How prediction markets work

A prediction market is a marketplace where people buy and sell shares in the outcome of a future event: an election, a sports result, a crypto price, an economic number. Each possible outcome — usually YES or NO — is its own share.

When a market resolves, a winning YES share pays out exactly $1 and a losing one pays $0. Everything interesting happens in between: while the event is still uncertain, those shares trade somewhere between 0¢ and 100¢, and that price is the whole point.

Here's the one idea that unlocks everything else: the price of a YES share, in cents, is the market's implied probability of YES, in percent. A share trading at 34¢ means the market collectively thinks there's roughly a 34% chance the answer is YES.

Why? Because each winning share is worth $1. If enough traders thought YES was more likely than 34%, they'd buy at 34¢ for the expected payout, pushing the price up until it matched their belief. The price is the crowd's live estimate, updated with every trade.

Every serious market publishes resolution rules — the exact wording that decides YES or NO. Good research starts here, because the rules, not the headline, determine who gets paid. A market titled "Will Team X win the championship?" might hinge on a specific date, a tie-breaker, or what counts as "winning."

When the event happens, the market resolves against a named source, winning shares pay $1, and the market closes. Reading the fine print before you commit is the single most underrated beginner skill.

Prices move because traders disagree and act on it. News drops, someone with an edge takes a position, volume flows in, and the price re-settles at a new consensus. That's why a prediction market often reacts faster than a poll or a pundit — money, not opinion, sets the number.

It also means the crowd can be wrong, herd, or overreact. Your job as a researcher isn't to assume the market is always right — it's to spot the moments where the evidence and the price disagree.