Author: Don Dailey
Date: 20:23:31 05/14/98
Go up one level in this thread
On May 14, 1998 at 18:41:22, Brian McKinley wrote: > >On May 14, 1998 at 17:10:37, Don Dailey wrote: > >>On May 14, 1998 at 05:32:13, frank phillips wrote: >> >>>Can anyone point me in the direction of explanations of the techniques >>>used in chess programming. Having learnt a bit of C this year and with >>>the help of articles found on the Net about search algorithms managed to >>>write crude connect4 and othello programs, I would now appreciate some >>>pointers to simple but thorough and complete explanations of bitboards >>>and hashtables. Several articles I have found partially explain these >>>topics but usually assume that the reader has more knowledge than I do, >>>perhaps because they assume a basic programming background which I do >>>not have. The source code to Crafty (thanks Bob) is well documented but >>>beyond me at the moment. On a specific issue, I would be grateful for >>>an explanation of how to get a pv out of the (alpha/beta negamax) >>>search. Everything I have tried so far based on storing the best move >>>in an array pv[depth] has failed, so I seem to be missing a fundamental >>>point - or am just plain stupid :-/ >> >>Frank, >> >>Another idea that some use is to "walk the hash table" if you have one. >>Each hash table entry has a best move associated with it so you simply >>walk through them. I don't like this method as well but it's easier >>to implement and has the advantage of not slowing down the program. >>PV's take a small toll on the search speed. >> >>- Don > >I take it you do this once at the end of the search. ? You do it every time you report a score change. You just keep making moves as you find them in the table and then unmake them all in sequence. - Don
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.