De-vigging removes the sportsbook's margin from a market's prices to reveal the fair odds underneath. It's how you turn two prices into one honest probability.
The method, with real numbers
Take a total priced -110 over, -110 under. Each -110 implies 52.4%. Together that's 104.8% — impossible, and the extra 4.8 points are the book's hold.
Normalize: 52.4 ÷ 104.8 = 50% each side. The no-vig fair price is +100. That's what the market actually believes, and it's why grinding -110 markets without an edge loses exactly the hold, forever.
A lopsided example: -200 / +170. Implied: 66.7% and 37.0%, total 103.7%. De-vigged, the favorite is 64.3% — fair price about -180. If another book posts -170 on that favorite, that price is better than fair: +EV.
One book lies, the market doesn't (much)
De-vigging one book gives you that book's opinion minus its cut. De-vig every book and take the median and you get a consensus that's hard for any single stale line to fool. When one book's price beats the consensus fair, the gap is measurable — that's the number in the Edge column on [the board](/edge).
The limits
De-vigging assumes the hold is split evenly between sides, which isn't always true (books shade popular sides). More sophisticated methods exist, but the simple normalization gets you most of the way, and it's the version you can do in your head at the window.
