Author: Mark Young
Date: 14:59:09 07/01/01
Go up one level in this thread
On July 01, 2001 at 17:20:18, Hristo wrote: >On July 01, 2001 at 15:38:11, Mark Young wrote: > >>On July 01, 2001 at 15:20:12, Hristo wrote: >> >>>On July 01, 2001 at 14:19:39, Mark Young wrote: >>> >>>>On July 01, 2001 at 13:57:22, Otello Gnaramori wrote: >>>> >>>>>On July 01, 2001 at 13:25:28, Mark Young wrote: >>>>> >>>>>> >>>>>>After that the Elo rating is how Grandmasters are ranked. >>>>>> >>>>>>2500 to 2599 Elo: standard or average Grandmaster rating. Most Grandmasters fall >>>>>>in this group. >>>>>> >>>>>>2600 to 2699 Elo: What some call the Elite Grandmaster ranks. Usually less then >>>>>>100 Grandmaster fill the Elite ranks. >>>>>> >>>>>>2700 & Up Elo: World Class Grandmaster ranks and usually includes the World >>>>>>Chess Champion(s) and challengers. Usually contains less then 10 players, but >>>>>>has gone up in recent years. >>>>> >>>>>Great reply Mark! >>>>>But now a subtle question arises ... Which of the three above mentioned >>>>>categories does the top programs belongs to ? >>>>>Probably this is the second step in definition ... >>>> >>>> >>>>The data suggest strongly that the computers are in the Elite Grandmaster ranks. >>>>Tiger like Junior is confirming this data. >>>> >>>>It must be noted every player above 2580 is a Grandmaster on the current Fide >>>>Elo list. >>>> >>>>All data suggest that computers are above the 2580 elo rating, or average >>>>Grandmaster class. >>>> >>>>That is why the GM argument is like arguing if the world if flat or round. >>>> >>>>Only subjectively could you argue that the world is flat, “Can’t you see by >>>>looking the world is flat”. >>>> >>> >>>This only makes cense if you have a "flat head" ... ;-))))) >>>If you don't I'm going to "crunch your head" and make it flat ... so >>>you argument is valid! ;-)))) ... >>>I refer to " The kids in the hall" ... thing ... just so you don't get >>>offended. >>> >>>Whats the point of claiming 2600 .... How valid is it?! Why don't you play >>>COMPS vs COMPS and tell me when you have the winner. After how many games? >>>What are the conditions for the COPMS vs COMPS tournament?! >> >>Why don't you think that strongest program are not 2600+? The data playing >>humans and human grandmasters shows the best programs to be this strong. >> >>What data do you have that counters this....other then what you would like to be >>true. >> > >:-))) >You might be missing my point! >Lets say all COMPS are over 2600 ... Whatever that might might mean! You need to find out what a rating is so you can frame a question that makes sense. BTW Its two different issues: What strength are computer programs today? What is the strongest computer program and how do you prove it. >NoW! How do you calculate the strenght of individual COMPS?! >If I tell you "Run an 11 rond tornament with all programs over 2400." >you are going to reject my sugesion, because the results might >not be correct! Fair enough! But why do you accept a 9 round tournament >,against humans, as a better measurement of COMPS strength?! >(even if it is three of them) >If you do ... then this is at best a hypocrisy ... ;-) >What does it take to prove that one COMP is better than another COMP??! >How do you do it?! Assuming human raitings have nothing to do with it! >You have won all possible human tournaments! Now what?! >Your logic must extend beyond the fact that COMPS can win against >humans! Can your COMP win aganst another COMP?! This other COMP, however, >might not have as many wins against humans as your COMP. .... >Well ... what do you do then .... >:-) >Mind, you, I'm not challenging the fact that COMPS are very (VERY) strong >... just, what is the meaning of all of this for me?! A simple human?! >For the most part, as a human, I can not gain anything from the fact that >COMP are over 2600. Simple as that. It doesn't matter ... >does the "neutrino" have mass? Who, the hell, cares! > >What other proof do you need! I can give you examples of the total stupidity >of those 2600+ COMPS, even after 10+ hours of thinking. Do you think that >1+ billion nodes will change your mind?! ;=| >You'll pick what is best to prove your point! However, the COMPS very often >consider lines that are, at best, silly! > >best regards. >hristo > > > >>> >>>regards. >>>hristo >>> >>> >>>>All objective evidence is showing the best computer programs are above 2500+, >>>>and strongly suggests them above 2600. >>>> >>>>More data coming...
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.