Author: Robert Hyatt
Date: 05:22:24 10/22/98
Go up one level in this thread
On October 22, 1998 at 03:03:19, Dave Gomboc wrote: >On October 21, 1998 at 17:29:00, blass uri wrote: > >> >>On October 21, 1998 at 16:29:01, Robert Hyatt wrote: >> >>>On October 21, 1998 at 09:18:57, blass uri wrote: >>> >>>> >>>>On October 21, 1998 at 08:07:53, Robert Hyatt wrote: >>>> >>>>>On October 21, 1998 at 04:37:38, Nouveau wrote: >>>>> >>>>>> >>>>>>On October 20, 1998 at 12:13:16, Dann Corbit wrote: >>>>>> >>>>>>>On October 20, 1998 at 10:37:36, Nouveau wrote: >>>>>>>>On October 20, 1998 at 01:36:22, Jouni Uski wrote: >>>>>>>> >>>>>>>>>Here's result for 20 games match with 60/5 time limit (under Winboard): >>>>>>>>> >>>>>>>>>Comet 0.5 0 1 0 0 0 1 1 0 1 0 0 0.5 0 0.5 1 0.5 0 1 0 = 8 >>>>>>>>>Wcrafty 0.5 1 0 1 1 1 0 0 1 0 1 1 0.5 1 0.5 0 0.5 1 0 1 = 12 >>>>>>>>> >>>>>>>>>So they are very close to each other in playing strength. >>>>>>>>> >>>>>>>>>Jouni >>>>>>>> >>>>>>>>12-8 is very close ?????????? >>>>>>>> >>>>>>>>When can we say : Crafty is better than Comet ? 18-2 ? >>>>>>>> >>>>>>>>I don't understand these statistical stuff : I can't imagine a 12-8 result in a >>>>>>>>match between 2 GM with a conclusion like "They are very close in playing >>>>>>>>stregth". >>>>>>>> >>>>>>>>Why do we need hundreds, maybe thousands of games between computers to evaluate >>>>>>>>relative strength, when few dozens are more than needed for human GMs ? >>>>>>>Any strong conclusion from a single match is faulty. It could be that Comet is >>>>>>>500 points above Crafty, or 500 points below (although both of these are >>>>>>>statistically very unlikely, really, very little has been demonstrated at this >>>>>>>point from a single set of games). >>>>>> >>>>>>Just imagine : the match between Kasparov and Chirov takes place and the result >>>>>>is : Kasparov-Chirov = 12-8. >>>>>>Maybe Kasparov is 500 points above Chirov or 500 points below...Show me any >>>>>>chess magazine that would print such an affirmation. >>>>>>I know, those chess journalists don't have a clue on science and stats ;o) >>>>>> >>>>>>> The international chess bodies like FIDE >>>>>>>have definitely got it right in the way that they perform evaluations using the >>>>>>>ELO method. Also, in requiring a long period of excellent results to become a >>>>>>>GM. >>>>>> >>>>>>Can someone make the math for this : a player has a 2600 level but no rating, >>>>>>how many games against a 2500 opposition does he need to reach 2600 ? >>>>>> >>>>> >>>>> >>>>>easy here. one game. his rating would be 2700 after that one game, since >>>>>the first N games uses the usual "TPR" type calculation. >>>> >>>>after 1 game you have no rating >>>>you need at least 9 games to have a rating (not important if it is 2005 or 2700 >>>> >>>>players who have 2005 rating need at least 30 games if you assume they cannot >>>>earn more than 20 elo in one game. >>>> >>>>Uri >>> >>> >>>No idea about FIDE rules about ratings, but the USCF publishes ratings after a >>>single tournament. I have known players with ratings like this: 2244/4, which >>>means provisional with only 4 games played so far. And during the provisional >>>period, the rating can fluctuate dramatically because the formula is simply >>>the sum of the ratings of the opponents you beat (+400 for each one) plus the >>>sum of the ratings of the opponents that beat you (-400 for each one) plus the >>>sum of the ratings of the opponents that you drew, divided by the total games >>>counted. IE performance rating. And in that light, beating a GM gives you his >>>rating+400 after one game... >>> >>>And I assume most use K=32 nowadays (at least the ratings I have seen do this) >>>which means you can go up/down up to 32 points in one game... >> >>I remember that in fide rules K is not constant and it is bigger for players >>that did not play many games. >> >>I do not remember the exact rules but I do not know about K=32 in fide rules and >>I think that K is smaller for everyone. >> >>I know that I needed at least 9 games against rated players to get a fide >>rating. >> >>I needed 2 tournaments for a fide rating because in the first tournament I had >>not 9 games against fide rated players. >> >>Uri > >USCF minimum might be 3 or 4 games. CFC minimum is 4 games before they'll >publish a provision rating. FIDE wants 9 games. > >If I recall correctly, FIDE's K factor was 12 back when it was constant. This >is a far cry from the USCF's 32. (The CFC used 32 when <2300, then 16 when >>2300. Recently they diddled with their formulas, I didn't pay much attention >to exactly how, but they did move the cutoff to 2200.) > >At any rate, all interesting modern hardware and software combinations are >strong enough that if you were going to treat them as players that could vary in >skill over time, a K factor of at most 12 would be appropriate. Since the >variablility of computer play is so low relative to the variability of human >play, perhaps a K of 4 (hand-wave) would be more realistic. > >Dave Gomboc Actually, I would change K differently. It was originally set at 32 because this gave reasonable rating changes when people might play one rated event a month, at most. Not bad... Now, particularly on servers, people can play 10-20 rated games in an evening... and a K=4 or whatever would definitely be an improvement, since your rating won't fluctuate nearly so wildly, and it shouldn't when you think about it. IE I think K should be set (like the Glicko system) based on the number/ frequency of games...
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.