Author: Bruce Moreland
Date: 17:33:07 09/29/00
Go up one level in this thread
On September 29, 2000 at 19:11:17, Bo Sjögren wrote: >Some time ago, maybe a year or so, there was an interesting >discussing about the minimum number of bits necessary to store an >arbitrary chess position. What was the conclusion of the >discussion? Can I find the discussion in the archive, or a >summary somewhere on the net? > >Regards, >Bo S It is somewhere in the archives, and you could find it in an evening of downloading and using grep, if nobody remembers what month the discussion took place. I think the most interesting thing about that line of research is that it attempts to approximate the number of legal positions, since if you knew for sure how many bits it would take to store every position, you would know how many positions there are. A major problem is that you don't solve chess if you have a value for every legal position. You solve it if you begin play from the root position, but if you have to pick up at an arbitrary point (for instance if you are doing analysis), the play that preceds the starting position affects repetition and 50-move values, possibly affecting the theoretical result or at least the path taken to that result. So there are big problems involved with attempts to analyze chess positions by solving chess via retrograde analysis. bruce
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.