Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: likelihood instead of pawnunits? + chess knowledge

Author: James Swafford

Date: 11:17:49 10/25/02

Go up one level in this thread


On October 25, 2002 at 13:15:03, Sune Fischer wrote:

>On October 25, 2002 at 12:39:38, Ingo Lindam wrote:
>
>>I would really like to see the computers measure a position rather in a
>>set of probabilities e.g. (P+,P=), where
>>
>>P+ = Probability in the position to evaluate white/player to move will
>>win and
>>P= = Probability that position will end in a draw
>>P- = Probability in the position to evaluate white/player to move will
>>lose
>>
>>with P+ + P= + P- = 1
>>
>>(also a confidation measure about the Probabilities might be useful)
>
>The pawn scale is equivalent so it doesn't matter, you can see the mapping in
>eg. the article on TD(lambda) for the KnightCap engine (I think Dann has it on
>his ftp).
>

It is a hyperbolic tangent function, so the probabilities are in
the closed interval [-1..1].  So small changes in the "pawn units"
evaluation yield a more dramatic shift in the probablility if the
score is drawish (vice if you're already losing or winning by a queen).

>I prefer the pawn scale since that is in integers, also the mate scores are
>easier to work with I think.
>One could print out the score in probability terms, but there is no tradition
>for that either in human chess or computer chess.

I agree, but only because I'm so used to looking at the pawn scale that
I tend to try to "reverse map" a probability back to the pawn scale.

>
>>Ofcourse out of the set of probabilities a single measure could obtained
>>to be optimization criteria in an search algorithm. A simple one would
>>be P+ + 1/2P=, but also different formulas considering strength of
>>opponent, standing of the match or just an increasing influence of P=
>>when position is weak might be interesting.
>>
>>Even more important seems to me to demysticize terms like "chess
>>knowledge", "experience", "plans", "positional criteria".
>>
>>There is such a huge amount of chess games and analysis in a computer
>>readable/usable format and what else should be a source of chess
>>knowledge than games and results? Yes, there are books and ideas of
>>great human chess thinkers as Nimzowitsch. But also his ideas are
>>experiences from his own analysis and games and should also be
>>verifyable by modern pratical chess. And where not, they might be no
>>longer of any use.
>>
>>A chess engine that is able to calculate 3 Million positions per second
>>should have no problems with dealing with less than 2 Million. As more
>>as a lot of conclusions out of the "experience" of 2 Million chess games
>>may be drawn rather in preperation of a match than during a game.
>>
>>"Positional pattern" (another mysticized term reserved for human beings
>>especially GMs) may easily formulated and efficiently retrieved on the
>>basis of low level chess position items and clusters of those. Computer
>>scientists may argue that there is a too huge amount of possible
>>patterns. But a chess engine as well as a GM (not less a normal human
>>chess player) should first of all be interested in patterns that often
>>apear in practical chess.
>>
>>I expect that a CD (or DVD) full of positional chess patterns drawn out
>>of a suitable number and choice of chess games (out of a permanently
>>growing number) will have a much greater effect on the play and results
>>of a chess knowledge using chess engine than 4 or 5  pieces tablebases
>>have nowadays on the results of tablebases using chess engines.
>
>Patterns are used in chessprograms, connected passed pawns or rook on open file
>for instance. Consider that computers are too small and slow to be using a
>neural net with 100 billion neurons for the entire postion like humans. Besides
>how would you train it? You need scores on every position, which you do not
>have.

Agreed about the size of the neural net, though that may not be true
forever.  I think you should be able to derive a score from game outcomes.


>
>For "probing" a CD you first need to design some clever index scheme and if you
>plan to use it at every node it's going to slow your program to a crawl (of
>course that wouldn't matter much if it was extremely accurate).

True, but ... :)  computer scientists are often not interested in
hardware limitations.  Instead they concern themselves with what
is theoretically possible.





>
>-S.



This page took 0.01 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.