Author: Mike S.
Date: 09:39:46 03/05/00
Go up one level in this thread
On March 05, 2000 at 11:32:07, John A. Gonzales wrote:
[D]6k1/2p2r1p/2PrQ3/P7/8/4K3/1P5P/8 w - - 0 1
>
>In the above position, with white to move, I'm wondering if any of the top
>programs give Qxd6 as winning (...)
I've tested some ChessBase engines, none of which chooses Qxd6 within 1 minute
on K6-2/400. Relatively best was Little Goliath 2000a; after entering Qxd6
manually, it came up fast with a big white advantage in the evaluation ("T" for
the Rooks here):
Analysis by LiGo2000a: (K6-2/400 10 MB Hash)
1...cxd6 2.b4 Tc7
² (0.45) Tiefe: 6 00:00:00 1kN
(...)
1...cxd6 2.b4 Ta7 3.b5 Tc7 4.a6 Tc8
+- (1.97) Tiefe: 8 00:00:00 10kN
1...cxd6 2.b4 Ta7 3.b5 Tc7 4.a6 Tc8
+- (1.97) Tiefe: 9 00:00:00 81kN
1...cxd6 2.b4 Tf5 3.c7 Tf8 4.b5 d5 5.b6 Tc8 6.b7
+- (3.15) Tiefe: 9 00:00:05 81kN
1...cxd6 2.b4 Tf5 3.c7 Tf8 4.b5 d5 5.b6 Tc8 6.b7
+- (3.15) Tiefe: 10 00:00:05 1323kN
1...cxd6 2.b4 Tf5 3.c7 Tf8 4.b5 d5 5.b6 Tc8 6.b7
+- (2.72) Tiefe: 11 00:00:14 4206kN
1...cxd6 2.b4 Tc7 3.b5 Kf7 4.Kd4 Ke7 5.b6 Txc6 6.b7 Ta6 7.b8D Txa5
+- (3.74) Tiefe: 11 00:00:42 4206kN
1...cxd6 2.b4 Tc7 3.b5 Kf7 4.Kd4 Ke7 5.b6 Txc6 6.b7 Ta6 7.b8D Txa5
+- (3.74) Tiefe: 12 00:00:43 14949kN
Shredder 3, usually one of the best endgamers, takes more time to evaluate that
high for white. CSTal II was better here, -3.x after 4 seconds (on the black
move).
The prog's feel fine with the queen (around +2.x), but I think due to the
remaining presence of the queen on the board, they may not be able to reach the
necessary depth to sacrifice it. It's often a problem in endgame positions; the
queen has just too much possible moves, and the selective methods sometimes
can't filter enough out as it seems.
Regards,
M.Scheidl
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.