Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Crafty Classic 2005

Author: Uri Blass

Date: 04:52:51 08/28/05

Go up one level in this thread


On August 28, 2005 at 04:50:35, Harald Lüßen wrote:

>On August 28, 2005 Uri Blass wrote:
>
>>Fritz gui for some reason do not understand movei's mate score
>>and when movei writes 32764 it translates it to mate in 1 and
>>when it writes 32766 it translate it to mate in 0.
>>
>>It has no problem to understand movei's score when it writes mate
>>against itself.
>>
>>I wonder if there is some bug in the chessbase gui that translate
>>score of winboard engines to mate score wrong.
>
>How do GUIs detect mate scores if there is no fixed format?
>Do they look at the score and guess the mate range like this:
>
>    0.. 9000: normal
> 9001..10000: mate is 10000, mate in one move is 9999 and 9998, ...
>10001..15000: normal
>15001..16000: mate is 16000, mate in one move is 15999 and 15998, ...
>16001..16384: mate is 16384, mate in one move is 16383 and 16382, ...
>16385..31000: normal
>31001..32000: mate is 32000, mate in one move is 31999 and 31998, ...
>32001..32767: mate is 32767, mate in one move is 32766 and 32765, ...
>Or even string output of the engines:
>M0: is mate
>M1: is mate in one move or ply
>M2: is mate in one move or mate in two plies?
>...
>other?
>And negative values mirrored. (-M1?)
>
>What if the GUI sees a score of 16380 and later a score of 20000
>and then 32760?
>
>Are there definitions in the Winboard protocol or in the UCI
>protocoll? Are there differences?
>
>Harald

I think that they should use the definition that 32767 is mate score
32767-(2n-1) is mate in n for the side to move(distance of 2n-1 plies to mate)
2n-32767 is mate in n against the side to move(distance of 2n plies from mate).

Uri



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.