Author: Dann Corbit
Date: 10:18:24 11/09/05
Go up one level in this thread
On November 09, 2005 at 09:48:03, ALI MIRAFZALI wrote: >On November 08, 2005 at 12:16:45, Tony Hedlund wrote: > >>On November 08, 2005 at 11:28:24, ALI MIRAFZALI wrote: >> >>>Will SSDF test this program anytime soon? >> >>No. >Why is this? Cm 9000 was tested. Probably, they learned from their mistake. I am pretty astonished that they bothered to do any of the CM engines because of the staggering work involved. ChessMaster does not have an autoplayer. Look at the average game length when played at 40/2 (it's several hours). Look at the time invested for 300 games. Will you volunteer to sit for 2000 hours and manually input chess moves? I guess that the SSDF testers are all professionals whose time is worth at least $50/hour. That's a $100,000 project. Just do ten games yourself with CM verses Fritz or Junior and then imagine scaling that up to 300 or even 1000 games, as some of the engines get eventually. We have other places to measure the engine strength of CM (Graham Banks tests, CEGT tests). Of course, those tests use CM as a Winboard engine and probably do not use the CM books so they do not measure the strength of the system. But the system is simply badly designed for automated testing.
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.