Polymarket Launches Academic Division to Study Prediction Market Accuracy

Crypto-based forecasting platform Polymarket has established a dedicated research branch designed to advance the scientific study of market-driven forecasting and information discovery. The newly formed Polymarket Institute will operate as an independent research unit aimed at examining how decentralized probability contracts aggregate information and quantify real-world risk at scale.

Led by Polymarket Head of Data Kai Brusch as managing director, the institute has named Brian Jabarian, an assistant professor of economics and technology at Carnegie Mellon University, as its scientific director. To prevent conflicts of interest, Jabarian will serve without personal compensation. The centerpiece of the initiative is the Polymarket Science Fellowship, a one-year program targeting doctoral researchers in fields such as market design, information economics, and predictive modeling. Applications for the inaugural 12-fellow cohort open on August 18, with participants receiving a $10,000 grant and expanded access to platform data.

Although physical workshops will be hosted in New York, fellows may participate either locally or remotely. Grants will be disbursed using a university gift structure rather than a sponsored research model, ensuring Polymarket retains no editorial control or jurisdiction over the findings. Academics retain unconditional rights to publish their findings in any scientific journal, even if the results are critical of Polymarket.

To safeguard research integrity, fellows must sign legal agreements prohibiting them from trading on Polymarket or competitor platforms throughout their tenure. Scholars are also required to publish all underlying code and analytical methods to public GitHub repositories for independent peer verification. The research agenda will center on how prediction markets process risk, integrate with artificial intelligence, and intersect with broader policy and cryptocurrency frameworks.

Unlike traditional opinion polling, which relies on periodic statistical sampling, prediction markets use real-money incentives to continuously aggregate public sentiment into dynamic price signals. The establishment of the research institute follows rapid growth for the platform, which recorded 586,000 monthly active users in July according to analytics provider Token Terminal. Media organizations, including The Wall Street Journal and Yahoo Finance, have increasingly incorporated platform probabilities into their coverage of sports, political elections, and central bank interest rate decisions.