Author: Larry Griffiths
Date: 02:35:08 02/25/99
Go up one level in this thread
On February 24, 1999 at 22:19:04, William Bryant wrote: >On February 24, 1999 at 21:44:34, Larry Griffiths wrote: >>William, >> >>Very nice output! May I inquire as to how you are able to retain >>the principle continuation when using hash tables? Mine seems to >>get messed up when using a hash table. >> >>Larry :-} > >I don't very often at the lower plies. Note that at 18 plies on the first >example, the principle continuation is only 5 moves long. The longer PC(or PV >as case may be) occurs when the search goes beyond the previous hashed search >and starts to back up the moves. > >I am trying a suggestion of Dr. Hyatt's and not hashing the QSearch code, trying >to keep the QSearch lean. > >He also suggest rebuilding the the PV from the hash table entries if they are >not overwritten. I may try this soon (on the list, a long list :) ) but I am >concerned about that reduction in speed it may require. > >If you come up with a good trick or suggestion, let me know. > >William >wbryant@ix.netcom.com Thanks William, At least I know that it is a problem now. I ran that first position thru my program and it does the first 16 or 18 plys quickly but then hits the wall at 20 to 22 plys. If I try to play the position, then it gets even worse as I am assuming one of those pawns queens. My program still beats me as I have not figured out how to win the opposition when playing that first King and Pawns position. Larry :-)
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.