Business Finance Media Technology Policy Wealth Insights Interviews

Art Art Fairs Art Market Art Reviews Auctions Galleries Museums Interviews

Lifestyle Nightlife & Dining Style Travel Interviews

Power Lists Nightlife & Dining Art A.I. PR

About About Observer Advertise With Us Reprints

The World Cup Is Proving the Business Value of A.I. Prediction Models

Before every major tournament, A.I. models simulate thousands of possible outcomes that shape how broadcasters, sportsbooks and millions of fans experience the competition in real time.

Before a ball was kicked, the Opta supercomputer had already played the tournament 25,000 times. Its verdict: Spain were the most likely winners, winning 16.1 percent of the simulations, followed closely by semifinalists France, England and Argentina, each winning more than 10 percent. 

Sign Up For Our Daily Newsletter

Thank you for signing up!

By clicking submit, you agree to our terms of service and acknowledge we may use your information to send you emails, product samples, and promotions on this website and other properties. You can opt out anytime.

The prediction quickly became far more than a percentage on a page. It appeared on television screens, influenced betting markets, dominated social media discussions and gave millions of fans a new way to follow the tournament. What began as a forecasting experience became part of how the World Cup was experienced in real time.  

As Jonathan Whitmore, director of analytics at Stats Perform, explains, the model combines Opta’s team rating with betting market odds. The team ratings are built on an Elo system, the same family of models used by FIFA. Elo weighs not just wins and losses but the stakes behind them. “Each respective team has a score, and if you beat a team with a better score, you have the ability to win more points and boost your score, whereas if you lose against a weaker team, they effectively win those points back off you, so it adjusts over time,” Whitmore explains. 

Germany’s shock penalty-shootout defeat to Paraguay in the Round of 32 illustrates how the system adapts. The upset doesn’t just cost Germany a place in the draw, it actively transfers rating points to Paraguay, reshaping both teams’ odds in every subsequent simulation. 

 Ratings alone, however, cannot capture everything that shapes a match, so the model leans on a second data source to fill that gap: betting markets. “We indirectly account for information such as........

© Observer