Author: Christophe Theron
Date: 10:34:31 11/03/03
Go up one level in this thread
On November 03, 2003 at 09:33:53, Tord Romstad wrote:
>On November 02, 2003 at 15:26:06, Christophe Theron wrote:
>
>>In the same vein, the following position has always been a nightmare for Tiger:
>>
>>[D]8/1KP5/3q2k1/8/6p1/8/8/8 b - -
>>
>>I am interested in results of other (amateur and commercial) programs.
>
>My engine plays Qxc7 after 6 plies and 0.29 seconds on an Athlon 2.0 GHz. The
>reason it likes Qxc7 is that I have a really big bonus for an unblocked passed
>pawn on the 7th when the SEE thinks the pawn can safely advance.
I'm not sure it is the reason.
The bonus for the pawn can be symetrically added to both pawns. Actually Tiger
knows in evaluation that both pawns are unstoppable.
So their respective bonuses almost anihilate each other and it all boils down to
evaluating KPKQP vs KKQ correctly (KKQ being better).
The problem is that Tiger knows that after c8=Q Qxc8 the black pawn is also
unstoppable, and has almost the value of a Queen.
But the resulting ending should not be evaluated as KKP with value(P)~value(Q).
It should be evaluated to KKQ, and that's not the same thing.
Consistency between endgames evaluations is really a nightmare... :(
Christophe
>
> 1 8.31 0.00 272 Qd7
> 2 8.31 0.01 1593 Qd5+ Ka7 Qf7
> 3 8.31 0.03 5504 Qb4+ Kc6 Qc4+ Kd6 Qd4+ Kc6
> 4 8.30 0.09 17339 Qd5+ Kb8 Qb5+ Ka7 Qc5+ Kb8 Qb6+ Kc8 g3
> 5 8.31 0.20 48370 Qd5+ Kb8 Qb5+ Ka7 Qc5+ Kb7 Qb5+ Ka7
> 6 >= 8.31 0.21 49279 Qd5+ Kb8 Qb5+ Ka7 Qc5+ Kb7 Qb5+ Ka7
> 6 >= 10.17 0.29 75878 Qxc7+ Kxc7 g3 Kc8 g2 Kc7 g1Q
> 6 <= 10.22 0.30 82279 Qxc7+ Kxc7 g3 Kc8 g2 Kc7 g1Q Kb7 Qh1+
> 6 <= 10.19 0.30 84595 Qxc7+ Kxc7 g3 Kc8 g2 Kb7 g1Q Ka6 Qf1+
> 6 <= 10.17 0.31 86984 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Qb1+
> 6 10.17 0.31 86984 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Qb1+
> 7 >= 10.17 0.31 87530 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Qb1+
> 7 <= 10.17 0.36 109634 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Qb1+ Ka5
> Qg1
> 7 10.17 0.36 109634 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Qb1+ Ka5
> Qg1
> 8 >= 10.17 0.36 110604 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Qb1+ Ka5
> Qg1
> 8 >= 10.19 0.36 112016 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5
> 8 >= 10.25 0.36 112798 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5
> 8 <= 10.36 0.46 157766 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5 Kc6
> Qh1+
> 8 <= 10.34 0.54 208738 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5 Kc6
> Qh1+
> 8 <= 10.30 0.63 262980 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5 Kc6
> Qh1+
> 8 <= 10.25 0.73 320287 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5 Ka5
> Qe1+
> 8 10.25 0.73 320287 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5 Ka5
> Qe1+
> 9 >= 10.25 0.73 321574 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5 Ka5
> Qe1+
> 9 <= 10.25 1.07 476030 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5 Ka5
> Qe1+ Kb5 Qg1
> 9 10.25 1.07 476030 Qxc7+ Kxc7 g3 Kd7 g2 Kc6 g1Q Kb5 Kf5 Ka5
> Qe1+ Kb5 Qg1
>
>Tord
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.