Last night I coded a weighted roulette system of sorts with Math.Random() and let the program run for over a million spins.
Here is the result in comparison to the theoretical numbers on the notepad
Whilst I was making a spin system for another game I got a mythical class twice in a row which has a 4E-8 chance of happening so I assumed that Math.Random() may have bias but ig not. Probabilities are quite fascinating.