Sportsbook lines bake in a commission (the vig, or juice). Strip it out and you get the book's best estimate of each side's true probability — a much better number for comparing against your own handicapping or against other books.
The standard formula
For any two-sided market:
- Convert each side's odds to implied probability.
- Add the two implied probabilities together. The amount this sum exceeds 100% is the hold.
- Divide each side's implied probability by the total. The result is each side's no-vig probability.
In one line:
no-vig prob = implied prob ÷ (side A implied + side B implied)
Worked example: −110 / −110 (standard juice)
Both sides priced at −110.
- Implied probability per side: 110 / 210 = 52.38%
- Sum: 52.38% + 52.38% = 104.76%
- Hold: 104.76% − 100% = 4.76%
- No-vig probability per side: 52.38 / 104.76 = 50.00%
- No-vig American odds per side: +100 (a true coin flip)
So a −110/−110 spread is just the book pricing a 50/50 game and taking ~4.76% on top.
Worked example: −120 / +100 (asymmetric line)
Favorite is −120, underdog is +100.
- Favorite implied probability: 120 / 220 = 54.55%
- Underdog implied probability: 100 / 200 = 50.00%
- Sum: 104.55% → hold ≈ 4.55%
- Favorite no-vig probability: 54.55 / 104.55 = 52.17%
- Underdog no-vig probability: 50.00 / 104.55 = 47.83%
- Favorite no-vig American odds: ≈ −109
- Underdog no-vig American odds: ≈ +109
So the book's view of the true matchup is favorite ~52% to win, underdog ~48% — a slight edge, but not as steep as the −120 line suggests at first glance.
Why bother removing vig
To estimate true probability for EV math. Sharp books (Pinnacle, Circa) are widely regarded as the most accurate market pricers in the industry. Take their two-sided line, strip the vig, and you have a defensible estimate of the true probability for each side. Plug that into the EV calculator against any other book's odds to see if there's a positive-EV bet.
To compare books fairly. Two books with different prices may have similar no-vig views — or wildly different ones. No-vig math normalizes for the commission and lets you see what each book actually thinks the game is.
To find arbitrage. If the implied probabilities across two books sum to less than 100%, that's an arb. (The arbitrage calculator handles the stake split.)
A few caveats
The simple division method above is the standard and works well for most two-sided markets at small-to-medium hold. It assumes the book applies the vig proportionally — which is mostly true for spreads and totals, but less true for heavy favorites where books deliberately overprice the favorite side. For markets with extreme asymmetry, methods like Shin or the power method give better estimates, but they're overkill for typical handicapping.
For three-way markets (soccer 1X2), apply the same formula to all three sides — sum all three implied probabilities, divide each by the total.