Author: Peter Berger
Date: 13:02:31 12/16/03
Go up one level in this thread
On December 16, 2003 at 06:12:18, Amir Ban wrote: >On December 15, 2003 at 20:31:29, Slater Wold wrote: > >>Did you think you'd be playing this many games?! ;) >> >>I am sooo tempted to do my 'own' version of the Berger challenge!! >> > >There is still the 100:1 challenge to be played. Are you volunteering ? > >Anyone else ? > >This will be a slower affair. I would like Junior to have the equivalent of >30min/game on decent hardware, so maybe 25+10 vs. 2500+1000. 10 games should >take about 2 weeks. > >If you have a dual, it will take half the time, or maybe play twice the games. > >I think it should be played with books, but if there is objection to that, I'm >ok with symmetric positions that are not too deep in the game. > >I'm thinking end of January. I may put up a beta version, and Bob should also >have time to put up a "best and final version". > >Amir As I have been following these threads very closely and to anticipate some possible issues: a.) Bob doesn't trust the autoplayer, Junior doesn't support any public protocol - good that Slater volunteered to do the match on ICC. This sounds like an ideal solution given that there is a trustworthy setup for Junior (with sth like Ebbi's autoplayer e.g.). It might be further improvement to let the Crafty part be run by Dr Hyatt himself btw, but this is controversial. With the time control that has to be very long to make any sense, there has to be some automatic solution. b.) Given that Slater probably doesn't own two perfectly identical systems some benchmark is necessary. In fact this is probably an improvement in case a reasonable bench can be agreed on, because if Crafty gets the faster machine some very ununusual time control, like the 2500+1000 you suggested can be avoided. c.) The match probably has to be ponder=off. In case this is to the disadvantage of one of the contenders, it should be discussed. d.) Time control again: if the match is played on ICC your suggestion won't work if I am not completely mistaken. You can setup a blitz/fisher time control and give one of the opponents more time but you can't have different increments. So this probably means something like 35+1 versus 3500+1 should be used (just a suggestion). (e.) Junior will get blasted IMHO). 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.