Author: Dann Corbit
Date: 11:33:05 08/30/01
Go up one level in this thread
On August 30, 2001 at 14:29:12, Roy Eassa wrote: >Don't bother having any program ponder for hours or days on Nolot #9. Unless >you can let them ponder for a few HUNDRED THOUSAND hours. Because the solution >is so tactically deep, that's what you'd have to try (or maybe MILLIONS of >hours). Even if you "force" today's top PC programs to see the first several >plys, they still don't see that white is winning, at least within a few hours. Different programs have different scoring on eval, and different shaped search trees. Clever extensions and different null-move strategies might make a gigantic difference. >And doesn't each additional ply take more time than all the previous plies >combined? Usually. Doesn't have to, though. Could be that some move is forced and the next ply is free, for instance. Usually, it will be a rough multiplication of the branching factor in time from the previous move, of course.
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.