Author: James Robertson
Date: 15:03:43 12/11/99
Go up one level in this thread
On December 11, 1999 at 15:32:54, Frederic Friedel wrote:
>>Fritz 6 placed a question mark after 29. ... Rxe3? giving this analysis:
>>29... Rxe3? {-#6 Fritz 6: 29. ... Bxe3 30. Kg2 Bc5 31. Kf1 Re2+ 32. Kh1 Rxb2 33.axb7 Rxb7 -+ -14.22/7}
>>
>>I don't understand why a move that mates in 6 (Rxe3) would be worse than a move
>>that merely provides 14.22 points (Bxe3). Can anyone help me figure out how to
>>give this game a correct blunder check?
>>
>I ran Fritz6 on your game with all kinds of blundercheck settings -- small
>threshold, large threshold, three seconds per move, 30 seconds per move, etc. It
>never disagrees with 29...Rxe3 and mate in six. In fact whatever the time it
>breezes through the last moves at high speed, seeing the mate at each move,
>until it reaches moves where the mate is not forced. Then it starts computing.
>
>So I can't duplicate your report. Can you try once more, just to make sure?
>Watch out that your hard disk is not running and you are getting 4 ply searches
>(hash tables too large).
I have tried it too many times with many different settings. It works correctly
when blunderchecking white, but I see this problem with black. It may not be
searching deeply enough, but if so it is not a hard disk swapping problem... I
know enough about chess programs to avoid something like that....
James
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.