Prediction markets explained: how Polymarket event contracts work, how prices show probability, and how they differ from sportsbook odds and casino betting.
Prediction markets explained, in one sentence
Let’s kill the usual myth first: “prediction markets aren’t gambling, they’re trading.” That line is marketing. If you put money at risk on an outcome you can’t control, you can lose it, and no amount of financial vocabulary changes that. But the mechanics genuinely are different from a sportsbook or a slot, and those differences are worth understanding before you form an opinion either way.
So, prediction markets explained properly: a prediction market is a venue where people buy and sell contracts tied to whether a specific real-world event happens, and the price of each contract moves with what traders collectively believe the odds are. Buy a contract on “Candidate X wins” at 40 cents and you’ve effectively paid 40% of the full payout for a claim that settles at $1 if you’re right and $0 if you’re wrong.
The stated purpose is information aggregation. Because traders put money behind opinions, prices tend to reflect the pooled view of everyone willing to bet on it, which is the wisdom of crowds idea applied with skin in the game. Economists have used that property for decades, from academic election markets to internal corporate forecasting. Whether any individual market is accurate depends on how many informed people are trading it and how much money is actually in there.
How Polymarket works, step by step
Polymarket is the best known of the current crop of decentralized platforms, so it’s the clearest example of how event contracts work in practice. Here’s the flow.
- Find a market with a precise question. Good markets are written so the answer can’t be argued about: a named event, a hard deadline, a defined resolution source. Vague wording is where disputes start.
- Read the binary outcome structure. Most markets are Yes/No. Yes and No shares together cost about $1, because exactly one of them will be worth $1 when the market resolves. Multi-option markets (think “which bank fails first”) are usually a bundle of separate Yes/No contracts.
- Buy shares at the current price. Prices are quoted in cents, from near 0 to near 100. You’re trading against other users through an order book, with market makers and liquidity pools filling the gaps. Trading is funded with a dollar-pegged stablecoin on a public blockchain rather than a casino wallet.
- Hold or sell before resolution. This is the part that has no sportsbook equivalent. If news moves your market from 40 cents to 65 cents, you can sell and bank the difference without ever learning the final outcome. You can also sell at a loss to cut exposure.
- Wait for settlement. When the event concludes, the outcome is proposed against the market’s stated resolution source and can be challenged through a dispute process before it’s finalised. Winning shares pay $1 each; losing shares pay nothing.
Two practical warnings. Liquidity varies enormously, and a thin market can be expensive to enter and nearly impossible to exit at a fair price. And access depends on where you live: Polymarket’s offshore platform has barred residents of several jurisdictions including the UK, the EU and Canada, and reporting suggests some users have got around those blocks with VPNs, which breaches the terms and can cost you your funds. Rules and fee schedules change, so read the current ones rather than an article.
Prediction market vs betting: the differences that actually matter
On the surface, backing “Yes” at 40 cents and backing a 2.50 shot at a bookmaker look like the same bet. The structure underneath is not the same.
Event types and scope
Sportsbooks cover sport, with a side order of novelty markets. Prediction markets lean toward anything with a verifiable answer: elections, central bank decisions, company announcements, weather, awards shows, and sometimes uncomfortable financial questions. Casino.org reported that more than $77,000 had been traded on a Polymarket market asking which major banks would fail before the end of 2026, with HSBC, Lloyds, JPMorgan Chase, BNP Paribas, Bank of America and Goldman Sachs among the names listed, citing reporting by The Guardian. The assigned probabilities stayed in the low single digits, and there was no suggestion any of those banks was in trouble. No sportsbook would list that market. A prediction market will, because the premise is that any question with a clear answer can be priced.
How the odds get set
A bookmaker sets a price. It starts with a model, adds a margin, then adjusts for how money comes in. You take the number offered or you don’t bet. In a prediction market no house publishes a line at all: the price is simply the last price two users agreed on. If you think 40 cents is wrong, you can post your own order at 33 or 48 and wait for someone to take it. That makes you a price setter rather than a price taker.
Where the profit comes from
A sportsbook earns from its margin on both sides of the book, and a casino earns from the house edge baked into every game. A prediction market platform earns from trading or settlement fees, not from being on the other side of your position. Your counterparty is another user who thinks you’re wrong. That removes the built-in overround, but it does not make the game easy: you’re now competing with people who may know the subject far better than you do, and spreads plus fees still eat into returns.
| Feature | Prediction market | Sportsbook / casino |
|---|---|---|
| Stated purpose | Information aggregation and speculation | Entertainment and wagering |
| Who sets the price | Traders, via an order book | The operator, with a margin added |
| Your counterparty | Another user | The house |
| Built-in edge | No overround; fees and spreads instead | Bookmaker margin or house edge |
| Exit before the result | Yes, sell your shares any time there’s a buyer | Only if cash-out is offered, at the operator’s price |
| Typical events | Elections, economics, policy, sport, culture | Sport and casino games |
| Regulation | Treated variously as derivatives, binary options or gambling depending on country | Gambling licensing regimes |
How to read a market price as a probability
A contract that pays $1 and trades at 35 cents implies roughly a 35% chance, because the market is willing to pay 35% of the payout for the claim. Converting that into familiar betting language is one division: decimal odds equal 1 divided by the price. So 35 cents is the equivalent of 2.86, and a win returns 65 cents of profit per share.
| Share price | Implied probability | Profit per share if correct | Equivalent decimal odds |
|---|---|---|---|
| $0.10 | 10% | $0.90 | 10.00 |
| $0.25 | 25% | $0.75 | 4.00 |
| $0.50 | 50% | $0.50 | 2.00 |
| $0.65 | 65% | $0.35 | 1.54 |
| $0.90 | 90% | $0.10 | 1.11 |
Now compare that with a two-way sportsbook market priced at 1.91 on each side. Each price implies about 52.4%, and the two add up to roughly 104.7% rather than 100%. That extra 4.7% is the overround, worth about a 4.5% margin to the bookmaker, and it’s the reason sportsbook prices are systematically worse than the true odds. In a prediction market, Yes and No should add to about 100 cents, so the headline price is closer to an honest probability estimate.
Closer, not perfect. Prices still drift away from fair value when liquidity is thin, when one trader holds a large position, or when a market gets flooded with people betting their preferences rather than their analysis. Long-shot contracts priced at 2 or 3 cents are notoriously sticky, partly because there’s little incentive to tie up capital to squeeze out the last cent. Treat a market price as a useful signal, not a forecast with authority.
Why prediction markets are drawing attention and scrutiny
Several things arrived at once. Election cycles gave prediction markets a public scoreboard and a lot of media coverage. Blockchain settlement made it possible to run a global order book with instant payouts and no account manager in the middle. Volumes grew enough for institutional money and mainstream financial firms to start taking the category seriously as a data source. And the product itself is sticky: a market that updates every few seconds as news breaks is a different experience from a pre-match line that barely moves.
The scrutiny has grown at the same speed, and the bank-failure markets show why. Bobby Dean, a Liberal Democrat MP on the UK Parliament’s Treasury Committee, told The Guardian that regulators should not dismiss the risk that a fast-moving market on a bank’s collapse could amplify rumours and nudge depositors toward withdrawing their money. Banks are uniquely exposed to confidence: even a solvent one can buckle if enough customers demand their cash at once, as the 2023 failure of Silicon Valley Bank demonstrated when concerns spread through social media. Polymarket’s counterargument is that it simply makes the kind of information professionals have traded for years, through instruments like credit default swaps, visible to everyone else.
Regulators are drawing lines differently by country. The UK’s Financial Conduct Authority has said the financial prediction-market products it examined amount to binary options, which have been banned from sale to UK retail consumers since 2019 on consumer-harm grounds, and it has been discussing the category with overseas regulators. The Bank of England has been monitoring developments too. Elsewhere, similar contracts are handled as regulated derivatives, and in some places as gambling. If you’re trying to work out whether a platform is legal where you are, the answer depends on your own jurisdiction’s classification, not on what the platform calls itself.
A short checklist before you put money on an event contract
- Read the resolution criteria and the named source. If you can imagine arguing about the result, don’t trade it.
- Check the order book depth, not just the headline price. Thin markets are easy to enter and hard to leave.
- Confirm the platform accepts users in your country, and accept that using a VPN to bypass a block risks your funds.
- Do the probability maths before you click. If you wouldn’t take 1.54 on it, don’t pay 65 cents for it.
- Remember your downside is the full amount you paid for the shares.
Whatever label a platform uses, money placed on an uncertain outcome can be lost in full. Stake only what you can comfortably afford to lose, set your own deposit and loss limits, and treat speculation as spending rather than income. If it stops feeling optional, support services such as BeGambleAware and GamCare offer free, confidential help.
FAQ
What are prediction markets?
Markets where users trade contracts on whether a defined real-world event will happen. Each contract settles at full value if the event occurs and at zero if it doesn’t, and the trading price reflects the crowd’s estimate of the probability.
How do prediction markets work?
You buy shares in an outcome at a price between 0 and 100 cents, from another user rather than from a house. You can sell before the event concludes or hold to settlement, when winning shares pay out $1 each.
How is Polymarket different from betting?
There is no bookmaker setting prices or taking the other side. Prices come from an order book, Yes and No add up to about 100 cents instead of carrying an overround, and positions can be traded in and out before the result is known.
Are prediction markets gambling?
Legally it depends where you are: regulators have variously treated these contracts as derivatives, as binary options, or as gambling. Practically, you’re risking money on an uncertain outcome, so the same caution applies regardless of the classification.
