You've seen the number in a headline. A prediction market "gave the candidate a 78% chance," or a platform "traded $2 billion on the election." Then you go to cite it, and the questions start. A 78% chance according to which market, at what moment, with how much money behind it? Two billion dollars counted how, on which venue, over what window? Prediction market data is unusually easy to quote and surprisingly easy to quote wrong.
This is a guide to the numbers themselves. Not how to trade on them, but how to read them, sanity-check them, and cite them the way a careful analyst or reporter would. If you want the concept first, start with what a prediction market is. This piece is about the data it produces.
What prediction market data actually is
Strip away the interface and a prediction market emits a small, well-defined set of numbers. Four of them do almost all the work, and knowing what each one measures — and what it does not — is most of the battle.
Price, which reads as a probability
The headline number. Because a standard contract pays $1 if the event happens and nothing if it doesn't, its price in cents doubles as the market's estimate of the odds. A contract changing hands at 78¢ is the crowd, money on the line, pricing the outcome at about 78%. That's the implied probability, and it's the single most-cited figure in the space. The thing to remember when you quote it: a price is a snapshot of one moment. It moved before you read it and it will move again, so a responsible citation carries a timestamp.
Trading volume, which measures attention
Trading volume is how many contracts changed hands over a period — a day, a week, a market's whole life. It's the closest thing the space has to a Nielsen rating. High volume means a question is pulling in real money and real attention. It's the number you want when the story is "how big is this," and it's why a prediction market volume tracker is the tool most people actually reach for. But volume is a flow, not a level. It tells you how much traded, not how much is currently at stake.
Open interest, which measures money at stake right now
That second question is what open interest answers. It's the number of contracts currently held open — positions entered but not yet closed or settled. If volume is the traffic through a store all day, open interest is the number of people standing inside it right now. The two move independently, and conflating them is the most common error in prediction-market reporting. A market can post enormous single-day volume from frantic in-and-out trading while its open interest barely moves. Another can have modest daily volume sitting on a large, patient pile of open positions. Volume is not open interest. When you cite one, know which question you're answering.
Notional, which puts the numbers in dollars
Notional is the dollar size a figure represents. Two million contracts, each with a $1 face value, is $2 million in notional volume. It's how you translate contract counts into a dollar figure a general reader understands — and it's also where cross-source comparisons quietly break.
Here's the trap. "Volume" can be counted two ways. One is face value: every contract counts as its full $1, regardless of what it traded at. The other is cash: a contract that traded at 30¢ counts as 30 cents of money that actually changed hands. For a market full of coin-flip contracts trading near 50¢, those two methods differ by roughly a factor of two. Neither is wrong, but they are not the same number, and not every platform or data provider reports the same one. The practical rule: before you set one venue's volume next to another's, confirm both are counting the same way. If you can't confirm it, say so in the citation.
Provisional versus settled: data has a lifecycle
A number you pull mid-event is not the number you'll get after it resolves. While a market is live, today's figures are provisional — they update continuously as trades arrive and won't be final until the day's books close and, eventually, the event settles. A volume figure quoted at 2 p.m. can grow by dinner. A price is provisional by nature; it's a live reading, not a verdict.
For anyone citing a figure, this matters more than it sounds. "Kalshi traded $X on this market" is a moving target until the market is settled and the day is closed. When precision counts, anchor the claim to a settled, closed period rather than a live one, and note when you pulled it.
Why cross-platform is the hard part
A single, trustworthy number across venues is the genuinely difficult and genuinely valuable thing in this field, because the major platforms don't share conventions. Three anchor the landscape today.
- Kalshi is a CFTC-regulated U.S. exchange, with a live trade stream and deep coverage of economics, politics, weather, and sports.
- Polymarket is a large blockchain-based market on the Polygon network, denominated in the USDC stablecoin, known for very high-volume election and news markets.
- Polymarket US is a separate CFTC-regulated U.S. exchange, distinct from both global Polymarket and Kalshi.
Three differences make combining their data real work. First, day boundaries. The platforms don't agree on when a "day" ends, so a naive "daily volume" comparison can be comparing slightly different windows. Second, volume conventions, as above — face value versus cash. Third, data structure and vocabulary. What one platform calls a "series," another calls a "market," and what it calls a "market" is a single contract. Line the data up wrong and you'll double-count or under-count without ever seeing an error. Doing this correctly — one comparable number per topic, per venue — is exactly the problem TickerTracker exists to solve. Our head-to-head, Kalshi vs. Polymarket, walks through where the two genuinely diverge.
How to read the data like an analyst
Numbers in hand, a few habits separate a careful reading from a credulous one.
Always read a price next to its volume. A probability is only as trustworthy as the market behind it. A busy, liquid market — lots of resting orders, a tight spread between buyers and sellers — produces a price that's hard to push and reflects real consensus. A thin market can swing on a single order. The same 65% means very different things in a market that traded ten thousand contracts today and one that traded ten. When a price looks surprising, the volume beside it usually explains why.
Treat a volume spike as a question, not an answer. A sudden jump in activity tells you attention arrived; it doesn't tell you why. Sometimes it's news breaking. Sometimes it's a single large participant, or a thin market being nudged. The spike is the start of the reporting, not the end of it. Trace it back to what happened.
Watch volume over time, by category, and by ranking. A single figure is a data point; the shape is the story. Volume over time shows whether interest is building or fading. Volume by category shows where the money is concentrated — politics, sports, economics, crypto. And a volume ranking surfaces which specific questions are drawing the crowd right now. That triangulation, more than any lone number, is what reading prediction market data well actually looks like. For the odds-focused version of this, see how to read prediction market odds.
Where prediction market data lives
So where do you actually get this? The landscape is more fragmented than the polished headlines suggest, and each source involves a trade-off.
The platforms themselves publish data through their own APIs and dashboards. This is the primary source, and it's authoritative for that one venue. The cost is that you're on the hook for the plumbing — pulling, normalizing, and reconciling it yourself — and you get one platform at a time.
On-chain analytics, chiefly Dune, are powerful and, for crypto-native markets, superb. Because Polymarket settles on a public blockchain, its every trade is queryable, and the community has built rich Dune dashboards on top of it. The catch is real: these tools generally want a wallet or SQL fluency, they lean heavily toward on-chain venues, and coverage of an off-chain exchange like Kalshi is thin or bolted on. Great for a Polymarket deep-dive, awkward as a neutral cross-platform view.
Single-platform trackers and press data tables fill part of the gap with clean, readable figures. They're convenient, but you're trusting someone else's methodology, and when a tracker blends venues with different volume conventions, the combined number can mislead in ways that aren't visible on the page.
A neutral, cross-platform view is the harder thing to build and the reason TickerTracker exists. It's a prediction market data aggregator in the working sense of the phrase: it pulls Kalshi, Polymarket, and Polymarket US into one normalized model, so the same topic can be lined up across venues on comparable terms. The stance is deliberately analytical rather than promotional — the goal is to help you understand how big a market is, how active it is, and what its money-weighted crowd is forecasting, not to route you toward a position. You can watch the aggregate move in real time on live vitals, or browse the markets index to see where volume is concentrated right now.
A closing caution that outlives any single figure: prediction market data changes by the second, and the conventions behind it differ by venue. For a number you intend to publish, pull it live, note when you pulled it, and know which of the four metrics — price, volume, open interest, or notional — you're actually quoting. Get those right and the data is genuinely one of the sharpest, fastest forecasting signals available.
Common questions
Where do I get prediction market data?
From four kinds of source: the platforms' own APIs, on-chain analytics like Dune (strong for Polymarket, thin on Kalshi), single-platform trackers, and neutral cross-platform aggregators. Each trades convenience against coverage and methodological control. For a normalized view across Kalshi, Polymarket, and Polymarket US, TickerTracker's live vitals and markets index are built for exactly that.
What's the difference between volume and open interest?
Volume is how many contracts changed hands over a period — a measure of activity. Open interest is how many contracts are currently held open — a measure of money at stake right now. A market can have huge volume and modest open interest, or the reverse, so cite the one that answers your actual question.
Why is a prediction market price a probability?
Because a standard contract pays $1 if the event happens and $0 if it doesn't, its price in cents reads directly as the market's estimate of the odds. A contract at 64¢ implies about a 64% chance. See implied probability for the full explanation.
Why don't two sites report the same volume for the same market?
Usually because they're counting differently. "Volume" can mean face value (every contract counts as its full $1) or cash actually spent (contracts times their traded price), and the two can differ by roughly half in coin-flip markets. Different day boundaries across platforms add to the gap. Before comparing, confirm both sources use the same convention over the same window.
Is prediction market data the same as sports-betting data?
No. A sportsbook's odds are set to build in the house's margin. A prediction market's price is set by participants and reads as a probability, which is why researchers and reporters treat it as an information signal to study rather than a line to act on.