Author: Tom Likens
Date: 04:51:40 04/07/04
Go up one level in this thread
On April 07, 2004 at 07:05:01, Renze Steenhuisen wrote: >On April 07, 2004 at 06:55:31, Andrew Williams wrote: > >>On April 07, 2004 at 06:49:59, Renze Steenhuisen wrote: >> >>> >>>Hi all, >>> >>>could someone give me some numbers that are common with hashkey collisions? >>>Because I guess my % is little too high... >>> >>>I'm getting like 0.03% [which is 1 every 3000, if I'm not mistaken] >>> >>>This is when using TT=32MB (haven't got the exact number of entries) >>> >>>If you think it is an error, any suggestions on where to start looking? >>> >>>Thanks! >>> >>> Renze >> >>One in 3000 seems very high. How many bits are there in your hashkey? >> >>Andrew > >I use 64-bit keys, random numbers, and I am using the same random number >generator as Crafty uses. I have checked the randomness by counting the number >of zero's and one's, they are about equal. I counted the number of zero's and >one's per bit-position, and they are also about equal... > >I have two counters for calculating my hashkey-collisions: > hashmove_requested: every time I have a hashmove and I am going to check > whether it is a legal move. > hashmove_illegal: suggested hashmove from the TT was not legal, and > therefore a hashkey collision has to be occured. > >100*((double)hashmove_illegal/hashmove_requested) gives my number... 0.03 > >I'm clueless... > >Renze This does seem high. Are you positive you're not corrupting the hash key somewhere in the search? If you don't already have it, you might want to write a debug function to generate the hash key from scratch at each node and compare it to the incremental key. At the very least you would know that your key is good. regards, --tom
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.