Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: 10 hash collisions in 2.3 billion nodes

Author: Dann Corbit

Date: 10:55:07 03/24/05

Go up one level in this thread


On March 24, 2005 at 13:45:20, Alvaro Jose Povoa Cardoso wrote:

>Hi,
>do you think 10 hash collisions in 2.3 billion nodes is something to worry
>about?

With a 64 bit hash (and given ideal dispersion) you would see one collision
every 8 billion operations due to the birthday paradox. ~sqrt(2^65)

Your rate is 30 times higher than that, but not unreasonably outside of
expectations.

If you have a 32 bit hash, then it is miraculously good.



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.