26 March 2012

Premature Draw

Fabror the Guru posted a game played during a chess exhibition put on by his local club (see "Simul at the Office"). He states, "[w]e agreed to a premature draw" and notes that White appears better in the final position. I concur. It seems to me that White's pieces are better coordinated, and despite superficial appearances, White's king is less vulnerable.

White to move
After 23...Rxb2 Draw agreed
In order to test my conviction that not only did White have an advantage, but a substantial one, I set the position up and played against Rybka 4 at game 25 time control. With none of my frequent take-backs during such exercises, I was able to provoke Black's resignation after less than fifteen minutes of play. The machine used two-thirds of this time thinking while I used the other one-third. Half of my time was spent on a single move.

I began with a move that looked good to me while playing through the game on Fabror's blog, and that was discussed there.

24.h6 Rb8

I expected 24...g6

25.hxg7 Kxg7 26.Bd3

Rybka expected 26.Rh4. Running Hiarcs 12 as I annotate this exercise, I find that it prefers my move.

Black to move

26...Rg8 27.g4 Kf8 28.Bf5

Now that the pony on a2 is undefended, I'm looking around for a lasso.

28...Bxf5 29.gxf5 Rg2+ 30.Kd3 Rg3+

White to move

31.Kc4

Rybka expected 31.Kd4. Hiracs concurs that this move is best until it thinks longer, then it chooses my move. I used a mere three seconds to play my move. Several moves later, there was a moment where I perceived a checkmate threat that I needed to avoid.

31...h6 32.Rxh6 Kg8

White to move

33.Ne8!

I love it when the computer gives an evaluation of the position that then takes a jump up several pawns after I make an unexpected move. Rybka had +4.49 before 33.Ne8, but +9.66 afterwards. Hiarcs found this move after several minutes, and its evaluation jumped from ~+5.00 to ~+7.50 and continued to climb while I typed this comment. I spent twenty seconds finding the move. Of course, I do most of my thinking while the engine thinks, and 32...Kg8 was easy to predict.

33... Kf8 34.Nf6

I spent 2:12 on 34.Nf6 because the second best move was tempting. Here I saw the nasty disappointment that could have occurred. 34.Rh8+ Ke7 35.f6+ Ke6 36.Ng8+ Kxe4 37.Ra8?? Rc3#.

34...Rc3+ 35.Kd4 c5+ 36.Kd5 1–0

Black to move
Black gave up

Rybka 4 resigned. I foresee hiding my king behind Black's c-pawn where he helps stitch together a mating net for the enemy king. It's always nice to humble the silicon beast, even from a set-position that is theoretically winning.

2 comments:

  1. Replies
    1. Glad you like it Patrik. Interesting exercises come from interesting positions, and I had some good raw material to work with in this case.

      Delete