Author: KarinsDad
Date: 09:14:45 04/21/99
Go up one level in this thread
On April 21, 1999 at 11:31:25, Bruce Moreland wrote: > >On April 21, 1999 at 10:16:24, KarinsDad wrote: > >>In the real world, there are FIDE rules, there are USCF rules, and there are >>chess rules for a lot of other countries. >> >>However, is there an official body for chess rules for computer programs? >> >>Or, should a computer program be made aware by the operator as to which type of >>tournament it is playing under (and the rules thereof) and therefore should play >>accordingly? >> >>Is there any governing body on this? Has this been addressed before? >> >>Is there a standard by which the programs should be programmed? > >The governing body for computer chess is the International Computer Chess >Association (ICCA). > >http://www.dcs.qmw.ac.uk/~icca/index.htm > >This organization was founded in cooperation with FIDE. > >The rules for live computer vs computer games are patterned after the rules for >blind players. > >Before each WCCC they typically announce the rules in their journal. > >Basically you have a chessboard with pieces, a clock, two operators, and two >computers. The monitors are placed where both operators can see them. The >clock is the official time, and the board is the official position. Both >operators record moves as in a human vs human game. The operators serve as >proxies for the computers, they do what the computers tell them to do, and they >tell the computers the condition of the board and clock (they relay moves and >keep times current). > >A basic goal is that the operators not influence the game either way. > >There are rules about exactly when you can correct the computer's clock, but I >think they are archaic and need a rewrite. > >If a program offers a draw and the other one accepts, it's a draw. If a program >resigns, the game is over. The operators can offer draws, accept draws, and >resign, with the approval of the tournament director, who typicall insists that >things be pretty clear. > >Operator errors are corrected, as much as possible, without penalty. Some >number of crashes or other stoppages are allowed, but if there is too much >crashing the clock keeps running while the computer is restarted. > >An informal rule is that the clock is not started until one of the programs is >out of book. That avoids a meaningless clock scuffle at the start of the game. > >Adjudications based upon best play are typically done after some period of time >has elapsed. > >bruce Great. This explains computer vs. computer. There are still a few grey areas I suspect. For example, FIDE put into effect the rule that both sides need not make all 50 moves in order to claim a draw by the 50 move rule, the claiming side need make only 49 moves. However, if one computer claims the 50 move rule based on 99 moves and the other denies it, I'm sure the tournament director in an ICCA tournament can handle this. What is the governing body for computer vs. human? I assume it would be based on the tournament (i.e. USCF rules in a USCF tournament, etc.). If this is the case, then how does one make sure that their program obeys all of the rules? Do most programs have "variant rules" programmed into them so that the operator can pick and choose which rules are applicable to a given tournament? And what happens with down time in a human versus computer game in non-ICCA events? And how do discrepancies get handled on the chess servers such as ICC? 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.