Famous roulette wins, and what each one really shows

Joseph Jaggers, Monte Carlo, 1873

A British engineer hired clerks to record every result on the casino's six wheels for several days. One wheel showed nine numbers coming up far more often than chance allowed. He bet them, won heavily over several days, and the casino eventually found the cause, a wheel out of true, and rebuilt its wheels. Jaggers left with a large sum. The lesson: a mechanical defect can be found with enough data. The follow-on: casinos have checked their wheels ever since.

Hibbs and Walford, Reno, 1947

Two students did the same thing at the Palace Club, recording thousands of spins on a biased wheel and betting the favoured numbers until the casino replaced it. Same method, same cause, same ending.

The Eudaemons, late 1970s

A group of physics students built computers into shoes to time the ball and rotor and predict the sector where the ball would drop. On the slow, deflector-light wheels of the era it worked often enough to show a profit, though hardware failures limited the take. Casinos responded with wheel changes and rules against devices. The lesson: physics can beat a physical wheel if you can measure it before betting closes. Online, betting closes before the launch.

Gonzalo Garcia-Pelayo, Madrid, 1990s

Recorded results at the Casino de Madrid, found biased wheels, won a reported million euros with his family before the casino barred them and sued. A Spanish court ruled in his favour: observing a flawed wheel is not cheating. Casinos replaced the wheels. Same story, four generations on.

Ashley Revell, Las Vegas, 2004

Sold everything he owned, took $135,300 to the Plaza, put all of it on red. Red came in. He walked away with $270,600, in front of a television crew. The odds were 47.37% on the American wheel he chose. The lesson is the one he drew himself: do it once, and never again.

Mike Ashley, London, 2008

The businessman reportedly staked £480,000 across 17 and its surrounding bets at a Mayfair casino and won about £1.3 million when 17 came in. One spin, one number, a payout a little under 3:1 on the whole spread. It is the most expensive endorsement the number 17 has ever had, and it was a 1-in-37 event like any other.

What the two kinds of story prove

The bias stories prove that a physical wheel can fail and that data finds the failure. They also explain why modern wheels are maintained, rotated and logged, and why online RNG tables, with no wheel at all, have nothing to find.

The big-bet stories prove nothing about roulette and something about survivorship. For every Revell there are players who put everything on red and left with nothing, and no camera crew came. A single large bet at 47% or 48% is a coin flip with a house fee; winning it is not a method.

The honest place these leave you

A fair wheel cannot be beaten by any amount of recording, timing or nerve. The stories that look like exceptions all involve a wheel that was not fair, found by people who did the work. The wheels online today have been built by people who read the same stories.