A prediction market lets you put real money on whether a specific event will happen. You buy a YES contract for between 1 cent and 99 cents; if the event happens, the contract pays out $1 — if it doesn’t, it pays $0. The price you paid is the market’s collective estimate of how likely the event was. That single mechanism — buying probability and selling probability — is the entire product. The rest of this guide walks through how that mechanism becomes a usable platform, the maths you need to size positions, and where to actually trade.
The core mechanic: YES, NO, and a price between $0.01 and $0.99
Every prediction-market contract is binary. There are two sides — YES and NO — and exactly one of them pays out $1 per share at resolution. The other pays $0.
The price of YES at any given moment is the market’s collective probability estimate that the event will happen.
- YES at $0.62 = the market thinks there’s a 62% chance the event happens.
- NO at $0.38 = the market thinks there’s a 38% chance the event doesn’t happen.
- YES + NO = $1.00 (in theory; in practice slightly more — that gap is the platform’s edge).
You can buy either side. If you buy YES at $0.62 and the event happens, you get $1 per share — a profit of $0.38 per share, or about 61% on your money. If the event doesn’t happen, you lose your $0.62.
You can also sell out before resolution. If you bought YES at $0.62 and the price moves to $0.80 because new information makes the event more likely, you can sell now for $0.80 and book a $0.18 profit per share without waiting for resolution.
Implied probability and odds: the same thing in three forms
Prediction-market prices are just probabilities in a different costume. If you’ve used a sportsbook, you’ve already done this maths — you just used different numbers.
| Prediction market price | Decimal odds | American odds | Fractional odds | Implied probability |
|---|---|---|---|---|
| $0.10 | 10.00 | +900 | 9/1 | 10% |
| $0.25 | 4.00 | +300 | 3/1 | 25% |
| $0.50 | 2.00 | +100 / -100 | Even | 50% |
| $0.62 | 1.61 | -163 | 8/13 | 62% |
| $0.80 | 1.25 | -400 | 1/4 | 80% |
| $0.95 | 1.05 | -1900 | 1/19 | 95% |
A YES contract trading at $0.50 is the same probability as decimal odds of 2.00, American odds of +100, or “even money” — they’re all “50% likely.”
If you’re more comfortable thinking in sportsbook odds than prediction-market prices, our implied-probability and odds converter translates between formats. Most prediction-market traders learn to read prices directly within a few sessions.
How a market is created and resolved
A prediction market needs three things to be tradable:
- A clearly defined question. “Will the FIFA World Cup 2026 final go to penalties?” is good. “Will the World Cup be exciting?” is not — there’s no objective resolution.
- A resolution date. When does the contract pay out? For some markets that’s the moment the event happens (a championship final). For others it’s a cutoff date (“Will Bitcoin trade above $200,000 by 31 December 2026?”).
- A resolution source. A public, verifiable reference that determines the outcome — an official sports score feed, a regulator filing, a published index, an executive order on a federal register. This is the single most important field on a market page. If the resolution source is fuzzy, the market is fuzzy.
When the event resolves, the contract pays out $1 per share to the winning side and $0 to the losing side. Payouts hit your wallet automatically — there’s no manual claim.
Two ways markets handle liquidity: AMM vs orderbook
Two main models power the platforms you’ll encounter. They behave differently and matter for how you trade.
Orderbook (Roobet, Kalshi, Smarkets, Limitless)
Looks like a stock exchange. Buyers and sellers post limit orders at specific prices, and trades execute when a buy and sell match. You can see the depth of the book at each price level, and you’ll get filled at the best price available — or post your own limit order if you want a better price than the current best bid/offer.
Trade-off: Tighter spreads on liquid markets, transparent depth, slippage on big orders if the book is thin.
Automated market maker (Polymarket)
A liquidity pool with a pricing curve. Every trade moves the price along that curve — buying YES makes YES more expensive, buying NO makes YES cheaper. Liquidity providers seed the pool and earn fees from every trade. There’s always a price, even with no other traders active, but the price you get worsens with the size of your trade (slippage is built into the curve).
Trade-off: Always-available pricing, no need to wait for a counterparty, but slippage on large orders is more significant than on a deep orderbook.
For most casual trading, the difference is invisible — both feel like clicking “buy” and getting filled. For larger trades or thin markets, orderbooks usually give better fills.
Resolution sources matter more than they look
The single biggest beginner mistake is not reading the resolution source carefully. A few worked examples of where it bites:
- “Will candidate X win the 2026 election?” — Resolves to: AP race call. Not your local newspaper, not Twitter consensus, not your gut. If AP hasn’t called the race by the resolution date, the market may extend or resolve NO depending on the rules.
- “Will Bitcoin trade above $X by date Y?” — Resolves to: a specific exchange’s published price at a specific timestamp. A flash spike on a different venue doesn’t trigger resolution.
- “Will [company] IPO by date Y?” — Resolves to: SEC filing of a specific document type. A “private listing” or “direct listing” may or may not qualify depending on the wording.
Read the resolution source before you size a position. The platform’s interpretation, not yours, decides what happens.
How you actually make money
Three sources of edge in prediction markets:
1. Information edge
You know something the market hasn’t priced in yet. This is rarer than people think — markets aggregate information fast — but it does happen, especially on niche markets where mainstream news cycles are slow to react.
2. Probability calibration edge
The market is mispricing probability because of cognitive biases (recency, availability, narrative). Markets often over-react to dramatic news and under-react to slow grinding facts. If you can tell when consensus is wrong, you have an edge.
3. Liquidity provision edge
You make markets — post both YES and NO at slightly wider than mid — and earn the spread. This is how AMM liquidity providers and active orderbook market makers make money. Requires more capital and active management.
For beginners: focus on category 2. Trade markets you genuinely understand, look for places where consensus pricing seems wrong, and size small while you calibrate your judgement.
How to size positions: Kelly, in plain English
If your edge is real, the maths of optimal sizing is the Kelly criterion. The full formula is:
bet fraction = (edge × odds - (1 - edge)) / odds
In plain English: if you think the true probability is 60% and the market is offering YES at $0.50 (50% implied), your edge is 10 percentage points. Kelly says bet 20% of your bankroll. Most professionals bet half-Kelly — 10% in this case — to manage variance.
Three practical rules:
- Never bet on probabilities you can’t justify. If you can’t write down why you think the true probability is X, you don’t have an edge.
- Half-Kelly or quarter-Kelly is plenty. The optimal full-Kelly bet has high variance even when right. Cutting size in half gives up almost nothing in expected return and dramatically smooths the ride.
- One bet at a time on a thin market. Big positions move the price against you. If the book is thin, scale in over several smaller orders.
Our bankroll calculator covers similar maths for casino games — the principle is identical: stake size is determined by your edge, not your gut.
Where to trade: a quick map
Five mainstream venues, each suited to a different user:
- Roobet — crypto-native (BTC, ETH, USDT and others), no bridging, prediction markets + sportsbook + casino in one wallet. Best for international users who want the easiest experience. Restricted in the US, UK, France, Germany and several other countries.
- Polymarket — deepest book in the world, especially on US politics and macro. Requires USDC on Polygon and a self-custodied wallet. Officially geo-restricted from the US and UK.
- Kalshi — CFTC-regulated US-domestic exchange. The only legal option for US residents who want regulated event contracts. Bank-grade KYC, USD funding only.
- Smarkets — UKGC-licensed exchange for UK and EU users. Fiat funding, peer-to-peer matching.
- Manifold — play-money only, no real cash. Best for learning the mechanics without risking money.
For a deeper breakdown see our best Polymarket alternatives, best Kalshi alternatives and best crypto prediction markets rankings.
Common mistakes to avoid
- Treating $0.95 contracts as “almost free money.” A 95% probability still misses 1 in 20 times, and you’re risking $0.95 to win $0.05. Expected return is bad and variance is brutal when it goes wrong.
- Buying narratives, not probabilities. “X is definitely going to happen” is not a price; it’s a feeling. Buy when the price is wrong, not when the story is exciting.
- Ignoring fees and spreads. A 5-cent spread on a $0.50 market is a 10% effective fee on a round-trip trade. Liquid markets have tighter spreads.
- Holding to resolution unnecessarily. If the price hits your fair-value estimate before resolution, sell. Free expected value isn’t a reward for waiting.
- Sizing flat instead of by edge. Putting the same dollar amount on every trade ignores that some bets have 3x the edge of others.
Prediction markets vs sportsbooks: when to use which
Same outcome, very different prices. Sportsbooks set lines and adjust them based on bet flow; prediction markets discover prices through trading. The two often disagree — and when they do, that’s a signal.
We compare the mechanics directly in prediction markets vs sportsbook — useful if you’re already a sports bettor wondering whether predictions add anything. Short answer: yes, mostly for season-long futures and non-game outcomes the sportsbook doesn’t list.
Frequently asked questions
What is a prediction market?
A prediction market is a platform where you buy and sell YES/NO contracts on real-world events. Each contract pays $1 if its side is correct and $0 if not. The current price between $0.01 and $0.99 is the market’s collective estimate of how likely the event is to happen. You can hold contracts to resolution or trade them before resolution if the price moves.
How do prediction markets make money?
Most prediction markets earn revenue through a small spread between the buy and sell price (orderbook venues), through liquidity-pool fees on every trade (AMM venues), or through a transaction fee on each trade. There is no traditional “house edge” in the casino sense — the platform makes money on volume rather than on bettor losses.
Are prediction markets gambling?
The legal classification varies by jurisdiction. In the US, the CFTC regulates them as event contracts traded on designated contract markets (e.g. Kalshi). In Europe and elsewhere, regulated platforms typically operate under gaming or financial licences. The mechanics of trading are similar to financial derivatives, but the regulatory framing depends on country.
Can I lose more than I deposit?
No. Prediction-market contracts are fully collateralised — your maximum loss on a YES contract bought at $0.62 is $0.62 per share. There is no leverage, no liquidation, no margin call. The most you can ever lose on any given contract is the amount you paid for it.
What is implied probability in a prediction market?
The implied probability of a YES contract is just its price expressed as a percentage. A YES contract at $0.62 has an implied probability of 62% — the market collectively thinks there’s a 62% chance the event happens. The same applies to NO contracts: a NO at $0.38 implies a 38% chance the event doesn’t happen.
How is a prediction market different from a sportsbook?
A sportsbook sets odds for individual game outcomes (moneyline, spread, total) and adjusts them based on bet flow. A prediction market lets users trade contracts that resolve based on a verifiable real-world outcome — often longer-dated futures, non-sports events, or outcomes the sportsbook doesn’t price. The two often disagree, especially on season-long outcomes, which is sometimes a tradable signal.
Where can I trade prediction markets?
The mainstream venues are Polymarket (deepest book, USDC on Polygon, no US/UK access), Kalshi (CFTC-regulated, US-only), Roobet (crypto-native, international ex-US/UK), Smarkets (UKGC-licensed, UK/EU) and Manifold (play-money). The right one depends on where you live and how you want to fund.
Do I need crypto to trade prediction markets?
Not for all venues — Kalshi and Smarkets accept fiat. But the deepest crypto-native books (Polymarket, Roobet, Limitless) all use crypto funding. Roobet accepts BTC, ETH, USDT and other major coins directly; Polymarket and Limitless require USDC on specific networks (Polygon and Base respectively).
