Connect with us

AI

World Cup Predictions Built a Billion-Dollar AI Data Business

Opta’s World Cup supercomputer called the final four correctly, but the real payoff is a billion-dollar data industry facing a new federal rules fight.

Published

on

Opta ran the 2026 World Cup 25,000 times before the opening kickoff and gave Spain the best shot at the trophy, in 16.1 percent of simulations. France, England and Argentina each cleared 10 percent. All four reached the semifinals exactly as forecast.

That accuracy is not just a bragging point for a football website. It is the foundation of an industry now worth well over a billion dollars a year in combined revenue, one that is racing into a new, legally contested frontier: prediction markets that regulators in Washington have not finished deciding how to police.

The Model Nailed the Final Four

Opta’s own analytics site confirmed the record once the semifinal lineup was set: the supercomputer correctly called the final four, with the top four pre-tournament favorites all still standing.

The model kept moving after that. Heading into the France-Spain semifinal, Opta’s projections made France favorites to lift the trophy with a 34 percent chance, eleven points clear of Spain. Then Spain won 2-0 in Dallas on goals from Mikel Oyarzabal and Pedro Porro, and the simulation flipped. Spain became the new favorite, given better than a 56 percent chance of the title regardless of opponent, according to TNT Sports’ breakdown of the updated numbers.

Jonathan Whitmore, director of analytics at Stats Perform, the company that owns Opta, has described the mechanism behind those swings as a blend of two inputs. Team strength comes from an Elo rating system, the same family FIFA itself uses, where beating a stronger side transfers points from loser to winner. Germany’s penalty-shootout exit to Paraguay in the round of 32 is the clean example: the upset does not just eliminate Germany, it hands rating points to Paraguay and reshapes every simulation run afterward.

The second input is betting market pricing, which Whitmore says folds in information the ratings alone cannot see. “We indirectly account for information such as injuries, squad selection from the betting odds,” he said, describing how the model stays current on team news between matches.

A Forecast Became a Broadcast Product

What started as a pre-tournament curiosity is now stitched into how the World Cup gets watched. Win probabilities run beside the scoreline. Momentum graphics move with the run of play. Qualification scenarios update the instant a result lands elsewhere in the group.

BBC Sport used the live version of the model during Scotland’s decisive group-stage match, showing viewers in real time how many goals Scotland could concede and still advance, a margin that Brazil’s goal erased before the final whistle. That is the pattern broadcasters, sportsbooks and apps are now built around: a single trusted number that updates with every kickoff, tackle and card.

The same feed does several jobs at once. Among the places one predictive model now shows up:

  • Broadcast graphics – live win probabilities and route-to-final overlays during match coverage
  • Betting products – in-play odds and market pricing built directly on the same simulations
  • Second-screen apps – real-time updates for the roughly 93 percent of Gen Z viewers who watch sport with a phone in hand
  • Editorial content – the recurring supercomputer articles that outlets republish after every round
  • Push notifications and emerging AI assistants – automated alerts tied to probability shifts rather than just goals

Each of those is a separate customer relationship, and increasingly a separate revenue line for the companies that supply the underlying data.

Genius Sports Turns Forecasting Into a Billion-Dollar Line

Stats Perform, Opta’s parent, is privately held by Vista Equity Partners and does not disclose revenue. But its publicly traded rival shows what the category is worth. Genius Sports, the New York and London-listed data supplier to leagues including the NFL and NCAA, reported first-quarter 2026 revenue of $188 million, up 31 percent year over year.

After closing a $1.2 billion acquisition of Legend, an affiliate-marketing network for the betting industry, Genius Sports raised its full-year 2026 revenue guidance to a range of $990 million to $1.01 billion, up from a prior forecast of roughly $810 million to $820 million, with adjusted earnings margin targets climbing to about 28 percent.

Company Ownership Latest Reported Revenue Core Business
Genius Sports NYSE: GENI $188M Q1 2026 (+31% YoY); FY2026 guidance up to $1.01B Official betting data feeds, odds, Legend affiliate network
Sportradar Nasdaq: SRAD FY2025 revenue of €1.29bn ($1.48bn), up 17% Betting data and streaming, IMG Arena acquisition
Stats Perform (Opta) Private, Vista Equity Partners Not disclosed Owns the Opta brand and World Cup supercomputer model

The consolidation has been fast. Sportradar’s own full-year 2025 revenue rose 17 percent to that €1.29 billion figure in its first year folding in IMG Arena, a rival it acquired outright. Every major supplier in this space is buying its way into more sports, more leagues and more of the pipeline between a stadium and a betting app.

Revenue Climbs, Losses Do Not Go Away

The growth numbers sit next to a less flattering pattern. Genius Sports’ net loss widened to $111.6 million in 2025 even with revenue up 31 percent for the year, its fifth straight annual deficit since going public through a SPAC deal in 2021, according to trade outlet SportBusiness.

Investors reacted hard to the Legend deal itself. Shares dropped as much as 28 percent in February on valuation concerns, before climbing 8.6 percent the day of the first-quarter earnings report in May, according to Legal Sports Report’s coverage of the call. Bryan Castellani, the company’s chief financial officer, told analysts the acquisition was “immediately margin accretive and accelerates our path to our previously stated long-term revenue and margin targets by two years.”

The bigger business point holds even where a company like Genius Sports is not yet consistently profitable. This is exactly the kind of AI use case most enterprises have failed to find, echoing a broader gap where nearly 90 percent of businesses use AI but only a small share see real financial payoff. Sports data has become one of the few corners where an AI product visibly changed how customers pay for a service, even if the bottom line has not caught up yet.

Washington Has Not Settled the Rules

The next growth market Genius Sports is chasing is not a sportsbook at all. On the same earnings call, CEO Mark Locke told analysts the company sees prediction markets, the exchange-style platforms where users trade contracts on real-world outcomes, as a coming source of data revenue, and said Genius had already onboarded several market makers using its low-latency feeds. He compared the long-term economics to those of major sportsbook operators.

That bet lands in the middle of an active regulatory fight. Trading volume on U.S. prediction markets grew from under $1 billion in June 2024 to nearly $24 billion by April 2026, according to a legal analysis from law firm Ropes and Gray. On June 10, 2026, the Commodity Futures Trading Commission published a proposed rule on event contracts that would generally permit sports outcome contracts like moneylines and point spreads while barring bets on injuries, officiating and single in-game plays. Public comments are due July 27, 2026.

Well, if Opta says so.

That was FIFA president Gianni Infantino’s reply when asked about Spain being installed as pre-tournament favorite, a light line that nonetheless captured how much weight a single data brand now carries across broadcasting, betting and even football’s own governing body.

Not every stakeholder agrees the CFTC should be the one setting these rules at all.

  • The CFTC’s position – Chairman Michael Selig’s agency argues sports event contracts serve legitimate price-discovery functions and fall under its exclusive federal jurisdiction over derivatives.
  • The casino industry’s position – The American Gaming Association says prediction markets offer gambling outside state and tribal regulatory frameworks and avoid licensing and taxes that sportsbooks must pay.
  • The anti-exchange coalition’s position – Mick Mulvaney, executive director of the group Gambling is Not Investing, argues Congress never intended federal derivatives law to become a backdoor for sports wagering oversight.

Courts are already involved. A federal appeals panel ruled in June that the Commodity Exchange Act preempts a New Jersey law targeting sports event contracts, while a separate Nevada case over the same question remains on appeal. Several people close to the fight expect it eventually reaches the Supreme Court.

None of that has slowed the tournament itself. Spain now waits in the final for the winner of England’s semifinal against Argentina, a matchup that briefly reopened old rivalries on the pitch even as, off it, Argentina’s own federation dealt with a separate data breach tied to its dispute with Egypt over World Cup officiating. Fans without cable access can still follow the run through the knockout rounds using existing streaming options for the tournament’s final weekend.

The final is set for Sunday, July 19, in New York. Whatever the scoreline, the supercomputer will already have run the outcome 25,000 times, and somewhere between a broadcast graphic, a sportsbook app and a prediction market contract, that number will already be for sale.

Disclaimer: This article discusses publicly traded companies, sports betting data and prediction markets for informational purposes only; it is not investment or gambling advice, and figures are accurate as of publication.

Logan Pierce is a writer and web publisher with over seven years of experience covering consumer technology. He has published work on independent tech blogs and freelance bylines covering Android devices, privacy focused software, and budget gadgets. Logan founded Oton Technology to publish clear, no nonsense tech news and reviews based on real hands on testing. He has personally tested and reviewed dozens of mid range and budget Android phones, written extensively about app privacy, and built and managed multiple WordPress publications over the past decade. Logan holds a bachelor's degree in English and studied digital marketing at a certificate level.

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending