A contract sits on Kalshi at 63 cents. Another one, on Polymarket, sits at 41 cents. Neither price has a percent sign next to it, neither price is described anywhere on the screen as a probability, and yet every serious trader looking at either one is reading it as exactly that. Sixty-three cents means a 63% chance. Forty-one cents means a 41% chance. The prices are the forecast.
This is the single mechanic that makes prediction markets legible. Once you understand that a price in cents is a probability in percent, the whole apparatus opens up: you can compare a market to a poll, you can spot when two venues disagree, you can work out whether a trade is worth the risk. Miss the mechanic and the screen is just a wall of numbers.
Deposit $20, get $50 to trade
Welcome bonus on the world's largest prediction market.
18+ · New users only · Availability and terms vary by region · Trading involves risk of loss · BeGambleAware.org
The dollar that anchors everything
Start with what a contract actually pays. On a binary prediction market, a Yes share pays out exactly $1 if the event happens and $0 if it does not. A No share is the mirror: $1 if the event does not happen, $0 if it does. That fixed dollar payout is the anchor for the entire pricing system, and it is why every price you see is trapped between 0 and 100 cents.
So if a Yes share is trading at 63 cents, what is that price telling you? It is telling you what the marginal buyer is willing to pay for a coupon that returns $1 in one scenario and nothing in the other. Rearrange it. To be indifferent about paying 63 cents for a shot at $1, you have to believe the shot is worth at least 63 cents on average. Which means you have to believe the underlying event has at least a 63% chance of happening.
That is implied probability. The price, read as a percent, is the market's collective estimate of how likely the event is. Our guide to how prediction market odds work walks through the cents-to-probability translation in more detail, with worked examples on both sides of a market.
Why the price is the forecast, not a description of one
Here is where the intuition often slips. Traditional forecasters produce a probability first and then, sometimes, publish it. Nate Silver's model spits out a number; a pollster reports a share; a betting analyst posts a percentage. The probability is the output.
On a prediction market, the probability is the input, held in place by money. Someone thinks Yes is underpriced at 40 cents and buys. The price ticks up. Someone else thinks 65 cents is too rich and sells. The price ticks down. The final resting number is not a forecast attached to the market; it is the point at which no trader with capital thinks either side is worth taking. That equilibrium, dollar-weighted and updated tick by tick, is the forecast.
Which is why a market price is different in kind from a poll. A poll asks respondents what they think. A market asks respondents to back their thinking with cash. The difference matters, and we have written about the tradeoffs in our piece on whether prediction markets beat polls.
The maths, kept honest
The translation itself is trivial. A price of $0.63 implies a 63% probability. A price of $0.08 implies 8%. A price of $0.50 implies a coin flip. Multiply by 100 and you have the percent.
The more useful skill is reading the pair. On a binary market, Yes plus No should sum to $1.00, because between them they cover every outcome. If Yes is 63 cents, No should be 37 cents. If it is not, something has gone slightly wrong: the spread is wide, the book is thin, one side is stale, or the market maker has parked prices apart to earn a fee. On Kalshi you will often see Yes at 63 and No at 39, for instance, because the bid-ask spread is baked in. That extra two cents is not free money on either side; it is the cost of trading. Beginners sometimes see the gap and imagine an arbitrage. It usually is not one, for reasons we walk through in our guide to prediction market arbitrage.
The honest number is the mid-price. If Yes bids at 62 and offers at 64, the implied probability is 63%, not 62 or 64. The bid and the offer bracket where the market thinks the truth is; the mid is the best single-number read.
Why this is cleaner than bookmaker odds
Bookmakers build a margin into their odds. A sportsbook quoting a coin-flip event might price both sides at decimal 1.91, which implies roughly 52% on each side and adds up to 104%. Those extra four points are the overround, the house's built-in edge, and you have to subtract them before the number means anything.
A prediction market has no overround to back out. Traders match against each other, so the two sides of a binary contract sum to about a dollar by construction rather than by the venue's choice. The price is the probability: a 30-cent contract is a 30% chance, with no margin to remove and no fractional-to-decimal gymnastics. The format is closer to a thermometer than a betting slip.
That does not make it free. Polymarket now charges a taker fee on most retail markets, and a fee changes what a price implies in the same direction an overround does, just by less and from a different place in the transaction. The honest version is that a prediction market price is a more direct read on consensus belief than bookmaker odds, not a costless one.
Where implied probability distorts
Markets are not oracles, and the price is only as good as the traders willing to move it. Three distortions are worth knowing.
Thin books. A market with $5,000 of total volume can be pushed twenty points by a single stubborn buyer. The implied probability is real in the sense that it is the current price, but it is fragile. It does not carry the same weight as a price set by millions in volume. Our primer on liquidity walks through why thin markets are treated with a pinch of salt.
Long-shot bias. Contracts priced under 5 cents tend to trade rich. A 3-cent contract on a genuine long shot is often implying more probability than the actual base rate would justify, because small-stakes traders enjoy buying lottery tickets and market makers have limited incentive to push the price to zero. The same effect, in mirror image, shows up at the 97-cent end.
Resolution risk. If there is any ambiguity about how the market resolves, price will drift away from the true probability of the underlying event. A contract on whether a politician "visits" a country, with no clean definition of visit, will settle on a price that reflects both the event and the disagreement about what counts. That is not a bug in the implied-probability model; it is a reminder that the price is the market's read on the contract, not the read on the underlying question. Our primer on market resolution covers the mechanics.
Reading a market like a trader
Once the mechanic clicks, a screen full of contracts becomes a set of forecasts you can act on. A Kalshi market on the next Fed decision at 78 cents is telling you the market's collective view is a 78% chance of that outcome. A Polymarket contract on a US election at 41 cents is a 41% forecast. Two venues priced differently on the same question is a disagreement worth investigating, though not always worth trading.
And the read cuts both ways. If your own estimate of an outcome sits far from the implied probability, you have a choice: you have found an edge, or the market knows something you do not. Newcomers assume the first. Experienced traders assume the second and demand evidence otherwise. That default is worth adopting. The market's implied probability is not always right, but it is almost always informed, and every trade you place is a bet that you are seeing something the price is missing.
iPredicta is a UK-based discovery platform for prediction markets, tracking prices across venues like Polymarket, Kalshi, and Smarkets and translating them into the forecasts they actually represent. Our editorial work centres on that translation: what the price is really saying, what it is missing, and where the read is worth acting on.
Watch the tails for the same reason. A favourite drifting from 92 cents to 88 cents looks like a small move on the surface. In probability terms the long-shot outcome just went from 8% to 12%, a 50% relative increase. The price moved a little and the forecast moved a lot, and the second is the one worth reading.
Frequently asked questions
How do I convert a prediction market price to a probability?
Read the price in cents as a probability in percent. A Yes share at 63 cents implies a 63% chance the event happens; a No share at 37 cents implies a 37% chance it does not. The maths is that simple because binary contracts pay exactly $1 if they resolve in your favour and $0 if they do not, so the price is the market's average expected payout. If the Yes and No prices do not sum cleanly to $1, use the mid-price of the bid and offer to strip out the spread. The mid is the honest read; the outer quotes are what you actually pay to trade.
Is implied probability the same as odds?
Implied probability is the underlying idea; odds are one of the formats it gets displayed in. A 63% implied probability is the same forecast as decimal odds of about 1.59 or fractional odds of roughly 4/7. Bookmakers tend to quote odds, prediction markets tend to quote cent prices, and both are describing the same probability of the same outcome. The difference is that traditional odds usually contain a bookmaker's margin baked in, whereas a prediction market's cent price is set by traders and often, though not always, sits closer to a clean probability read.
Why should Yes and No prices add up to $1.00?
Because between them, Yes and No cover every possible outcome of a binary market. If the event happens, Yes pays $1; if it does not, No pays $1. There is no third possibility, so the combined value of the two contracts must equal the guaranteed payout of $1. When you see Yes at 63 and No at 39, the extra two cents is the bid-ask spread, which is the cost of trading. That is not free money, though; you cannot buy both sides at those quotes simultaneously, because the offer is what you pay and the bid is what you receive.
Does implied probability mean the market is always right?
No, and treating it that way is a common newcomer error. Implied probability is the market's current collective estimate, weighted by the money behind it, and markets are often better than individual forecasters at aggregating information. But they are not oracles. Thin markets can be pushed by a single trader, long shots often trade rich because people enjoy buying lottery tickets, and any ambiguity in how a contract resolves will pull the price away from the true underlying probability. Read the implied probability as informed and worth respecting, not as truth. Every trade you place is a bet that the price is wrong.
Why do the same event's prices differ across platforms?
Different venues attract different traders, different volumes, and different rules on how a contract resolves. A US election contract on Polymarket might sit at 41 cents while the closest equivalent on Kalshi sits at 44, because the two pools of capital are only loosely connected and each is setting its own price. Sometimes the gap is a genuine mispricing worth investigating; more often it reflects a real difference in liquidity, in resolution language, or in who is allowed to trade. Cross-venue gaps are useful signal, not automatic arbitrage, and the smaller of the two venues is usually the one with the softer price.