Actually, I created a script just for the heck of it to test how well the RND functions.. turns out pretty well. I haven't done a proper statistical analysis, but... this script at least tests that there is convergence when taking the average as the number of rolls approaches infinity.
code:
break on
$avg1 = 0
For $g = 1 to 100
SRND(@MSECS)
$x = 1000
DIM $Y[$x]
$Z = 1
For $n = 0 to $x
$j = RND($Z)
$Y[$n]=$j
Next
$j = 0
For $n = 0 to $x
$j = $j + $y[$n]
Next
$avg = (CDBL($j)/($x+1)) * 2
$avg1 = $avg1 + $avg
Next
CDBL($avg1)/100
Brian
