Author: Joshua Lee
Date: 13:55:13 11/12/01
Go up one level in this thread
>I believe that there are programs that may find the move in less than 96 hours >and if I am not wrong Junior7 could do it with number 3. > >The only problem with 3,6,9 is that we do not know if there are correct >It may be interesting if you can convince yace after learning the lines that you >teach it to choose the right move in them but to be fair give yace also to learn >the main line that program suggest(I remember that in number 3 the score of the >computer move and the score of the move that is supposed to be improved are >getting up every iteration). > ><snipped> >>Goliath Light 1.5 :18 11 CMB on (chessbase should make this default) > >No >The default version should be the version that is playing better > >Uri I believe That Goliath Light with Combination ON is better for Solving Test Positions Also would be best for long games if given a Fast Enough Computer but i will gladly re test all the positions with Combinations OFF. I decided to run these Nolot Positions on my faster 1Ghz to decide what to run all the other set's at. I was doing 15Minutes per problem but very few were solved no matter the program, and untill i get Deep (Fritz 7) , Junior 7 Deep Junior 7 , Shredder 5.32, Nimzo 8, I believe that the same would hold true that some will be solved but not many out of the 1700 or so. Since the "Easier" Nolot positions seem to take atleast 90 Minutes i may run every other problem that long, it depends on trying to solve the harder nolots. I wasn't about to give Fritz , Hiarcs, GT2 96 Hours each on 1700 Problems hehe. Is Deep Fritz 7 the same as Fritz 7 except for the SMP? Are they equal on a Single Processor?
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.