Author: paul bedrey
Date: 13:22:25 01/25/04
Go up one level in this thread
On January 25, 2004 at 13:38:40, Michel Langeveld wrote: >> >>It would be nice to have a subset of 500 epd's (that are not in other testsuites >at the moment) and that have enough difficulty to distinquise the performance of >today's chess programs. > >Michel Agreed. Its been my goal to try to either find or create such a test. So far the WM-test is the closest I've seen to predictability. Assuming all answers are correct I think a test set should have at least 1000 epds. The more the merrier to avoid randomness. Just like games the small sample size could sqew the results and lead to answer chasing. So far I'm leaning towards 50% positional, 35% tactical and 15% endgame. Test set should contain a mix of easy, medium and hard. I'm unsure as to the time. I'm thinking maybe 5-6 minute for the positional and 1-2 minutes for the rest.
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.