Author: Dave Gomboc
Date: 13:24:13 03/28/99
Go up one level in this thread
On March 28, 1999 at 16:16:30, Jeroen van Dorp wrote: >That's an interesting thought. > >Same rpm with different transmission ratio is hardly interesting, you are >absolutely right. But how do you compare transmission ratio with chess engine >parametes? > >Let's say we do a brute search at 8 plys or so. With the same hash table size. >Why isn't it possible to compare performance of two chess engines? > >I think it tells you something about the efficiency and the strength of the >chess engines, because their "transmission ratio" is the same. Using different >hash tables or brute force against selective search you are right. But if those >parameters are the same, doesn't it say something about engine performance? > > >Jeroen ;-} Even if there is an option to tell it to use "Brute Force", the software will still probably use its quiescence search. Of course, the q-search will vary from program to program. Also, some people detect checkmate within a node, some people wait to see if a king gets captured. That sort of thing mucks these kind of comparisons up too. The best way I know how to assess the relative ability of the programs is to play many games, like the SSDF does. Dave Gomboc
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.