Author: Joshua Lee
Date: 16:59:10 11/16/01
The match will wake a lot of people up look at Junior's game against Kramnik has anyone attempted to see exactly how long it takes Junior to make moves that would've kept it from loosing? I really think that even on a 320Ghz machine any chess program will still be able to be beaten but the correct positions have to be found from games that test positional play. I have a decent test taking all main epd test sets and solving a great deal of these problems originally about 10-15,000 including duplicates now removed and i was left with around 1700 which i previosly posted. I am now testing instead of 15minutes per position 1Hr so given a few months i will have cut this hopefully in half. I had looked 15minutes as this would be Tournament level on a 4Ghz but after looking at Nolot positions i noticed that most are solved within 90minutes some take days (3,6,9) #10 and 5 which are harder are solved within 18 hours on a 1Ghz machine. so i figured the problems not solved within 1hr are Worthy of being included in a revision of a new Nolot Like Test Set. Provided after this 71+ days i can isolate the problems with incorrect solution or position. Already there are a few that are not solved in under an hour by Gambit Tiger and one of which that takes 8+ hours. Still more programs will have to be tested.
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.