No-Vig Fair Odds Calculator
Strip the bookmaker's margin out of a market to see the fair price behind every outcome.
This estimates the fair price behind the prices you entered, so it measures one bookmaker's opinion with the margin taken out, not the true chance of the outcome.
Bookmaker margin
3.93%
Offered pricesadd up to 103.93%
Past the line: the margin, 3.93%
Proportional de-vig keeps the ratio of implied probabilities and just scales them to sum to 100%. A solid default for balanced markets.
The market
Every odds field reads a decimal price (2.50). Switch the format above to enter prices the way your book quotes them, or open the odds converter to see one price in all of them.
The four methods on this market
Spread across the four methods
3.1%
Out 2's fair price runs 3.014 to 3.108 depending on which method you pick.
Out 2 fair price, by method
| Method | Out 1 | Out 2 |
|---|---|---|
| Proportional | 1.49766.8% | 3.01433.2% |
| Power | 1.47467.8% | 3.10832.2% |
| Shin | 1.48267.5% | 3.07532.5% |
| Additive | 1.48267.5% | 3.07532.5% |
No column here is the right answer. All four are estimates of an unobservable fair price; the point of the table is that they can disagree, not which one to trust.
Written and reviewed by Alexander Korsager · last reviewed September 26, 2026 · corrections
The idea
What this tool shows
Every bookmaker's market is priced so the implied probabilities add up to more than 100%. That excess is the margin (the overround, the vig, the juice) and it is the house edge on that market. A two-way market at 1.90 / 1.90 implies 52.63% + 52.63% = 105.26%, a 5.26% margin.
De-vigging removes that margin to estimate the bookmaker's true probabilities, and the four methods here differ only in where they take it from.Proportional divides every implied probability by the book total, which keeps their ratios intact and so removes the margin in proportion to how likely each outcome already was. Additive subtracts an equal absolute share from each outcome, which is proportionally more off a longshot than off a favorite. The power method solves for an exponent that makes the probabilities sum to 100%. Shin's method infers a share of informed money in the market and backs the fair prices out of that. The last two shorten favorites and lengthen longshots relative to proportional, which is the direction the favorite-longshot bias in real markets points.
One property is worth knowing before you pick between them: in a two-outcome market, Shin's method and the additive method return exactly the same numbers.That is not a rounding artefact, it falls out of the algebra. Shin's condition rearranges to fair probability squared, plus the insider share times fair probability times one minus itself, equals the raw implied probability squared over the book total. With only two outcomes that middle term is identical on both sides, so the difference between the two fair probabilities has to equal the difference between the two raw ones, which is exactly what subtracting an equal share from each produces. They are usually presented as separate methods with different theoretical foundations, which they are, but on a two-way market the answers coincide, and they only part company from three outcomes up.
How much the method matters at all depends on how lopsided the market is.At 1.90 / 1.90 all four return 2.00 and the choice is irrelevant. On a moneyline of -400 / +310 (decimal 1.25 / 4.10, a 4.39% margin) the underdog's fair probability comes out at 23.36% under proportional, 22.20% under Shin and additive, and 21.55% under the power method. Taking that underdog at +330 measures as +0.5% against the proportional fair price, -4.6% against Shin, and -7.3% against the power method. The gap between the widest and narrowest reading is larger than the margin being removed, so on a lopsided market the method is a bigger decision than the de-vig itself.
The fair odds this produces are only as good as the market you feed it.A de-vigged price from a sharp, high-limit bookmaker is a strong fair-value estimate; one from a soft book with a big margin and slow prices is not. The main use is comparison: de-vig the sharp price, then see whether anywhere else is offering clearly more.
FAQ
Questions players ask
Which de-vig method should I use?
On a balanced two-way market it makes no difference, because all four return the same fair price. The choice only starts to bite as the market gets lopsided, and the evidence there favors Shin: Strumbelj (2014) tested it against plain proportional normalization across every bookmaker and sport pair in the study and found Shin probabilities more accurate in all of them, with the advantage shrinking as the market got larger. The one thing that reliably produces a bad answer is picking the method after seeing which side it favors. Settle on one before you look, and treat a bet that only reads positive under one of the four as one where the method, not the market, is doing the work.
Does a bigger margin mean the de-vig method matters more?
No. What decides how much the method matters is how lopsided the market is, not how big the margin is. A 1.90 / 1.90 market carries a 5.26% margin and all four methods return the exact same fair price, 0% spread, because the market is perfectly balanced. A 1.28 / 4.00 market carries a smaller margin, 3.13%, and spreads 5.3% across the four methods, because it is lopsided. Margin down, disagreement up. The two are correlated in most real markets, since the most lopsided prices are often also the highest-margin ones, which is likely why the two get run together, but they are not the same axis, and a heavily vigged balanced market is one where the choice genuinely is free.
Why do Shin and the additive method give me the same answer?
Because on a two-outcome market they are the same calculation. Shin's condition reduces to a form where, with only two outcomes, the difference between the two fair probabilities must equal the difference between the two raw implied probabilities, and that is precisely what subtracting an equal share of the margin from each one does. So the switch between them changes nothing on a moneyline or any other two-way market. They give genuinely different answers from three outcomes up, where the equality no longer holds.
Does this work for three-way markets like 1X2?
Yes, enter all three prices. Proportional and power handle any number of outcomes cleanly. Shin still returns an answer, but its model was derived for a two-way market, so the insider-share reading of the number it solves for is no longer the one the model was built for. Additive is the one to watch here: on a lopsided market of three or more outcomes it can drive a longshot's fair probability to zero or below, and when that happens this tool falls back to the proportional result rather than printing an impossible number.
What does it mean if the margin comes out negative?
It means the prices you entered add up to less than 100%, which no single bookmaker's market does. Almost always the market was assembled from more than one book, or from prices that moved between copying the first one and the last. Watch what the tool does next, because it is easy to misread. With a book under 100% there is no margin to remove, so the de-vig runs in reverse and every fair price comes back shorter than the price entered beside it. Proportional, additive and power all do this, on every outcome, without exception. Paste 2.05 and 2.00 from two different books (a 98.78% book) and the fair prices return as 2.025 and 1.976, so both sides read as value against the prices you can actually get. Both sides of a market cannot be underpriced at once, so what that reading measures is how you assembled the market, not the market. Shin's method returns nothing at all here: it solves for a share of informed money inside the overround, and with no overround there is nothing to solve, so the tool substitutes the proportional result and labels it.
How do I use fair odds to find value?
De-vig the sharpest price you can find, then compare it with the price you can actually get. Longer odds than the fair price means positive expected value of roughly that gap, and that gap is usually small: on the -400 / +310 example above it is smaller than the spread between the four methods on this page, which is why the method matters. A positive estimate is also not a return you collect. It is computed from the same market you are measuring against, it still loses most individual bets on a longshot, and the variance around it is what decides whether a bankroll lasts long enough for any edge to show up at all.
Methodology
How this is calculated
Each offered price becomes an implied probability (1 divided by the decimal odds). Those sum to the book, and the book minus 1 is the margin. Each method then rescales them to sum to exactly 1. Proportional divides each by the book. Additive subtracts the margin divided by the number of outcomes from each.
The power method solves by bisection for the exponent k that makes the sum of each implied probability raised to k equal 1.
Shin's method solves by bisection for the informed-money share z under which each fair probability equals the square root of (z squared plus 4 times (1 minus z) times the implied probability squared divided by the book), minus z, all over 2 times (1 minus z), with the results summing to 1.
Fair decimal odds are 1 divided by the fair probability. Two markets have no answer under the method asked for, and the tool substitutes the proportional result in both, labeled on screen rather than applied silently.
If the additive method produces a probability at or below zero, which can only happen from three outcomes up, the proportional result stands in.
Shin's method has no solution unless the book is above 100%, because the bisection it runs needs the book total to exceed 1 before a root exists to find, so a market priced at or under 100% substitutes as well.
Common misreadings5
- A fair price is an estimate of one bookmaker's opinion with the margin removed, not the probability itself. Two sharp books de-vigging to different fair prices is the normal state of a market, not evidence that one of them is mispriced.
- The gap between a fair price and a price you can get is an estimate of edge, not money you can expect to collect. On a lopsided market it is routinely smaller than the spread between the four methods here, which means it can be entirely an artefact of the method selected.
- Choosing the method after seeing which side each one favors turns this page into a way of confirming a bet rather than pricing one. A selection that only reads positive under one of the four is not an edge that has been found.
- Taking the best price on each side from different bookmakers and de-vigging the result. The book total that produces is nobody's margin, and once it falls under 100% the fair prices come back shorter than the prices offered, so every outcome reads as value. The order that works is the other way round: de-vig one book's own market, then compare that fair price against what other books are offering.
- A positive estimated edge does not make a bet safe. It still loses on most individual longshots, and the spread of outcomes around it, not the edge, is what decides whether a bankroll survives long enough for the edge to matter.
Assumptions5
- The prices entered cover every outcome of one market and were all available at the same moment. A market with an outcome missing, or assembled from prices captured minutes apart, produces a book total that does not mean what this reads it as.
- Every odds field reads whichever format the odds format control is set to, and only that one. The control exists, rather than a field that takes anything, because 150 is a real decimal price as well as the American price +150: a field reading both would have to guess which you meant on exactly the input where guessing wrong is invisible, since a wrong price of that kind still prints a complete and correctly formatted panel. A price that does not read in the selected format shows nothing rather than a figure, and the page names the format it would read in when there is one. Switching the format rewrites the prices already entered, so a field never holds a number in a format the page is no longer reading, and a fractional price is rewritten to the closest fraction with a denominator of 100 or less.
- The margin is the only distortion being removed. Deliberate shading of one side, a stale price, or a limit small enough that the price is not a real offer all stay in the answer.
- Proportional assumes the margin sits on each outcome in proportion to its implied probability, and additive assumes an equal absolute amount on each. Power and Shin assume it falls more heavily on longshots, which is the direction the favorite-longshot bias points but not a fixed quantity.
- Shin's model assumes the margin exists to protect the bookmaker against a fraction of informed money, and that this is what sets the spread. Where a margin is set by something else, the z it solves for is still a number but no longer an insider share.
Limits6
- On a balanced two-way market all four methods return the same fair price, so the choice only matters as the market gets lopsided. On a heavy favorite it can matter more than the margin being removed.
- In a two-outcome market Shin's method and the additive method are arithmetically the same calculation, so switching between them changes nothing. They diverge only from three outcomes up.
- Shin's model is derived for two-way markets. It returns a value for three or more outcomes, but the interpretation the model supplies does not carry over.
- The additive method can drive a fair probability to zero or below on a lopsided market of three or more outcomes. When it does, this tool substitutes the proportional result, so a reading labeled additive is not always an additive reading. It cannot happen with two outcomes, because an implied probability taken from decimal odds above 1 is always below 1.
- A book under 100% reverses the de-vig. There is no margin to take out, so proportional, additive and power all return fair prices shorter than the prices entered, on every outcome, and the whole market reads as value at once. No single bookmaker prices a market that way, so what that reading measures is how the market was assembled, not the market.
- A fair price is only as sharp as the market it came from. Removing the margin from a soft book's slow price produces a precise number that is not a fair-value estimate.
Evidence
The margin arithmetic is the definition of the book total. The two solvers implement published models: Shin's is from H. S. Shin, Measuring the Incidence of Insider Trading in a Market for State-Contingent Claims, Economic Journal 103 (1993), pages 1141 to 1153; the comparison that made it the sharp default is E.
Strumbelj, On determining probability forecasts from betting odds, International Journal of Forecasting 30 (2014), pages 934 to 943, which found Shin probabilities more accurate than plain normalization for every bookmaker and sport pair it tested, with the advantage narrowing as market size grew. math.test.ts pins the two-outcome equivalence of Shin and additive across a grid of markets, checks that additive cannot go negative with two outcomes and that it can with three, pins Shin substituting on exactly the markets that are not over-round, pins the reversal a book under 100% produces in the other three, and pins both the -400 / +310 example and the 2.05 / 2.00 two-book example quoted on this page.
The margin-versus-asymmetry finding in the FAQ above (a bigger margin does not predict a bigger spread across the four methods, a more lopsided market does) is sourced to this page's own engine rather than a third party: math.test.ts pins the five markets the FAQ cites, plus that the highest-margin one of the five produces zero method disagreement while a lower-margin, more lopsided one produces the largest.
A passing test means the code matches these models, not that any of them is the right description of the market you are pricing.
Gambling can cause serious harm. The National Problem Gambling Helpline is free, confidential and open 24/7, call or text 1-800-MY-RESET (1-800-697-3738), or chat at ncpgambling.org/chat.
Read up
Go deeper
- Our researchWhat a prediction market actually chargesPrediction markets are sold on not charging vig. We snapshotted every quotable contract on Kalshi and priced the two-way cost the way a sportsbook's hold is priced. On 78,119 markets the median came to 9.9c per dollar of payout, against 4.8c at a standard sportsbook line. On the markets people actually trade, the answer flips.
- Our researchThe longshot gap, and the spread that swallows itThe academic literature reports that prediction-market longshots are mispriced. We checked it on 12,755 settled Kalshi contracts, priced a day before they closed, and then did the part nobody had done: measured what reaching the mispricing would have cost at the same price, at the same moment, out of the same rows. In 10 price buckets out of 10, the cost was larger than the gap.
- GuideBetting Odds Explained: Decimal, Fractional, American and Implied ProbabilityHow decimal, fractional and American odds work, how to read the implied probability behind any price, and how to spot a bookmaker's margin eating your value.
Related tools
Try next
- Odds ConverterConvert odds to any format
- Expected Value CalculatorCheck if a price has value
- Arbitrage CalculatorCheck whether two or more prices on the same event lock in a profit, and get the stake for each side.
- Parlay CalculatorPrice a multi-leg bet
- Middle CalculatorPrice a middle on both sides of a moved line: what it pays, what a miss costs, and how often it has to land.
- CLV CalculatorWork out your closing line value, and whether the sample you have is large enough to mean anything yet.
- Pari-Mutuel Payout CalculatorEstimate a win, place or show payout from the pool sizes, and see what the takeout and your own stake cost you.
- Each-Way Bet CalculatorWork out the win, place and lose returns on an each-way bet, and when the place part is the value.