Computer Chess Club Archives




Subject: Re: Non power of two hash table sizes

Author: Heiner Marxen

Date: 06:23:18 02/19/02

Go up one level in this thread

On February 19, 2002 at 02:47:41, Dan Andersson wrote:

>There was an interesting observation about random number generators, recently. A
>weakness was found concerning the distribution of first n numbers, where n was
>small. If one looked at the distribution of those for multiple runs of a
>generator a distribution pattern emerged. I think it was true for most
>MvH Dan Andersson

Yes, that sounds about right.  IIRC, Marsaglia himself posted a correction
in some newsgroup.  But then, this deficiency is relevant only for a small
number of applications.  In the context of computer chess I do not expect
this to be of any relevance.  In cryptography the standards are much higher
than for simple monte carlo experiments.


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.