Author: Dann Corbit
Date: 10:28:19 06/04/99
Go up one level in this thread
On June 04, 1999 at 13:07:29, Wes Groleau wrote: >I have a chess program which came with data files and instructions >for tests called BT2630 and CCR. Unfortunately, the instructions >for BT2630 make no sense at all, and the instructions for CCR say >that you check the computer's answer at certain times during the >run and give it a point if it is correct. It doesn't tell me how >to know whether it is correct. > >I can't find anything about either of these on the Web. Can someone >educate me a little? > >Also, the instructions gave formulae for computing a rating for the >program. Are these ratings comparable to the ratings used for humans? >That is, if it said 1425, does that mean the program could be expected >to beat a 1425-rated human half the time? You can find test suites on Manfred Rosenboom's web page or on my ftp site, for instance. These test suites show an approximate measure of the tactical ability of a chess program. However, you cannot effectively extrapolate this measure to play against humans. You can assume some sort of relationship between tactical ability and playing ability. Manfred's page: http://www.geocities.com/CapeCanaveral/Launchpad/2640/chess.htm Follow the obviousl links. My ftp site: ftp://38.168.214.175/pub/ Go to the "epd" area. HTH
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.