Die roll generator - how random is it?

I am fairly new to Vassal – having played 6-10 games. Some opponents use the Vassal dice system – some do not, as they don’t think it is random enough. So because I am a wierdo I counted the Vassal dice rolls in a recent game – the following is the results from about half way through when I quit counting:
The 1st number is the number is the average, the 2nd is the actual. So for DR1 the expected average (11.5) is the same as we actually rolled (11); whereas for DR4, we rolled higher than average (19 rolls v expected 11.5)
D6 (69 rolls): (16.7 is average)

  1. [11.5, 11]
  2. [11.5, 15]
  3. [11.5, 10]
  4. [11.5, 19]
  5. [11.5, 11]
  6. [11,5, 12]
    D12 (144 rolls; expected avg in parins)
  7. [4.0, 8]
  8. [7.9, 8]
  9. [11.9, 14]
  10. [16.0, 13]
  11. [19.9, 23]
  12. [24.0, 26]
  13. [19.9, 17]
  14. [16.0, 12]
  15. [11.9, 12]
  16. [7.9, 5]
  17. [4.0, 4]

So what do you think, how random is it?
For my part, it seems that overall it may be ok, but it seems like it rolls similar numbers in bunches.

Thus spake korvessa:

I am fairly new to Vassal – having played 6-10 games. Some opponents
use the Vassal dice system – some do not, as they don’t think it is
random enough. So because I am a wierdo I counted the Vassal dice rolls
in a recent game – the following is the results from about half way
through when I quit counting:
The 1st number is the number is the average, the 2nd is the actual. So
for DR1 the expected average (11.5) is the same as we actually rolled
(11); whereas for DR4, we rolled higher than average (19 rolls v
expected 11.5)
D6 (69 rolls): (16.7 is average)

  1. [11.5, 11]
  2. [11.5, 15]
  3. [11.5, 10]
  4. [11.5, 19]
  5. [11.5, 11]
  6. [11,5, 12]
    D12 (144 rolls; expected avg in parins)
  7. [4.0, 8]
  8. [7.9, 8]
  9. [11.9, 14]
  10. [16.0, 13]
  11. [19.9, 23]
  12. [24.0, 26]
  13. [19.9, 17]
  14. [16.0, 12]
  15. [11.9, 12]
  16. [7.9, 5]
  17. [4.0, 4]

So what do you think, how random is it?
For my part, it seems that overall it may be ok, but it seems like it
rolls similar numbers in bunches.

This far from the sample size one needs to judge the quality of a random
number generator.

If you do an analysis on a few million rolls—which I have, several times
—you find that the pseudorandom number generator we use produces results
which are nearly indistinguishable from a true random number generator.


J.


messages mailing list
messages@vassalengine.org
vassalengine.org/mailman/listinfo/messages