Friday, December 9, 2011

Engine Analysis of Positions from Sao Paulo

Aside from the fact that this tournament featured “Grandmasters Under Glass”…they played in a soundproofed glass room…Ivanchuk and his wife being robbed at gun point as they entered a taxi on their way back to the airport and Nakamura’s stupid loss to Vallejo, there were a couple of interesting games that were fun to check out with engines.
About Nakamura’s loss: He was a P up when Vallejo completed his 40th move with 5 seconds left. Nakamura asked the arbiter, “Is that move 40?” Of course the arbiter was not allowed to inform Nakamura one way or the other, so he said nothing. Nakamura thought the arbiter made an involuntary body movement in the form of a nod that Nakamura interpreted as a yes. So what did he do? He got up and went over to the refreshment table and calmly poured himself a glass of orange juice while his last 45 seconds ran out and he was declared forfeited on time. Nakamura became extremely upset and appealed, but it was quickly rejected.
This first position is from Aronian-Anand after 17.Qb3.  Anand played 17…Rab8 and lost in 25 moves.  It was noted in Chess Life that computer analysis preferred returning the P immediately with 17…Rad8, but that Anand probably feared reaching an ending with a bad B and wrecked P’s after 17…Rad8 18.Qxb7 Nd4 19.Nxd4 Bxd4 20.b3  but he overlooked the tactical shot 20…Bxf2+ which would have given him the advantage.  I set this position up and let the engines have a crack at it and this time they did better.

All of them, with one exception, chose 17…Rad8 with an evaluation of around 0.20 for White.  Spike 1.4 preferred Anand’s move and evaluated it 0.39.

The second position is after Black’s 24th move between Nakamura and Anand and was very interesting so I was curious to see how some different engines evaluated the position as compared to the evaluations of the GM’s.

3 comments:

  1. Hi! Do you have a FEN? I will try it with Hiarcs, Zappa, and Chiron to see what they say.

    Did I just write "say"? Talk about anthropomorphizing ;)

    ReplyDelete
  2. Well, they really are not "thinking" either but that is what we all call it! I only let the engines evaluate the position 2-3 minutes each so of course a longer analysis might have produced different results especially in this complicated position. Here’s the FEN

    3r4/1Q1nkp2/4p2p/1pq1b1p1/2p5/5BB1/1P3PPP/3R2K1 w - - 0 1

    ReplyDelete
  3. I guess thinking sounds more intelligent than calcuating!

    All three of those engines after 15 minutes each came up with the same Bc6. Most enthusiastic about that particular move was Hiarcs at +1.23

    ReplyDelete