Author: Rick Bischoff
Date: 17:58:59 09/29/04
Go up one level in this thread
>I have never heard of this behavior. Are you talking about from iteration to >iteration? I would like to see an example of this. I assume you are not >talking about getting a hash hit at ply one? Here is a hypothetical example... White has a forced win beginning with Nf6+ Kmove Ne4+ Kmoveback <Some other sequence of moves> This position, during the original search, is stored into the hash table with a depth of, say, n. Program plays Nf6+ King moves Program plays Ne4+ King moves back Program then searches all legal moves at ply 0... ply 1 after Nf6+ returns immedietaly because it has an entry of depth n already stored... So unless the computer can search deeper then "n" in the allotted time, it will play Nf6+ and go into a hash loop. >> >>I replace always EXCEPT where the search has the same age and less than equal >>depth to the stored entry. >> >>***** 11201 ms elapsed 1247030 nodes 111332 nps. 181809 probes %13.6242 hits, >>%0.0440022 exact, %2.67203 lower, %1.0577 upper. nulltry=127769 nullsec=84970 >>ext:(mate=63,check=29699,pawn=143) > >Your numbers seem low. I can get 16-25% exact+lower+upper hash cutoffs in the >opening position. > >Michael
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.