I lived out of rental cars for a decade. Counting cards, moving between Nevada casinos. I know what it means to beat a game within the rules.
Gonzalo Garcia-Pelayo did something I never could. He beat a game that was not supposed to be beatable.
The Setup
Garcia-Pelayo was a Spanish musician. In 1989, he was visiting the Casino de Madrid with his sons. He brought a laptop computer. Laptops in 1989 were heavy and rare, but he had one.
He started recording the results of the roulette wheel. Every spin for weeks. He kept data on where the ball landed.
European roulette has 37 numbers (0 through 36). Perfect randomness means each number comes up 1/37 of the time. Over 10,000 spins, you would expect each number to hit approximately 270 times. If a number hit 350 times, it is biased.
Garcia-Pelayo was looking for bias.
What He Found
After 10,000 spins, his data showed that certain numbers on the Casino de Madrid's primary wheel appeared more frequently than probability dictated. The number 9 was hitting at a 1.6 times greater frequency than expected. Numbers 14 and 17 were also elevated.
The cause was mechanical. Roulette wheels must be manufactured with perfect precision. A deviation of 0.1 millimeters in any pocket creates bias. The Casino de Madrid's wheel had experienced years of wear. The bias was real and measurable.
The Approach
Garcia-Pelayo began placing bets on the biased numbers. He started small, $50-$100 per spin. He tracked his results. After 15 days, he was up $5,000.
He expanded the operation. His sons helped track the wheels. They coordinated with casino staff who provided data. He expanded to other casinos: Casino de Estoril in Portugal, Casino de Grand Marin in France.
Each casino's wheel had biases. The wheels were old. The biases were exploitable.
The Mathematics
If you know a number will hit 45 times per 1,000 spins instead of 27 times, the expected value calculation changes.
European roulette pays 35 to 1 on a single number bet. Your expected value on a biased number is: (45/1,000 * 35) - (955/1,000 * 1) = 1.575 minus 0.955 equals 0.62.
On a $100 bet, you have an expected value of $62 per spin if the bias holds. Place $100 per spin for 30 spins and you expect $1,860 profit.
The Timeline
From May 1989 to August 1992, Garcia-Pelayo won approximately $1.5 million from Spanish, Portuguese, and French casinos. He was not cheating. He was exploiting a mechanical flaw.
Casinos eventually figured out what was happening. They understood that wheel bias was the issue. They began servicing wheels more frequently. They began rotating wheels between tables. They started monitoring for statistical deviations.
The Casino de Madrid eventually discovered the bias when their staff noticed Garcia-Pelayo's consistent wins. They examined the wheel and found it was worn. They serviced it. The bias disappeared.
Why This Matters
Garcia-Pelayo proved that even games with an inherent house edge can be beaten if a real physical flaw exists. Roulette was designed to be unbeatable at perfect randomness. The flaw was not the game design. The flaw was the physical implementation.
For a card counter, the edge comes from information: knowing which cards remain in the shoe. For Garcia-Pelayo, the edge came from physics: knowing which pockets the wheel favors.
The Legal Status
He was not prosecuted. He did not cheat. He exploited a flaw. This is legal. Casinos reserve the right to refuse service to advantage players, but they cannot prosecute them for beating a flawed game.
European casinos eventually implemented safeguards. Wheels are now serviced monthly. Wheels are rotated between tables daily. Wheels with measured bias of above 5 percent are retired.
The Broader Lesson
Mechanical randomness is not perfect. Perfection requires precision manufacturing and regular maintenance. Without these, bias emerges. Garcia-Pelayo found the bias and exploited it.
For online casinos, the randomness is software. Software randomness can be tested and certified. It cannot develop bias from wear. The threat Garcia-Pelayo represented to casinos is nearly impossible in the online environment.
This is not a strength of online casinos. It is a difference. The flaw Garcia-Pelayo exploited was a mechanical limitation of the 1980s. Online casinos have no mechanical wheels. They have different vulnerabilities.
Garcia-Pelayo's run lasted three years. It ended when the casinos understood the problem and fixed it. The game was not unbeatable. It just required a specific window of time, a specific flaw, and the patience to exploit it correctly.
Wings for this entry
