Nature study puts Polymarket scientific forecasts against expert assessments
A Nature study published in early June examines whether Polymarket prediction markets can forecast scientific developments more accurately than subject-matter experts. The research, published June 1, 2026, compares market-based forecasts against expert assessments on outcomes spanning climate change and quantum computing. Researchers quoted in the article raise concerns about prediction market accuracy on technical scientific topics. Separately, an arXiv preprint titled 'Polymarket-v1 Database' also surfaced on June 4, describing prediction markets as foundational infrastructure for real-time probability estimation across politics, sports, and finance.
Polymarket's reputation rests heavily on electoral forecasting accuracy. PNAS could follow Nature's lead and commission a systematic review of its science and health track record, possibly before the university consortium terms of reference are finalized in late 2026.
Brings the count of mainstream-media legitimization attempts to three this week — a16z crypto's non-sports usage chart, the BBC podcast, and now Nature — as prediction-market platforms press beyond sports into science and information infrastructure.