Author: John Merlino
Date: 12:25:02 06/30/01
Go up one level in this thread
On June 30, 2001 at 13:44:33, Divy wrote: >On June 30, 2001 at 08:12:09, Jason Williamson wrote: > >>Thanks John. >> >>Now there is yet another bit of proof that CM8k is not weaker then CM6000. >>Jeeze where are all the skeptics now? >> >>JW > > >I'm conducting a test of my own using CM6 (personality:Utzinger) vs. CM8 >(personality:default). > >CM6 book:Kostick / CM8 book:Default; (2)700Mhz PIII; G60; 32M hashtables. > >If the CM8 engine is inherently stronger, shouldn't it do just as well as >J.Merlino's testing? I'll update the results. Not necessarily. The randomness involved in playing a relatively small number of games can produce almost ANY result. Also, the fact that you are testing the Utzinger personality for CM6000 against the default for CM8000 makes your test very different from mine. The default personality is believed to be the best for all time controls on all supported hardware (the minimum spec for CM8000 is a PII-233, after all). The Utzinger personality is, I gather, optimized for faster machines and longer time controls. Additionally, nobody has shown with any reasonable certainty that the CM8000 engine "is inherently stronger" than the CM6000 engine. Personally, I believe that it is stronger, but it is very difficult to determine. But several people have stated that they believe that the CM8000 engine is WEAKER than the CM6000 engine. It was that theory that my test intended to disprove (although MANY more games will have to be played by people other than myself to truly disprove it). So, your results might be indentical to mine, but they instead might be VERY different. But, I suspect that the DATA will certainly be useful to somebody.... jm
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.