Author: Dann Corbit
Date: 13:27:48 01/18/01
Go up one level in this thread
On January 18, 2001 at 15:45:22, Scott Gasch wrote: >Hi, > >Wow, I ran this suite and barely got 15 right! Actual result was 20/80 @ 20 sec >per move. It seemed like the ones my engine was able to get were mostly >tactical wins (score jumped up when it got the right one). Are most of these >positions largely positional in nature or are they just deeper tactics than my >engine could search? I suspect the former. I wonder how "subjective" these >positional "best" moves are? Would it be a good idea in general to tune the >eval to see such moves (i.e. would doing so likely increase strength and not >just cook the engine to get a better score on a suite?). How do other amateur >engines do on this suite? This is a really tough test suite, intentionally. If your engine gets 80/80 at 20 seconds per position, I am ready to announce it as a GM on the spot, despite the unscientific nature of any such announcement (barring a cook, that is). Some are positional moves. Others are simply deep tactics (but positional moves become tactical moves if 'deep' is deep enough).
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.