How to De-Vig Odds: Find the Fair Probability Behind a Line

Posted odds are not fair odds. Every sportsbook bakes a margin (the vig, or juice) into its prices, which is why the implied probabilities of a two-way market add up to more than 100%. De-vigging strips that margin back out so you can see the fair probability the line actually implies.

What is the vig, and why don't odds add up to 100%?

Convert any moneyline to an implied probability and add the two sides of a market together. On a fair coin you'd expect them to sum to exactly 1.0 (100%). They never do. A typical two-way line sums to something like 1.02 to 1.05. That excess is the overround, and it's the book's margin: the vig.

The vig is why beating the closing number matters more than winning any single bet. To find real value you first have to remove the margin and recover the underlying fair probability.

Step 1: convert the odds to implied probability

American odds convert to probability directly. A favorite priced at -152 implies 152 / (152 + 100) = 60.3%. An underdog at +138 implies 100 / (138 + 100) = 42.0%. Together they sum to 102.3%, and that extra 2.3% is the overround you need to remove.

Step 2: remove the overround

The simplest method, proportional (or multiplicative) de-vigging, just rescales both sides so they sum to 100%: divide each implied probability by the overround. It's fast and transparent, and it's the right first tool.

Proportional de-vigging assumes the margin is spread evenly across both sides. In reality books shade favorites and longshots differently (the favorite-longshot bias). The power method and Shin's method both correct for this. The power method raises probabilities to a fitted exponent; Shin's models the margin as insider money. On lopsided lines they can move the fair number by a full point or more versus the proportional estimate.

Which de-vig method should you use?

For a quick read on a roughly even market, proportional is fine. For favorites, heavy underdogs, or any time you're comparing a soft book against a sharp one to hunt for value, prefer the power or Shin method. The bias correction is exactly where the edge hides. The calculator below runs all three side by side so you can see how much the method choice actually changes the answer.