Author: KarinsDad
Date: 20:44:13 07/02/99
Go up one level in this thread
On July 02, 1999 at 19:49:02, Marc Plum wrote: [snip] > >I hope all the editing doesn't create confusion, but this was getting rather >unwieldy. > >What interests me is that there is ample precedent for humans winning one >tournament to qualify for another much later. For example the winner of the >Hasting Challengers would be invited back next year for the main event. I don't >know, however, if organizers have given much thought to applying this to >computer programs. As you have been pointing out, the process of improving a >program is *not* identical to the way humans improve themselves, although it may >have similarities. > >It is up to organizers how they want to handle this, but perhaps it would be >good for them, or FIDE, or the national federations, to think of some ground >rules for upgrades of programs. For example, suppose a program were entered in >the FIDE World Championship under the format to be used in Las Vegas. Fritz >6.01, in the first round of matches, might be Fritz 6.03 in the quarterfinals, >and 6.07 halfway through the finals. That is, if upgrades are allowed to take >over when it was an earlier version that qualified. Theoretically, the program >might even be upgraded before every single game. > >It isn't that I'm arguing for what the rule *should* be. I just think that, if >computers are going to participate regularly with humans, there need to be some >clear rules on when, and how, an upgraded program is allowed to fill in for a >previous edition. If this kind of thing isn't spelled out in advance, I predict >extended litigation. > >Marc Marc, One of the reasons I reply often after making my point on posts like this is to get people to think beyond their normal bounds. One way to do this is to be contentious and to hit upon multiple areas on the same subject. I am glad that you did not join the bandwagon of disagreeing with me, but rather used the conversation to discuss issues that most people do not think about. Bravo! You may be correct about the litigation at the highest levels of chess if programs are ever allowed on a continuing basis. I doubt it will occur below the GM or IM level, but it's possible at the superGM level where some of them have millions of dollars socked away with which they can hire lawyers. As for the organizers, you are right. Each organizer does whatever he feels and just attempts to abide by the main FIDE rules where possible. Since there are no FIDE rules on computer chess, they do whatever seems right at the time. Even the ICCA has an extended set of rules, but is uncertain on how to proceed in some cases. Take for example the accelerated pairings controversy and the controversy on Shredders placement. When you do not use consistent programs on consistent machines (i.e. you do not have semi-legitimate ratings), even the ICCA gets confused on placement and they are suppose to be the guys that have this all down pat. I find that it is interesting that the ICCA had a WCCC and a WMCCC title in order to make it a little more fair for those programs which run on a single processor and computer chess hobbyists like that. But some computer chess hobbyists also consider it fair for programs to upgrade from one tournament to the next (when invited due to winning the first tournament). If upgrading at any time is totally fair, then there should not have been a WMCCC title. Or, another way to put it is that in the Frankfurt tournament, there should have been a human title and a computer title since there is more of a difference between humans and programs than between single processor programs and multi-processor programs. I'm sure that you can appreciate my tongue in cheek here (and that I am not being totally serious), but it is interesting what we as computer chess buffs take for granted and what we find strange. KarinsDad :)
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.