How Prediction Markets Work in Sports: Pricing a World Cup

A France contract trading at 22¢ on Polymarket is not an opinion that France will win the World Cup. It is a claim that France wins 22% of the time — and that the marginal dollar on both sides of that price agrees. Understanding how prediction markets work in sports starts with that reframe: the number on the screen is a probability, priced by people with money at risk, not a pundit's hunch.

With WC26 kicking off June 11 in Mexico City, the outright board is liquid for the first time in four years. France sits around +450 (≈18% true, ≈22¢ after fees and overround), Spain +500, England +600, Brazil +700, Argentina +850. Before you trade a cent of it, you need to know where those prices come from, what moves them, and where a sharp actually finds edge. That is the whole job below.

How prediction markets work in sports: contracts that resolve to $1

Every market we care about is a binary contract. "France to win the 2026 World Cup — YES" resolves to $1 if France lifts the trophy and $0 if not. Buy YES at 22¢ and you risk 22¢ to make 78¢; your breakeven is exactly the price.

That is the elegant part: price equals implied probability. A 22¢ contract is a 22% market. No odds conversion gymnastics — the cents are the percentage. To go the other way, decimal odds are simply 100 / price_in_cents, so 22¢ is 100/22 ≈ 4.55 decimal, or about +355 American after you strip the platform's cut.

Across a clean YES/NO pair the two prices should sum to roughly 100¢. When France YES is 22¢, France NO should be near 78¢. The gap between the best bid and best ask — and any premium over 100¢ across a full field — is where the house and the liquidity providers get paid.

Order books vs AMMs: two ways a price gets made

There are two machines that turn opinions into a price, and they behave differently when you size up.

Polymarket: a central limit order book (CLOB)

Polymarket runs a central limit order book. Makers post resting bids and asks; takers cross the spread. If the best ask on France YES is 22¢ for 5,000 shares and 23¢ for the next 10,000, your 8,000-share market order fills part at 22¢ and the rest at 23¢. You walk the book and your average price drifts up — that is slippage, and it is mechanical, not bad luck.

The book's virtue is price-time priority and transparent depth: you can see exactly how much sits at each level before you click. Settlement is on-chain in USDC, and resolution is handled by an optimistic oracle (UMA) rather than the exchange itself.

Kalshi: a regulated matching engine

Kalshi is a CFTC-regulated exchange that also matches buyers and sellers in an order book, but inside a US regulatory wrapper. Same core mechanic — limit orders, a bid/ask spread, depth that you eat into — with dollar settlement, KYC, and exchange-defined rules that specify the resolution source up front. For US traders it is the compliant venue; the microstructure intuition carries straight over.

AMMs: liquidity from a curve, not a counterparty

Some venues (and earlier Polymarket designs) use an automated market maker. Instead of matching you against another trader, you trade against a pooled curve: buying YES pushes its price up along a deterministic function. There is always a quote, but your price impact grows with size and the curve, not with who happens to be resting an order. Practically: a CLOB can show a tight 1¢ spread with thin depth; an AMM shows a continuous price that gets expensive fast. Know which one you are in before you size a position.

Where the wisdom-of-crowds price actually comes from

The reason these prices are sharp is adverse selection working in reverse for the crowd. Anyone who thinks France is mispriced can push the line by trading. If France "should" be 20¢ and sits at 22¢, sellers appear and drag it down; if it should be 25¢, buyers lift it. The equilibrium is where the marginal informed dollar is indifferent — and that point tends to sit close to the true probability.

This is why closing prices beat almost everyone. The line at kickoff has absorbed every public model, every team-news leak, and every sharp's position. Beating the closing price consistently is the actual definition of skill in this game.

The bars above show the trade thesis in one picture. Where your model column sits below the market column (France, England), the favourite looks rich — a sell. Where it sits above (Brazil, Argentina), the market is underpricing a side you'd buy. Prices are illustrative snapshots; verify live before trading.

How news, lineups and injuries move the line

Outright markets are slow-moving until information hits. Then they repriced in seconds.

• Lineups (≈1 hour pre-match): a confirmed starting XI is the single biggest scheduled catalyst. A rested first-choice forward vs a rotated B-team can swing a match line 8–12¢ the instant the sheet drops.
• Injuries: a tournament-ending injury to a France talisman might knock their outright from 22¢ to 18¢ — a ~18% relative haircut — within minutes of confirmation. The edge is being certain of the source before the herd.
• Red cards and in-play state: during knockouts, a 70th-minute red card can move a live match contract 15–20¢ on one event. (We break the mechanics of trading those swings in in-play trading the World Cup knockouts.)
• Draw and bracket structure: when the group stage resolves and the knockout bracket fills in, outrights snap to the new path. A favourite handed a soft quarter gets bid up; one drawn into the group of death leaks value.

The pattern: scheduled, certain information (lineups, the bracket draw) is tradeable if you are fast; rumour is a trap, because the market has often already moved by the time you can confirm it.

Resolution: who decides who won, and why it matters

A contract is only as good as its resolution rule. This is the most underrated risk in prediction-market trading.

On Kalshi, the exchange specifies the settlement source in the contract terms — typically the official FIFA result — and the exchange settles. On Polymarket, resolution runs through an optimistic oracle: a proposer asserts the outcome, and if no one disputes within a challenge window, it settles; a dispute escalates to token-holder voting. For "France wins the World Cup" that is trivially clean. For edge cases — an abandoned match, a shootout result, "first goalscorer" when an own goal opens scoring — read the exact rule text, because two venues can resolve the same real-world event differently.

Where a sharp trader's edge actually comes from

It is not from "liking France." Real, repeatable edge in WC26 markets comes from four places:

1. A better probability than the market. Your own model — Elo, xG-based, squad-strength — disagrees with the line by enough to clear the spread and fees. That gap, sized by Kelly, is the whole game.
2. Speed on certain information. Being first to act on a confirmed lineup or injury before the book reprices.
3. Cross-venue dislocations. Kalshi, Polymarket and a sportsbook rarely show the identical implied probability. When France is 22¢ on one and 19¢ on another, there is a structural arb or at least a better entry — see how to hedge your World Cup futures.
4. Resolution and structure knowledge. Reading the rule text and the bracket better than the field, so you are never surprised at settlement.

How to actually trade this at WC26

Start by devigging the board so you compare true probabilities, not raw prices padded with overround. Build or borrow a model and only act when your number clears the price plus the spread plus fees — a 2¢ "edge" inside a 2¢ spread is no edge. Price the size you intend to trade, not the top-of-book 100-share quote. And before any position, read the resolution rule once, out loud. Do that, and you are no longer betting on France — you are trading a mispriced probability, which is the only thing that pays over 48 teams and 39 days.