Operating System:Windows 7
Operating System Bit:64bit
Software Version:3.8.9
Problem: Random Processing generates random sources, but not randomly enough.
Test1: I ran 10 trials, with 10 different sources, to generate 4 Columns of Random Data.
Column1 = Words
Column2= Words
Colum3= Words & Numbers
Column4= Words & Numbers
There was mostly double repeats per column of data.
I repeated the process and had the same results.
Then for the final test,
I increased the sources to 20, and ran 10 trials, to generate 4 Columns of Data
(for a potential work around, to generate results more randomly til this BUG is fixed)
Result:
Column1 had 1 repeating word 3 times (firstword x3)
Column2 had a set of 3 duplicate words (firstword x2, secondword x2, thirdword x3)
Column3 had a set of triples, 2 doubles of numbers (#1 x3, #2 x2, #3 x2)
Column4 had a quad and double (#1 x4, #2 x2)
What this means is each column had 20 potential different values to be selected.
The odds of picking the same value has a 1/20 probability.
The bot randomly picked the same number at least 33% percent of the time, which is a failure, considering each number had a .05% chance to be picked. In this final test, no number should have been randomly picked twice, and even then, it should have been only 1 occurrence at maximum
Solution: Random Number Processing needs to be highly optimized for more randomly generated data.
Note: The results does show that letters/words have a lesser chance of being selected repeatedly, whereas numbers have a much higher chance of being selected repeatedly.