Author: Mark Young
Date: 20:04:38 03/27/99
Go up one level in this thread
On March 27, 1999 at 21:51:40, Lawrence S. Tamarkin wrote: >But lets not forget that we would like to continue to attract the GM's to >playing these matches at places like the ICC. As long as there is enough money >from somewhere to pay them to play, its probably no problem. (At least for >getting them to play). Another aspect is that of the interest of the matches >themselves; if the programs are always just going to win, (at the time controls >used), then the matches become pointless. But if some compromises can be worked >out that continues the questions of what type of play the GM's can use to beat >the wretchid machines, than the interest in seeing these things come about >continues. The compromise is for the GM's to play longer time controls, as I said before most GM's are no match for the better chess programs at G/30 or faster. Giving just GM's increments would not have changed the outcome of this match. GM Wilder lost the last game because of a positional misjudgement, not becasue of a time scramble. GM Wilder lost the game because he went for the win when he only had a draw. > >Another question - Wasn't there more of a sustained interest in seeing >Karpov-Kasparov matches, when it was quite unclear who was gonna win? Now >Kasparov says the only really interesting match would be against #2, Vishy Anand >& himself. He says there just wasn't enough interest in Shirov for the financing >to happen. (Of course I hope that Kasparov is wrong). > >I think everyone agrees that even if a reunification match between Karpov & >Kasparov could happen, it would be very boring indeed. (unless by some miricle >Karpov pulled ahead to an early lead and didn't then lose it - fat chance of >that happening!) So by this logic most matches are not even very interesting >if they become too pridictable, and it makes sense to consider ways to conduct >future events. But of course the most important opinions are by the GM's >themselves, and the programmer's, like Bruce Morland. Does not your argument scream out for standard time control matches, and leave the faster time control games behind us. I don't think it screams out for handicapping computers by playing time odds matches. Looking at the games, if Ferret had a 2 sec penalty pre move, ferret still would have won the match, as Ferret had almost 7 min left on its clock. It must be made clear, Ferret did not win the matches at G/30 because it can blitz off moves fast. That is a disservice to Ferret, Ferret won the G/30 matches because it is clearly stronger then the Grandmasters it has play at G/30 time controls. It is now time to go to a slower time control for both sides for the Grandmasters to have a chance at winning. > >mrslug - the inkompetent chess software addict! > >On March 27, 1999 at 19:56:14, Mark Young wrote: > >>On March 27, 1999 at 19:37:41, Charles Milton Ling wrote: >> >>>On March 27, 1999 at 19:07:30, Mark Young wrote: >>> >>>>On March 27, 1999 at 18:34:14, Lawrence S. Tamarkin wrote: >>>> >>>>>Ferret won its match against Wilder 2-0 on the ICC on 3/27/99. Wilder appeared >>>>>to have some winning possibilities in the 2nd game as White. He declined to >>>>>take the draw that was available by force, and wound up losing in the time >>>>>scramble that ensued. I agree with all the people who were suggesting that >>>>>computer programs have become such killer's at this G/30 stuff that some kind of >>>>>increment is in order for the human when this kind of match takes place. >>>>> >>>>> >>>>>mrslug - the inkompetent chess software addict! >>>> >>>>I don't understand the logic, why should the human player get increments when >>>>Playing g/30. If computer are now stronger then the human GM's at G/30, its time >>>>for the Grandmaster to agree to play at standard time controls. It has been >>>>clear for some time now that only a very few and best grandmasters can play the >>>>best micro computer programs at fast time controls such as G/30 and below and >>>>have any hope of winning a match. >>>> >>>>Again nice job Ferret and Bruce, and its time for the Grandmaster to play "real >>>>chess" >>> >>>My suggestion (and I doubt it is new) is as follows. The computer profits from >>>being able to move "in no time at all" (in the second game, I believe Ferret >>>used about 1 second for the dozen moves before it was out of book). Give the >>>human 2 seconds per move to compensate for this. (No increment for the >>>computer, of course. Probably difficult to set up on the servers, though, I >>>just realized. Hmm.) This won't really change anything fundamental, I honestly >>>believe, but it seems a bit fairer to me. (E.g. compensation for the phenomenon >>>just mentioned, human cannot lose dead-drawn endings...) >>>$0.02 >>>Charley >> >>There is nothing fair or unfair about a computer being able to blitz off moves >>fast. It is just a strength and the nature of a computer. >> >>Humans have there strengths too, but I never heard of making things more fair >>for computers when GM's were winning, so why should this fairness issue only >>come up now that the programs are winning.(Did programmer say it was unfair to >>use anti-computer tactics) When the programs were being beat by the Grandmasters >>at G/30...this fairness thing about a program being able to blitz off moves fast >>was never an issue, and should not be now.
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.