Some years ago around 1985 we had people who were using primitive computer software to predict the outcome of roulette. This was not their idea, Ed Thorp the noted blackjack player and stock market speculator was coming up with a similar idea twenty years earlier. But what Thorp suffered from was not a lack of knowledge to make it work, it was a lack of software and hardware at that time. Even in the late seventies and early eighties the problem wasn’t all that much better.
So when you consider the massive advancements of computer software and hardware over the past ten years especially then I can only believe in my own mind that there are people out there who are using highly sophisticated computer programs to gain an edge on roulette by predicting where the ball will land. But do not misunderstand what I am trying to say here, when I am referring to “prediction” I am certainly not saying that if a ball lands in numbers 17, 21, 32 and 8 in succession that a computer could predict that.
There is too much randomness in the results of roulette wheels for that to ever be possible. But just like with stock market specialists, some sort of order can be made from chaos. Prediction means long term profitability and not 100% accuracy. If you make money over time in the stock market then this does not mean that you have or even can predict ever possible movement or every top and every bottom, that simply isn’t possible. But in roulette prediction then being able to predict that just two numbers out of 37 wouldn’t arrive on a certain spin would be enough to guarantee an edge. If a computer predicted that the result of a spin would be one of 35 numbers in a 37 number wheel then you are approaching it wrongly if you fail to make money in that situation.