Convert both sides of a sportsbook market into implied probability and you will notice something right away: they add up to more than 100%.
That is not a math error. It is the price of the market.
The extra is the sportsbook's margin — the vig, juice or hold. It is built into every posted price, and it is the reason raw odds are not fair odds. If you want to know what a market is really saying about two outcomes, you have to take the margin out first.
The sportsbook doesn't need to be wrong about the team. It only needs to charge you a bad price.
This article shows how to see that price, remove it, and use what's left.
What the vig actually is
A sportsbook makes money the way most marketplaces do: it charges a little more than fair on both sides.
The cleanest way to see it is a perfectly balanced -110 / -110 market. Suppose one bettor risks $110 on Side A and another risks $110 on Side B.
- The book takes in $220.
- Whichever side wins collects its $110 back plus $100 profit: the book pays out $210.
- The book keeps $10, no matter who wins.
$10 out of $220 is about 4.55%. That is the book's theoretical margin on a balanced book.
Real books are rarely balanced, so actual results vary with the mix of bets, line moves, limits and promotions. But the margin is visible in the prices before a single bet is placed — and you can measure it.
Why the probabilities add up past 100%
Convert -110 into implied probability:
110 / (110 + 100) = 52.38%
Both sides are 52.38%:
52.38% + 52.38% = 104.76%
Two complementary outcomes can't truly total 104.76%. The extra 4.76 percentage points is the market's overround — the figure most bettors and most calculators, including ours, call the hold.
Posted prices vs no-vig market
OutpriceTheBook · Education
The margin sits past the 100% line
- Overround (hold)
- 4.76%
- Balanced-book margin
- 4.55%
- Fair Side A
- +100
- Fair Side B
- +100
Hold also tells you nothing about whether a single bet is good or bad. It tells you the cost you have to overcome.
Removing the vig: proportional normalization
The standard simple method for a two-way market has three steps:
- Convert both prices into raw implied probability.
- Add them together.
- Divide each side by the total.
No-vig probability = side's raw probability / total of both raw probabilities
The result always adds up to exactly 100%. That's the no-vig market.
Then, if you want prices instead of percentages, convert back:
Fair odds (favorite, above 50%) = -100 × p / (1 - p)
Fair odds (underdog, below 50%) = +100 × (1 - p) / p
Proportional removal assumes the margin is spread evenly across both sides. Other methods can shade heavy favorites and long shots a little differently, but proportional is the usual starting point and the one the No-Vig Calculator uses.
If converting odds into raw implied probability is new to you, start with How Implied Probability Works — everything here builds on it.
Where proportional removal can mislead
Books don't always charge both sides equally. On lopsided markets — a heavy favorite against a long shot — a bigger share of the margin is often built into the long-shot price, because that's the side recreational money tends to chase.
When that happens, proportional removal leaves the long shot looking a bit more likely than it really is, and the favorite a bit less. For close markets the difference is small. For a -600 / +450 market, treat the no-vig number as approximate, and be extra skeptical of any "edge" you find on the long shot.
Three full calculations
1. The standard market: -110 / -110
- Raw: 52.38% and 52.38%, total 104.76%, hold 4.76%
- No-vig: 52.38 ÷ 104.76 = 50.00% each
- Fair odds: +100 each
The book offers -110 on both sides. A fair market would be +100 on both. The gap is the price of playing.
2. An uneven spread price: -105 / -115
Books often shade one side of a spread. Say Side A is -105 and Side B is -115.
- Raw: -105 → 51.22% (105 ÷ 205); -115 → 53.49% (115 ÷ 215)
- Total: 104.71%, hold 4.71%
- No-vig: 51.22 ÷ 104.71 = 48.92%; 53.49 ÷ 104.71 = 51.08%
- Fair odds: about +104 and -104
The hold is almost the same as a -110 / -110 market. What changed is where the market is centered: it leans slightly toward Side B.
3. A favorite and an underdog: -150 / +130
- Raw: -150 → 60.00%; +130 → 43.48%
- Total: 103.48%, hold about 3.48%
- No-vig: 60.00 ÷ 103.48 = 57.98%; 43.48 ÷ 103.48 = 42.02%
- Fair odds: roughly -138 and +138
That is a much cleaner reference than comparing your opinion with -150 and +130 directly.
Bonus: a three-way market
Soccer's full-time result has three outcomes — home, draw, away — and the same method works with three prices. Say the market is Home +150, Draw +230, Away +190.
- Raw: +150 → 40.00%; +230 → 30.30%; +190 → 34.48%
- Total: 104.79%, hold 4.79%
- No-vig: 38.17%, 28.92% and 32.91%
- Fair odds: about +162, +246 and +204
Every outcome gets a little less likely and a little longer once the margin is gone. The No-Vig Calculator has a three-way mode for exactly this.
From sportsbook price to your decision
OutpriceTheBook · Education
The full path for the -150 / +130 market, with a 61% estimate on the favorite
1Sportsbook price
Favorite-150
Underdog+130
2Remove the hold
Raw total103.48%
Hold3.48%
3No-vig market
Favorite57.98% · -138
Underdog42.02% · +138
4Compare your number
Yours (Favorite)61.00%
vs no-vig+3.02 pts
vs -150 price+1.00 pts
Two numbers, two jobs
That last step is where most of the value is.
The no-vig probability tells you where the market is centered once its margin is removed. It is the market's cleaner opinion.
The break-even probability of the price you can actually bet tells you whether that specific price is good enough for your estimate.
A bettor who makes the favorite 61% is about a 3-point difference versus the no-vig market estimate of 57.98% — but only 1 point above the raw 60% break-even probability at -150. The first number says you disagree with the market. The second says how much of that disagreement survives the price.
The same idea gets sharper with bigger moneylines. Take a market of -220 / +180, where it is hard to eyeball the real opinion. The favorite's raw break-even is 68.75%, the underdog's is 35.71%, and the two total 104.46%.
For that -220/+180 example, the favorite's no-vig market probability is about 65.81%, while the -220 price requires 68.75% to break even. If your own estimate is 71%, you're about 5.19 percentage points above the no-vig market but only 2.25 points above the price you actually have to beat.
That difference is the whole point: no-vig probability tells you where the market is centered after removing margin; break-even probability tells you whether the specific price in front of you is good enough for your estimate.
No-vig is a benchmark, not your number
The word "fair" in "no-vig fair odds" causes a lot of confusion.
It means the sportsbook's margin has been mathematically removed. It does not mean the result is the true probability of the event. The market can still be wrong. Late news can still change the game.
There are really two different fair numbers:
| Market no-vig fair | Your modeled fair | |
|---|---|---|
| Where it comes from | The sportsbook's own prices, margin removed | Your handicap: strength, availability, environment, matchup |
| What it tells you | Where the market is centered | Where you think the price should be |
| Can it create an edge? | No — it is the market | Only if it's built from better information or better judgment |
If the no-vig market says an underdog is 42% and your work says 48%, you have a real disagreement worth investigating. If your work says 42.5%, you basically agree with the market — and agreeing with the market at a price that includes margin is a losing position.
Removing the vig is not a handicapping model. It doesn't know who's injured, what the wind is doing or who has the matchup edge. Building your own number is a separate job; How to Build Your Own Fair Betting Line walks through that process.
Why books post different prices on the same game
Two sportsbooks can hang different numbers on the same market at the same moment. They have different customers, different exposure, different risk appetites and different timing — one may move on news a few minutes before another, or shade a popular side to manage its liability.
For you, that's an opportunity. Here's one game across three books:
Same game, different books
OutpriceTheBook · Education
One game, three books
| Book | Favorite | Underdog | Hold | No-vig Favorite |
|---|---|---|---|---|
| Book A | -150 | +130 | 3.48% | 57.98% |
| Book B | -140 | +120 | 3.79% | 56.20% |
| Book C | -155 | +135 | 3.34% | 58.82% |
| Best available | -140 | +135 | 0.89% | shopping both sides |
Each book on its own charges a margin of 3.34% to 3.79%. Combine the best price on each side, and the two prices total only 100.89% — a hold of 0.89%.
You only bet one side, of course. But the point holds: shopping doesn't change your opinion of the game, it changes how much of the margin you pay.
Hold changes across markets
Not every market costs the same to bet.
High-volume main markets — spreads, totals and moneylines in the major sports — tend to carry lower margins. Alternate lines, player props, derivatives, niche sports and same-game parlays usually carry wider ones.
The math is the same everywhere: convert, total, normalize. Doing it on the markets you bet most often tells you where your edge has the most room to breathe. A small edge in a high-hold market can simply be too expensive.
Common mistakes when removing the vig
Calling raw implied probability "fair"
Raw probability includes the margin. It is the break-even rate of that price, not a no-vig estimate.
Removing vig from one side only
You need every outcome's price to normalize. A single price can't tell you the margin.
Using stale prices
A no-vig calculation from an old screenshot can be worse than a raw price from a live market. Use current numbers.
Treating the no-vig market as your model
If your only number is the market's number, you haven't created an independent opinion — you've just measured the market.
Assuming every market holds the same
A prop and a main spread from the same book can carry very different margins.
A practical process
For a two-way market:
- Record both current prices.
- Convert each to raw implied probability (Odds Converter).
- Add them to measure the hold.
- Divide each side by the total to get the no-vig market.
- Convert back to fair odds if that helps you read it.
- Compare your number with both the no-vig market and the break-even of the price you can actually get.
If you have an estimate, the EV Calculator turns that comparison into expected value at the real price — and Expected Value vs Probability Edge explains why that number differs from the raw probability gap.
The takeaway
The vig isn't hidden. It's right there in the prices, and once you convert both sides you can measure it.
Removing it gives you a cleaner view of where the market is centered. It doesn't give you an answer. That still comes from your own work — and from the discipline to compare it against the price you can actually bet, not the price you wish you had.
Is the probability I'm being offered cheaper than the probability I believe is actually there? That's the only question no-vig math is for.