Author: Peter Fendrich
Date: 15:17:03 04/19/98
Go up one level in this thread
I think you missed my point... These kind of questions always comes down to philosophy or even religion. And hopefully not mainly to political or statistical issues. Let's stop the philosophy and slide over to computer chess. Now we are in the middle of two subjects that I really like! There are numerous test sets for chess programs. Suppose we want to measure how strong a chess program is just by doing a test with 150 positions. (About the same number positions as questions in the IQ test.) You probably agree with me that no current test is reliable and will never be. It's impossible to cover all aspects of strength in chess just by 150 positions, furthermore it's very hard to compose the positions so that different aspects have the exact weight in accordance to the other aspects of chess strength. Of cource we will find a correlation between strong programs and good performance in the test but we can never rely on the test for each individual chess program. Exactly as for the IQ test. Stength in chess is just a minature of intelligence which is far more complex so the same should apply to IQ test but worse. My opinion in a nutshell... But if you still want a test set for chess programs, a good way to design it is to do it the same way as as the IQ tests are made. Mathematically not bad really! Has anyone tried this? I still wont believe in these tests but if someone does the job anyway, why not doing it right! Peter
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.