Roulette problem

[Galteeth]

Ok. Say the minimum bet on a roulette wheel is five. Assume an American roulette wheel with 38 numbers, two of which are zeros (non-colored).

You're going to be betting on a color so the odds of winning an individual spin is 9/19.

If you win, you reset to a minimum bet. If you lose, your next bet is double your total losses, or past the first bet, triple the previous. In other words, the bets look like this: 5, 10, 30, 90, 270, etc...

What are ways to model the odds of winning money in say, 100 bets, at some maximum bet limit where losing said bet will also cause you to revert to your minimum bet?

[zli034]

You can use data modeling and simulation, on R or Excel.

[bpet]

Sounds like a variant of the Martingale betting strategy - the articles on that might give you an idea how to work out things like expected winnings, probability of bankruptcy etc.