Author: Heiko Mikala
Date: 12:43:26 10/30/99
Go up one level in this thread
Hi Dave! On October 30, 1999 at 00:14:43, Dave Gomboc wrote: >[...] >He provided a mathematical foundation for >his static evaluation algorithm, the intuition for which Jonathan described as >"brilliant" -- and Jonathan doesn't throw that word around loosely. He >commented about constraints for his specific implementation. He generalized >experimental results that he had obtained regarding the tradeoff between the >quality of static evaluation and the depth of the game-tree search. [...] Talking about mathematical foundation, constraints, generalization and so on, I think there is an important point we should talk about. Just before reading your post, I read a post by Bob Hyatt in which he replied to Ed Schröder, that if commercial programmers feared the public in a group like CCC, they should and could avoid it and should instead publish their ideas in (scientific) journals. I read that and thought, yes, Bob has a good point here. Then I read your post, found it very true too, but the above paragraph reminded me on something I had thought about before. I guess a very important reason for commercial programmers not to publish their ideas in print may be, that it is normally common to not only present your ideas and give some experimental results, but also to give mathematical proof for your ideas. Now, I'm very sure, that most of the commercial chess programmers are not scientists. They may not even have visited a university. So what should be done in my opinion is to encourage the commercial programmers to publish their ideas anyway; to tell them that we don't expect mathematical proofs from them and that we don't even expect them to write their ideas down in a scientific way. I guess everybody agrees, that we all would more than enjoy reading about Mr. Langs, Mr. Schröders, Mr. Hirschs etc. ideas, and that in such a case we wouldn't care at all about a scientific approach in presenting the ideas. Experimental results would be more than enough, since they have already proven that their ideas work. To be honest, what I found out about myself is, that I normally skip articles about chess programming, which look too mathematical to me, although I would well be able to understand them of course. My experience during the last 10 or so years was, that I learned much more from articles written by the "practical" guys, those who had really implemented their ideas succesfully in working chess programs. For example I enjoyed reading the articles about Cray Blitz, Deep Thought and Hitech, even if there was no mathematical proof given for anything. On the other hand, I read some very mathematical, sometimes hard to understand articles which gave me - absolutely nothing. I often had the feeling, that the presented ideas were good in theory but never had been tried in practice. I don't know much about the ICCA journal, because (unfortunately, shame on me!) I'm not a member. All the articles I read had been published in books or AI journals and procedings. If the ICCAJ is different that's fine, but all the other publications are (of course) scientific publications made by scientists. This may discourage some very talented people from publishing their ideas. Greetings, Heiko.
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.