Re: [computer-go] Statistical Significance

[Gunnar Farnebäck <gunnar@xxxxxxxxxxxxxx>]

Chrilly wrote:
> There is general consensus  in chess to use the Elo system.

>From what I have seen there is very little concensus about rating
systems in go.

> What is - concerning the rating system - the difference between Go
> and chess?

The main difference is that go has handicap stones. While rating
differences in both chess and go can be used (according to model) to
predict win probabilities, in go they should also tell how many
handicap stones are needed to even out the win probabilities. Of
course, for this to work the data set needs to include handicap games.
Otherwise I don't see much difference with respect to rating between
chess and go, except possibly that draw is a much more common result
in chess than in go, in case that matters.

/Gunnar
_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/