HOW TO REMOVE
THE VIG

Convert every outcome in a market to an implied probability, add them up, then divide each one by that total. The result is a set of probabilities that sums to 100% instead of 104% — the book's view with its margin taken out.

Two -110 prices become 50% and 50%. That is the whole method, and for symmetric markets it is genuinely all you need.

The complications start when the two sides are not priced symmetrically, when the market has more than two outcomes, and when you start treating one book's number as the truth. Those cases are the rest of this page.

The margin hides inside the price

A sportsbook does not charge a visible commission. It shortens the odds on both sides, so the implied probabilities add up to more than 100%. That excess is the margin, also called the vig, juice, or overround.

Because it is baked into the price rather than shown separately, it is easy to compare two books and not notice you are comparing two different-sized commissions.

Why -110 on both sides is not a coin-flip price

-110 implies 52.38%. Two of them sum to 104.76%, so the market carries about 4.76 percentage points of margin.

In practice that means you need to win 52.38% of your -110 bets just to break even. Betting a genuine coin flip at -110 loses money steadily. Nothing about the outcome is unfair; the price is.

How much margin is normal

Standard two-way markets are usually the tightest a book offers. Player props, alternate lines, and same-game parlays carry noticeably more. Multiway markets like touchdown scorers carry the most of all, sometimes by a wide margin.

Working out the margin before you bet is worth doing on its own, separately from any devigging: it tells you which markets are expensive to play in.

This is the standard method, called multiplicative or proportional devigging. Everything else on this page is a variation on it.

Step 1 - Convert each price to an implied probability

For a negative American price, divide the number by itself plus 100: -110 becomes 110 / 210 = 52.38%. For a positive price, divide 100 by the number plus 100: +140 becomes 100 / 240 = 41.67%. For decimal odds, use 1 divided by the decimal.

Step 2 - Add every outcome in the market

Add the implied probabilities of every outcome, not just the one you care about. Two-way markets have two; a three-way has three; a touchdown scorer market has as many as the book lists. A -110 / -110 market totals 104.76%, and the amount above 100% is the sportsbook's margin.

Step 3 - Divide each outcome by the total

Dividing rescales the set so it sums to 100%. In the -110 / -110 market, 52.38 / 104.76 = 50.0% for each side. Those are the no-vig, or fair, probabilities implied by that book's prices.

A receiving-yards line priced -110 on both the over and the under.

Now a market priced -160 on one side and +140 on the other. Here the three devigging methods stop agreeing, which is the part most explanations skip.

Raw implied probabilities are 61.54% and 41.67%, totalling 103.21% — a margin of 3.21 points.

All three remove the same total margin. They differ in how they assume the book distributed it between the two sides.

Multiplicative

Divide each probability by the total. This assumes the margin is applied proportionally, so the favourite carries more of it in absolute terms. It is the default almost everywhere, including in most free calculators.

It is simple, it is fast, and on tight two-way markets it is close enough that the method choice barely matters.

Additive

Subtract an equal share of the margin from each outcome. This assumes the book spread its commission evenly across the sides rather than proportionally.

On the example above it moves the favourite's fair probability up about a third of a point relative to multiplicative — small, but it points the opposite way from what most people expect.

Power

Raise each probability to a power chosen so the set sums to exactly 100%. On the asymmetric example the exponent works out near 1.05.

This is the method that takes favourite–longshot bias seriously: it shrinks longshots harder than favourites, because longshots are systematically overpriced relative to how often they win.

Which one to use

On a tight two-way market, multiplicative is fine and the disagreement between methods is smaller than your own estimation error.

On lopsided prices and on multiway markets, the choice starts to matter, and multiplicative is the least defensible of the three. If a tool does not say which method it uses, assume multiplicative.

A three-way example

Take a race-to-10-points market priced -140 / +130 / +2000. Raw implied probabilities are 58.33%, 43.48% and 4.76%, totalling 106.57% — a margin of 6.57 points, noticeably fatter than the two-way market above.

Multiplicative devigging gives 54.74%, 40.80% and 4.47%. The power method gives 55.76%, 40.55% and 3.69%.

Look at the third outcome. The two methods differ by nearly a full percentage point on a 4% longshot, which is a relative difference of about 17%. On the favourite they differ by one point on 55%. The method you choose matters most exactly where the payouts are biggest.

Never devig a slice of a bigger market

This is the most common serious mistake. If you take two players out of a touchdown-scorer market and normalize just those two to 100%, the answer is meaningless — those two outcomes were never a complete market, and the other thirty players hold most of the probability.

Devigging requires the full set of mutually exclusive outcomes. For player props that is naturally satisfied, because over and under are the whole market. For anything with more than two outcomes, you need every one of them.

Favourite–longshot bias

Longshots win less often than their prices imply, consistently, across sports and decades. Multiplicative devigging preserves that distortion, because it scales every outcome by the same factor.

That is why a multiplicative fair price on a big underdog tends to be too generous, and why treating it as truth will make genuinely bad longshot bets look positive.

Compare it against your own number

A fair price is only useful next to an independent estimate. If the devigged market says 50% and you also think 50%, you have learned that you agree with the market — which is worth knowing, and is not a bet.

The value is in the disagreements, and only if your own number came from somewhere other than the line.

Compare it across books

Devig the same market at several books and you get several fair estimates. Where they cluster is a more robust reference than any single book, and a book sitting well outside the cluster is either wrong or knows something.

A worked expected-value check

Suppose your projection says 55% and the fair market probability is 50%. At an offered price of -110 the break-even rate is 52.38%, so the expected value is about +$5.00 per $100 staked. At +100 it is +$10.00. At +110, +$15.50.

The same 5-point disagreement is worth three times as much at +110 as at -110. That gap is what line shopping is actually buying you.

The arithmetic is easy and reliable. The assumptions around it are where the damage happens.

Treating one book as the truth

A devigged price is that book's opinion with the margin removed. It is not the true probability. If you devig a soft book and bet against a sharp one on the strength of it, you have carefully computed the wrong number.

Stale prices

Devigging a line that has not moved since an injury report tells you what the book thought an hour ago. The bigger the price gap you find, the more likely the explanation is that the line is stale rather than wrong.

The margin is not the only cost

Removing the vig from the price does not remove it from your results. You still pay it on every bet you place. Devigging tells you what a fair price would be; it does not get you one.

The no-vig calculator does the arithmetic on this page for you, and the EV calculator compares a fair probability against an offered price.

For an independent number to compare the fair price against, the NFL prop projections hub and the player index are where WinForge's own estimates live.

If you want to know whether those estimates deserve any weight, read how WinForge works and the public accuracy ledger before trusting them.

• Favourite-longshot bias — background on why longshots are systematically overpriced
• Vigorish — definition and background
• National Council on Problem Gambling