Quote Originally Posted by Archvaldor View Post
Even on an unbiased wheel the numbers will rarely have equal frequencies due to random fluctuation except over an extremely large sample size.

Assuming that all wheels are biased-which is questionable this isn't that interesting in itself. A weakly biased number with probability of occurrence of say 1/34 isn't worth betting on because of the high variance associated with a 35-1 shot even if you could actually verify it existed. A sector might be-but mainly you are looking for strong bias of <1/30.
Yup, bias sounds a strong word. The roulette wheels always show different frequencies. Hamburg Spielbank table1 showed frequencies varying from 1 to 10 in 220 spins (one day). Did the top 12 numbers show the same trend the following day? I don’t know. I might take a chance and play them the following days. Frequencies from 7 to 10. There is a real chance some numbers will repeat sooner. I not saying it’s a bias.

Lets forget mechanics for a while. Look at random generators. All generate roulette numbers with different frequencies. Every random run looks a lot like a physical roulette wheel. The wheels are random too. But being mechanical, there is wear and tear.

I have a theory. The most physical roulette dealers have the greatest impact on wheel bias. The OP should bet on his observation for a while, with the special dealers spinning. Increase the bet a bit after a loss. He should quit if losing more than he can afford to. Live to play another day.