Author: Bob Durrett
Date: 13:14:10 09/12/03
Go up one level in this thread
On September 12, 2003 at 15:54:40, Gian-Carlo Pascutto wrote: >On September 12, 2003 at 15:42:54, Bob Durrett wrote: > >>I really believe that the "tests" I'm running are quite revealing and far more >>relevant to my interests. I see no need to resort to counting wins and losses. >>That would be a waste of time for me since I don't play against chess engines. >>Why should I care whether or not one engine can whip another??????????? > >It was *YOU* that stated: > >------------------ >In such a competition, I suspect that Crafty would come out on top of the rating >list simply because Hyatt says Crafty does very little pruning. [This assumes >that the Crafty operator does not shut the engine off due to boredom. : )] >------------------- > >If you can't validly substantiate your claims and aren't willing to defend >them, THEN PLEASE DON'T MAKE ANY. Well, in the first place, I didn't make any claim except to say that I "suspected" whatever. That's a far cry from what you are talking about. It remains true that I "suspect" what I said I "suspected." Surely you cannot challenge that! Do you never "suspect"? > >And as for your 'relevant' tests, are those the ones where you look at the >score of multiple engines and conclude that the one that is 'out of line' >must necessarily have 'something wrong' with it? No, I never said "must necessarily have 'something wrong' with it." If you review the thread, you will see that those were your words, not mine. Lets put this on a more serious note: I claim that the results I'm seeing cause me to tentatively conclude that Shredder needs more work insofar as it's displayed position evaluation scores [and evaluation symbols] are concerned. Do you know of any evidence to the contrary? I have already said that I have been allowing the engines to reach a minimum depth of at least 16, whatever that is, and that I have been allowing all five engines to evaluate the same positions. Ten games times 40 moves per game is 400 moves. I also do the same in evaluating positions which occur in evaluation lines. My guess is that I have looked at over 1000 evaluations [depth > 15] produced by these engines since I acquired them. This is an ongoing process. I will continue doing that over the coming years! Only GOD knows how many positions I will look at. The bottom line: I wish to know how to make the MOST sense out of the items of information that the engines are giving me, in the context of serious post-mortem game analysis. Comparison of the engine outputs has been very revealing! [They are all different. They would make a terrible Choir!] Bob > >...Riiiiight > >-- >GCP
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.