Re: this sounds like a whine I know.....

Rayola said on 11/26/2003 9:45 AM:

> Can anyone put their hand on their heart and say that fibs is a truly random
> program.  Seems to me there are far too many doubles thrown and weird throws
> which favour the defending player, especially coming off the bar.  This
> happens both ways so it's not a whine it's just a bit freaky.
>
> Ray.
>
>
>  
==========
I've started keeping a spreadsheet in Excel with the analysed results of 
matches on FIBS.

I analyse them with gnubg, of course. Then I enter the opponents name, 
guestimated FIBS RATING, luck ( per move ), my guestimated FIBS rating, 
my luck ( per move ) and whether I won or lost and the score.

I'm doing it for 1,3,5, and 7 point matches.

I can then get an average for my opponents FIBS rating ( guestimated ), 
my FIBS ( guestimated),  and opponents luck and my luck.

Over a range of matches, if FIBS is truly random, everyone's luck factor 
should equal 0 ( zero ). I don't know how many matches it would take to 
be a statistically valid sample though. But I would think that over a 
100 matches average luck should be pretty close to zero, but it might 
take a larger sample than that for luck to even out, I'm not sure. But 
if it doesn't even out it will give me a good idea if I've been lucky or 
unlucky. I used Luck Per Move instead of Luck Overall because it gives a 
larger, more understandable number. Luck Overall gives a lot of numbers 
in decimal points instead of integers. I think Luck Per Move is easier 
to understand also as well all know all about bad unlucky rolls :-(  

At this point I have only 33 FIBS matches entered and my GNUBG 
guestimated FIBS rating is 160 points higher than my actual FIBS rating. 
It may synchronize over a larger sample.

--
Clinton lied ... and ... Hillary cried.

Bush lied ... and ... thousands died.