De-Vig Calculator: How to Strip the Margin From Any Betting Line

What a de-vig calculator does and why you need one

Every sportsbook line includes a built-in tax called the vig. A -110/-110 market implies 52.38% probability on each side. Add those up: 104.76%. Real probabilities sum to 100%. The extra 4.76% is the sportsbook's margin.

A de-vig calculator (also called a no-vig calculator, vig-free calculator, or fair odds calculator) removes that margin. It takes the odds from both sides of a market and outputs the true implied probability underneath. That true probability is the number you need for every downstream calculation: expected value, Kelly sizing, closing line value.

Without de-vigging first, you are comparing your probability estimate against a biased number. You will overestimate your edge on favorites and underestimate it on longshots. The de-vig calculator handles all 7 methods described below and works for markets with 2 to 20+ outcomes.

Step 1Input odds from both sides

→

Step 2Calculator removes the vig

→

Step 3Get true probability for each side

→

Step 4Compare to your own estimate

→

Step 5Calculate EV on any difference

→

Step 6Size with Kelly if +EV

The 3-step de-vig process

De-vigging follows the same structure regardless of which method you use. The steps are mechanical. The only variable is how Step 3 distributes the margin removal.

Step 1: Convert odds to implied probability

Use the standard American odds formulas. (The odds converter handles this for any format.)

• Negative odds: |odds| / (|odds| + 100)
• Positive odds: 100 / (odds + 100)

Step 2: Calculate the book sum

Add up the implied probabilities for all outcomes. The total will exceed 100%. The excess is the overround, and the hold calculator converts it into the hold percentage.

Step 3: Remove the overround

This is where the 7 methods diverge. Each one makes different assumptions about how the sportsbook distributes its margin across the outcomes.

Worked example: NFL moneyline at -200/+170

Step 1: Convert.

• -200: 200 / 300 = 66.67%
• +170: 100 / 270 = 37.04%

Step 2: Book sum. 66.67% + 37.04% = 103.71%

Overround: 3.71%. The hold is 1 - (1 / 1.0371) = 3.57%.

Step 3 (multiplicative method):

• Favorite true probability: 66.67% / 1.0371 = 64.28%
• Underdog true probability: 37.04% / 1.0371 = 35.72%

Sum: 100.00%. The vig has been stripped. The sportsbook was pricing the favorite at 66.67% implied, but the de-vigged line says the market's true estimate is 64.28%. That 2.39 percentage point gap was entirely vig.

7 de-vig methods compared

Not all methods give the same result. For tight, two-way markets like NFL spreads, the differences are small (often less than 0.5%). For multi-way markets with heavy longshots, the method matters significantly.

Multiplicative: the default choice

Divides each implied probability by the book sum. Simple, fast, accurate for two-way markets.

When to use it: NFL spreads, NBA totals, any market with two roughly equal sides. This is the right default for 80% of bets.

Shin method: the longshot correction

The Shin model assumes sportsbooks charge more vig on longshots because those lines are where informed bettors are least likely to trade. A 50-1 longshot carries proportionally more margin than the 1-2 favorite.

Worked example: horse racing market.

Three runners priced at -250, +300, +800.

Implied probabilities: 71.43%, 25.00%, 11.11%. Book sum: 107.54%.

The Shin method assigns more vig to the +800 longshot (reducing its true probability from 10.33% to 9.45%) and less to the favorite. In practice, this matches how sportsbooks actually set their lines. If you are betting the longshot, the Shin estimate gives you a more accurate baseline for calculating EV.

Power method: the mathematical middle ground

The power method finds a constant k such that raising each implied probability to the power k produces a set summing to 100%. It naturally assigns more vig to extreme probabilities without the specific insider-trading assumptions of the Shin model.

When to use it: Multi-outcome markets with 3+ runners. Horse racing, futures, award markets.

Worst-case: the conservative filter

Assigns the maximum possible vig to whichever side you are betting. If the book sum is 104%, worst-case assumes your side carries all 4% of the overround. This gives you the lowest possible true probability for your side, meaning any edge you calculate is a conservative floor.

When to use it: When you want to stress-test whether a bet is still +EV under the most pessimistic de-vig assumption. If a bet shows positive expected value even under worst-case de-vig, it is a strong play.

When the method matters (and when it does not)

For a -110/-110 market (book sum 104.76%), all 7 methods produce true probabilities between 49.8% and 50.2%. The difference is irrelevant. You will not make or lose money based on which de-vig method you choose for a standard NFL spread.

The method starts to matter when:

The market is lopsided. A -500/+380 line has a book sum around 108%. The multiplicative method puts the favorite at 78.7%. Shin puts it at 79.4%. That 0.7% gap changes whether a +400 line at another book is +EV or not.

The market has 3+ outcomes. A 10-runner horse race with a 120% book sum produces wildly different de-vigged probabilities depending on the method. The longshot might range from 2.1% (Shin) to 3.8% (multiplicative). At +4000 odds, that difference is the entire edge.

You are comparing across books. If you use multiplicative de-vig on one book and Shin on another, your cross-platform arbitrage calculations will be inconsistent. Pick one method and use it everywhere.

The de-vig calculator runs all 7 methods simultaneously so you can see the range. For most bettors, the multiplicative method is the right default. Switch to Shin or power when the book sum exceeds 110% or the market has 3+ outcomes.

De-vigging prediction market prices

Prediction markets like Kalshi and Polymarket use order books instead of fixed odds. The de-vig concept applies identically: if Yes trades at $0.58 and No trades at $0.47, the combined price is $1.05. The overround is 5%.

The multiplicative de-vig gives:

• True Yes probability: 58% / 105% = 55.24%
• True No probability: 47% / 105% = 44.76%

But prediction markets add another layer. Kalshi charges 7% on profits. Polymarket's vig lives entirely in the spread. You need to de-vig the market price AND account for platform fees to find your true breakeven.

The prediction market fee calculator handles both layers. First de-vig the spread, then subtract the fee impact. The number that comes out is your true cost basis for expected value calculations.

Common de-vig mistakes

Mistake 1: Using implied probability as true probability. This is the most common error. If a line is -150, the implied probability is 60%. But the true probability (after removing vig) might be 57.5%. Betting as if the true probability is 60% means you are systematically overvaluing favorites. Every EV calculation downstream will be wrong.

Mistake 2: De-vigging one side only. Some bettors strip the vig from the side they want to bet and ignore the other side. De-vigging requires both sides (or all sides in a multi-outcome market). You cannot calculate the book sum from one number.

Mistake 3: Using the wrong method for longshot markets. Multiplicative de-vig on a 20-runner horse race with a 140% book sum will produce meaningfully different probabilities than Shin. If you are betting longshots in multi-runner fields, the method selection changes your edge calculation by multiple percentage points.

Mistake 4: Forgetting to de-vig before comparing lines. Line shopping without de-vigging is comparing apples to oranges. Book A at -108 and Book B at -112 on the same side look different, but they might imply the same true probability if Book B has a tighter overall market. De-vig both books, then compare the true probabilities.

De-vig as the foundation of every edge

De-vigging is not an isolated calculation. It is the first step in a pipeline that determines every betting decision.

Step 1: De-vig to find the market's true probability estimate. Use the de-vig calculator.

Step 2: Compare your estimated probability against the de-vigged line. If you think the true probability is 55% and the de-vigged line says 50%, you have a 5% edge.

Step 3: Calculate EV using that edge and the available odds. The EV calculator turns edge into dollars.

Step 4: Size the bet with the Kelly Criterion. Full Kelly is aggressive. Most sharps use quarter or half Kelly.

Step 5: Track closing line value to verify that your edges are real over time. If you consistently beat the closing line after de-vigging, your probability estimates are better than the market's.

Skip de-vigging and the entire pipeline breaks. Your EV calculations are inflated, your Kelly sizing is too aggressive, and your CLV tracking compares against the wrong benchmark. The de-vig calculator takes 10 seconds. Use it on every line before you calculate anything else.

Frequently asked questions

What is a de-vig calculator?

A de-vig calculator removes the sportsbook's built-in margin (vig) from betting odds to reveal the true implied probability. It converts biased odds into fair odds that sum to exactly 100%.

What is the best de-vig method?

For two-way markets (spreads, totals, moneylines), the multiplicative method is accurate and simple. For multi-outcome markets with longshots (horse racing, futures), the Shin or power method produces more accurate results by accounting for favorite-longshot bias.

Why do I need to de-vig before calculating EV?

Implied probability includes the sportsbook's margin, which inflates the probability of every outcome. Without removing it, your EV calculation compares your estimate against a biased number, causing you to overestimate edge on favorites and underestimate it on longshots.

Do prediction markets have vig?

Yes. On order-book platforms like Kalshi and Polymarket, vig appears as the bid-ask spread. If Yes and No prices sum to more than $1.00, the excess is vig. Some platforms also charge explicit fees on top of the spread.

How much difference does the de-vig method make?

For standard -110/-110 markets, less than 0.5%. For lopsided markets or multi-runner fields with 110%+ book sums, the method can shift true probabilities by 1-3 percentage points, which is enough to flip an EV calculation from positive to negative.