No-Vig Fair Odds
WeBall3 min readUpdated Aug 15, 2026Reviewed Aug 15, 2026
This is the calculation that separates people who evaluate bets from people who have opinions about games.
The problem it solves
Posted odds always imply more than 100% across a market, because the excess is the bookmaker's fee. Compare your own probability estimate to the posted number and nearly everything looks like a bad bet — not because it is, but because you are comparing against a price that includes a tax.
The fair probability strips that tax out and gives you the market's actual belief. That is the number worth arguing with.
The method
fair(A) = P(A) ÷ [P(A) + P(B)] · fair(B) = P(B) ÷ [P(A) + P(B)]
A moneyline at −200 / +170
- 01−200 implies 200 ÷ 300 = 66.67%.
- 02+170 implies 100 ÷ 270 = 37.04%.
- 03Total is 103.70%, so the hold is 3.70%.
- 04Normalise: 66.67 ÷ 103.70 = 64.29%, and 37.04 ÷ 103.70 = 35.71%.
The no-vig calculator runs this for any two prices, and shows both the fair probabilities and the fair odds.
Three-way and multi-way markets
The same logic extends to any number of outcomes: convert everything, sum, divide each by the total. This is how soccer's home/draw/away markets get de-vigged, and how a 30-team futures field gets normalised.
The larger the field, the more the simple method distorts. A championship market with 130% total implied probability is hiding a lot of margin, and proportional removal will systematically make the long shots look better than they are.
Where the simple method breaks down
Multiplicative de-vig assumes the bookmaker spreads its margin evenly across outcomes. Real books do not. They load more margin onto long shots, because that is where casual money goes and where they carry the most risk.
The consequence: on a lopsided market, simple normalisation understates the favourite's true probability and overstates the underdog's. Two corrections exist.
The power method raises each probability to an exponent chosen so the set sums to one, which shifts more of the correction onto the long shots.
The Shin method models the margin as protection against informed bettors, which produces a similar shape with a different theoretical basis.
Both matter most on markets with a heavy favourite or a wide field. On a −110/−110 spread, all three methods give essentially the same answer, which is why the simple version is fine for most day-to-day work.
Using it in practice
Once you have the fair probability, everything else follows. Your edge is your probability minus the fair probability. Your expected value is that edge converted into money at the price you can actually get. Your stake follows from the edge via Kelly.
Skip this step and every number downstream is wrong in the same direction — in the book's favour.
Common questions
Is the simple method accurate?
It is accurate enough near even money and slightly overstates heavy favourites. Removing margin proportionally assumes the book applies it evenly, but in practice books load more margin onto long shots. The power and Shin methods correct for this.
Which book's prices should I use?
The sharpest one you can, or a consensus across several. A market-making book with low hold and high limits produces a much better fair estimate than a recreational book that shades its lines toward public money.