Prediction Markets Miss Key Election Calls Worldwide
Polymarket and Kalshi priced several recent candidate odds inaccurately, raising questions about how well markets forecast elections.
Updated on Nov. 3, 2026 in Elections
Live Poll
Do you trust prediction markets to accurately forecast election results?
On 25 October, Polymarket gave Flávio Bolsonaro roughly 80% odds on election day in Brazil's presidential runoff, but Luiz Inácio Lula da Silva won by a narrow margin. Both major platforms also gave Francesca Hong about a 95% chance in Wisconsin's Democratic governor's primary, which she lost by less than half a percentage point.
Why it matters
Prediction markets price the chance of winning on election day rather than current polling margins, and often amplify small polling gaps into large probability swings. Their performance is a growing concern as the U.S. midterm elections approach, the most significant credibility test they face.
Polymarket and Kalshi, the two largest prediction market platforms, reached a record $21 billion in October trading volume. Polymarket's 80% forecast for Bolsonaro in Brazil's runoff and the roughly 95% chance given to Hong in Wisconsin both missed the actual outcomes, and it is unclear what drove those misses.
The players
Polymarket
Polymarket is one of the two largest prediction market platforms, where users trade on the outcomes of real-world events including elections.
Kalshi
Kalshi is one of the two largest prediction market platforms and a regulated exchange for trading event contracts.
The details
Polymarket gave Aleksandar Vučić a 70% probability in Serbia's snap parliamentary election, where he received 39.2% of the vote and Student List won with 52.6%. In Nebraska's Senate race, a 2% polling margin between Pete Ricketts and Dan Osborn coexists with a market price of 70% for Ricketts.
Timeline
October: Polymarket and Kalshi reached record monthly trading volume of $21 billion.
25 October: Brazil's presidential runoff, where Polymarket's 80% Bolsonaro forecast missed.
Season Trajectory
The U.S. midterm elections are the most significant credibility test prediction markets face, and the platforms' recent misses in Brazil, Serbia and Wisconsin will be weighed against their record $21 billion October volume. Proponents argue that requiring traders to stake money on predictions harnesses collective intelligence, a claim these recent results put under scrutiny.
Readers following midterm forecasts should note that market odds reflect election-day win chances and can diverge sharply from polling. The recent misses suggest treating high market probabilities with caution when judging a race.
The takeaway
A market probability is a forecast, not a result, and the recent misses show that sharp odds can still be wrong. Weigh market prices alongside polling and the margin between candidates before drawing conclusions.
Further reading
For more on how candidates and races are covered, see Elections.
Source note: This article includes information reported by RocketNews.
Live Poll
Do you trust prediction markets to accurately forecast election results?
