Author: Daniel Clausen
Date: 06:46:52 12/05/01
Go up one level in this thread
On December 05, 2001 at 08:22:40, Gerd Isenberg wrote: >About the pawn hash key discussions: Why using Zobrist keys for pawn hashing at >all, at least in BitBoard programs? Isn't it smarter to use a unique 46 Bit >(Rank2-7) Difference of two colored Pawn-BitBoards instead of zobrist keys for >pawns? No key collisions and key and index (key mod nEntries) calculation on the >fly. > >Gerd Note: It's 48 bit, not 46. One problem is that similar positions lead to similar hashkeys. (similar in the sense that the Hamming distance is small) Now let's further assume that you use the lower 16 bit as index (and the remaining 32 for the collision test) Since the hamming distance between hashkeys is small, the index will not be very well spread of the whole table and you will get a lot of hash collisions. I think that a similar position (read: only one piece at a different square) should switch about half the bits in the resulting hashkey, which leads to optimal hashtable usage. Sargon
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.