Playing over lost games, or games where I played poorly, provides plenty of lessons for my chess improvement. The following game has a shelf-full of lessons for both me and my opponent - especially when it comes to middle game play.
perrypawnpusher - genericme
blitz, FICS, 2015
1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Nc3 Bc5 5.Bxf7+
The Italian Four Knights Jerome Gambit. I have 57 examples of this position in my games in The Database, scoring 73%. (That seems pretty good, but it's below the 81% I have scored with the regular Jerome Gambit; the 89% with the Semi-Italian Jerome Gambit; and the 92% with the Blackburne Shilling Jerome Gambit.)
5...Kxf7 6.Nxe5+ Nxe5 7.d4 Nd3+
This is a creative idea from my opponent, and a TN according to The Database. It is the kind of "shocking" move that can be found in blitz games, and I admit that I was expecting the regular (and stronger) 7...Bxd4.
Black moves the threatened Bishop, and puts pressure on White's e-pawn. Ceteris paribus, such an idea should work - but "all things" are not "equal" here, as White's sacrifice of a Bishop at f7 has moved Black's King. Concrete analysis is important, even in blitz.
After 8...Bb6 9.e5 Ne8 10.O-O d6 11.f4 White could feel that his "Jerome pawns" would balance out Black's extra piece.
Playing over this game afterward, I am sure my opponent will decide that 9...d5 10.Qxb4 Nxe4 would have been better than exposing his King further.
I had expected 10...Nxe4 11.Nxe4 Re8 when I planned on castling, although now, after the game, it seems that 12.f3 d5 13.O-O dxe4 14.Bf4 is the way to go.
After the game, I looked at 11...Nxe4 with Stockfish, which recommended for White something I had not considered: 12.d5!?
The strength of the move - among other things, threatening to win the Knight at e4 - slowly becomes apparent.
For example, Black cannot simply hope to exchange with 12...Nxc3, as 13.Qg4+ Kf7 14.Bg5 would allow White to win the exchange (as well as recapture on c3).
Protecting the Knight with 12...Qe7 seems best, but, again, after 13.Qxe7 Rxe7 14.Re1 Nxc3 15.Rxe7 Nxd5 16.Re5 Nf6 White has won the exchange.
And if Black simply retreats the Knight with 12...Nf6, Stockfish recommends the consistent 13.d6!?
It will be a long while before I play chess like that, though!
I am afraid that me playing the Jerome Gambit (giving "Jerome Gambit odds") seemed to have convinced my opponent that I was even less skilled than I actually am. Here he goes ahead with an attack that I had already prevented with my last move. Instead, 12...d6 would have been better, but White would have the advantage.
13.Bg3 Qh5 14.f3
The text move consolidates White's position - which was the psychological need I had at the time - while Stockfish, later, suggested that 14.Nd5!? would have deeply troubled Black's. The closer you look, the stronger it appears.
This is about as good as any other move. Black's game unravels from here, as he seems to give up hope.
15.Rf2 Nxc2 16.Rxc2 a5 17.Qc4 d6 18.Nd5 Be6 19.Nf4+ Kg5 20.Nxe6+ Kg6
Everything leads to material loss and checkmate.
21.Nf4+ Kg5 22.Nxh5 Black resigned.