Author: Dieter Buerssner
Date: 10:15:44 08/31/01
Go up one level in this thread
On August 31, 2001 at 11:53:44, Gian-Carlo Pascutto wrote: >You are tuning for a position. > >If I hardcode it into Sjeng, it finds it in 0 sec at ply 1. Well, I didn't hardcode any solution. But you are right, I tunded. I changed a parameter, that is not user accessable. However, not for one position, but for all those "very forced mate postions". Much like Bruce has explained. BTW. I just tested an 3 years old version of my program, of which I think, it was much more rough, and it found the solution at depth 8. And, my parameter change (which I have done some times earlier already and tested), is not really overtuning for this position. In few testgames, I really did not find much of a difference, but I have the fear, that some positions in practical games can yield in search tree explosion and weaker play. In test positions, it finds few earlier and few later. >But that won't help me in actual play. > >>John Merlino even has shown a better result. > >I wont contest that. But it still means nothing about DB. I agree. That actually was my point. Neither does finding this at ply 8 by mean much, IMHO. Bruce has explained this very well. Regards, Dieter
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.