No Vig Fair Odds: How to Remove the Margin

No vig fair odds are betting prices calculated after removing a market's built-in margin. To find them using the proportional method, convert every outcome's odds into an implied probability, divide each probability by their total, then convert the results back into odds.

That gives you a margin-free estimate, not proof of an outcome's actual chance. The distinction matters: removing vig fixes the market's probability total, but it doesn't fix a bad assessment of the teams or players.

How do you calculate no vig fair odds?

Start with a complete set of mutually exclusive outcomes: exactly one must win. For a two-way market without a push, that's both sides. For a three-way soccer result, you need the home win, draw, and away win.

Convert each price into decimal odds, then calculate its raw implied probability:

Implied probability = 1 ÷ decimal odds

If you're starting with American odds:

• Negative odds: probability = absolute odds ÷ (absolute odds + 100)
• Positive odds: probability = 100 ÷ (odds + 100)

Add the probabilities. Any amount above 100% is the market's overround, a measure of the margin embedded in those prices. It isn't the sportsbook's guaranteed or realized profit rate.

The proportional method, also called multiplicative normalization, removes that excess:

Fair probability for an outcome = its implied probability ÷ sum of all implied probabilities

Finally, fair decimal odds = 1 ÷ fair probability. Keep full precision during the calculation and round only the displayed results.

Worked example: remove vig from odds

Say a hypothetical two-way market lists both sides at -110, with no possibility of a push.

1. Convert each side: 110 ÷ (110 + 100) = 11/21 ≈ 0.5238095, or about 52.38%.
2. Add both probabilities: 11/21 × 2 = 22/21 ≈ 1.047619, or about 104.76%.
3. Find the overround: (22/21 − 1) × 100 = about 4.76 percentage points.
4. Normalize each side: (11/21) ÷ (22/21) = 0.50.
5. Convert back: 1 ÷ 0.50 = 2.00 decimal, equivalent to +100.

Both sides have a 50% proportional fair probability. That happens because the starting prices are identical, not because every two-way market is a coin flip.

For an uneven hypothetical market, suppose Side A is -150 and Side B is +100:

The steps are 0.60 ÷ 1.10 = 6/11 for Side A and 0.50 ÷ 1.10 = 5/11 for Side B. Their fair American prices are −100 × (6/11) ÷ (5/11) = -120 and 100 × (6/11) ÷ (5/11) = +120.

Which devig methods should you use?

Different devig methods make different assumptions about how the margin is distributed. The original prices alone don't reveal a unique set of true probabilities.

Proportional method

Divide every implied probability by the total. It's simple, transparent, and preserves the ratio between outcomes' probabilities.

This is a useful starting point for a manual calculation. It also gives you a baseline for comparing other methods.

Additive method

Subtract an equal amount of probability from every outcome:

Fair probability = raw probability − [(total probability − 1) ÷ number of outcomes]

In the hypothetical -150/+100 market, subtract 0.05 from each side. The results are 55% and 45%, rather than the proportional method's approximately 54.55% and 45.45%.

The limitation is that an outcome with a very small raw probability can become negative. A negative probability isn't usable.

Power method

Raise each raw implied probability to the same exponent, chosen so the adjusted probabilities sum to one:

Fair probabilityᵢ = raw probabilityᵢᵏ, where Σ raw probabilityᵢᵏ = 1.

Finding the exponent usually requires numerical calculation. Unlike proportional normalization, this method changes the probability ratios between outcomes. If your estimated edge disappears when you change methods, treat it as fragile rather than choosing whichever output favors your bet.

How do fair odds help identify a value bet?

Fair odds give you a benchmark to compare against an available price. A better payout than that benchmark suggests positive expected value, but only if the estimated probability is sound.

Use Side B's proportional probability from the table, 5/11, and suppose another hypothetical market offers +130. For a $100 bet:

1. Profit if it wins: $100 × 1.30 = $130.
2. Loss if it loses: $100.
3. Expected profit: (5/11 × $130) − (6/11 × $100) = $50/11, or about $4.55.
4. Expected return on stake: ($50/11) ÷ $100 × 100 ≈ 4.55%.

That's an average expectation under the probability assumption, not a prediction for the individual bet. The bet can still lose.

Once you've calculated a probability estimate, you can check the arithmetic with an expected-value calculator using that estimate and the offered price.

What should you check in a fair odds calculator?

Check which method it uses and whether it accepts every outcome in the market. A calculator shouldn't silently treat a three-way market as two-way or mix prices for different totals.

Compare like with like: the same event, line, settlement conditions, and odds snapshot. Markets involving pushes, voids, or partial wins need extra care because the simple win-or-lose calculation may not describe their payouts.

For a two-way market where a push returns the stake, normalized probabilities estimate the relative win chances conditional on no push. Calculating expected profit per original bet also requires accounting for the push probability.

Common mistakes when using fair odds

• Calling fair odds “true odds.” Devigging produces estimates based on prices and assumptions, not verified probabilities.
• Using only one side. You can't measure a complete market's overround from one price alone.
• Mixing timestamps or lines. Unmatched quotes can create misleading probability totals.
• Assuming the reference market is correct. Devigging doesn't remove pricing errors or stale information.
• Judging a process by one result. A winning bet doesn't validate its estimated edge, and a losing bet doesn't disprove it.

Look for records that include losses as well as wins, rather than relying on isolated winning screenshots. When reviewing our track record or today's picks, check the prices and assumptions yourself. A listed pick isn't an automatic bet.

18+ only. Only bet what you can afford to lose.

Frequently asked questions

Are no vig fair odds the same as true odds?

No. No vig fair odds remove the margin according to a chosen method, but the underlying market can still misjudge an outcome's probability. Treat them as a benchmark rather than a verified forecast.

Can you remove vig using only one side's odds?

Not without making additional assumptions. You need prices for all mutually exclusive outcomes to calculate the market's total implied probability and normalize it.

What is the best devig method?

There isn't one method that reveals the true probabilities for every market. Proportional normalization is a useful baseline because it's transparent, while comparing other methods can show how sensitive an estimated edge is to your assumptions.

Does getting better than no vig fair odds guarantee a profit?

No. It suggests positive expected value only if the fair-probability estimate is accurate and the settlement conditions match. Individual bets can lose, and repeated bets can still produce losses.