Author: Vincent Diepeveen
Date: 09:32:35 12/03/99
Go up one level in this thread
On December 03, 1999 at 08:52:30, Daniel Clausen wrote:
>Hi
>
>On December 03, 1999 at 08:32:39, Vincent Diepeveen wrote:
>>What i predict is a much closer to 50% score for the 6 against 6 ply
>>match, and a complete annihilation of amateur when both searching at
>>10 ply.
>>I assume here already of course that DIEP's eval is better than
>>Amateurs.
>
>Note: I don't have this whole threat in my head anymore, so if I
> miss something, I'm sorry. :)
>
>So what is this test all about? Since both machines are limited to
>the same depth, the only difference between the engine is their
>eval.
>
>If both evals have a similar quality, then I'd expect 50-50, for
>all plies. So the experiment makes sense only when the evals don't
>have the same (or similar) quality.
>
>You say that it's obvious, that the deeper both engines are allowed
>to search, the bigger the difference in strength. (Correct me if
>I'm wrong.)
Partly true.
I say that 6 ply depth searches are missing so much tactics
that we cannot take any list based upon that serious, as most
games you just need to be lucky with a certain tuning to do
a certain move.
A 10 ply search is a lot more serious than a 6 ply search,
the root score is influenced a lot less by tactical insufficiency.
The difference between 10 and 14 ply is already a lot more vague
to define.
>I don't disagree with you, but it's not that obvious - at least to
>me. Afterall, if both programs calculate inf. plies (or in other
>words: solve the game) then eval is of no importance. (except for the
>fact, that mate is won) This indicates a trend in the opposite direction
>you claim is obvious.
This is again something theoretical which we both will not see soon to happen.
If each 2 years hardware gets 2 times faster then to search from
the current say 10 million nodes you search you need to search like
10^10 more *at least*.
10^10 = 2^33
that is expressed in years: 33 * 2 = 66 years from now.
That is 2065. I'm getting 92 years old by then.
>Maybe (prolly) I missed other things which move the trend in your
>direction, but is this only un-obvious to me and Will? And what
>are these things I'm missing?
>Kind regards,
> -sargon
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.