Author: macaroni
Date: 21:18:57 06/26/03
Go up one level in this thread
On June 26, 2003 at 23:30:07, Peter Stayne wrote: >Not always, one example: on the playchess server, there was a guy running a >version of the king against my Deep Fritz 7. I was getting ~2500kn/s and was >having a tough time beating him. > >His kn/s? 83. > >It's all about how much knowledge is being used to eval the position to smartly >prune out inferior lines. If mine studies 2.5m nodes in a second but a higher >percentage of those weren't worthy of studying in the first place, a slower, >more accurate one that studies more 'worthy' nodes per second can overtake it. > >Also see Shredder 7.04 vs Deep Junior What are the most used methods for sorting out the 'unworthy' lines? I currently use the history heuristic, but my main search seems to be doing ok anyway, it's the quiescence search that's having problems, with the q search I use biggest capture, smallest attacker, is there something much better?
This page took 0.01 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.