Re: [Bug-gnubg] grandmaster with 3ply

[Louis Zulli]

Hi,

If I recall, for a binomial distribution with n=100, p=.5 and q=.5, the 
standard deviation is sqrt(n*p*q) = 5. So having gnubg start first 80 times 
represents 6 standard deviations above the mean. The probability of this 
happening by chance is very, very small! 

Maybe you were simply very, very unlucky. Or maybe your memory is fooling you.

For my dice generation, I use a random number generator (Mersenne Twister). I 
have no sense that gnubg wins the first roll more than I do, and I have played 
hundreds (maybe thousands) of games.

Louis


On Jan 25, 2010, at 7:07 AM, Filipa Andrade wrote:

> Hi,
> 
> I am playing against gnubg in grandmaster level with 3ply.
> I find it strange that only around 20% of the times I start. 80% is gnubg 
> that starts. Is it normal? I have made a sample of more than 100 games.
> I a normal 7 games match it happens often that I never get to start first. 
> Which are the best dice settings?
> 
> Thank you for your help and thank you so much for gnubg, I am learning a lot!
> _______________________________________________
> Bug-gnubg mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/bug-gnubg