Increasingly, gamblers are placing bets on the projected number of measles cases in the United States. In January alone, wagers exceeded $9 million on platforms like Calci and Polimarket. Evidence shows their forecasts can effectively model the spread of infection.

Prediction markets operate by allowing users to buy and sell stocks tied to specific outcomes. Each market presents a question regarding future events, with participants betting on “yes” or “no” outcomes. The cost of each bet is set according to the majority opinion in the market.

For instance, if 86% of bets predict a “yes” outcome, each “yes” stock costs 86 cents. If the event occurs, successful betters receive $1 per share, while unsuccessful bettors lose their investment.

The concept of prediction markets originated from scientific research. In 1988, economists Robert Forsythe, George Newman, and Forrest Nelson at the University of Iowa aimed to forecast federal elections, leading to the creation of these markets where small stakes could predict outcomes.

Their predictions proved remarkably accurate. In 2003, infectious disease researcher Philip Polgreen urged economists to expand these markets to include disease forecasting, emphasizing an ethos centered on education and public welfare.

In recent years, companies like Kalshi and Polymarket have commercialized prediction markets, operating legally in the U.S. under Commodity Futures Trading Commission regulations, albeit facing growing scrutiny from government entities.

These markets have faced criticism for enabling bets on sensitive subjects like the Iran and Ukraine conflicts. Some observers deem it morally questionable; for instance, a trader known as Magamiman profited $553,000 by accurately predicting the removal of Ayatollah Khamenei from power on February 28, 2026. This success raised concerns among U.S. lawmakers about potential insider trading.

As measles cases rise in the U.S., a similar betting market has emerged for the disease. Although the ethical dilemma surrounding these bets is complex, it may offer valuable data insights. According to Spencer J. Fox, a professor at Northern Arizona University who specializes in predictions for COVID-19 and other respiratory viruses, the measles prediction market could serve as an innovative data source.

The June 2025 prediction market projected around 2,000 measles cases by year-end, a figure closely matching the actual data of 2,288 cases. Fox noted, “Our model anticipated far worse scenarios.”

To forecast disease, epidemiologists utilize various data types, including vaccination rates, genomic information, and climate considerations. “Everyone is searching for an edge in infectious disease prediction, constantly exploring new data streams,” Fox remarks. However, measles poses a challenge for predictions due to its “highly stochastic” nature.

Emile Servan Schreiber, the CEO of prediction market firm Hypermind, suggests that the accuracy of measles forecasts may stem from harnessing the “wisdom of the crowd,” where non-experts contribute diverse perspectives that balance out gaps in specialized knowledge.

Nevertheless, Fox argues that prediction markets cannot fully replace scientific models employed by epidemiologists. These markets often lack the comprehensive predictions and granularity that scientific approaches provide. He emphasizes, “We would need to make thousands of bets each week on all possible predictions.”

Moreover, he emphasizes that only seasoned experts are adept at predicting rare events. “If we neglect to cultivate expertise in infectious disease prediction now, we risk being overwhelmed by the next pandemic,” he warns.

Neither Kalshi nor Polymarket responded to a comment request from New Scientist.