The Guardian in 2004:
It is understood that the three suspects, two Serbian men aged 38 and 33, and a 32-year-old Hungarian woman, made two visits to the casino earlier this month. On the first night they walked out with £100,000. They returned on a subsequent evening and appeared to disprove Albert Einstein's conclusion - that the only way to win on roulette was to pocket the money when the dealer was not looking - by walking out with £1.2m.
Detectives are investigating claims that a gang used a laser scanner hidden in a mobile phone and linked to a computer to help beat the roulette wheel.
Two men and a woman were able, it is claimed, to place their bets in the area the computer had pinpointed as the ball's most likely resting place.
Kit Chellel for Bloomberg, in a long article (practically a novella) bursting with delightful detail (the underlying problem is casinos make money off suckers who bet after the ball is motion, but an elite few can use that time to exploit the system):
I spent six months investigating the clandestine world of professional roulette players to find out who Tosa is and how he beat the system. The search took me deep into a secret war between those who make a living betting on the wheel and those who try to stop them—and ultimately to an encounter with Tosa himself. The British press got plenty wrong in their reports about what happened on the night of March 15, 2004. There was no laser. But the newspapers were right about one thing: It is possible to beat roulette.
the way Tosa and his friends played roulette stood out as weird even for the Ritz. They would wait until six or seven seconds after the croupier launched the ball, when the rattling tempo of plastic on wood started to slow, then jump forward to place their chips before bets were halted, covering as many as 15 numbers at once. They moved so quickly and harmoniously, it was “as if someone had fired a starting gun,” an assistant manager told investigators afterward.
Tosa and his companions reacted to being arrested with the same surreal calm they’d shown at the roulette wheel. At the station, they were interviewed separately through an interpreter. Tosa was robotically unhelpful, declining to answer questions. Marjanovic was more talkative but just as confounding.
When Jovanovic and Fadel arrived at the Colony, they were led to a private roulette chamber to find not only police, as they’d expected, but also half a dozen casino security bosses in dark suits. Most were former soldiers like Wootten, some had visible scars or warped knuckles, and all looked hostile.
Eventually, I realized the different addresses Tosa had given casinos over the years were clustered along the same stretch of Croatian coast, south of Dubrovnik. They were tiny villages, mostly. I hoped someone might have heard of him, so
The article also goes deep into the history of trying to cheat the casinos:
J. Doyne Farmer, a physics student at the University of California at Santa Cruz, took up the challenge. Farmer dreamed of creating a utopian community of hippie inventors funded by gambling profits. He and his partners called their venture Eudaemonic Enterprises, after Aristotle’s term for the fulfilling sensation of a life well lived. Like Thorp before him, Farmer learned that roulette was more predictable than anyone imagined, and also that making the science work amid the sweat and noise of a real casino was almost impossible. His device used a hidden buzzer that told the wearer which of eight sections, or “octants”