How to devig odds: turning betting prices into honest probabilities
By George Boyle · Updated 2026-08-21 · The Sport Stack
Devigging is removing the sportsbook’s built-in margin from a price to recover the probability the market actually implies. Raw odds always sum to more than 100% across a market; devigging rescales them back to 100% so you can compare a book’s true opinion against your own, or against a sharper book’s.
Why raw odds overstate every probability
Convert both sides of a typical -110 / -110 spread to implied probabilities and you get 52.4% + 52.4% = 104.8%. The extra 4.8% is the vigorish — the book’s margin. Every posted price overstates the true probability of its outcome, because the book charges you to play.
Using raw implied probabilities in a model or an EV calculation therefore bakes the book’s margin into your math. Devigging removes it: rescale the implied probabilities so they sum to 100%, and what remains is the market’s actual opinion.
The basic method (and when it isn’t enough)
The simplest approach is proportional normalization: divide each side’s implied probability by the sum. For -110/-110 that returns 50%/50%, which matches intuition. For moneylines with a big favorite, proportional devigging is a reasonable start but known to be slightly biased — longshots carry disproportionately more vig (the favorite–longshot bias), and methods like the power or Shin adjustment handle that better.
One rule matters more than the method: devig per book, then aggregate. Averaging raw American prices across books before removing the vig produces a distorted consensus — the correct order is to recover each book’s fair probabilities first, then take the median. Our NFL win-probability consensus is built exactly that way.
What devigged numbers are for
A devigged consensus is the reference point for everything else: whether your model disagrees with the market (an edge claim), whether a stale price at a lagging book is actually value, and whether you beat the closing line. Without devigging first, all three comparisons quietly measure the book’s margin instead of your skill.
Written by George Boyle, who builds The Sport Stack — the models, the public ledger and these explainers. Corrections and questions: hello@thesportstack.io. Who runs this.