Author: Albert Silver
Date: 15:03:52 02/08/00
Go up one level in this thread
On February 08, 2000 at 17:38:24, Colin Frayn wrote:
>On February 08, 2000 at 16:33:52, Michael de la Maza wrote:
>
>>Earlier today I recorded a 20-25 minute interview with Professor Robert Hyatt.
>>You can listen to the interview by:
><snip>
>
>Just listened to it - very interesting.
>
>Does anyone seriously believe what he said about computers not playing better
>with increased ply, or at least asymptotically slowing down?
It's both true and not true, it depends on the strength of the player. You have
to realize that a human player rated 2500 as opposed to a player rated 2200
isn't simply about calculating a little further. Sure, they most likely play
more consistently, but the biggest difference is in strategical play and the
game phase transitions. Notably the transition from the middlegame to the
endgame. There are certain elements of knowledge that can't be simply
compensated for with a few extra plies, thus if you get a program to destroy 50
opponents rated around 2200 and perform a rating of 2600, you are absolutely not
guaranteed to achieve that same rating performance against 2500+ players,
because these last are fighting with different weapons than their 2200
colleagues. The only way to find out how a computer will do against GMs is by
having it play GMs. Lesser players won't cut it.
Albert Silver
> I have 600 games
>against Crafty which prove the opposite.
>
>I never thought about that connected passed pawns thing either. Damn that's
>obvious.
>
>Thanks for the advice, Prof. Hyatt. That world title's mine next time :)
>*grin*
>
>Cheers,
>Colin
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.