece on move 21 leads eventually to the draw.
Again I am struck by the usefulness of analyzing your own games as an improvement practice. Had I been serious about this earlier in my career, it would have led more quickly to better performance. In this case, the neglect of the e-file should have led to a loss and meant that Black was happy to end up with a draw. My tendency to neglect this necessary defensive aspect of the position was evident in the previously analyzed game, but the lesson had not been learned.
[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Class B"]
[Black "ChessAdmin"]
[Result "1/2-1/2"]
[ECO "B18"]
[Annotator "ChessAdmin/Fritz/Houdini"]
[PlyCount "74"]
{B18: Classical Caro-Kann: 4...Bf5 sidelines} 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4.
Nxe4 Bf5 5. Ng3 Bg6 6. Nf3 Nf6 7. c4 {there are only a handful of games in the
database with this move. It is premature in several respects, as White should
keep the c-pawn in reserve to later drive away a knight posted on d5 or block
the e1-a5 diagonal.} e6 8. Bd3 $146 {now out of the database.} Bxd3 9. Qxd3 c5
{while the c5 break is a basic idea for Black in many lines, here it is very
premature and cedes control of d5, in the process making White's earlier c4
useful.} (9... Bb4+ {would most likely lead to an exchange of bishops on d2,
removing a key attacking piece for White. ...Nbd7 would also be a useful
developing move here.}) 10. d5 $14 Be7 11. O-O O-O 12. Re1 {White's advantage
is now clear, with his pieces well positioned to attack down the e-file and on
the kingside. Meanwhile, Black has no counterplay to speak of and must
concentrate on defense.} exd5 13. Nf5 Nc6 (13... Re8 {would be more solid.})
14. Bg5 {a conventional use of the bishop.} (14. Bh6 {is what Houdini likes,
immediately putting Black's king in jeopardy.} gxh6 15. Qe3 Ne8 16. Qxh6 {
with a strong attack.}) 14... d4 $2 {Black fails to bring up the necessary
defensive reinforcement.} (14... Re8 15. Ne5 Nxe5 16. Rxe5 Bf8) 15. Nxe7+ $2 {
White loses his attacking nerve and simply exchanges down.} (15. Ne5 {is the
win spotted by the engines.} Nxe5 16. Rxe5 $18 {and now in contrast with the
previous variation with 14...Re8, the hanging bishop on e7 is fatal for Black.}
Bd6 17. Rd5) 15... Nxe7 $11 16. Bxf6 gxf6 {now Black's extra pawn is
compensation for his weak kingside structure. White no longer has sufficient
attacking forces to break through there.} 17. Nh4 (17. b4 {is preferred by the
engines, in order to undermine Black's queenside pawns.}) 17... Ng6 18. Nf5 Re8
{the rook finally gets into the action.} 19. h4 {White evidently still has
hopes for a breakthrough on the kingside.} Qc7 20. h5 Ne7 21. Nxe7+ {White
keeps exchanging off his attacking pieces.} Rxe7 22. f4 $6 {this loses a pawn,
although Black declines to take it.} (22. h6 {is the logical continuation,
given the previous h5 push. Now Black would have to pay close attention to his
defenses, although they are adequate.} Rxe1+ 23. Rxe1 Qf4) 22... h6 {Black
thinks safety first.} (22... Rxe1+ $5 23. Rxe1 Qxf4 24. Rf1 $17 {I recall that
I didn't like the optics of this - White's pieces and the h-pawn look rather
menacing, given the air gap in front of Black's king. However, it would be
easy for Black to simplify down into an ending where White had all the losing
chances.} Qe3+ 25. Qxe3 dxe3) 23. Qg3+ Kh8 24. Rxe7 {essentially forced,
otherwise Black will double rooks on the e-file.} Qxe7 25. Re1 Qd6 26. Qf3 Qc7
27. Qe4 Rd8 {the rook enters the game very usefully. Black finally has some
threats of his own involving the passed d-pawn.} 28. g3 Kg7 29. Kf2 Qd7 30. Qe7
b6 {safe but drawish.} (30... Qc8 $142 $5 $17 {would preserve the queen, which
has the potential to help Black support a breakthrough. For example} 31. Re2 d3
32. Rd2 b5 33. cxb5 c4) 31. Qxd7 $15 Rxd7 32. Re8 {correctly getting to the
8th rank in order to tie Black up with potential threats in his rear. The king
will be able to stop the d-pawn.} f5 {controls e4 and g4 and gives the king an
opening on f6.} 33. Ke2 f6 {Black is out of ideas on how to make progress, but
White in any event should be able to easily hold the draw.} (33... d3+ 34. Kd1
$11) (33... Kf6 34. Ra8) 34. a3 $11 Kf7 35. Rh8 Re7+ 36. Kd2 Kg7 37. Rb8 Rd7
1/2-1/2
8...Bxd3 looks like a mistake to me. 8...Bb4+ is very embarrassing for White. Even so, after 9.Kf1 Shredder gives only a slight advantage for Black. On general grounds, the exchange is questionable because it develops White's Q for free.
ReplyDeleteHello and thanks for the comment.
Delete...Bb4+ on either the 8th or 9th move is certainly an excellent choice, probably best.
I don't consider the Bxd3 exchange a mistake, even if it's not necessarily best. (Some people consider anything not the best move a mistake, but that's a philosophical debate best discussed elsewhere.) The bishop exchange is normally done in the Classical Caro-Kann in order to prevent Bxg6 from doubling Black's g-pawns and giving White better attacking chances if Black castles kingside. In this particular case, Black missed the opportunity to play ...Bb4+ first, an option not normally available in this variation. Black should have recognized what had explicitly changed in the nature of the position, versus the normal variations, after White played c4.
In terms of computer evaluations, Houdini shows 8...Bb4+ as even, while 8...Bxd3 is shown as a slight advantage to White (0.15) that is still within standard range for equality. After Black's glaring 9...c5 error, the engine shows a +0.43 evaluation, with White's 10. d5 correctly exploiting the mistake.