Author: Janosch Zwerensky
Date: 06:42:22 01/26/02
Go up one level in this thread
On January 25, 2002 at 16:46:38, ALI MIRAFZALI wrote: >I would like to know the opinion of the readers of this forum on the following >questions. >1.What would be the Elo rating of the perfect chessplayer? Depends on the perfect player and the population of rated opponents his rating is to be measured against. I can imagine perfect players who would get very near to winning every single game against a human GM population just as well as perfect players who often don't get more than a draw against the average club player. The difference between these two types of perfect players would be that the former would chose *aggressive* and *complicated* perfect lines while the latter ones might not do anything to get more than draw when a draw is possible. >2.Are there natural limits to the strenght that can be achieved in chess >for a computerchess player? There certainly are limits, since the set of all physically implementable computer players is finite. But where these limits are or even how the limit playing strength should be quantitatively described, I do not know. To be honest, I don't even think we know exactly what computer playing strength can be achieved *on currently available hardware* (though educated guesses are possible on that one). If there is no substantially better way algorithmically to do chess in computers and no substantial improvements to be expected from new types of computing devices (like quantum computers), then one could of course try to determine how fast a last generation supercomputer might possibly run, considering realistic physical constraints on supercomputer design, and plug that into one of the formulas around which try to describe how speed improvements correlate to strength improvements. I wouldn't trust such an approach too much though.
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.