Author: Dann Corbit
Date: 16:18:47 11/07/05
Go up one level in this thread
On November 07, 2005 at 15:31:31, Kurt Utzinger wrote: >On November 07, 2005 at 15:15:30, Paul Jacobean Sacral wrote: > >>>It is not a good test position because Rxh5 is not the only way to win the game. >>>It is the simplest from human point of view but it is not important for games >>>and I believe that engines can win this position as white with different moves. >> >>Do you want engines to find the best moves? Or is any move which wins, equally >>good for you... (including very stupid simple moves)? >> >>I prefer the best moves! :) >> >>Yours truly Paul J. Sacral > > > You are in principle right. On the other hand one > should whenever possible avoid test positions in > which there are different ways to win the game. > Is it not very human like for an engine finding > a win to no longer change its mind? In my opinion, we do not know what the best moves are util we prove formally the final game outcome after making that move. Once we have proven that a given move leads to checkmate, then it is a best move. If (on the other hand) it leads only to draw, then we must prove that no other move can lead to checkmate. And if it leads to a loss then we know it is not a best move but not whether a best move exists. For instance, if a move leads to a mate in 19 moves and another leads to a mate in 7 moves, one might argue that the shorter mate is a best move and the longer mate is not. But it wins the same as the shorter mate and so it is also a best move (has the same game state output of +1). When we talk about pretty moves, then that is another matter. But we all may have different tastes. I find 1. b4 to be a beautiful move and others will find it truly attrocious.
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.