Prediction markets that allow participants to place bets on conditional events and Boolean combinations of base events, not just individual outcomes. By enabling richer betting vocabularies, they can elicit joint probability distributions over related events rather than only marginal probabilities.
Cluster: Mechanism Design
Prediction markets that allow participants to place bets on conditional events and Boolean combinations of base events, not just individual outcomes. By enabling richer betting vocabularies, they can elicit joint probability distributions over related events rather than only marginal probabilities.
Referenced in 2 articles
Kalshi's multivariate-event interface doesn't ship a catalogue of parlays; it mints market objects on demand from whatever legs a request names. Over one registered seven-day window the author pulled 7,611,594 unique REST tickers, created at roughly 1.09 million per day with heavy hourly bursts. The hierarchy then compresses hard: 5.26 million distinct event keys sit under just three observed collection keys, one of which holds 94.82 percent of all objects, and 66 million leg occurrences reduce to 83,701 primitives with an effective count near 720. Only 35.38 percent of objects had any volume or open interest at retrieval, REST and WebSocket surfaces overlap on a mere 276,177 tickers, and every observed 24-hour volume reading was zero. On-demand instantiation is a market-architecture phenomenon, not a storage artifact — ticker count is not economic breadth.
Investigates combinatorial prediction markets, which extend the standard model to support forecasts on conditional events (e.g., A given B) and Boolean combinations of events rather than only base events. Reports experimental results comparing combinatorial versus flat market structures on forecasting accuracy and calibration. Co-authored by Robin Hanson, whose LMSR underpins most automated prediction markets.