Author: Dann Corbit
Date: 10:52:08 10/31/03
Go up one level in this thread
On October 31, 2003 at 06:48:22, Francesco Di Tolla wrote:
>On October 31, 2003 at 06:28:00, Francesco Di Tolla wrote:
>>If the original book cannot be converted, my solution to this problem is very
>>simple:
>
>Apparently the whole argument is pointless, since the book can be converted:
>
>http://www.talkchess.com/forums/1/message.html?324458
>
>so why not using it?
Even this may not really be a solution. What I mean is that programs do not
just blindly follow their book. They look at statistics in the book and often
at some point deviate, even though something is in the book.
Won/Lost/Drawn figures. Elo of the player making the move. The centipawn
computer evaluation. All of these things can be considered. Would the
decisions TheKing makes be identical to those made by the CB API?
I think the experiment is interesting. I do have one small problem with it,
which is that it is highly likely to be misinterpreted by the buying public.
On the other hand, if someone does have both products, they would be able to set
up a system that plays like the SSDF. So they would have measurements for that
sort of thing.
This is the problem that I see. If CM does spectacularly, we will see it on the
CM 10K box. If CM does poorly, they have an out ("It wasn't our book.").
I think that they will probably be honest enough to state the conditions
somewhere in those tiny weasel words, but nobody reads them anyway.
Did you know that the drain cleaner's weasel word collection says on it (among
other things): "Do not ingest this product."
I firmly believe that anyone smart enough to know what ingest means knows that
you aren't supposed to eat the drain cleaner.
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.