Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Hash collisions in chess programs?

Author: Inmann Werner

Date: 11:45:59 02/18/99

Go up one level in this thread


On February 18, 1999 at 06:37:34, Rémi Coulom wrote:

>On February 18, 1999 at 06:32:12, Rémi Coulom wrote:
>
>>I just calculated the probability to have two identical values in a list of p
>>values chosen at random among q possible values. Maybe everybody knew this
>>already, but I was ignorant of the result and I find it is interesting. The
>>result is, for large values of q and p < q, equivalent to exp(-(p*p)/(2*q)). It
>
>Sorry, the result is 1 - exp(-(p*p)/(2*q))

What does this mean in particular for chess programming?

I use one hashingadress between 0 and 2000000 for the hash adress and one
extra calculated long int for verification. Each of the both hashcodes result
out of 2 different sets of 64 long int numbers.
How high is the probability of a collision in this case, and do you use the
same algorithm?

Greetings

Werner



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.