A prediction market is a marketplace where people trade contracts whose payoff depends on whether a specific future event happens. If the event occurs, one side of the contract pays out (often $1 per share); if it does not, the other side pays. Prices move as traders buy and sell, and those prices are widely interpreted as market-implied probabilities.
Unlike a sportsbook quote shown only by a house, prediction markets are usually continuous and two-sided: you can enter and exit positions as news arrives, and the price you pay or receive reflects what other participants collectively believe.
The basic building blocks
Every prediction market lesson starts with the same vocabulary.
An event is the real-world outcome being forecast—say, “Candidate X wins the 2028 nomination.” A contract or share is a tradable claim tied to that event. Binary contracts are most common: YES pays if the event happens, NO pays if it does not. The price is what you pay today for a YES or NO share. On many platforms a YES price of $0.62 is read as roughly a 62% chance the event happens, before fees and microstructure effects. Resolution is the formal rule that decides the winner. Clear resolution rules are what make a market trustworthy.
If resolution is vague (“wins the election” without specifying which office, date, or data source), prices become noisy and disputes explode. Good market design nails resolution before trading goes live.
How trading works (conceptually)
You do not need to memorize every platform’s UI to understand the mechanics. The exchange or protocol lists an event and opens YES/NO (or multi-outcome) contracts. Traders post bids and offers, or trade against an automated market maker. News, polls, models, and private information push flows to one side. The price adjusts until marginal buyers and sellers agree. After the event, the market settles: winning shares redeem at face value, losers go to zero.
Your profit or loss is the difference between what you paid to open the position and what you receive at settlement (minus fees). That is why prediction markets are often described as event derivatives or forecasting securities.
You can also exit before resolution on liquid markets by selling your shares to another trader. That ability turns a forecast into a position you can manage as evidence changes—unlike a poll you cannot “sell” once fielded.
Binary vs. multi-outcome contracts
Most introductions focus on binary YES/NO markets because the probability story is cleanest. Multi-outcome markets list several mutually exclusive results (three candidates, five policy options). Each outcome has its own price; in theory the prices should sum to about 100% once you account for fees and any “Other” bucket.
Multi-outcome markets are harder to read but more expressive. A nomination race with four serious candidates is often more honest as four contracts than as four separate binary markets that might contradict each other.
Markets as a time series
A single price is a snapshot; the chart is often the product. Organizations track how implied odds moved after a debate, a court ruling, or a earnings pre-announcement. A jump from 40% to 55% in an hour is itself information—it signals that marginal traders repriced the event, whether or not you know the headline yet.
That time-series view is why journalists cite “market odds” alongside polls: the level matters, but so does the direction and speed of change.
Prediction market vs. poll vs. “gut feel”
| Approach | What it measures | Updates | Typical weakness |
|---|---|---|---|
| Opinion poll | Stated intentions at a snapshot | Weeks | Sample bias, non-response, “shy voter” effects |
| Expert forecast | Judgment of specialists | Irregular | Groupthink, reputation risk, slow to revise |
| Prediction market | Price traders will accept today | Continuous | Thin liquidity, manipulation risk, legal limits |
Polls ask “who would you vote for today?” Markets ask “what odds are people willing to risk money on?” When participants are informed and liquidity is real, that financial skin-in-the-game filter can surface information polls miss—but markets are not magic. They fail when liquidity is fake, rules are muddy, or trading is restricted to a biased crowd.
Where you see them in the wild
Prediction markets today span several domains. Politics and policy cover elections, legislation, cabinet appointments, and regulatory decisions. Macro and business markets price rate cuts, recession timing, earnings beats, and product launches. Sports and entertainment overlap with regulated wagering on championships, awards, and release dates. Science and tech include smaller but growing markets on trial outcomes, launch dates, and climate milestones. Crypto-native platforms list on-chain event contracts with global participation and varying compliance postures.
The use case is the same across domains: convert dispersed beliefs into a single number you can track over time.
Why organizations care
Companies, researchers, and governments watch prediction markets because they can aggregate information from people with different data and models, update quickly when headlines land, expose disagreement when spreads are wide or prices volatile, and support decision-making by hedging operational risk or stress-testing plans against market odds.
Traders care for a simpler reason: if your forecast is better than the price, you can earn by buying mispriced contracts—subject to fees, competition, and risk limits.
Core concepts to remember
Implied probability. For a standard $1 binary YES contract, market price approximates implied chance of YES, adjusted for fees, spread, and whether $1 is the true redemption value.
Efficiency is conditional. Markets are only as smart as their liquidity, participant pool, and resolution quality. A $500 market on an obscure tweet is entertainment, not science.
Arbitrage links markets. When the same event trades on two venues, sophisticated traders align prices minus transfer costs. Dislocations often mean different resolution rules—not free money.
Regulation shapes design. In the U.S., federally regulated event contracts coexist with state gambling law, offshore crypto venues, and outright bans. The legal wrapper determines who can trade what.
A minimal numeric example
Suppose a YES share on “Policy Y passes by December 31” trades at $0.40. You pay $0.40 per share now. If the policy passes, you receive $1.00—profit $0.60 per share before fees. If it fails, you receive $0 and lose $0.40 per share.
Break-even thinking: at $0.40 you need the event to happen more than 40% of the time in expectation to justify buying YES, ignoring fees and risk preference. That framing is the bridge between trading and forecasting.
Common misconceptions
“The market is always right” is false—it is the best available consensus given current traders and liquidity, which can be wrong or manipulated. A higher price does not mean the event will happen; it means traders currently assign higher probability, and tomorrow’s news can flip that. Prediction markets are structurally similar to wagers in many cases, but the information aggregation purpose and institutional use cases differ from pure entertainment bets—regulators still debate where the line sits.
Another mistake is treating every platform quote as comparable. The same headline event on two venues can trade at different levels because resolution text, fees, and who is allowed to trade differ. Comparison is a skill you will build later in this module; for now, remember that a “market probability” is always probability under a specific contract.
Who participates (and why it matters)
Retail traders, political junkies, crypto natives, hedge funds, market makers, and sometimes the organizations being forecast all share the same order book. Their goals differ—entertainment, hedging, research, manipulation, or liquidity provision—but the price you see is the marginal trade among whoever actually showed up.
That is why “the market” is shorthand. In practice it means this pool of capital, under these rules, at this moment. Later chapters on crowds, law, and criticism unpack when that pool is informative and when it is misleading.
What comes next in this module
This chapter defined the object. The rest of Foundations of Prediction Markets builds the toolkit: why prices track probabilities, when crowds beat experts, history from academic experiments to modern platforms, regulation, ethics, and known failure modes.
Once those pieces click, you can read any live market quote as more than a number: it is a compressed argument about the future, continuously renegotiated by people willing to back their beliefs with capital.
The chapters ahead in foundations-of-prediction-markets unpack each layer: probability math, crowd logic, history, comparisons to polls, applications, law, and honest limits.
Next: Why "Price = Probability" – The Information Aggregation Thesis