No-vig odds calculator
Quoted odds imply more than 100% of probability. This removes the margin under three published methods at once, shows the hold and the overround separately, and prints how far the methods disagree.
Worked example — Standard -110 on both sides is a hold of: 4.55%. a two-way market priced -110 on both sides
Remove the margin from a real market
Enter the price on every outcome, from two up to six. The tool returns the hold, the overround, and the fair probability each of the three published methods produces, side by side, because the choice of method moves the answer more than most people expect.
Solver diagnostics 1
| Outcome | Quoted | Implied | Proportional | Equal margin | Power | Fair odds, power |
|---|---|---|---|---|---|---|
| Side 1 | -160 | 61.54% | 58.60% | 59.03% | 59.25% | -145 / 1.69 |
| Side 2 | +130 | 43.48% | 41.40% | 40.97% | 40.75% | +145 / 2.45 |
Proportional
Divides every implied probability by their sum. Removes margin in proportion to probability, so it takes the most from favourites and the least from longshots. It is the method most free calculators implement, usually without saying so.
Equal margin
Subtracts the same number of percentage points from every outcome. Simple, and defensible on a two-way market, but it breaks down as the field grows: on a high-margin market it can push a longshot below zero.
Power
Raises every implied probability to a common exponent k chosen so the results sum to one. It shrinks longshots more than favourites, which matches the observed favourite-longshot bias in posted prices better than the other two.
The three methods disagree by 0.66 percentage points on this market. This tool computes from the odds you type; it does not read prices from any venue.
Hold, overround and break-even at every standard price
A two-way market with the same price on both sides. This is the reference that decides whether a book is competitive before you look at a single line, and the break-even column is the win rate a bettor needs at that price just to stay level.
| Price both sides | Decimal | Implied each side | Overround | Hold | Break-even win rate | Kept from $200 taken |
|---|---|---|---|---|---|---|
| -105 | 1.9524 | 51.22% | 2.44% | 2.38% | 51.22% | $4.76 |
| -110 | 1.9091 | 52.38% | 4.76% | 4.55% | 52.38% | $9.09 |
| -115 | 1.8696 | 53.49% | 6.98% | 6.52% | 53.49% | $13.04 |
| -120 | 1.8333 | 54.55% | 9.09% | 8.33% | 54.55% | $16.67 |
| -130 | 1.7692 | 56.52% | 13.04% | 11.54% | 56.52% | $23.08 |
| -150 | 1.6667 | 60.00% | 20.00% | 16.67% | 60.00% | $33.33 |
| -200 | 1.5000 | 66.67% | 33.33% | 25.00% | 66.67% | $50.00 |
At -105 the two figures sit 0.06 points apart. At -200 they sit 8.33 points apart. Comparing one book's overround against another book's hold is therefore not a comparison at all, and it is a mistake that appears in published margin tables regularly.
- Overround, how far the two prices exceed 100%
- Hold, the share of total stakes kept
A margin you can shrug off on a single bet stops being ignorable the moment it is multiplied. Each leg of a parlay keeps 95.45% of its fair value, so five legs keep that figure raised to the fifth power and give up 20.75% of the stake. The parlay payout chart, with the hold at every leg count.
Every figure on this page is computed from the odds you type in; the page does not read prices from any venue and holds no market data. How to de-vig a market, the long version.
Which method to use, and what it costs to pick wrong
One quoted market, three published ways to strip the margin out of it, three different answers. Removing margin requires a story about how the book distributed it, and nothing in the posted prices tells you which story is true.
| Method | What it assumes | Who it favors | When it breaks |
|---|---|---|---|
| Proportional | Margin was loaded onto every outcome in proportion to its size. | Neither side relatively: every outcome loses the same percentage of its own probability. In percentage points it takes the most from the favourite. | Never fails to produce a valid probability, but is the least realistic on markets with a strong favourite-longshot bias. |
| Equal margin | A flat number of percentage points came off every outcome. | The favourite, relatively: the same flat point-cut is a much bigger share of a small probability. | A market with a real longshot: on this page's own three-way example it can drive an outcome to zero or below, which is why it is refused rather than printed when that happens. |
| Power | A single exponent bent the whole book, shaped like the favourite-longshot bias posted prices actually show. | The favourite: probabilities shrink proportionally more the smaller they are, by construction of the exponent. | No hard failure mode, but it has no claim to being the true mechanism, only a curve fit. |
Take a three-way market implying 60%, 30% and 15% of probability, a 5.00% overround. Here is what each method says the same market is worth.
| Method | Favourite | Middle | Longshot | Longshot as a price |
|---|---|---|---|---|
| Proportional | 57.14% | 28.57% | 14.29% | +600 |
| Equal margin | 58.33% | 28.33% | 13.33% | +650 |
| Power | 58.36% | 28.10% | 13.53% | +639 |
Same input, three defensible methods, 0.96 points of disagreement on the longshot and +600 against +650 as a price on it. Equal margin is harshest on the longshot while power is softest on the favourite, so which method flatters your number depends on which side of the market you are backing.
Power, on any market longer than two outcomes, and proportional on a straight two-way where the three barely differ, is what JMM Research uses. Posted prices consistently overprice longshots relative to how often they win, and power is the only one of the three that removes margin in a shape resembling that bias. It has no claim to being the mechanism. Equal-margin is worth keeping off any market with a real longshot in it, because the failure is not gradual: it hands you a negative probability, and a calculator that prints one anyway is worse than no calculator. To carry a fair probability into an expected-value and stake calculation, use the odds calculator.
Hold vs overround, what's the difference
At -110 on both sides, each price implies 52.38% and the two sum to 104.76%. The 4.76% above 100 is the overround. The hold is smaller: on $100 taken each way the book pays the winner $190.91 out of $200 and keeps $9.09, which is 4.55%. Both get called the vig, and articles regularly print one while naming the other. The gap widens as the market gets longer, so a margin comparison across books is meaningless unless both sides use the same definition.
Vig questions with numeric answers
What is the vig on -110?
4.55% as a hold, 4.76% as an overround. The first is the share of total stakes the book keeps when money is split to guarantee the margin, the second is how far the two implied probabilities exceed 100%. Both are correct answers to different questions.
How do I calculate the hold from two prices?
Convert both to decimal, add 1 divided by each, and call that sum S. The overround is S minus 1 and the hold is (S minus 1) divided by S. At -110 both sides, S is 1.0476, so the overround is 4.76% and the hold is 4.55%.
What is a good hold to bet into?
Under 3% on a two-way market is competitive, which is roughly -105 both sides at 2.38%. The price you are actually being quoted is -110, a 4.55% hold, and that is the number worth sitting with: it means 52.38% is break-even rather than the 50% most people picture, and a bettor hitting 55% earns 5.00% on turnover instead of the 10.00% the same record would earn at a fair price. The hold takes half the return of a good handicapper before anything goes wrong. Past -115, a 6.52% hold, the price consumes an edge that almost nobody has.
Can I de-vig a single posted price?
No, and any tool that claims to is guessing. Removing margin requires every outcome of the market, because the whole calculation is about how the excess above 100% is distributed. One side tells you the sum of a probability and an unknown share of margin, which is one equation with two unknowns.
Why do the three methods disagree?
Because each assumes a different rule for how the book spread its margin. Proportional takes more from favourites, equal-margin takes the same number of points from everyone, and power takes proportionally more from longshots. On a tight two-way market the three land within a few hundredths of a point. On a long market with a big longshot they can differ by several points, which is the difference between a bet and a pass.
Does a lower hold mean a better price for me?
On the side you want, usually, but not always. A book can post a low overall hold while shading one side heavily, so check the fair price on your side rather than only the market's total margin. That is why the table in the tool prints a fair probability per outcome instead of one headline number.
The next question after this one
Odds calculatorCalculatorTake a fair probability from here into an expected-value and Kelly calculation.Open nextOther questions
- Arbitrage calculatorCalculatorWhen the de-vigged book falls below 1.00 across two accounts, this splits the outlay in stakes a book will accept.
- Closing line value calculatorCalculatorThe same margin removal applied to the closing price, which is the only benchmark that scores a bet you already placed.
All 130 calculatorsHow JMM sources and checks its numbersWhat stays free, and what early access would add
More calculators for the same question
The odds calculator carries a single price into an expected-value and Kelly stake, and the parlay calculator shows what the hold does once you multiply several legs together.