Author: Robert Hyatt
Date: 20:26:59 10/11/02
Go up one level in this thread
On October 11, 2002 at 17:12:35, Rolf Tueschen wrote: >On October 11, 2002 at 16:36:59, Matthew Hull wrote: > >>On October 11, 2002 at 16:24:22, Rolf Tueschen wrote: >> >>>On October 11, 2002 at 15:05:26, Matthew Hull wrote: >>> >>>>Rolf, >>>> >>>>In the interest of clarity, if it was up to you, what would be the rules for a >>>>human versus computer chess match? >>>> >>>>Matt >>> >>>First and basic Law: >>> >>>Chess programming does always mean being honest about the actual strength of the >>>program/machine entity. So impostering should be forbidden already out of >>>self-respect. All tricks which are meant to exercise psychological confusion >>>should be regarded as insult against computer chess itself. >>> >>>Consquence: >>> >>>That means by logic that all the Wch shows and super GM hype "bye-bye" from now >>>on. There are plente of strong IM and experts who would be glad to play the >>>machines. >> >> >>It is not clear what you mean by "show" and "hype" as applies to the rules of a >>match. > >Rule of honesty. Yes, I see how that surprises you. > > >> >> >>> >>>Second Law: >>> >>>Opening books should never contain lines a program can't "understand" simply >>>because the key is too deep in the tree. So also such pretension should be >>>regarded as insult against the moral of CC. >> >> >>This one seems impractical. I play openings all the time that I don't >>understand. Should I not be allowed to play them in a match? Anyway, how could >>such a thing be policed? How would an arbiter know if the program "understands" >>an opening or not? You want the program to play with no book? > >Did I write "no books"? You say impractical. Because we should give up all hopes >that CC should be able to return to "normal" after the period of unbelievable >hype? First of all we should understand that the ChessBase plans with their GM >shows is hype, nonsense and danger. Commercial engines are not yet strong enough >for GM. If we restrict ourselves to honest CC. > >Rolf Tueschen > You keep bringing this up, so here's a challenge: Formulate a rule (or rules) governing opening book knowledge. The rule has to be fair to both players (computer and human) and the rule _must_ be enforcable or it will be useless. What would you like to see and why? \ > >> >> >>> >>>(NB that I'm not being asked for giving my specific ideas for new chess topics >>>in CC programming. But I think that the ideas of Roay are fantastic. I saw that >>>Ed already agreed although he's the famous addict of night long autoplayer games >>>instead of human vs comp test.) >>> >>> >>>Rolf Tueschen
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.