Hideki Kato wrote: > Hi Don, > > There are not enough evidence to believe this. > I realize that - and I don't expect it to be perfect. But it's a starting point. I would like to put something up pretty soon just to have something to debate about :-) Seriously, if we make some kind of stab at it there is no harm done. I would of course put a disclaimer on the main web page and make it clear that this formula is highly speculative and that it will improve as we get more information.
- Don > Tast-3k has too few matches against each program, less than ten games > and has no matches against strongest programs including Crazy Stone, > MoGo and greenpeep. In addition, there seems some bias, that is, > his winning rate against gnugo-3.7.10(s) is less than 50% but is > 100% against FatMan while their ratings are almost the same, about > 1800 ELO. > http://cgos.boardspace.net/9x9/cross/tast-3k.html > > -Hideki > > Don Dailey: <[EMAIL PROTECTED]>: > >> I'm going to estimate that 100 ELO is roughly 1 rank based on this: >> >> http://en.wikipedia.org/wiki/Go_ranks_and_ratings >> >> This may not hold for 9x9. If a 1 kyu beats a 2kyu about 64% of the >> time in an even game at 19x19, it doesn't imply that he will do the >> same at 9x9, but until I have a reason to believe differently I will >> set this as our first guess in the formula. We can always come back >> and modify this later. >> >> We may be able to borrow KGS data of well established players playing >> 9x9 games against each other to estimate this. Would anyone like to >> volunteer to do this? >> >> So as it stands, our formula is: >> >> 3ky = 1942 cgos_elo >> >> dan = (your_elo - 2142) / 100 >> kyu = -dan + 1 >> >> 1 Dan = 2242 cgos >> 1 Kyu = 2142 cgos >> 2 Kyu = 2042 cgos >> 3 Kyu = 1942 cgos >> >> The strongest bot on CGOS all time list seems to be greenpeep0.5.1 >> <http://cgos.boardspace.net/9x9/cross/greenpeep0.5.1.html> with a rating >> of 2621. >> >> This implies it is almost equal to a 5 Dan player - which doesn't sound >> right to me. However, this could be fluky since it is as at the >> extreme end of the scale. It would be great if this same program >> could play some strong humans at the equivalent time control on KGS at >> 9x9 and we could adjust the difference between ranks accordingly. >> >> I suspect there is more than 100 ELO between ranks at 9x9. >> >> - Don >> >> >> >> Don Dailey wrote: >> >>> Christoph, >>> >>> Your bayeselo rating is 1942 on CGOS. I compiled a table that has >>> all players with 50 games or more which can be found here: >>> >>> http://cgos.boardspace.net/9x9/hof2.html >>> >>> >>> - Don >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> Christoph Birk wrote: >>> >>> >>>> On Tue, 11 Dec 2007, Don Dailey wrote: >>>> >>>> >>>>> Christoph, >>>>> Let me know when you are finished, what name you are playing under and >>>>> I will do the bayeselo thing to get a better figure. >>>>> >>>>> >>>> I am playing using the 'tast-3k' account. Right, now I have 71 games >>>> and a rating of 1979 ELO. >>>> >>>> >>>> >>>>> Also, I can >>>>> throw out any games that were irregular if you can identify them, such >>>>> as if a match started when you were not looking or your interface got >>>>> glitchy or something. >>>>> >>>>> >>>> Since I added the GUI I lost no games due to software problems. >>>> Only a few "won" games lost to human stupidity :-) >>>> >>>> I will take a break over the holidays, maybe playing a few more >>>> games in the new year, but I guess for my purposes a zero-point >>>> "3k-AGA ~=~ 2000 CGOS-ELO" >>>> is close enough. Unless we get some other (AGA or KGS) rated >>>> players it not make sense to get a more precise rating for the >>>> scale. >>>> >>>> Christoph >>>> _______________________________________________ >>>> computer-go mailing list >>>> computer-go@computer-go.org >>>> http://www.computer-go.org/mailman/listinfo/computer-go/ >>>> >>>> >>>> >>> >>> >> _______________________________________________ >> computer-go mailing list >> computer-go@computer-go.org >> http://www.computer-go.org/mailman/listinfo/computer-go/ >> > -- > [EMAIL PROTECTED] (Kato) > > _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
