Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The question is why is the same engine finding different moves?

Author: Ingo Bauer

Date: 10:17:14 02/08/05

Go up one level in this thread


On February 08, 2005 at 13:11:18, robert flesher wrote:

>Here is all learning and hash cleared.

I was not able to reproduce it with 128 in a reasonable time as well (think on
the other comp was some learning involved) but here it is with 256 MB Hash.

[D]3r1r2/pb3pkp/1pn2Np1/2p1PP2/8/b2BQ3/qP4PP/3R1R1K w - -

Engine: Shredder 9 UCI (256 MB)
by Stefan Meyer-Kahlen

11/27  0:01   +2.11    1.fxg6 fxg6 (743.752) 560

12/30  0:03   +2.11    1.fxg6 fxg6 (1.789.543) 553

13/33  0:11   +2.11    1.fxg6 fxg6 (6.036.281) 541

14/36  0:25   +2.11    1.fxg6 fxg6 (13.626.167) 538

14/45  1:31   +2.12++  1.Nxh7 (47.863.105) 522

14/45  1:35   +2.47++  1.Nxh7 (49.940.569) 522

14/46  5:39   +3.62    1.Nxh7 Bc8 2.Nxf8 Bxf5 3.Bxf5 Rxd1
                       4.Rxd1 Bxb2 5.Nd7 Qa4 6.Rf1 Qd4
                       7.Qg5 Nxe5 8.Qf6+ Kg8 9.Nxe5 Qxe5
                       10.Qd8+ (174.846.818) 515

best move: Nf6xh7 time: 5:43.484 min  n/s: 514.366  CPU 97.7%   n/s(1CPU):
526.474  nodes: 176.660.749

But even if it could not solve it, who cares. I know that in testpositions there
are allways some that are solved and some that arent. Thats the way it goes.
What is important is the "overall" score.

Bye
Ingo




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.