Author: Robert Hyatt
Date: 05:09:26 08/11/98
Go up one level in this thread
On August 11, 1998 at 07:50:57, Francesco Di Tolla wrote: >Hi guys, >as far as I've understood in the match the operator did some choices about the >very first moves to be played. If this is true, I'd like to know what do people >think about this fact. >IMHO I think I would be better to have programs which choose alone, and if the >want, they have access to databases and select which openings non to play. > >I think this is only a minor issue, but important in terms of principles. > >regards >Franz Think about what you are asking, first. No "program" can choose openings on its "own". It has to have some algorithm that chooses from the openings in its "book". Some use weights that are entered by humans, others use win/lose frequency counts which were produced by humans playing the games, some use small opening books to get things started on a good track before switching to the large book when the small book runs out. So it really doesn't matter whether the operator selects opening moves at the instant the game is played, or months before, as the human is still making the choice. It would be "easier" to do it OTB, of course, because you only have to be concerned with the opening being actually played rather than with all of the potential openings that might come up. But humans influence the game in many ways... In most events, after the first move is played, the human can not participate any further, ie it would be illegal for me to do anything to Crafty after white plays his first move. Prior to this, I could certainly enter commands that say "if he plays e4, play e6, or if he plays c4 play Nf6", but after the game is started, rules generally do not allow the human to take any active role in the game. I'd hope the Rebel vs Anand games were played in this way, otherwise it certainly is at odds with how such matches have been played in the past. As I have mentioned many times, this is the hardest part of a chess server interface, because programs that are manually operated can be influenced by humans in many ways, from overriding a dumb move, to help with time allocation, to you-name-it, while a fully-automatic program must do *everything* by itself, from offering/accepting/declining draws to choosing the openings and allocating time, and recognizing when it is in trouble and needs to look further...
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.