Measles Prediction Markets Hit 2,087 Cases Right on Target

Online gamblers just proved they can predict disease outbreaks with surprising accuracy, and scientists are paying attention.

Prediction markets on platforms like Kalshi and Polymarket saw nearly $9 million in wagers on measles cases since January. In June 2025, bettors predicted around 2,000 cases by year's end. The actual number? 2,087 cases.

"I've seen many worse forecasts from our models," says Spencer J. Fox, who forecasts infectious diseases at Northern Arizona University. He thinks these betting markets might become a valuable new data stream for disease tracking.

Prediction markets work like this: people buy shares betting "yes" or "no" on future events. If 86 percent of people bet "yes," a "yes" share costs 86 cents. Winners get $1 per share when their prediction comes true, paid by those who bet wrong.

The concept started in 1988 when three University of Iowa economists created a market to forecast elections. The idea worked so well that in 2003, researchers added disease tracking "in the spirit of education and public good." Now commercial platforms have taken over, though they face growing criticism for allowing bets on wars and political events.

The Bright Side

Disease forecasters are always hunting for better data. They already use vaccination rates, climate patterns, and genetic information to predict outbreaks. Prediction markets could add another layer to these models.

Emile Servan-Schreiber, CEO of prediction platform Hypermind, credits the "wisdom of crowds." He says amateurs bring "cognitive diversity that replaces what they lack in expertise." Thousands of people weighing risks and information can sometimes spot patterns that traditional models miss.

Fox cautions that betting markets won't replace scientific expertise. Epidemiologists make thousands of detailed forecasts weekly, covering different regions and scenarios. Prediction markets can't match that granularity. Plus, only trained experts can predict rare events that might trigger the next pandemic.

The approach works especially well for measles because the disease is "very probabilistic," making it harder for standard models to forecast. As measles cases rise across the US, having another accurate prediction tool could help health officials prepare and respond.

The ethical questions remain murky, but the data speaks clearly: sometimes the crowd gets it right.