What are prediction markets?

A beginner's guide · Updated August 2026

A prediction market is a marketplace where people trade contracts that pay out based on the outcome of a future event — an election, a game, an interest-rate decision, an awards show. Because traders put money behind their opinions, the market's price ends up being a live, numerical estimate of how likely that outcome is.

In everyday life we get our forecasts from pundits, polls and gut feeling. Prediction markets replace the punditry with a price. Instead of asking one expert "who's going to win?", they let thousands of people effectively vote with their wallets — and then aggregate all of those bets into a single number that anyone can read at a glance. That number is what Prediction Headlines collects, ranks and summarizes across sports, entertainment, politics, finance and the internet's stranger questions.

The core idea: opinions with skin in the game

An ordinary poll asks people what they think. A prediction market asks them what they're willing to risk money on — and that difference matters. If you genuinely believe an event is more likely than the current price suggests, you have an incentive to buy and push the price toward your view. If you're just talking, you stay out. Over many traders, this filters noise and rewards people who are actually informed, because being wrong costs them and being right pays them.

Economists call the result "information aggregation": the price quietly pools together everything the crowd collectively knows — polls, insider knowledge, models, hunches — into one continuously updating figure. It's the same "wisdom of crowds" effect that lets a room full of people guess the number of jellybeans in a jar better than almost any individual, except here the guesses are weighted by conviction and money.

A worked example

Suppose there's a market on "Will it rain in Central Park on July 4th?" Each "Yes" share is designed to pay $1 if it rains and $0 if it doesn't. If forecasters are split, "Yes" might trade around 45¢. Read that price as a 45% implied probability. A big storm system appears on the radar the night before; buyers pile into "Yes" and the price jumps to 80¢ — the market now implies an 80% chance. When the day resolves, "Yes" settles at either $1 or $0, and whoever held the correct side is paid. The price was never a promise; it was the crowd's best running estimate, revised in real time as facts changed.

A short history

Betting on public events is centuries old, but the modern, research-grade version started with the Iowa Electronic Markets in 1988 — a university-run market that often predicted US elections more accurately than opinion polls. The 2000s brought Intrade, a popular offshore exchange that became famous for tracking elections and Oscar races. Regulatory crackdowns closed several early venues, and the space fragmented into tightly regulated, small-stakes platforms like PredictIt and research markets. The current era is defined by two very different players: Polymarket, a large crypto-settled exchange with global reach and huge volumes, and Kalshi, a US federally-regulated exchange (a CFTC-designated contract market) that offers event contracts to Americans. Prediction Headlines pulls from these public markets to show what's active right now.

What people trade on

• Politics & government — elections, nominations, legislation, and central-bank rate decisions.
• Sports — championships, individual games, player milestones and season awards.
• Economics & finance — inflation prints, jobs numbers, and where an index or a crypto price lands by a deadline.
• Entertainment & culture — awards, box-office numbers, casting, release dates and viral moments.
• Oddball questions — the long-shot, only-on-the-internet markets about aliens, records and the improbable that we highlight for fun.

How they differ from polls and pundits

Polls are snapshots taken every few days from a sample of people and reported with a margin of error. A prediction market updates continuously, incorporates far more than survey answers (money, models, breaking news), and hands you a single probability instead of a spread of percentages you have to interpret. That's why markets often move within minutes of a debate, an injury report or an economic release, while the next poll is still days away. They aren't magic — a market can only be as smart as the people trading it — but as a fast, self-correcting consensus they're hard to beat.

How they differ from a sportsbook or casino

At a casino the house sets the odds and profits from a built-in edge. In a true prediction market, traders set the price by trading against each other, and the venue simply matches buyers and sellers (usually taking a small fee). Prices can move in your favor before an event even resolves, and you can often sell your position early to lock in a gain or cut a loss — more like a stock exchange than a betting slip. That structure is exactly what makes the price informative rather than just a bookmaker's line.

Real money vs. play money

Some markets use real money or crypto; others (and many academic markets) use play-money or reputation points. Real-money markets tend to be sharper because mistakes are costly, but even play-money markets can forecast well when participants care about their score. Whichever it is, the reading is the same: the price is a probability.

The limits — when not to trust the price

Prediction markets are powerful but not infallible. Be skeptical when a market is thinly traded (a handful of dollars can swing the price), dominated by a few large traders, driven by hype rather than information, or asking about something genuinely unknowable. Very-long-shot markets also tend to be mispriced — people overpay for tiny chances, the same way lottery tickets are overpriced. We rank by trading volume precisely because the busiest markets are usually the most trustworthy; treat the quiet ones as entertainment.

Want the mechanics? Read how prediction markets work for prices, order books, resolution and reading the odds — or head back to the live markets.