Author: Uri Blass
Date: 13:52:57 09/20/00
Go up one level in this thread
On September 20, 2000 at 15:53:39, Robert Hyatt wrote: >On September 20, 2000 at 14:26:26, Uri Blass wrote: > >>On September 20, 2000 at 13:06:20, Robert Hyatt wrote: >> >>>On September 20, 2000 at 01:24:30, Jouni Uski wrote: >>> >>>>On September 19, 2000 at 10:22:03, Robert Hyatt wrote: >>>> >>>>>On September 18, 2000 at 17:42:06, odell hall wrote: >>>>> >>>>>>Hello CCC >>>>>> >>>>>> >>>>>> How many think the new SSDF List is relatively Accurate? Personally I commend >>>>>>SSDF for doing a outstanding Job, Based on my observations of the grandmaster >>>>>>Challenge and other 40/2 events, I think the list is very reliable. >>>>>>I believe it is safe to say that any top program running on a K62-450 is 2500 >>>>>>elo, or very near. I think that now that the rating has been significantly >>>>>>lowered, this list will be taken far more seriously in determining Fide rating >>>>>>for Modern Programs. I am curious if some skeptics of the List in the Past, >>>>>>consider the list still to high, or Just about right? Opinions Welcome >>>>> >>>>> >>>>>For anyone that understands the statistics in the Elo system, the correction >>>>>makes zero sense. There is no way to take two totally different rating pools, >>>>>and adjust one set of ratings to make them comparable to the others. As has >>>>>been pointed out so many times here, the _difference_ between two ratings is >>>>>the _important_ piece of data. Not the raw ratings themselves. All this >>>>>correction does is attempt to correct highly over-rated numbers on the SSDF >>>>>list. Adjusting _all_ ratings on their list changes _nothing_ of course. And >>>>>it still doesn't mean that their ratings have anything to do with FIDE ratings, >>>>>other than the top programs are now a bit closer to "reality ratings"... >>>> >>>>But are You still claiming programs are NOT at GM level? >>>>In reality they 100% sure are. >>>> >>>>Jouni >>> >>> >>>100% sure based on what? 100% opinion? >>> >>>The crux is this: >>> >>>If a GM plays "normal chess" vs a computer, then the computer will do OK. >>>It will _not_ run the GM off the board, but the computer will certaily play >>>and produce results as if it were a GM. >>> >>>If a GM plays "anti-computer chess" then this changes. And the machines will, >>>in general, look silly. They will fall into silly draws. They will fall into >>>silly losses. They will still win a few. But they won't be confused with a GM >>>in any way. >> >>I think that if they play with super GMs of 2700 then they can score less than >>50% and still get the GM norms. > > >Sure they will. On occasions. Statistics not only predicts this, but demands >it. But they won't come often, IMHO. > >> >>Deep Junior did 4.5 out of 9 against average rating of more than 2700 and part >>of the GM's tried to play anti computer chess. > >Some tried, most didn't. Of the ones that tried they had more successes than >those that didn't try. > > >> >>Part did not succeed to win. >>Anand tried to play something that works against Fritz but Junior is different. >>Adams tried to play something that worked in blitz against Junior but tournament >>time control was a different story. >> >>I guess that part of the GM's prefer to play the board and they will never play >>anti computer chess. > >For the time being. But when they begin to get dislodged from finishing where >they want at a tournament, I'll bet they begin to "study the problem" a bit more >seriously. > > > >> >>3.5 out of 9 in 2 similiar tournaments is going to be enough to get the GM >>title. > >I haven't seen a tourney where 3.5 of 9 would be a GM norm, unless you only >include the top 8 GMs in the world (maybe). I agree that usually 3.5 out of 9 is not enough but 3.5 out of 9 is enough if the average rating of the players is above 2700. This was exactly the case in durtmond when kasparov did not play. You need to play against some of the best players of the world but you do not need to include only the best 8 players. However, 2 norms isn't the _only_ >requirement. They have to maintain a GM rating in the FIDE list for the time >they are working on the norms. I think that this task is easier. They need rating of at least 2500 when for a GM norm they need performance of 2600. They can earn the 2500 rating in tournaments against weaker players even if they do not get the GM norm against them. > > > > >> >>It may have problems to get the GM norm in tournaments against weaker players >>who play for a draw inspite of getting more than 50% but players can choose the >>tournament they play to get the GM norms and they can play only in tournaments >>when their rating is lower than the average rating. >> >>Uri > > > >Do you think top-level tournaments will allow computers that have not competed >in the regular tournaments to play? They wouldn't allow a human in. But they >might allow a computer, _if_ the sponsor contributes enough cash to make it >"interesting" I suppose. I think that they are going to allow humans who never compete in regular tournament if they show evidence that they are strong enough(I believe that sponsors are going to let humans who prove that they can beat chess programs to play in strong tournaments but I do not know about humans who avoid playing in regular turnaments and prefer to beat programs). Uri
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.