Author: Roger D Davis
Date: 14:50:21 07/14/00
Go up one level in this thread
All caveats about sampling (small number of games) aside, Dann, the remarkable thing about DJ's performance is that this is no ordinary GM tournament...there are some super GM's here. The FIDE world champion is playing, and so are players who are stronger than the FIDE world champion. Perhaps we should approach the question this way: What is the rating at which GM strength begins? Now ask yourself what number of games that Junior should have won against its opponents based on that bare minimum GM rating. Now compare that number to the actual number of games DJ won. The point is that DJ should have been stomped in this tournament, and it wasn't...it's performance has been stellar. Yeah, it's GM strength... Roger On July 14, 2000 at 15:01:03, Dann Corbit wrote: >On July 14, 2000 at 14:27:05, Jerry Adams wrote: > >>Is Deep junior Grandmaster strength? > >If the message is only to one person, then use email. > >DJ played extremely well in this tournament. One person tried anti-computer >strategy and it worked. Given enough games, will the GM's figure out ways to >beat it? > >Nobody knows. > >I think it very likely that DJ is GM strength. However, just by using the >qualifications for a regular GM, DJ clearly does not qualify as a GM. > >However, I do feel that Dortmund could be considered as earning one GM norm >(whether the authorities recognize it or not). > >To become a GM, many games must be played. Suppose that someone memorizes a >particular opening, having discovered two or three novelties, and uses it to >breeze through a tournament. However, subsequent analysis of the games shows >the future opponents what these novelties are, and also, they try to avoid his >favorite opening lines. Will he fair as well in the next tournament? Probably >not. There is a reason that it takes a lot of games to become a GM. That is >because a few games may or may not have a strong connection with a player's real >strength. > >Is Deep Junior of GM strength? >For this tournement the TPR says 'yes'. >But we don't know the answer to that question. Not scientifically anyway. > >There are some people who feel strongly that by looking at a single game, they >can accurately gague the true strength of a chess player. > >Maybe they can, but not with scientific measure. And (I feel) they can be >wrong. At any rate, eventually -- if DJ is allowed to play enough games -- we >will certainly know if it is of GM strength. > >For now, it is still "How many angels can dance on the head of a pin?" > >In other words, lacking sufficient data you can draw any conclusion you like. >It might be the right conclusion. It might be the wrong conclusion. Time will >tell. > >If I were a GM and facing computers in a future chess match, I would read the >Anti-computer chess page, and study the games. I would try strategies against >computer opponents and learn how to beat them. I think eventually computers >will be the best players in the world. Whether they are GM's or not, we cannot >be certain, but evidence is mounting. Will they hold their lofty spot once the >GM's learn to play them? I think not. They will be toppled, only to rise again >and eventually surpass them. But that is not a scientific judgement, only my >impression.
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.