Author: Kai Lübke
Date: 02:38:24 10/16/97
Go up one level in this thread
On October 15, 1997 at 11:08:02, Vincent Lejeune wrote: >Strange Behaviour of Genius5 (WIN95, P120, 8 MByte Hash) while analyzing >my game : [snip] > >On move 39 instead of 39. f4 Genius5 propose 39. Rg1 with the line : > +1.48 08 03:17 c1-g1 g7-g6 f5xg6 h7xg6 h5xg6 h8-g7 g1-g2 >but when i'm playing the 6 first moves the analyse give : > +0.42 09 02:27 h2-g3 f8-h8 f2-f4 d8-c7 f4xe5 h8-h6 g3-f2 c7-c2+ >that mean that the previous analyse give a score 1 pawn up >to "reality" !! > >Bug in analyse engine, inacurrate analyse ??!! What's your opinion ? That's perfectly normal. By playing the "6 first moves", you advance 6 ply deeper down that line, so the program starts seeing things it did not see before simply because they were too far down the tree. And if you ever should find a program that _never_ drops from eval N to N-1 during a game, that program would be nearly invincible. :-) --- Shep
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.