Author: John Coffey
Date: 12:28:07 02/25/00
Go up one level in this thread
I think that mabye you are missing my point. I am suggesting storing just the base position not the searched positions (because I thought it would take too much disk access to do otherwise.) So in a 50 move game we would only store 50 positions. If we retrieve a position that was played in a previous game, the only things we might have (unless we try to store more) are the chosen move, depth of search, and the evaluation. The question becomes is this really useful information? It could be useful if we just wanted to skip searching altogether and play the move, but can this actually help our search? Can we start searching at N ply from where we left off? I don't think so because we would be missing all that useful information in the transposition table that helps us to order the moves for faster searches. A N ply search without this information might actually take longer than if we had done the itterative deepening first. John Coffey
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.