062. Computer Moves

Black: remyrey - thematic tournament, ChessWorld.net, 2020

“I don't think I'll try 8 Bd3 again” was my conclusion to Game 60. However, I was playing another game with it at the same time.

In this one I opted for 9 h4, followed by 9...h6 10 Qh5 Qf6 11 Nh7!?. As I said to my opponent: “I just played it because it looked so silly”. Of course it's a complete patzer's move. Black just takes the knight and then boxes in the bishop with ...g7-g6. Why would White allow anything like that?

Of course it's a computer move. 11 Nh7 is the type of startling, anti-intuitive idea very typical of an engine. And because we all have engines now, it's been seen several times before. White generally ends up with rook and two pawns for two minor pieces, which is okay, if no more than that.

Concrete computer analysis continues 11...Rxh7 12 Bxh7 g6 13 Qe2! Nf4 14 Qf3 Bb7 15 Nc3 c5 16 Ne4 Qg7, assessed as equal after any of 17 h5, 17 d4 and 17 d3.



And they've all been tried:

a) 17 h5 0-0-0 18 hxg6 fxg6 19 d3 Qxh7 20 Bxf4 exf4 21 Qxf4 was L.Simal Moreira-R.Cvak, correspondence 2017 (½-½, 34). Personally I'd rather have the two bishops here.

b) 17 d4 0-0-0 18 Bxf4 exf4 19 0-0-0 f5?! 20 dxc5! Rxd1+ 21 Rxd1 Bxe4 22 Qa3 was more promising for White in I.Popov-A.Moskalenko, Moscow 2018 (still ½-½, 74). I think Black's ...f7-f5 was mistimed; 17...f5 18 dxe5 fxe4 19 Qxf4 Qxh7 is preferable, and again the minor pieces seem superior to me.

c) 17 d3 f5 18 Qg3 0-0-0 19 Bxf4 exf4 20 Qxg6 fxe4 21 Qe6+ Kc7 22 Bxe4 Bc8 23 Qg6 Qxb2 was G.Legemaat-C.Steinert, correspondence 2018 (1-0, 43). The result is irrelevant, as Black abandoned the game in a favourable position.

I decided on line ‘c’, then deviating with 21 Qxg7. With the queens off White's extra rook has more chances to show its quality. Indeed, it was Black who chose to force the draw (½-½, 31). Still, a draw is still a draw.

I don't think I'll try 8 Bd3 again.