An alleged high-tech roulette scam that saw three people walk out of a London casino with £1.3 million recently sounds too implausible even for a movie plot.
But a physicist who developed a technology-based system that famously beat the wheel in the 1970s has told New Scientist that in theory it would have been fairly easy to carry out with a little know-how and the right tools.
Two men and a woman were arrested on 16 March after raking in a huge win over two evenings. Suspicious casino staff are said to have reviewed videotapes of the players and called in the police, whose investigation is continuing. The trio are now on bail and have not been charged with any crime.
Some media reports have suggested they used mobile phones fitted with laser scanners to measure the speed of the roulette ball when it was released, in order to calculate where it was likely to fall. The whole calculation would need to have been completed in just a few seconds, as the dealer cuts off betting after the ball has rolled three times around the wheel.
But the trick could be pulled off a lot more simply if the phones were used as stop watches, says Norman Packard, a physicist at the Santa Fe Institute in New Mexico, US.
Packard should know. In the late 1970s, he and a group of other physics graduate students set out to create computers that could compute the sector of the wheel a roulette ball would land in. They hid these computers under their clothes or in their shoes, clicking buttons with their toes.
"In the best circumstances, we could predict the quadrant correctly," says Packard. "We definitely got to the point where we were winning money, but we didn't continue long enough to make large amounts."
Such a caper is "fairly easy to set up", he says. Just two equations - one for the ball and one for the wheel, which move in opposite directions - predict the likely area where the ball will stop. These equations comprise only a handful of parameters, including the mass and size of the ball, the shape and roughness of the track, and the tilt of the wheel.
A scouting mission to the casino could give these values for a particular wheel and ball in advance, meaning the equations can be partially solved before attempting betting.
The cell phones reportedly used in the alleged London scam could have been used to determine the ball's speed if buttons on the phones were pressed when the ball was released and then after one revolution, Packard says. In fact, some cell phones have their own built-in stopwatches.
A remote computer, or perhaps even one in the phone, could then solve the equations "very rapidly", says Packard, "because you've done all that homework".
Even crude predictions could be profitable because of the way the game is set up. European roulette wheels have 37 resting positions for the ball. In one version of the game, you bet on a number between one and 36, so you have a one in 36 chance of winning.
If you win, the casino pays you 35 times your bet. So any time you increase your odds to better than 1 in 35, you win on average. "Even saying which half of the wheel is extremely powerful because the payoff is so good," Packard says.
Casinos could thwart such a method simply by spinning the wheel more quickly, which makes the ball bounce around unpredictably. "But croupiers don't like to do that because people like to watch the wheel," says Packard.
He stopped his own attempts partly because new laws in some US states barred computers from casinos. British gambling laws from 1845 are currently in the process of being redrafted to bring them up to date with 21st Century gaming.