Author: Dann Corbit
Date: 00:53:35 06/14/00
Go up one level in this thread
On June 14, 2000 at 03:33:59, Mogens Larsen wrote: >On June 13, 2000 at 20:21:25, Dann Corbit wrote: > >>This is completely and in all ways irrelevant. The *EXACT SAME THING* is true >>with the data used by the evaluation function. And the endgame database files >>are written (generally speaking) by pure programmer types and so are the least >>"chess genius" part of the system and (ironically) the only part that is >>perfect! > >You are more than welcome to pretend that everything is pure data. And from a >narrow point of view it probably is. But if you're unable to distinguish between >different types of data, then as a non-programmer I can't explain it to you. I am certainly able to distinguish between different types of data. Otherwise I could not do my job. However, your classification of this set as "special" is completely arbitrary. In fact, you are trying to hamstring the computer and to pretend otherwise is commical. Why should the computer not be allowed to know this? Why must the computer calculate it on the fly when any idiot knows that a human does not? >>You don't understand how programs work very well. A program operates on data. >>Without those "additons and attachments" there is no program. Period. The >>opening book data is *not different* than the eval data or the endgame data or >>any other data. > >Yes, it is. Where are your facts. Without facts, your argumentation is nothing more than: "Yes it is! No it's not! Yes it is!" Sure, it is not type same type of information. But it is nothing more than sophisticated information gathered from experts used to solve a problem. >>So, the algorithms are not from the author. The data is not from the author. >>What (exactly) does the author contribute? It is more or less a general >>knowledge of how to merge data with algorithms and tune such a system for >>performance. > >We can't disagree on this, even if I try. > >>We may as well disqualify the algorithms if you insist that the information must >>"come from the programmer only." > >No, that's not my position or my problem. If you want it in data terms it's >something like this: Why should a chess program be forced to use data it's >constructed to produce. Because it is the smart way to do it? Why should you memorize times tables when you could just as easily add up the summs each time. >A chess program is supposed to calculate a certain value > (the best) using an evaluation function. This value is then, I guess, >associated with the relocation of a certain object (a piece) on a particular >square through some kind of move generator. This is probably far from the actual >event, but in principle it really doesn't matter. The use of an opening book (or >endgame tables) is essentially an atttempt to bypass the purpose of the program. The purpose of the program is to win chess games. If you write a program, your purpose may be to write a nifty evaluation function. But for you to name the purpose of my program is ridiculous. If you want to exclude opening book information from any chess program you write, then that's fine with me. And I also understand that people are afraid of computer opening books (and pretty much everything else about them). But their irrational, hysterical fear is not a reason to abandon logic and hamstring the program. >The generation of moves. There's isn't even a comparison of data, ie. between >the generated data and the preexisting data. So there is indeed a difference >between data. I think it is time for me to stop trying to explain it. The data that the evaluation function uses to make decisions is pre-existing. Augh. What's the use? >>You are special casing the opening books. That's fine. But (like I said) you >>have not even chosen the most important band of data. If you really want to >>cripple the program -- go for the heart. Rip out the eval data and anyone can >>beat it. > >I'll just say it again. I'm not trying to cripple anything. There is no other conceivable reason for removing the information from the program. You are trying to make the program weaker. Admit this or not: "GM's have memorized opening lines." If you say 'false', then you are a liar, because no one is so naiive as to believe this. If you say 'true', then the computer has no advantage. Hence, you have no leg to stand on, either way. >And no, I don't have >anything against opening books, but in principle it shouldn't be necessary to >use one. You don't need an evaluation function either. Just search all the way to checkmate by position only. Sound stupid? It only slightly less ludicrous than your position. >Best wishes... >Mogens I think we'll just have to agree to disagree.
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.