Recently I started the thread called "Advanced Roulette Strategy," which looked at a graduated-bet method (no, NOT a Martingale) taken from some web site. Of course, it was easily broken, as I expected. The killer was that while you would consistently make small gains, eventually you would have the inevitable "big loss" that would wipe it all out. But I took the approach and re-built it in Excel, with different numbers. What you have to do is carefully calculate the bet size increment to PREVENT the Big Loss from happening, or at least make it statistically improbable to the point of insignificance. This was not difficult to do, if you're a decent Excel hacker. Instead of the original 20 games, I ran to 30 games. Since you're betting on 7 numbers, the odds of one of them winning is about 1 in 5. Going 30 spins without hitting one of them is extremely unlikely. I did it one single time in 5,000 simulated spins. Of course, going that far (30 spins) requires a bigger bankroll, a little over $15K, even if you're just starting with $1 per number. I also calculated a table going up to 46 spins. Going 46 spins without hitting 1 of 7 numbers is virtually impossible. And again, if you have the initial bankroll, you can prevent the Big Loss from happening. Of course, I'm using some common sense math here. It's statistically possible to have 1,000 spins without hitting one of those 7 numbers, in which case to survive, your bankroll would have to be equal to the entire United States GNP, but I can live with that improbability... It goes without saying that to use this approach, you'd be playing in the Whale Room. Of course there may be some limits on initial bankroll, or maybe the willingness of some casinos to give you $100,000 plaques, but these are logistical factors that have nothing to do with the system's statistical viability. It's quite obvious, though, that if you had a sufficient bankroll (probably in 7 figures), you could make a very nice 6-figure annual income playing roulette. There are no approaches to roulette that will enable a low-stakes player (e.g. $1 chips) to beat it and get rich off of it without self-destroying. That is not to say that roulette can't be beat, however. It can, if you have the resources to get your foot in the door. I'd be happy to share the Excel file with anybody who wants to discuss it. Send me a PM.