Prediction market glossary
Definitions as a cross-venue engine has to use them — where a term means something specific once you are comparing two venues rather than reading one, the definition says so. Every number quoted here is measured on our own data and is linked to the page that shows the working.
Prediction market
A market in binary contracts that pay $1 if a stated event happens and $0 if it does not. The price is therefore bounded by 0 and 1 and is usually read as a probability. Kalshi is a CFTC-regulated US exchange; Polymarket settles on-chain.
Contract
One unit that pays $1 on a yes resolution. Buying 100 contracts at 40¢ costs $40 and returns $100 if the event happens. Quantities on both venues are whole contracts.
YES and NO
The two sides of a binary market. A NO contract at 60¢ is the same position as a YES contract at 40¢, which is why comparing prices across venues requires checking which side each venue is quoting before subtracting.
Implied probability
The contract price read as a probability: a contract at 62¢ implies a 62% chance, before fees and before any adjustment for the cost of holding the position to resolution.
Cross-venue spread
The difference between the same outcome's price on Kalshi and on Polymarket, in percentage points. A spread is not automatically profit: fees, the two order books' depth, and differences in how the two venues resolve all sit between the gap and the money.
Arbitrage
Buying the same outcome cheaply on one venue and selling it dearly on the other so that the pair pays out regardless of the result. On these two venues most apparent arbitrage does not survive fees; MarketMaster reports the gap gross and after fees separately for that reason.
Edge
The gap expressed as a return on the capital the trade would tie up. Gross edge ignores fees; edge after fees subtracts the venue fee on the filled side. The second number is the only one worth acting on.
Settlement risk
The risk that two venues resolve what looks like the same market differently — different resolution sources, different cut-off dates, different treatment of an ambiguous outcome. MarketMaster flags a pair as divergent when the two contracts do not close within the same window, and a divergent pair is not arbitrage no matter how wide the gap.
Resolution source
The authority a venue names in advance to decide the outcome. Two markets on the same question with different resolution sources can both be correct and still pay differently.
Calibration
Whether prices mean what they say: of everything that traded at 70¢, roughly 70% should resolve yes. Measured on 938 of our own resolved markets, contracts at 20–35¢ resolved yes 13.0% of the time and contracts at 80–90¢ resolved yes 94.3%.
Kalshi taker fee
Charged on the filled side as fee_multiplier × 0.07 × contracts × P × (1−P), rounded up to the cent. Because of the P × (1−P) term the fee is largest on contracts trading near 50¢ and smallest at the extremes, which is exactly where cross-venue gaps look widest.
Maker and taker
A maker posts a resting order and waits; a taker crosses the spread and fills immediately. Cross-venue gaps close in seconds, so the trade that captures one is almost always a taker trade and pays the taker fee.
Order book depth
How many contracts sit at or near the best price. A 6-point gap on 40 contracts of depth is a different proposition from the same gap on 4,000, and depth is why a quoted spread and an executable spread differ.
Slippage
The difference between the price quoted at the top of the book and the average price actually paid once an order eats through several levels.
Volume and open interest
Volume is what changed hands over a period; open interest is how many contracts are outstanding right now. A market can show large volume and thin open interest if the same contracts are being traded repeatedly.
Whale
A single account whose position is large enough to move the price on its own. MarketMaster tracks these from on-chain Polymarket positions, where holdings are public.
Fixture matching
Deciding that a Kalshi contract and a Polymarket contract are the same event. For sports this needs the start time, not just the teams: two meetings of the same pair on the same day are different fixtures, and treating them as one produces spreads that do not exist.
Mid price
The midpoint of the best bid and the best ask. Quoted as a market's price when no trade has just printed, and the number most comparisons are built on.