Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: King safety evaluation

Author: Robert Hyatt

Date: 21:45:47 05/30/98

Go up one level in this thread


On May 31, 1998 at 00:33:02, Stuart Cracraft wrote:

>
>On May 30, 1998 at 16:21:30, Peter Fendrich wrote:
>
>>On May 30, 1998 at 10:03:21, Robert Hyatt wrote:
>>
>>- snip -
>>
>>>I'm not a big believer in mobility, based on years of experimentation
>>>with
>>>it.  It is not clear to me whether mobility is a *cause* of winning a
>>>game
>>>or a *result* of winning. You do *not* want to evaluate "result" terms,
>>>or
>>>else you'd see something like "in the last 10 games where Brett Favre
>>>was
>>>hurt in the 4th quarter of an NFL game, Green Bay *won*."  We need to
>>>win
>>>this game badly, so you linemen don't block and let's get Brett hurt so
>>>we
>>>can win.  Logic is wrong of course...  one easy example in chess is
>>>moves like
>>>a4-a5 or h4-h5, which increase the mobility of the rook, but which also
>>>disrupt or weaken the pawn that moved,plus the squares that it helps
>>>defend...
>>>
>>>I think that you have to rely on search for some of this, and on
>>>evaluation
>>>for the parts that you can clearly evaluate...
>>
>>I don't use mobility either, but have been thinking of using 'lack of
>>mobility'.
>>When mobility is below some certain treshold the program starts to
>>penalize for lack of mobility to prevent it from going into cramped
>>positions.
>>That would be more useful than just trying to get more space in general.
>>Haven't tried it yet, however.
>>
>>//Peter
>
>Shouldn't mobility be considered as part of centrality and king-attack?
>
>That is, what's the worth of controlling a few desirable squares in the
>center, when my opponent has multiple squares controlled by multiple
>pieces right next to my king!
>
>Seems to me you could cover mobility, centrality, and king attacks in
>just one call testing the attacks of all pieces on the boards against
>a pc/sq table that is initialized for each terminal node. Filling this
>pc/sq
>table should not be expensive at all, even at terminal nodes.
>
>It seems important to me to understand space, control, and attack with
>a calculation that rewards all.
>
>This was all covered by David Levy in The Joy of Computer Chess,
>pp. 18-22.
>
>--Stuart


I hate to say this, but check out how well his programs played over the
years...  then you might take some of his ideas with "a grain of salt."

and the expensive part of detecting kingside attacks is *not* the first
piece that attacks, it is the second or third piece lined up in battery.



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.