Author: Robert Hyatt
Date: 07:17:58 02/02/03
Go up one level in this thread
On February 02, 2003 at 03:38:00, Sergei S. Markoff wrote: >Hello Robert! > >>You can fudge around all you want, but there is no way to solve the problem >>short of storing the complete path played to reach a position in the table >>entry, Which would make the table useless and way too small... > >Is it proven? Yes. The game of chess is a graph, not a tree, which means the pathways are interlinked in odd ways. But the 50 move rule and repetition rules are written with the idea of a tree, because of the tranposition table. The only way to fix it is to include the "graph information" which is the complete path from the starting position to the position being stored... That's simply not doable. > >>You just have to ignore it. > >I see Crafty is trying to clear hash table in endgame to avoid this problem... >What you think about not accepting hash probe in some cases in endgame? > >Sincerelly yours, >Sergei It clears it as it nears the 50-move barrier, which is about all you can do to avoid walking into a 50 move draw in a winning position. It isn't optimal, but it helps. Note that it only "helps". It isn't a total solution at all. I'm not sure what you would do to "not accept some hash probes."
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.