I agree with the patterns and cycle idea. Is it 100% foolproof? Definitely not because nothing in roulette can be. Can it be reliable enough to at least double your income? Yes, I think so!
If we keep our analysis in the realm of reality (not theory), we have a practical advantage that we can rely upon!
If the odds of an event occurring are more than many millions to one, then it is likely that we will not ever see it in our roulette lifetime, much less having it to repeat before our income at least doubles!
If I can run a computer simulation that doesn't show a loss over many millions of spins, that is very significant!