From: Larry Kaufman COMCAST NET> Date: 30 may 2003 Subject: Re: Dual ratings ----- Original Message ----- From: "Daniel M. Toebbens" HMI DE> To: TECHUNIX TECHNION AC IL> Sent: Friday, May 30, 2003 5:38 AM Subject: Re: Dual ratings > Well, it seems like all people Larry knows about are dan-players. :-) > That of course makes any curve fitted to these data highly unreliable if > applied to the kyu-range. So, since I have a reliable rating in both > lists, I'll add them to the pile: > > Pan-Atlantic: 1138, Shogi Dojo: 336 > (BTW, my handle at the dojo is also Toebbens.) > > The difference here, while still very large, is of course lower then > calculated from fits derived from the previous data alone. According > to those my rating at the dojo should be negative. However, my dojo rating > at the moment is pretty stable, while my offical rating keeps rising (and > I'm pretty happy about that. :-) So, the difference might actually reach > the predicted 1000-and-something points. > > > Daniel Toebbens Thanks for the data point. We do need data from kyu players. I think that if we get more data from kyu players, we will find that the slope of the formula is higher than .5, maybe .6 - .7. That would put you right in line with expectations. We'll rerun it when we get a couple more datapoints. Larry Kaufman