Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Random keys and hamming distance

Author: James Swafford

Date: 06:20:38 08/17/02

Go up one level in this thread


On August 17, 2002 at 01:16:05, Russell Reagan wrote:

>On August 17, 2002 at 01:11:13, Uri Blass wrote:
>
>>hamming distance between 2 vectors is the number
>>of coordinates that they differ.
>>
>>Example d(010,100)=2 because 0!=1,1!=0 0=0 so we have
>>2 different digits(only the last digit is the same).
>>
>>In hash tables we have 64 digits so the hamming distance can
>>be at most 64.
>
>So how do you compute the "minimum hamming distance" as James said? Surely you
>cannot compare all values in the hash table with every other value in the hash
>table.
>
>Russell

Uri's comments are correct, but I just wanted to add that it's important
that the Zobrist numbers you use to generate hash keys are sufficiently
random.  If the hamming distance between them is too low, you'll get more
collisions, because it's more likely that two positions have the same key.

--
James



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.