Beating the Italian

Last night, I finished with a perfect 3-0 to win my section of the Turkey Quads. I had not played in this event since 2011, and then it started a nice run of wins (see "Eleven Consecutive Wins!"). That run propelled me to my lifetime peak USCF rating of 1982, so this year's success in the Turkey Quads could be a good omen. However, I don't believe in omens. I do believe in training. My approach to chess training has been a little more serious since the Eastern Washington Open, my fourth consecutive tournament that led to a rating drop.

In reply to Todd Bryant's challenge (see comments on "Good Blitz, Bad Blitz"), these annotations were produced without reference to databases and engine analysis, except for limited use of the Chess Openings app on my iPhone during the postgame analysis, confirming that my sixth move was unusual.

van Heemstede Obelt,Walter (1600) -- Stripes,James (1750) [C50]
Turkey Quads Spokane (3), 17.11.2016

1.e4 e5

Everyone in Spokane knows that I play the French, but I don't always.

2.Nf3 Nc6

I considered 2...Nf6 for about five seconds, as I'm playing the Petroff in a correspondence game right now after having won in twenty moves on the White side against the same opponent.

3.Bc4 Bc5

3...Nf6 is too risky for me with the clock running. I believe that Black is objectively fine against the Fried Liver Attack, but that in practical play White has a clear edge.

4.d3 Nf6

I thought that I would castle next move.

I considered 4...h6, but did not seriously calculate 5.Nxe5 Nxe5 6.d4 Bd6

 (6...Bxd4 7.Qxd4 Nxc4 8.Qxg7 Qf6 9.Qg3)

7.dxe5 Bxe5 8.Qh5 Qf6

5.Bg5 h6

I remembered some Paul Morphy games where he delayed castling and threw his pawns and better coordinated pieces at the enemy king.

6.Bh4 Qe7?

This move appears to be unusual, or at least my opponent thought so after the game, and the Chess Openings app on my phone confirmed his hunch.

a) 6...d6 seems simple enough.
b) 6...Nd4? invites 7.Bxf7+ Kf8 (7...Kxf7 8.Nxe5+) 8.Nxe5 with a clear advantage for White.
c) 6...g5 7.Nxg5 frightened me 7...hxg5 8.Bxg5.

7.Nc3

And now I thought for ten minutes, feeling that I had given up a little too much already. How am I going to get a playable game? I decided that I needed to push my kingside pawns and be prepared to castle long. Would an open b-file prove useful to my opponent? I decided that he could not organize forces there fast enough to balance what I hoped to accomplish on the kingside.

Black to move

7...Bb4

7...g5 looks even worse than before 8.Nxg5 hxg5 9.Bxg5 with the idea 10.Nd5.

8.a3 Bxc3+ 9.bxc3 Nd8?!

The knight is headed to f4, but maybe 9...d6 is better.

10.Nd2

10.d4 challenged Black's effort to reposition his knight. 10...d6

(10...exd4 11.cxd4 If my students were shown this position and told it was a Morphy game, they would say that Morphy has White. Black is clearly underdeveloped, as were many of Morphy's opponents.)

11.d5 Bg4 12.Qd3 holds little promise for Black.

10...Ne6 11.0–0

11.Bxe6 Qxe6 12.0–0 g5 13.Bg3 d6 White's pieces are playing; Black's are watching.

Black to move

11...d6

I was thinking that either White's f-pawn of his or h-pawn would move after I drove the bishop back. I wanted to be able to sacrifice my bishop on h3 if he pushed that pawn.

12.Bb3 Nf4 13.Nc4 g5 14.Bg3 h5 15.f3

I did not work out all the details after 15.h3 but thought that 15...Bxh3 was at least worth considering. 16.gxh3 Nxh3+ 17.Kh1

(17.Kh2 h4)

17...h4 18.Bh2 g4 19.Ne3

(19.Rg1 Nxf2+)

19...g3 20.Bg1 g2+ 21.Kxg2 Nf4+ 22.Kh1 Qd7.

15...h4 16.Be1

16.Bf2 g4 17.Be3

(17.Ne3 helps me 17...g3 18.Be1 gxh2+ 19.Kh1 (19.Kxh2 h3 20.g3 Ng2 21.Qe2 Nh5 22.Nxg2 hxg2 23.Kxg2 Bh3+ 24.Kxh3 Nf4+ is obviously not forced at every turn, but serves to reveal that Black can temporarily sacrifice material and earn a return on the investment)

19...h3 20.g3 N6h5 21.gxf4 exf4 also looks good for Black.

16...g4 17.Ne3

Here I again thought for a long time. I observed that I had not yet sacrificed anything and decided to give my opponent something else to think about in addition to my kingside attack.

17...c6

Walter thought for a long time here, and while he was thinking I discovered that my last move sacrificed a pawn. Then, I began looking for ways to get some use from the sacrifice.

I considered 17...g3 but was not convinced that I had anything after 18.h3 Nxh3+ 19.gxh3 Bxh3 looks good for Black, but this position was not clear in my head during the game.

18.fxg4

Black to move

18...h3 19.Rxf4 

This move caught me by surprise, and I think it was a mistake.

I expected 19.g3 Ng2 20.Nxg2 hxg2 21.Kxg2?

(After the game, we looked at 21.Rf2 Nxg4 22.Rxf7 concluding that White was winning)

21...Bxg4 22.Qd2 Bh3+ and Black seems better.

19...exf4 20.Nf5

I'm looking for the trap--White's reason for sacrificing the exchange, and do not see it.

20...Bxf5 21.exf5 0–0–0 22.Bf2

With this move, White's idea becomes clear.

Black to move

22...Rdg8 

I think that I have a clear advantage now.

23.Bd4 Rxg4 24.Qxg4

I saw this possibility, but was happy that if my opponent went for the two rooks for the queen, I would have a smother mate.

24.Bxf6 Rxg2+ 25.Kf1 Qxf6 seemed the alternative.

24...Nxg4 25.Bxh8 Qe3+

Now I discovered to my horror that my own pawn on h3 prevents the smother mate. Clearly, even a win reveals plenty of room for improving my play when I overlook such obvious problems with my plan. Pattern recognition in the absence of concrete analysis is a worthless skill.

26.Kh1

Black to move

hxg2+

I thought that I still had a checkmate by force here, and do, but not the checkmate that I imagined.

26...Nf2+! 27.Kg1 Somehow I missed 27...Nd1+ 28.Kh1 Qe1#.

27.Kxg2 Qf2+ 28.Kh3 Nxh2

28...Qxh2+, which I had planned, loses the knight.

0–1

Walter resigned as checkmate can be delayed, but not prevented. 29.Be6+ fxe6 30.Rg1 Qxg1 31.Kh4 Qg4#.