No-vig odds / fair odds
Updated November 10, 2026
No-vig odds (or fair odds) remove the bookmaker’s margin so implied probabilities sum to 100%. On sportsbooks, apply multiplicative normalization: divide each raw implied probability by the total across outcomes.
Raw implied probs: Outcome A 55%, Outcome B 55% (110% total). Fair (no-vig) probs: A 50%, B 50% (each 55 ÷ 110). Use the no-vig probability calculator for three-way lines.
- Raw implied total
- 110%
- 55% + 55%
- Fair probability, A
- 50%
- 55 ÷ 110
- Fair probability, B
- 50%
- 55 ÷ 110
On prediction markets: In a liquid $1 binary, yes and no mid-prices often sum near $1.00, already close to a no-vig probability split, unlike a sportsbook line. That gap is approximate: bid/ask spread, fees, and thin liquidity can push sums above $1 or leave stale quotes.
Sportsbooks vs prediction markets
A sportsbook price is never a clean probability; it carries vig by construction. A liquid prediction market mid-price on a binary is closer to a no-vig probability because traders are expressing belief rather than paying a built-in book margin.
Removing the vig from an uneven line
The 55%/55% example splits evenly. An uneven line shows what the method does.
- Convert each side to an implied probability
A two-way line at −150 / +130. For negative odds, divide the odds by the odds plus 100: 150 ÷ 250 = 60.0%. For positive odds, divide 100 by the odds plus 100: 100 ÷ 230 = 43.5%.
- Add them up
60.0% + 43.5% = 103.5%. The 3.5% above 100% is the vig.
- Divide each side by the total
60.0 ÷ 103.5 ≈ 58.0% and 43.5 ÷ 103.5 ≈ 42.0%. The two now sum to 100%.
The fair line works out to about −138 / +138. The margin came off both sides in proportion to their size, so the favorite gave up about two points and the underdog about one and a half.
Other ways to strip the margin
Dividing by the total, as above, assumes the book spread its margin in proportion to each side's probability. Other methods make different assumptions. The additive method subtracts an equal share of the overround from each side: 1.75 points off each in the example, giving 58.25% and 41.75%. Other models assume books load more of the margin onto longshots, and they move the underdog's fair probability lower still.
On a close two-way line the methods land within a fraction of a point of each other. They drift apart on lopsided lines and on markets with many outcomes, where small probabilities make up a larger share of the total.
FAQ
How do you calculate no-vig odds?
Convert each outcome's odds to an implied probability, add them, and divide each one by the total. The results sum to 100% and can be converted back to odds. The no-vig probability calculator does this for two-way and three-way lines.
Are no-vig odds the true probability?
They are one book's line with its margin removed, using an assumption about how that margin was spread. That makes them a cleaner estimate than the raw odds, and still an estimate.
What is the difference between fair odds and implied probability?
Implied probability is read from a single quoted price. Fair odds come after you normalize across every outcome so that the probabilities sum to 100%.