Author: Uri Blass
Date: 02:12:05 06/02/03
Go up one level in this thread
On June 02, 2003 at 03:15:57, Ryan B. wrote: >My tests against a wide range of engines all run in tournaments of random time >controls show List 5.04, Yace, and Amy to all be stronger than Ruffian running >on a Duron 800 all games with 24 mb hash. > >Ryan I think that something must be wrong in your tests because Yace and Amy are significantly weaker than Ruffian based on Leo's tournament and we are not talking about a small difference. Maybe Ruffian has problems when it does not run under winboard. I know that Ruffian impressive results in Leo tournament were achieved under wonboard. I found in tests with a version of movei that it may show a different behaviour under Fritz and under winboard in one of the nunn matches. I used versions that search fixed number of nodes so the time control is not an excuse. For playing 2 games of the nunn2 match I simply wrote movei a book with only one line(the line of the nunn2 match game 4). Under winboard I had no problem to play 2 games. Under Fritz the first game was finished in a mate for one program like under winboard but in the second game one of the versions showed an illogical pv and stopped to respond(I used 40 minutes/40 moves but the time control is not very important when the program searches only 81920 nodes). The behaviour is reproducable. The first game was not adjudicated but was finished in a mate so the tendency of Fritz to adjudicate games cannot be an excuse. Uri
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.