Prediction markets aggregate the wisdom of crowds into probabilities — here's the mechanics, the math, and the evidence they actually work.
You buy contracts that pay $1 if an event happens, $0 if it doesn't. The price you pay (between $0.01 and $0.99) is the market's consensus probability. If you think the real probability is higher, you buy YES; if lower, you buy NO.
We earn commissions via affiliate links — this never affects our rankings. Read our full methodology →
A prediction market is an exchange where users trade contracts whose payouts depend on the outcome of a future event. Each binary contract pays exactly $1 if the event happens and $0 if it doesn't — and the current price is the market's consensus probability.
If a contract pays $1 when an event resolves YES and you can buy it for $0.40, you're being offered a 150% return when the event happens. Rational traders will keep buying until the price reflects their best estimate of the probability — which is exactly $0.40 if they think it's a 40% event. With many traders, the equilibrium price converges to the crowd's consensus probability.
Price = probability. That's the entire mental model.
Modern venues like Polymarket and Kalshi use a continuous order book — you trade against other users, not against a house. Some markets also use automated market makers (AMMs) to seed liquidity. Either way, there's no traditional bookmaker margin: fees come from the bid/ask spread (Polymarket) or a take on profits (Kalshi at 5–7%).
The empirical evidence is strong. The Iowa Electronic Markets have outperformed major election polls in 74% of presidential races since 1988. Polymarket called the 2024 US presidential winner days before mainstream pollsters converged. Robin Hanson's work on information aggregation formalized why this happens: traders have skin in the game, so they reveal information that pollsters don't capture.
In the US, the two regulated venues are Kalshi (CFTC-regulated, USD via ACH, 47 states + DC) and Polymarket (CFTC-licensed, US app, 49 states + DC). Rebet and OG cover the sweepstakes and crypto-funded niches. See our ranked best platforms.
Markets can be thin, especially on niche events — wide spreads eat the edge. Resolution disputes happen (Polymarket had a notable Russia/Ukraine dispute in 2023). And like any market, prediction markets can be wrong, especially on low-probability tail events where the price is anchored too far from reality.
You buy contracts that pay $1 if an event happens, $0 if it doesn't. The price you pay (between $0.01 and $0.99) is the market's consensus probability. If you think the real probability is higher, you buy YES; if lower, you buy NO.
Yes — repeatedly more accurate than polls and expert panels. The Iowa Electronic Markets beat polls in 74% of presidential races since 1988, and Polymarket called the 2024 election winner before mainstream pollsters converged.
Through fees: either a bid/ask spread (Polymarket, ~0.5–1%) or a take on profits (Kalshi, 5–7%). They don't make money as a bookmaker — there is no house betting against you.
Yes, on CFTC-regulated venues. Kalshi (CFTC DCM) and Polymarket (CFTC DCM via QCX) are both legally available to US residents.
A prediction market matches users against each other on an exchange. Nobody sets a line and takes the other side, so the price is simply what other traders will pay — your cost is the spread plus a small trading fee.
Catie has spent 15 years covering iGaming, sports betting, and now prediction markets — first as a trade reporter, later as head of editorial at two major industry publications. She has tested every regulated US event exchange since Kalshi's 2021 launch and writes the bulk of PredictionWins' platform reviews and tax explainers.
Reviewed against our editorial policy.
The 5 most-moved prediction markets, every Monday.