Author: KarinsDad
Date: 12:15:18 06/11/99
Go up one level in this thread
On June 11, 1999 at 13:14:46, Dave Gomboc wrote: [snip] >> >>I think the FIDE rules indirectly imply that a move must be made (although >>Miles/Reuben got away with a draw with no moves in 1975), depending on >>interpretation: "The game is drawn upon agreement between the two players during >>the game. This immediately ends the game". Since this says "during the game", it >>implies that the game is actually being played (or one move has been made???). > >Yes. At www.clubkasparov.ru you can read something by V. Dvorkovich (sp?), an >International Arbiter, who discusses how he dealt with a recent "GM draw". Yes, I had read that article. It implies that the rule (see my other post): "An offer at any other time during play is still valid" really should read: "An offer at any other time during play is invalid" This makes more sense. Maybe the FIDE rules on the Internet have bugs in them. [snip] >>If at best black has a forced draw after Ng1, then Ng8 could lead to a fast draw >>as you claim. However, you also stated that maybe minimum draws are not >>necessarily best (your opponent has less of a chance of making a mistake). A >>more complicated or slower draw may be best. So, regardless of whether Ng1 leads >>to a win for black or a draw for black, Ng1 should not be moved since it lowers >>the chances of your opponent making a mistake (by trying to minimize the number >>of moves of a draw AND by giving your opponent two tempi). Some different >>drawing move that does not lead to a possible win for black or leads to a slower >>or more complicated draw should be made. > >Correct, Ng1 is not a good move when playing fallible opponents. This does not >refute its brilliance under the original conditions. <grin> If 1. Nf3 Nf6 2. Ng1 is a forced win for black, it would refute the brilliance of Ng1 under the original conditions. <grin> > >>KarinsDad :) >> >>PS. I think I got you with my logic this time Dave, but I'm sure you'll think >>of a way to squirm out of it. :) > >Squirm. > >>PSS. I guess you have convinced me that the minimal draw move may not be best >>(with the caveat that it may be best in time pressure). > >That was my objective. Does this mean that I got you with my logic? :-) For falliable opponents you did. Of course, that was not the original discussion. The minimal draw move is definitely not best against fallible opponents since the basic definition of a fallible opponent is one who will eventually lose against a perfect program, so why play for a draw when you can win. However, the definition of fallible could include a "perfect" program with a few minor bugs, so a perfect program should still play for a win and if the imperfect opponent is good enough to find a line that gives a draw, at least the perfect program drew. It would never (even against a fallible opponent) play a move that could lead to a loss (possible Ng1). Against infallible opponents, the minimal forced draw move is perfectly fine. But since that was what we were originally talking about, it seems okay that I thought that minimal forced draws could be the only ones in the tablebase. However, I doubt you could ever convince me that 1. Nf3 Nf6 2. Ng1 ever leads to a forced drawn position with perfect play by both sides. So, although I conceed your point for falliable opponents, I think your example is still poor for infalliable ones. Even if a perfect program knew that it's opponent was perfect, it would still make moves that ONLY lead to draws or wins and I doubt that Ng1 only leads to draws or wins (with perfect play). :) > >Dave 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.