Author: Bernhard Bauer
Date: 03:46:38 05/23/01
Go up one level in this thread
It's interesting to see the effords on these types of positions. So now
something is done with the problems of the null-move heuristic. Glad to see it!
Here is a similar position from a game of L.Szabo.
[D]3r1Bk1/pp1q2b1/6pp/PP1Np2n/2Pp1p2/3P2P1/3QPPbP/R4RK1 b
And here some output from crafty.
9-> 5.30 1.21 1. ... Bxf1 2. Bxg7 Kxg7 3. Rxf1 Nf6
4. Nb4 Qh3 5. gxf4 Qg4+ 6. Kh1 Qxf4
7. Qc2
10 8.50 1.08 1. ... Bxf1 2. Bxg7 Kxg7 3. Rxf1 Nf6
4. Nxf6 Kxf6 5. a6 b6 6. f3 Kf5 7.
g4+ Ke6
10-> 11.36 1.08 1. ... Bxf1 2. Bxg7 Kxg7 3. Rxf1 Nf6
4. Nxf6 Kxf6 5. a6 b6 6. f3 Kf5 7.
g4+ Ke6
11 16.30 1.08 1. ... Bxf1 2. Bxg7 Kxg7 3. Rxf1 Nf6
4. Nxf6 Kxf6 5. a6 b6 6. f3 Qf5 7.
g4 Qg5
11-> 24.68 1.08 1. ... Bxf1 2. Bxg7 Kxg7 3. Rxf1 Nf6
4. Nxf6 Kxf6 5. a6 b6 6. f3 Qf5 7.
g4 Qg5
12 38.35 1.18 1. ... Bxf1 2. Bxg7 Kxg7 3. Rxf1 Nf6
4. Nxf6 Kxf6 5. a6 b6 6. f3 fxg3 7.
Qxh6 gxh2+ 8. Kf2 e4 9. Qxh2 exd3 10.
exd3
12 1:36 1.17 1. ... Bxd5 2. Bxg7 Kxg7 3. cxd5 Qxb5
4. Rfc1 Qxd5 5. Rc4 Rd6 6. Rb1 Kh7
7. gxf4 exf4 8. e4 fxe3 9. fxe3 dxe3
10. Qxe3
12 9:14 ++ 1. ... Qh3!!
12 16:43 0.28 1. ... Qh3 2. Bxg7 f3 3. exf3 Bxf3
4. Ne3 Nxg7 5. a6 b6 6. c5 dxe3 7.
fxe3 Be4 8. c6 Rxd3
(2) 12-> 17:47 0.28 1. ... Qh3 2. Bxg7 f3 3. exf3 Bxf3
4. Ne3 Nxg7 5. a6 b6 6. c5 dxe3 7.
fxe3 Be4 8. c6 Rxd3
13 21:34 -0.06 1. ... Qh3 2. Bxg7 f3 3. exf3 Bxf3
4. Ne3 Nxg7 5. a6 b6 6. c5 dxe3 7.
fxe3 Rxd3 8. Qf2 e4 9. cxb6
13-> 22:15 -0.06 1. ... Qh3 2. Bxg7 f3 3. exf3 Bxf3
4. Ne3 Nxg7 5. a6 b6 6. c5 dxe3 7.
fxe3 Rxd3 8. Qf2 e4 9. cxb6
time=30:01 cpu=200% mat=3 n=687077289 fh=87% nps=381k
ext-> chk=30366598 cap=3601877 pp=3366190 1rep=4730562 mate=744380
predicted=0 nodes=687077289 evals=166517765
endgame tablebase-> probes done=0 successful=0
hashing-> trans/ref=0% pawn=1% used=99%
SMP-> split=648 stop=110 data=8/32 cpu=60:02 elap=30:01
Kind regards
Bernhard
On May 22, 2001 at 22:01:03, Dieter Buerssner wrote:
>On May 22, 2001 at 21:10:59, Frank Quisinsky wrote:
>
>>2 interesting positions from Leiden:
>>
>>Leiden: Yace - Patzer
>>3r1r1k/ppp1q1pp/4p3/2P1Pp2/3Pp3/PQ3bP1/1P2NP1P/3RR1K1 b - -
>>
>>Patzer won the game with a very strong attack move.
>
>I agree, that Patzer played this game extremely well. Here it moved f4. I think,
>after this move, this is a forced win. I tried many variations, to come to this
>conclusion. I believe, the game was lost for Yace even earlier, although I am
>not sure yet, what the decisive mistake was.
>
>Anyway, I think Patzer played extremely well in this game, and Yace was very
>naive. It toatall underestimated the danger of the bishop on f3. Also I think,
>this position can be a real problem for engines, that use nullmove. There are
>many mate threats, but they are introduced by more or less silent moves. This is
>not a total failure of nullmove using engines, but it can make them use much
>more depth, than one would assume, to see the real threat.
>
>I already made quite a few changes to Yace, that all were inspired by this game.
>
>I really would be interested in the scores, that Patzer showed in this move.
>Perhaps you can sen me those in email.
>
>My congratulation goes to Roland. This really was the fastest defeat of Yace in
>any game at a long time control.
>
>Regards,
>Dieter
>
>P.S. On my computer, Yace finds f4 in one minute.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.