Skip to main content
  1. Home
  2. Emerging Tech
  3. Web
  4. News

These scientists beat the bookies — until the online casino shut them down

Add as a preferred source on Google

It’s nearly impossible to win money betting on sports, and that’s by design. Whether it’s online or at a casino, the bookmakers who set the betting odds for sports gambling have an array of statistical tools at their disposal to ensure the numbers are always in their favor. But they also have to account for the human factor, and this is where a team of scientists came up with a way to use the bookies’ own calculations against them.

Researcher Lisandro Kaunitz of the University of Tokyo and a few of his friends from around the world devised a mathematical system that let them consistently make money betting on soccer games online. The MIT Technology Review has all the details on the system they used, but to understand it, you have to know how sports betting actually works.

Recommended Videos

When bookmakers set the odds on a particular match, they use historical data and sophisticated analysis to predict the most likely outcome, and then set the odds of a win, loss, or draw accordingly. Certain teams are more popular than others, of course, and tend to draw more betting action, especially on big events like the Super Bowl. As a result, the oddsmakers may adjust the betting line by a few points one way or the other to compensate for this bias.

Kaunitz and colleagues devised a system that consistently identified betting opportunities that favored them rather than the house. They tracked odds offered by online betting companies on soccer matches and calculated the average odds to discover any outliers. Then they analyzed whether a bet on the outlier matches would favor them or not.

To test their system, they analyzed the results of 479,440 soccer games played between 2005 and 2015. This simulation delivered a return of 3.5 percent. “For an imaginary stake of $50 per bet, this corresponds to an equivalent profit of $98,865 across 56,435 bets,” they said. A random simulation yielded a return of  negative 3.2 percent, or a loss of $93,000.

“At this point we decided to place bets with real money,” Kaunitz said. Over a five-month period, their 256 different $50 bets paid off 47.2 percent of the time, and they made a profit of $957.50, an impressive return of 8.5 percent.

But then down came the banhammer.

The online casinos would no longer accept their wagers, or would limit them to amounts as small as $1.25. “The sports betting industry has the freedom to publicize and offer odds to their clients, but those clients are expected to lose,” Kaunitz said. “If they are successful, they can be restricted from betting.”

One of the bookmakers used by the team told the New Scientist it’s the casino’s prerogative to restrict certain bets. “This can be for a number of reasons, including bonus abuse and taking proportionately more than their fair share of special offers and enhanced prices, which are designed for the many rather than a few,” said the spokesman.

What did they do with their winnings? Kaunitz and his wife splurged on a nice dinner in Tokyo. “We were excited, but it’s worth mentioning – you need to spend a lot of time to do it,” he said.

Mark Austin
Former Digital Trends Contributor
Mark’s first encounter with high-tech was a TRS-80. He spent 20 years working for Nintendo and Xbox as a writer and…
AI agent reportedly carried out an entire ransomware attack on its own
AI didn't just write malware. It apparently clocked in for work.
Cybersecurity

Cybersecurity researchers say they have documented what could be the first ransomware attack carried out almost entirely by an autonomous AI agent, marking a significant shift in how cyberattacks could be conducted in the future. According to cloud security firm Sysdig, they have uncovered a ransomware operation dubbed JadePuffer that appears to have relied on a large language model (LLM) agent to perform nearly every stage of the attack without continuous human intervention.

If confirmed, the incident suggests AI is moving beyond writing malicious code and into actively planning, adapting, and executing cyberattacks in real time.

Read more
The Washington Post predicted how tech will advance 50 years ago and the success rate is humbling
The Washington Post predicted 2026 tech in 1976. It got a lot right.
Representative Image

Fifty years ago, when floppy disks were cutting-edge and the personal computer revolution had barely begun, The Washington Post attempted a remarkably ambitious exercise: predict what life in 2026 would look like. Some of those predictions now read like science fiction. Others feel surprisingly ordinary because they have become part of everyday life.

In a retrospective published for America's 250th anniversary, the newspaper revisited science editor Thomas O'Toole's 1976 article Inventing the Future, comparing its forecasts with today's technological reality. The results reveal that while predicting exact timelines is nearly impossible, identifying long-term scientific trends can be remarkably accurate.

Read more
Australian government warns doctors over AI scribing tools as privacy and safety concerns grow
AI medical scribes face regulatory scrutiny in Australia amid safety concerns
Representative Image

The Australian government is urging healthcare professionals to exercise caution when using AI-powered medical scribing tools, as regulators examine whether stronger safeguards are needed around one of healthcare's fastest-growing technologies, according to a report by The Guardian.

AI scribes have rapidly gained popularity by recording, transcribing, and summarising doctor-patient conversations into clinical notes, reducing the administrative burden on healthcare workers. However, government officials now warn that the technology's rapid adoption has outpaced oversight, raising questions around patient privacy, informed consent, and the accuracy of medical records.

Read more