This post was originally written by Gregory Keener on 24 April 2007. It was revisited and re-annotated by Jeffrey Barke on 9 February 2008.
Jeffrey Barke vs Gregory Keener
Brooklyn, NY . 23 April 2007
Alekhine's Defense: Modern with Bg4 [B05]

Alekhine's Defense, Modern.
Alekhine's Defense, Modern [B05]. 922 games at ChessGames.com

If 6…Bxf3 7.Bxf3 Nxc4 winning the "free" pawn, either 8.Qa4+ winning the knight or 8.Bxc7 winning material.
6…c6 avoids both possibilities and makes it possible for Black to win the c4 pawn on the next turn.
7. Nbd2 is the last book move. Note—there are 25 games at ChessGames.com that follow this move order and White has a win rate of 76%!
| 7 |
|
g6 |
| 8 |
exd6 |
exd6 |
| 9 |
h3 |
Bxf3 |

White has the option to recapture Black's queen's bishop with either the queen's knight or the king's bishop. Better is 10.Nxf3, leaving the bishop to protect c4 and having both the knight and queen protect d4 in anticipation of 10…Bg7. Capturing with the knight also avoids a cheap Qe7 check.
I, of course, played 10.Bxf3
| 10 |
Bxf3 |
Bg7 |
| 11 |
Qe2 |
Kd7 |
| 12 |
0-0 |
Bxd4 |
| 13 |
Qd3 |
c5?? |

13…c5 is a blunder. While it protects the Black bishop, it opens a8-h1 diagonal and allows White to win the Black queen's rook.
Better is 13…Be5.
| 14 |
Bxb7 |
Nc6 |
| 15 |
Bxa8 |
Qxa8 |
| 16 |
Nb3 |
h5 |

16…h5 seems inaccurate. Right now, the action is occurring in the center and queenside. While it's often a good idea to answer a threat on one side of the board by generating a counter-threat on the opposide side, in this case h5 seems too slow.
At this point, I was an even worse attacking player than I am now! I believe I played 16.Nb3 to threaten Black's king's bishop and to free my queen's bishop. However, move 17 I play Be3? when the correct move is to capture the bishop and then play Be3.
| 17 |
Be3? |
Bxb2 |
| 18 |
Bxc5 |
Bxa1 |
Instead of giving the material back, I think 18.Rad1 is stronger.
| 19 |
Qxd6+ |
Kc8 |
| 20 |
Rxa1 |
Nxc4 |

20…Nc4 is incorrect due to 21.Qf6 Rd8 22.Rc1 N6e5 23.Be7 Qc6 24.Bxd8 Qxf6 25.Bxf6 Kb8 26.Bh8 g5 27.Nd2 leaving White with a bishop, knight, and rook vs two black knights.
Better for Black is 20…Rd8 21.Qf6 22.Qb7.
21…Rh7 tucks the Black rook safely away, but at the expense of removing it from the game. Given Black's pawn structure, what future does the rook have on h7?
For the remainder of the game, almost every move on both side is inaccurate.
I can't remember if this was a blitz game or not, but regardless the inaccuracies continue. 22. Rc1 is better, attacking the the c4 Black knight and ultimately threatening an absolute pin against the c6 Black knight.
I'm not sure what the rationale behind 22…a5 was, but it doesn't seem to accomplish anything other than losing the queen with 23.Re8+.
23.Nd4?? Why not the obvious 23.Re8+ winning the Queen for free when the king steps to d7? Note—if 23…Nd8 24.Qxd8+ Kb7 25.Qxa8+ losing both the queen and a knight. If 23…Kb7, attempting to pick up the rook after losing the queen, 24.Rxa8 Kxa8 25.Qxc6+ Kb8.
Better is 24.Nxc6 Nxc6 25.Re8+

And Black resigns after 24…g5, in light of 25.Qf5+ forking the king and rook.
→ Download PGN
Gregory Keener's original analysis:
The opening is well played by both sides, although Black is not quite able to equalize. On move 10, Black develops his king's bishop to g7 logically, preparing to castle kingside in keeping with general theory, however, overlooking that the Qe2+ causes him to lose his castle and chokes his position. Here, a queen exchange would have been advantageous to Black after playing Qe7, Qxe7, Kxe7, because although he would have lost his castle, the position would have been less volatile with the queens off the board. Once the king is pushed to d7 though, the castle is lost and all bets are off. Black was thinking here that the king's position was not weakened, and that after White's kingside castle, a pawn rush on the king side accompanied by pressure in the center from the bishop would be constricting for white, however, move 13 loses outright and makes the game an easy win for White, who now has not only the crushing material advantage, but the spatial and positional advantage to launch an attack as slowly and precisely as the position calls for.
Honestly, I wish I could say I learned something from this game, however, I already know that pushing the c pawn to the fifth rank loses the rook to the bishop parked on f3 after the logical bishop knight exchange in the modern variation. Ugh…one day, I will learn from my mistakes.
On a minor note, I have learned that 8. g5 generates sharp counterplay for Black in this variation, but loses to a specific response from White. If White doesn't respond appropriately though, the g5 push can create sharp counterplay for Black in the center after move 8. I'm too much of a wimp to risk it.
It's nice to note, though, that we now play chess well enough that a single mistake in the transition to middlegame leads to an instant victory. I would like to think that as a group, our chess was not this strong six months ago, and that such missteps as Kd7 as opposed to Qe7 wouldn’t have been dispositive in the outcome of the match.
Kudos Jeff…next one's on me.
—checkers is for tramps
Related tags: alekhines-defense, annotated-games