We are going to fit a curve to the ELO system to determine what works best for the actual players on CGOS. Steve Uurtamo is helping me with this.
I really like the idea of rating entities separately, but for now CGOS won't be doing this - it would require a good bit of reworking of CGOS to make this work right. So the system will use a formula based on rank. You might get a different gap between 2 stones and 3 stones and also you are likely to get a different gap between a weak player and a strong player. - Don On Sat, 2006-12-30 at 13:21 +0100, [EMAIL PROTECTED] wrote: > I think it is good to know that the EGF rating system is a modified > version of the ELO rating system. AFAIK the ELO rating system links a > 100 point difference to a winning percentage of 68% (corresponding to > 1 standard deviation in the normal distribution). The EGF rating > modifies this by rank dependent weights so that a 100 points > difference corresponds to 1 rank (dan/kyu) difference. I think the > world champion would have an EGF rating of about 3000 points, > corresponding to an amateur 10d. > > I am currently investigating the actual statistics. At 7d a rating > difference of 500 points corrresponds to a winning percentage of > 99% and near 10k it corresponds to a winning percentage of 75%. I also > have to add that these percentage are biased, because of the fact that > a tournament game took place in a european tournament which are mostly > paired in the MacMahon way. This means that the stronger player is > doing badly or the weaker player is doing well (or both) in this > particular tournament. But anyway, this difference is accounted for in > the EGF, because otherwise the corresponcence of 100 point to 1 rank > would be lost over time. > > Dave > > ----- Oorspronkelijk bericht ----- > > Van: terry mcintyre <[EMAIL PROTECTED]> > > Datum: vrijdag, december 29, 2006 10:16 pm > > Onderwerp: [computer-go] multiple entities may complicate things > > > Don proposed creating multiple entities for programs playing at > > different handicaps. > > > > That seems to complicate things. Is it possible to factor > > handicaps into elo-style ratings? > > > > We might start with some assumptions that, for example, 100 elo > > points is comparable to a one-stone handicap, test the > > assumptions, and make adjustments as appropriate. Given enough > > games, it might even be possible to answer the question of whether > > the handicap stones vs elo rating is best represented as a linear > > or nonlinear function. Some suggest that the answer would depend > > on the ability to use the handicap stones to good effect. > > > > There is some extensive discussion of existing rating systems here: > > > > http://en.wikipedia.org/wiki/Go_ranks_and_ratings > > > > Terry McIntyre > > UNIX for hire > > software development / systems admin istratio > _______________________________________________ > 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/
