Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How do you insure that Hash Table entries are correct?

Author: José Carlos

Date: 15:30:21 09/16/00

Go up one level in this thread


On September 16, 2000 at 17:04:23, Dan Newman wrote:

>On September 16, 2000 at 12:25:27, Larry Griffiths wrote:
>
>>I have tried using a 64-bit hash key and storing it in each hash entry to insure
>>that different positions that point to the same hash table entry can be
>>verified.
>>
>>I also have a 32 byte smallboard that I use for debugging.  This smallboard has
>>4bits for each piece color and type so I know for sure if the current position
>>matches the hash table entry.
>>
>>The 64-bit hashkey works most of the time, but sometimes different board
>>positions produce the same hash-key.
>>
>>Is there a fail-safe method other than my 32 byte smallboard for insuring that
>>different board positions with the same hash keys can be resolved?
>>
>>Larry.
>
>I once tried to measure the error rate and got no errors at all with 64-bit
>hashcodes over several hours of testing.  I was able to measure the error rate
>for 32-bit hashcodes--that was about 1 false match/second (at perhaps 100k
>probes/s).  I think someone came up with an estimate of (very approximantely)
>one error/day with a 64-bit hashcode at 100 knps--or was it 1 Mnps?  Anyway,
>the error rate is very low and can mostly be ignored.  I do try to make sure
>that such an error won't crash my program though...
>
>-Dan.

  If you got 1 false match per second with 32 bits hashcodes, then you shoud
have a false match every 2^32 seconds with 64 bits hashcodes, which is IMHO not
a problem at all.
  I simply ignore that problem in Averno, and the only thing I do is test if the
HT move is legal in the current position before trying it.

  José C.



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.