Actually getting access to a European roulette wheel (comparable in quality to what is in use in casinos) seems the hard part to me. I ended up buying a 70's wheel, but I don't feel my measurements on this antique are good enough to start betting real money.<p>One manufacturer told me they only sold to accredited casinos, and these had to return the wheels to them at the end of their useful life. Not sure if he was just playing me or if they were actively trying to keep the wheels away from roulette-hackers.
You can do this even without a computer. There are two approaches that still work (i.e., that don't focus on biased wheels, as the casinos fixed that bug since years): Wheel watching and distance counting. The first lets you watch the rhythm of the croupier when he spins the wheel and the ball. The second just lets you count the distance between the results in the wheel. The first works way better, but requires much more training.<p>But I guarantee you: This kind of work is mind-numbing. I did this for a while in 1999 and let it go, members of our group continued with it over years and really suffered from the boredom. The same with gaming the slot machines (they have certain regularities or bugs, depending on the model) or poker and other games. So if you're a hacker, the challenge and fun wears off pretty quick and you wish you had a cubicle job instead.<p>Anyway, there are books from the mathematician and physicist Pierre Basieux, unfortunately only in German.
Last month, I was having lunch with my friend, John Boyd, and he came out with this incredible story about using computers to beat roulette in Vegas back in the 70's. He was part of Doyne Farmer's original team from UCSC, and apparently plays a reasonably big role in the book. I've known him for ten years, and he never bothered to mention it. I was so riveted that I made myself late for class plying him for more details. I'm sending him this link now; I'm sure he'll enjoy it.
The Newtonian Casino is a book that tells Farmer's story. It is fascinating and I've read it twice: <a href="http://www.amazon.co.uk/The-Newtonian-Casino-Thomas-Bass/dp/0582057523" rel="nofollow">http://www.amazon.co.uk/The-Newtonian-Casino-Thomas-Bass/dp/...</a>
Also interesting is the book "Calculated Bets: Computers, Gambling, and Mathematical Modeling to Win" which talks about one man's quest to beat the odds at Jai alai (<a href="http://en.wikipedia.org/wiki/Jai_alai" rel="nofollow">http://en.wikipedia.org/wiki/Jai_alai</a>) by modeling the game and the players. It's a fun, humorous and technical book.<p><a href="http://www.amazon.com/Calculated-Bets-Computers-Gambling-Mathematical/dp/0521009626" rel="nofollow">http://www.amazon.com/Calculated-Bets-Computers-Gambling-Mat...</a>
<i>In the 1970s, Doyne Farmer, then a graduate student, used the world's first wearable computer to beat roulette tables in Nevada, but never revealed how he did it.</i><p>Wait, I thought Claude Shannon with some help from Ed Thorpe were the first to build a wearable computer back in the 60's. Also to beat roulette. They shelved it though, because there were too many practical difficulties and there were better ways to make money - the stock market.
I read about this years ago in The Eudaemonic Pie, I believe.<p><a href="http://www.amazon.com/The-Eudaemonic-Pie-Thomas-Bass/dp/0595142362" rel="nofollow">http://www.amazon.com/The-Eudaemonic-Pie-Thomas-Bass/dp/0595...</a>
It's no surprise that it should be possible to predict the trajectory of a roulette ball with today's technology, but doing it back in 1970 is seriously impressive.