“prediction markets are right about as often as their probabilities suggest, not as often as headlines imply”
WaPo analyzes 1,276 Kalshi markets and 829 Polymarket markets across 344 primary races and finds prediction markets are well-calibrated: candidates given 70-80% probability win about as often as predicted. High-profile misses (Spencer Pratt in LA, Massie in KY) are not failures of the market but the expected 25% outcome in a properly calibrated system where probabilities are often misunderstood as certainties.
No technical background needed
Platforms mentioned: Kalshi, Polymarket