Big Pawn: It seems like black would have an easier game to play after 9...Na5. His path to after 9...de is not as clear, in my view. After 9...de it seems that White has some initiative and black has trouble finishing his development.
10. Bb5 pins the Nc6 and threatens to win a pawn on e5. This enticed black to play 10...f6.
Now what?
11. Nc4 is a good move for white because it allows him to develop his queenside and puts the N on a good square, attacking the Bb6. It also opens the d-file for a rook to attack blacks queen with tempo later.
11...Nge7 black hurries to finish his development.
12. O-0 a good, natural move. Now black has to find a safe place for his king. Should he castle queenside? He tries to.
12...Qe6 making room for the LSB. Black has been behind since he lost a tempo defending his e5 pawn.
13. Qa4 putting more pressure on black's q-side. This pins the a-pawn to the Ra8, gets it off the diagonal with the black queen (avoiding exchanges) and adds more pressure to the pinned Nc6. In response, black can only hope to finish developing.
13...Bd7 striving to finish his development and run away with the extra pawn.
14. Nxb6 forcing black to capture away from the center. This is the result of putting the Q on a4. This capture is also a forcing move which doesn't allow black to finish development.
14...cb
White has some compensation for the pawn in better development and the fact that black has somewhat ruined his pawn structure on the last move.
How can white keep black from developing and consolidating his game?
15. Ba3?! Na5 looking to relieve the pressure. This seems to be a good move.
It seems as though 15. Ba3 lets black off the hook. A better way to keep up pressure may be 15. Rd1 0-0-0 16. Bc4 Qg4 17. Qc2 Kb8 18. Ba3. The computer says black is better by half a pawn so white has compensation. It's unclear and the battle rages on.
16. Rd1 this is a natural move but it seems as though black can castle long here and consolidate. White can also try 16. c4 but black consolidates after Nec6.
16...Bxb5?! trading some attacking pieces with a view to relieving the pressure but it doesn't work out so well. Black might have tried 16...0-0-0 17. c4 Bc6 and although white enjoys open files to the black king, black seems safe.
17. Qxb5+ Kf7
Black tries to connect his rooks since blocking the Qb5+ with either of the three pieces still wouldn't allow castling. Even a player near 2800 can miss the right path in the Evans Gambit. He should have castled long instead of trading bishops.
Kibitzing
