On Mar 10, 5:58=A0am, Stig Holmquist <stigfjor...@[EMAIL PROTECTED]
> wrote:
> On Sun, 9 Mar 2008 15:56:47 -0700 (PDT), Mensanator
>
> <mensana...@[EMAIL PROTECTED]
> wrote:
> >On Mar 9, 2:39?pm, Stig Holmquist <stigfjor...@[EMAIL PROTECTED]
> wrote:
> >> If you add up the numbers ein each position after they have been
> >> sorted in ascending order
>
> >Why would you do a silly thing like that?
>
> >Wouldn't doing stats on which ball is drawn first
> >be more meaningful?
>
> If you think stats based on order of draw are more meaningful
It could be. Suppose the two most frequently occuring balls
drawn are 17 & 18. But they always come up as the first ball
drawn and almost never in the same drawing. Thus, it may be
that 17 & 18 are competing for the same bucket (which could
happen if the balls are loaded into the drum in numerical order
and not stirred properly as was the case in Illinois in 1984).
> feel free to calculate them
Already have.
> and tell us what is so interesting about them.
Such as actually coming up with winning numbers?
> .Order of draw can be found at various sites. Go ahead.
I won the Illinois Lotto twice. Only match 4's however,
so I only got about $1000. And shortly thereafter they
modified the Lotto machines to make them much more random
and render such statistics meaningless.
>
> Complete set of means for 264 draws is 9.6-19.1-27.8-38-46.9
And if the stats say they're indistinguishable from random,
then what? Stats are only useful if the machine is not a
true random number generator where the non-randomness can
be exploited.
>
> Stig Holmquist
>
>
>
> >> you will find after the last 112 draws that
> >> the means are 9.6, 18.7, 27.4, 37.1 and 46.6 and tere appears to be a
> >> slow decrease in each mean from earlier draws.
>
> >> It is tempting to suggest that the ultimate means after millions of
> >> draws would be 9-18-27-36-45, but to verify this wild guess one would
> >> need to do a simulation or else apply some kind of probability
> >> calculation, which is beyond my ken.
>
> >> The 55 ball numbers for PB can be divided into five equal ranges:
> >> 1-11, 12-22, 23-33, 34-44 and 45-55 which have the means :
> >> 6-17-28-39-50. The fact that the observed means differ from the
> >> nominal means is due to many balls drawn for position 1 are over 11
> >> and there are no numbers below 1 to compensate and likewise there
will
> >> be many numbers below 45 drawn for the 5th position but no numbers
> >> over 55 to compensate.
>
> >> Would it be possible to predict by probability calculation what the
> >> ultimate means should be?
>
> >> ?Stig Holmquist- Hide quoted text -
>
> - Show quoted text -
|
