Author: Peter Kappler
Date: 09:15:17 07/02/99
Go up one level in this thread
On July 02, 1999 at 08:10:11, eric guttenberg wrote: >I'm sorry, I don't want to get into an exchange of unpleasantries but >I think you are mistaken here. We are looking at the same posts I believe. >The posts that show Hiarcs winning are all engine v engine in one computer >and there is a lot of evidence that these results simply are not reliable >in determining which is the stronger; in the results obtained when each >program plays on a separate but equal computer the results are very >different. Look at Jim Walker's results where in 200 games at g/5, >40 games at g/15 and 10 games at tournament times, F5.32 leads H7.32 >in each category. Do you recall any posts that show H7.32 leading F5.32 >in computer v computer games? >The reliability of engine v engine tests to determine playing strength >has been discussed before on this site many times. That is what I mean >when I say the results are dubious and prove nothing. Tania's engine >v engine posts between these two programs fairly consistently showed >H7.332 better by about 200 elo points. I don't think anyone really >believes that. Hiarcs ability to "ponder" in engine v engine seems to >skewer the results. If Hiarcs is really pondering in the "same-computer" matches, then I agree, the results are badly flawed. But, it is hard to believe that Chessbase would have shipped Hiarcs 7.32 with this bug. --Peter
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.