Author: KarinsDad
Date: 22:37:46 01/31/99
Go up one level in this thread
On January 31, 1999 at 16:54:47, James B. Shearer wrote: >On January 31, 1999 at 13:03:18, KarinsDad wrote: > >>On January 30, 1999 at 01:53:19, James B. Shearer wrote: > > <snip> >>> Even Kasparov has bad games. Have you looked at the game he just lost (as >>>black in 28 moves) in the Wijk tournament? White's first 24 moves were book. >>>White's next 4 moves were good but nothing a 1800 player couldn't find on a good >>>day. >> >>Everybody has bad games. No doubt about it. And 1800 players do find 2800 level >>moves. The problem is that 1800 players do not find 2800 level moves throughout >>the entire game. > > Except in this case the "entire game" contained just 4 nonbook moves the >last 2 of which were pretty easy. Yes, and who was Kasparov's opponent? What was his rating? BTW, just because 24 moves are in the book does not mean that both opponents knew the book to that level. > What do you think are reasonable scoring (ie counting wins and draws) >expectations for large rating differences. You may be right that the model >values are far too high but I am not entirely convinced. > James B. Shearer It's hard to say. Certainly not 1 in 317 at 1000 points. It would be nice to have some data with which to ascertain this. You cannot rely on computer data or ICC data. It would have to be from the USCF or some other federation's data to give a clearer picture. When Joel Benjamin was being interviewed during the recent US Championships, he basically stated the following: A 1700 rated player would probably have a slightly better chance against a 2200 player than a 2200 player would against a 2700. The reason for this is that a 2200 player can play erratically and his playing strength can fluctuate much more than that of a GM. Hence, a 500 point difference means different things at different levels. Do not quote me here, it's as accurate as I can remember. The point that I'm trying to make is that GMs have very few weaknesses. It takes other GMs to spot them (and no, using your computer doesn't count since you would never have found the weakness without it being pointed out to you). They have studied for years and understand a position to a much greater degree than even regular masters, let alone Class A players. A superGM to a class A player is like a class A player to someone who just learned how to move the pieces. Even if a class A player did study a given superGM's games and prepared for him and found a trap, the superGM would most likely take losing a minor piece in stride and still whup up on the class A player. Odds of 1 in 317 are just blatantly incorrect for a 1000 point rating difference. Maybe odds of 1 in 31,700 would be more accurate, it would be hard to say without the data. KarinsDad
This page took 0.01 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.