The Convergence Has a Tell: Prediction Markets and Sportsbooks Are Pricing the Same Games Differently
The defining structural question in US gaming right now is whether prediction markets and sportsbooks are converging into one market or competing as two. Most of that debate has been regulatory and theoretical. The World Cup is the first mass-attention event where it can be measured directly, on identical games, at scale, in real time. Four games into the tournament, the early read is that “the market price” is not one number, and the gaps between venues are neither random nor trivial.
Three layers of divergence in the first four games
Prediction markets versus sportsbooks. On the opening slate, prediction-market prices sat above the de-vigged sportsbook consensus on the favorite leg in three of four games: Mexico by 1.8 points, Canada by 0.9, the USA by 0.4. Real-money traders on Polymarket and Kalshi leaned harder on home favorites than the books did. Small, but consistent in direction.
Prediction market versus prediction market. By match day, the USA market produced the first genuine split between the two largest prediction markets: the blended consensus at 47.2% on the morning snapshot, with the full cross-venue spread reaching 2.5 points — Kalshi the most confident venue, the de-vigged sportsbook consensus the least, and Polymarket between them. It was the first genuine cross-platform disagreement of the tournament. For a tournament that opened with every game priced inside two points across all platforms, that is a structural first, not a rounding error.
Sportsbook versus sportsbook. Even within the books, individual operators sat as much as a point and a half off the field on specific legs in this slate — and in earlier basketball coverage, Tater recorded single-book gaps of two to three points on identical markets.
Why most comparisons of these venues are wrong before they start
The single most common error in comparing prediction markets to sportsbooks is comparing a traded prediction-market price to a raw book line. The opening-slate books carried an overround of 4.7% to 5.3%. Compared raw, every favorite looks roughly three points stronger than the market actually believes. Any divergence analysis that skips the de-vig is measuring the vig, not the disagreement. The numbers above are de-vigged, which is the only comparison that means anything.
Why this matters for operators, suppliers, and investors
If prediction markets systematically lean a direction relative to sportsbooks, there are only two explanations, and both are consequential. Either the prediction markets are sharper, in which case they are a pricing signal a book should be watching against its own line, or they carry a behavioral bias such as home-team enthusiasm, in which case they are a different animal with exploitable structure. For an operator, the gap is line intelligence. For an investor weighing the prediction-market thesis against incumbents, a persistent, measurable divergence is evidence these venues are not redundant feeds. For anyone modeling where regulated and event-contract liquidity flows next, it is the first hard data on how these markets actually behave side by side.
The honest part
This is four games. It is a directional signal, not a finding, and we will say so until the sample earns more. The discipline is the point: we are tracking the prediction-market-minus-sportsbook gap on every match through the Final. If the home-lean persists across the group stage, it is structural. If it washes out, it was noise. Either result is worth knowing, and you can only see it from a vantage that watches every platform at once.
That vantage is the product. Tater blends prediction markets and sportsbooks across sixteen platforms into one consensus and measures the gaps between them in real time. The running record is public at taterit.com/world-cup, with the full research archive at taterit.com/research.
Follow Tater’s World Cup, start to finish
Tater built the instrument. SCCG will publish what it finds, every day, through July 19.
If you read one thing from this tournament, make it the daily edition. One email each morning: every World Cup match priced across eleven venues, where the platforms agree, where they split, and by the Final, the answer to the question this industry has never settled with data. Did prediction markets or sportsbooks price the World Cup better?
Markets and odds shown are for informational purposes only. This is not gambling advice. Always gamble responsibly. 21+. If you or someone you know has a gambling problem, call 1-800-GAMBLER or visit ncpgambling.org.
