Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: a test position from hiarcs-tiger

Author: Uri Blass

Date: 09:20:23 09/27/00

Go up one level in this thread


On September 27, 2000 at 12:02:24, Vincent Diepeveen wrote:

>On September 27, 2000 at 10:49:13, Uri Blass wrote:
>
>>On September 27, 2000 at 10:39:33, Vincent Diepeveen wrote:
>><snipped>
>>>Which piece square tables?
>>>piece square tables won't help you much as they don't tell you the
>>>king is without protection at the other side of the board.
>>>
>>>Diep is evaluating this as +1.151 for black, it sees the mate in 1 in
>>>its qsearch however, so not even 1 ply needed to see mate in 1 here :)
>>
>>You are right that piece square tables do not tell that the king is without
>>protection but the point is that I believe that in most of the cases when the
>>king go to these squares it is without protection.
>>
>>I agree that it is better to see it by evaluation without piece square
>>tables(for example you can see that there are no pawn near the white king) but I
>>believe that if programs do not see it by better king safety evaluation then
>>they can at least use better piece square tables
>>
>>I saw that the position was the end of the main line of Yace and a better
>>evaluation could help Yace to avoid the mistake.
>>
>>Uri
>
>
>So you think yace isn't doing more as a piece square table evaluation?
>
>Programmer will be delighted hearing that i bet :)

I did not say that yace does not do more than piece square table.
I am not the programmer of yace and I do not know what yace do.
I said that I think that better piece square table without changing the
evaluation can help.
I did not say that there is no better solution.

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.