Author: Robert Hyatt
Date: 15:21:40 07/31/99
Go up one level in this thread
On July 31, 1999 at 17:14:18, blass uri wrote: > >On July 31, 1999 at 16:33:08, Robert Hyatt wrote: > >>On July 31, 1999 at 15:16:02, blass uri wrote: > >>>I looked in this position and the latest version of Junior can find Bxh6 even at >>>depthes 8 or 9 (eqvivalent to brute force depth of 4,5) >>> >>>It can find Bxh6 even at level 1 second per move on pentium200. >>> >>>I think that Cray blitz and crafty has problem in this position relative to part >>>of the commercial programs. >>> >>>Uri >> >>1. Crafty has difficulties with it because of null-move. >> >>2. Ed was talking about doing a _one_ ply search. You just said it >>takes junior to depth=8 or 9. See what is wrong with your argument? It >>has to see it at depth=1, not depth=anything > 1. > >The point is that other program if they use 1/3 of the regular time or even only >few seconds for moves that they believe that are forced could find Bxh6 at >tournament time control. > >You can find cases when it is wrong to do it but the question is what is the >probability that it happens. > one such move in a game is all it takes. In 1981, Cray Blitz was searching about 2-3K nodes per second... a few seconds wasn't enough... and it isn't enough for other programs either until you speed up the hardware to today's speed. But the point remains... there are sufficient positions where a shallow search says 'this is the only move' and a deeper search says 'that only move loses badly'. Do you _really_ want a program to save time here and there, and then make such a gross mistake? I don't... that is why my 'easy move' is so very restricted, and generally gets used once every few games or so, because it is always 'suspicious' that something is fishy. If it is wrong, it only wastes time my way. The way I used to do it would lose a game when it was wrong... >You discover cases when you made an horrible mistake because of playing fast for >moves that you believe are forced but you may earn more from cases >that you are winning because you save time or the opponent does a mistake >because the oppenent has less time to think on your time and find the right >move. > >Uri
This page took 0.01 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.