The crowd is well-calibrated on average — but systematically wrong in predictable ways. Prediction market strategy is about identifying those patterns.
Historical data from 130,000+ resolved Polymarket markets shows systematic patterns: certain categories, time horizons, and price ranges where the crowd consistently under- or overprices outcomes. High-probability markets are often overpriced. Low-liquidity markets have wider errors.
Markets priced at 85-95% often have meaningful NO edges. News-driven markets are slow to update — there's typically a 30-120 minute window where old prices persist after new information becomes public. Low-liquidity niche markets have wider gaps between price and true probability.
Once you've identified an edge, sizing matters. The Kelly Criterion gives the optimal bet fraction: edge / odds. In practice, use fractional Kelly (25-50%) — prediction market edges are uncertain, so conservative sizing protects against estimation error.
Most traders overestimate their edge because they remember wins and discount losses. Systematic tracking — every prediction timestamped before resolution — is the only way to know if you actually have an edge.
How much edge do you need to be profitable? Edges below 3-4 percentage points are too small to overcome transaction friction. WhenWin flags markets with 8%+ quantified edge as the minimum viable threshold.
What is the biggest mistake prediction market traders make? Overconfidence in inside-view analysis — overweighting specific event information and underweighting historical base rates.
View WhenWin track record → Get picks with quantified edge →