Every set of Pinnacle prices carries a hidden surcharge, and the overround makes it visible: convert each price to an implied probability, add them up, and the amount above 100% is the margin. Two sides at 1.91 each sum to 104.71%, so the overround is 4.71%. That number is more than an intermediate step in margin removal. Read over time, it tells you about the market itself.
What the number represents
The overround is the book's built-in edge on that market at that moment. A market priced at 102% leaves little room between the prices; one at 108% has a wide berth. Because the fair prices you compute are the posted prices with this margin removed, the overround sets how much material the removal has to work with. The odds converter handles the format side, turning decimal, American or fractional inputs into the implied probabilities the sum is built from.
That has a direct consequence for anyone consuming the feed. The larger the overround, the more the fair price depends on which removal assumption you apply. On a market priced at 102%, proportional and power removal barely disagree. At 108% on an uneven market, they diverge visibly. The margin's size is a sensitivity gauge for every fair price downstream.
Why it varies between markets
Main lines on major competitions tend to carry smaller margins than side markets and minor leagues, because volume and competition compress them. Live markets typically carry more margin than prematch ones on the same fixture: pricing in real time is riskier, and the margin is the cushion. Paste any complete set of prices into the no-vig calculator and it shows the overround for that market directly.
None of that is a rule to trust blindly for a given match. It is a reason to measure the overround for the markets you actually consume rather than assume a figure.
The overround as a data-quality check
When you read prices from the API, the overround is a free integrity test on every snapshot. A sum that suddenly jumps on a market you track usually means something changed in how that market is being priced. A sum that falls just under 100% is an underround: almost always a sign of an incomplete or mismatched set of outcomes, not a gift.
The calculator flags an underround without changing your inputs, and the same check belongs in a pipeline: compute the sum on arrival and quarantine rows that fail it before they reach a model. The calculator accepts up to twelve outcomes per market, as its page states, checked 2026-10-01.
Watch it move, not just its level
A single overround reading is a fact about a market. A series of readings is a fact about its life: margins often widen when news breaks and tighten toward the close as the book grows confident. If you store prices, store each snapshot's overround beside them. It costs one extra division and makes later analysis much easier to trust.
The margin is the price of the prices. Measure it per snapshot and per market, and it will tell you when to trust a line and when to look twice.