A prediction market is a place where you can buy and sell contracts that pay $1 if a specific event happens, and $0 if it doesn't. The price of that contract, at any given moment, is the market's best estimate of how likely the event is to happen.
That's it. Everything else is detail.
A concrete example
Say Kalshi lists a contract titled "Will the Fed cut rates at the May meeting?" that's currently trading at 28¢ for YES. That means:
- Someone buying a YES contract pays 28¢. If the Fed cuts, they get $1 back (72¢ profit on 28¢ risked). If not, they get $0 (lose 28¢).
- Someone buying NO contract pays 72¢. If the Fed doesn't cut, they get $1 (28¢ profit). If the Fed cuts, they get $0 (lose 72¢).
- The market is collectively pricing the event at roughly 28% probability.
Where does the price come from?
Every contract has an order book, just like a stock. Traders post bids (I'll pay up to X for YES) and asks (I'll sell YES for Y). When bids and asks match, trades execute. The "last price" is the most recent trade; the "midpoint" is the midpoint between best bid and best ask.
The price moves because information moves. If CPI prints hot, the Fed-cut market ticks down. If a candidate wins a primary, their 'nominee' market ticks up. The collective response to new information is what makes prediction markets an interesting data source, they're a real-time probability estimate built from actual dollar flow.
Prediction markets vs. sports betting
These are legally and structurally different:
- Prediction markets (Kalshi, Polymarket) trade contracts that settle at $1 or $0. You're buying a contract that derives its value from an event outcome. Regulated by the CFTC in the US (Kalshi) or unregulated offshore (Polymarket).
- Sports betting (DraftKings, FanDuel) offers fixed odds on specific outcomes. The bookmaker sets the odds; you either take or leave them. Regulated state-by-state.
The practical difference: prediction markets have transparent price discovery (the price IS the probability) and let you both buy and sell. Sports betting offers take-it-or-leave-it odds with a bookmaker's vig built in.
Are they accurate?
Academic research (Wolfers & Zitzewitz 2004, Arrow et al. 2008) consistently finds prediction markets beat expert forecasts for many kinds of events, elections, economic releases, weather, when liquidity is sufficient. The "wisdom of crowds" effect is real when the crowd has skin in the game.
But they have known biases. Longshot markets (under ~10¢) are slightly overpriced relative to true frequency, the "favorite-longshot bias." Thinly-traded markets can be moved by single large traders. And very short-dated markets can be inefficient as new information hits.
How to read a market page
On any market page on this site, you'll see:
- YES price: the probability the market is assigning
- Volume: total dollars traded over the market's life (a credibility signal)
- Open interest: dollars in positions currently held (deeper = tighter spreads)
- Price history: how the implied probability has moved
- Resolution rules: the exact condition for YES