Author: Frank Phillips
Date: 00:33:42 03/23/03
Go up one level in this thread
On March 22, 2003 at 17:39:44, Andreas Herrmann wrote: >On March 22, 2003 at 15:59:38, Russell Reagan wrote: > >>I am writing an amazons playing program, and I am wondering how I can determine >>whether I should use 32-bit or 64-bit hash keys. In chess, research has shown >>64-bit keys to be better, but in other games a 32-bit key would be more than >>sufficient. How do I determine what is the best key size (not only for amazons, >>but for any game)? Is it a trial and error process, or can a good estimate be >>made if (say) the approximate state space of the game in question is known? >> >>Thanks, >>Russell > >Hi, > >look to the paper of Dennis Breuker. Inside you will find some statistics about >hash collisions. >http://www.breuker.demon.nl/thesis/index.html > >Andreas Interesting. Page 38 seems to suggest a probability for a 64bit key of about 1 in 100,000 per 1 billion nodes searched. Modern desktops are not fast enough for this to be a problem yet - not sure about Deep Blue. From a quick scan, the calculation does not include the probability that the has collision will cause the pv to change erroneously, which must add another safety factor. Frank
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.