Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Practical Tablebases (much smaller) ?

Author: Uri Blass

Date: 23:42:43 11/14/01

Go up one level in this thread


On November 14, 2001 at 18:23:25, William Penn wrote:

>On November 13, 2001 at 10:09:30, Robert Hyatt wrote:
>
>>How would you handle all the common zugzwang positions?  black king at e6,
>>white king at e4, white pawn at e3.  White to move draws.  Black to move
>>loses.
>
>I would probably disregard them as being fairly infrequent in practical play
>between chess masters.

I disagree
zunzwnags in pawn endgames happen in games of chess masters.

I also do not see the problem of zunzwang because the program can probe the
tablebases only when it is the right side to move.

>
>>There are plenty of positions where underpromotion is the only way to avoid a
>>stalemate.  That would convert many wins into draws.
>
>Stalemate is quite rare in practical play between chess masters and can be
>disregarded.

I disagree that stalemate is rare but I believe that in most of these cases
promotion is not relevant.

>
>>7.5 gigs is a trivial amount of disk space today, with new machines usually
>>coming with at least a 60 gig drive.
>
>Well, my hard drive is 12GB, so the size of the TBs is a factor for me.
>
>Via simplifying assumptions based on the relative rarity of certain situations
>in practical play, I think the endgame tablebases could be reduced to no more
>than 1GB and still contain most of the strength which they presently offer to a
>chess program. In other words rather than an elegant and complete endgame
>solution, an incomplete (but sufficient) set of tables is probably do-able. But
>I don't know how to code it.
>WP

Nalimov did not do the tablebases for you.
He could reduce the time to generate tablebases by using the 50 move rule but
prefered to generate wrong tablebases that do not assume the 50 move rule
because the people who used them were not interested in the 50 move rule.

If you want to use incorrect smaller tablebases then you have to build them by
yourself but you can be sure that people are going to complain.

Old versions of Junior did not consider underpromotions in their search and
people complained so the new Junior knows about underpromotion.

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.