A prediction market price is a probability wearing a dollar sign — 27¢ means 27%. Reading those probabilities fluently, and knowing when they're biased, is the foundational literacy of event trading.
Price-to-Probability Fluency
The conversion is direct: price in cents = implied percentage. The useful extensions: a 27¢ Yes means the market pays 2.7:1 on the event happening; moving from 27¢ to 35¢ means the crowd upgraded its estimate by 8 points. Thin-market caveat: on wide spreads, the mid-point (not last trade) is the honest probability, and a stale last-trade price on a dormant market may reflect old information entirely.
Calibration: When to Trust the Crowd
Aggregate calibration is strong: across large samples of resolved markets, 70¢ events happen ~70% of the time. Individual markets earn less trust when: liquidity is thin, resolution criteria are ambiguous, one whale's position dominates the book, or the topic is emotionally charged (partisan politics, fandom sports) — the documented conditions where prices detach from probabilities. Volume and spread are your trust meters.
Beating the Price
Your edge is the gap between your estimate and the market's — so the work is building estimates: start from base rates (how often does this class of event happen?), adjust for case-specific evidence, and write the number down before looking at the price (anchoring is ruthless). Longshot bias means cheap contracts are chronically expensive per unit of real probability; certainty-aversion means 90¢+ favorites are chronically cheap. Those two standing biases fund the patient.
Frequently Asked Questions
Is the price ever just wrong?
Regularly, on thin or emotional markets — that's the profit source. The discipline is requiring evidence for WHY you know better, not just feeling it.
What's a calibration score?
A measure of whether your X% predictions happen X% of the time. Twenty logged trades give a rough read; tools like Brier scores formalize it.
