17...Nxe4 18.f6
As I wrote earlier, now "it will be interesting to see which one of us unveils his 'improvement' on the play first."
18...Nxf6!?
My opponent takes the pawn, with a move not seen in any of the games in The Database, in effect saying "Show me!".
I think the text move leads to an edge for White, despite Black being up a piece, but it is necessary to prove that - and my opponent had a defensive plan in mind.
19.Rf4 Kg7 20.Raf1 Be6 21.Qg3+
21...Ng4
A scientific idea - the proper way to deal with a gambit is to take the material, and then return it at a time when it will intefere with the attack.
22.hxg4
Later, I became curious and asked Stockfish what I should have decided upon, here. It cranked out 22.d5!?.
I could see some of that - capturing the pawn with 22...Bxd5 would allow White to swing a Rook over to the g-file, doubling with the Queen, i.e. 23.Rxg4+. It turns out that, then, White would have a mate in 30-something moves, but I would never have figured that out.
But, what if Black did what he sometimes does, protect his King while giving White the choice of which piece to grab back - as with 22...Kh8!? ? Stockfish was confident that after 23.dxe6 Ne5 White would still be better (almost 3 pawns better, according to the computer), but even now I can not become excited by the position.
All in all, I am happy with having just captured the Knight.
22...Qg5
Black's idea. It falls victim to my initiative, however.
23.Ne4 Qd5 24.Nf6 Qg5
25.Nxe8+
Later, I discovered that this was not Stockfish's choice, either. It liked 25.Re4!?, with White holding the advantage of almost a Queen. I still don't understand that.
25...Rxe8 26.Qc3 Kg8 27.Qxc7
Okay, White is the exchange and a pawn better. A win, right?
"Won" games don't win themselves.
27...Qe7 28.Qxe7+ Rxe7 29.Re1
Wrong Rook. I had better with 29.Re4 d5 30.Re5 Re8 31.Rf6!? when, after an exchange of Rooks, Black's pawns are too weak to save from attack.
29...Rf7 30.Rxf7 Kxf7
Black's King and Bishop are more active than they should have been allowed to be.
[to be continued]
No comments:
Post a Comment