Re: [computer-go] strange results

[Evan Daniel <evanbd@xxxxxxxxx>]

On 5/11/05, Christoph Birk <birk@xxxxxxxxxxxxxxxxx> wrote:
> t.cooper wrote:
> >At 20:28 11/05/2005, David Doshay wrote
> >>Acceptable and high are relative to the person asking the question.
> >>We often will run a set of games and calculate a 95% confidence
> >>interval, and run another set of games where we get a 95% confidence
> >>interval that does not even overlap. Our solution is MORE GAMES.
> 
> >On the face of it, this seems to contradict the rules of probability.  Is
> >there some reason why this might be happening, or at least why it is not
> >as strange as it at first seems?
> 
> The reason is that we are NOT dealing with gaussian distributed random
> numbers but the results of (computer) games.

I must admit I'm confused by this.  If things are measured by
win/loss, not score, and the games are independent, why isn't it a
perfectly standard binomial distribution, on which you can use
standard statistical techniques to test significance?

Is there some reason the games aren't independent?

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