Author: Robert Hyatt
Date: 10:31:49 01/05/00
Go up one level in this thread
On January 05, 2000 at 01:05:20, Peter Kappler wrote: > > >>2 ) 1/2 hour per player per game - ChessTiger would beat everyone but would lose >>some games >> > >I'd still take Kasparov or Anand in a G/30 match against any micro, though it >would certainly be competitive. > >FWIW, I think most people underestimate the talent gap between someone like >Kasparov and an ordinary GM. The rating difference is roughly 300 points, which >is a massive difference in strength. > > >>3 ) 1 hour per player per game - Things become closer >> > >I think the super GMs (top 10 in the world) would be clearly ahead here. > > >--Peter There is _absolutely_ a "rating dilation" at the top end of the scale. It is obvious as to why. If you are rated over 2800, you will _never_ play anyone with a higher rating than you. You will never play anyone with a rating even close to you, so your rating has a harder time moving up... if you are rated at 2600, there are _plenty_ of players rated higher than you so every loss doesn't destroy your rating to the tune of 32 points. IE IMHO if you compare a 2600 to a 2400, and then a 2800 to a 2600, I would claim that the 2800 player is farther ahead of the 2600 player than the 2600 player is ahead of a 2400 player. I see this on ICC regularly. It is _very_ difficult to maintain a computer rating that is > 3000, when you play GM players that are 2600. Because when you win, your rating doesn't change, but when you lose, -32. There is only _one_ way your rating can go under such circumstances...
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.