A PROMINENT mathematician famous the world over for successfully turning the odds of roulette against the house has broken his decades-long silence about how he achieved the coup.
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.
Now he has decided to break his long silence after a pair of researchers, inspired by his story, developed and published their own method of beating the house.
"I kept silent because I did not want to communicate any information that might prevent anyone from taking the casinos' money," writes Farmer, now at the University of Oxford, in a draft paper that he showed to New Scientist. "I see no good reason for staying silent any longer."
Farmer's paper is a response to recent research by Michael Small from the University of Western Australia in Perth and Michael Tse from Hong Kong Polytechnic University, submitted to the journal Chaos. They demonstrate that with a few measurements and a small computer or smartphone, you can indeed tip the odds in your favour. The trick is to record when the ball and a set part of the rotating wheel both pass a chosen point.
Their model divides the game into two parts: what happens while the ball rolls around the rim of the wheel and then falls, which is highly predictable, and what happens after the ball starts bouncing around, which is chaotic and hard to predict. Because the first part is predictable, Small and Tse were able to calculate roughly where the ball would begin its erratic bouncing and therefore in which part of the wheel it was more likely to land.
Using a subtle counting device similar to Farmer's, the pair was able to predict in which half of the wheel the ball would fall in 13 out of 22 trials. In three trials, the model predicted the exact pocket. That is equivalent to taking the odds from 2.7 per cent in the house's favour (on European roulette wheels) to 18 per cent in the player's favour. That is a very small number of trials, so they then confirmed their technique via 700 trials using an automated camera system, which would be too conspicuous to use in a casino (arxiv.org/abs/1204.6412).
Farmer says Small and Tse's model is very similar to his own, except that they assume that the main force slowing the ball down is friction with the rim, whereas he found that it is air resistance.
Small is confident that casinos are aware of the trick. Holger Dullin, an expert in chaos theory and mechanics from the University of Sydney in Australia, says they could guard against it by closing bets before the wheel has rotated enough times for sufficient measurements.
Small says people tell him they have tried it and it works: "One guy even sent me pictures of his toe with this little clicker thing on the end."