I have seen a few posts about not ever seeing 6 standard deviations. I do not quite understand why 6 standard deviations in roulette.
Because (when applied to a significant sample size) it's an extremely rare event.
Here you can see the normal distribution of random outcomes in terms of standard deviations (horizontal axis):
(https://www.vlsroulette.com/proxy.php?request=nolinks%3A%2F%2Fupload.wikimedia.org%2Fwikipedia%2Fcommons%2Fthumb%2F8%2F8c%2FStandard_deviation_diagram.svg%2F500px-Standard_deviation_diagram.svg.png&hash=3f3e4764a4b08e67cb09b419f722c4f7a3d47513)
34.1% fall between 0 and 1σ (standard deviations)
13.6% fall between 1 and 2σ
and so on.
I recall only using up to 4 std.devs when programming statistical apps.
In wikipaedia you will find this table:
1 0.682689492137
2 0.954499736104
3 0.997300203937
4 0.999936657516
5 0.999999426697
6 0.999999998027
68% of values (see line 1 of the table) will be within 1 std.dev. away from the mean.
The "usual" 4 std.devs. will cover the 99.99% of cases (ie. only 1 in 10,000 will be outside this range)
With 6 std.devs. the range is so wide that only 1 in 1 billion random values will fall outside.