Sportsbooks make money by pricing both sides of a market above 100 percent. Here is where the hold lives, why a parlay multiplies it, and what it costs.

Sportsbooks make money by pricing both sides of a market so they add up to more than 100 percent. That surplus is the vig, and it is charged whether you win or lose. Two sides at −110 add up to 104.76 percent of probability, which leaves the book keeping about 4.55 cents of every dollar wagered if the money arrives evenly.
The interesting part is what happens to that number when you change the bet. It gets bigger on a longshot, and much bigger on a parlay. Drag the leg slider below and watch it move.
Enter both sides of any market and see the margin the book priced in, what a fair price would have been, and how much of your stake the house expects to keep. Then move the leg slider from a straight bet up to eight legs.
Illustrative only, not a prediction of any result. Both prices add up using the same measure the SmartStake Arbitrage Finder and Promo Converter rank their boards on, the fair prices come from the same multiplicative devig the SmartStake Devigging Calculator runs, the assumed single-bet margin is the odds-scaled curve the product falls back to when only one side of a market is priced, and the odds convert with the same functions the SmartStake odds tools use. Hold is a long-run average over many bets at these prices, not what happens on any one slip: an individual bet can still win or lose. It assumes both sides are priced as entered, every leg is independent, and no promotion, boost, or push applies, any of which changes the result. Only bet with disposable income.
Everything below explains what those readouts mean and where each mechanism comes from. Every figure on this page is computed from the prices stated next to it, and every hold number is a long-run average over many bets at those prices rather than a prediction about any single one. An individual bet can win or lose whatever the margin is.
Every price is a probability in disguise. A decimal price of 1.91, which is −110 in American odds, implies a 52.38 percent chance because 1 divided by 1.91 is 0.5238.
Price both sides of a coin flip at −110 and those two implied probabilities add up to 104.76 percent. No event has a 104.76 percent chance of happening, so the extra 4.76 percent is not a forecast. It is the charge.
That surplus goes by several names. Vig, juice, margin and overround are all the same thing, and it is the single largest source of sportsbook revenue. Strip it out and you get the no vig fair odds underneath, which for two sides at −110 is +100 on each.
Here is where most explainers get sloppy. The 4.76 percent and the 4.55 percent are both real, and they measure against different denominators.
Overround is the excess over 100 percent. It is the number you get by adding the two implied probabilities, and it is measured against what the book pays out.
Hold is the share of every dollar wagered that the book expects to keep. It is the overround divided by the total, so 4.76 divided by 104.76 gives 4.55 percent.
Hold is the number that matters when you ask how a sportsbook makes money, because handle is what it collects. A book taking $1,000,000 of balanced action on a −110 market expects to keep about $45,450 of it. Regulators publish hold for exactly this reason.
A parlay pays the product of its legs. Four legs at 1.91 pay 1.91 to the fourth power, which is 13.28 in decimal or +1228.
The fair price is the product of the fair legs. Four fair coin flips at 2.00 pay 16.00, or +1500.
So the bettor gets back 13.28 divided by 16.00 of the fair value, which is 83.02 percent. The book's hold on that slip is 16.98 percent. It did not shade a price, invent a correlation rule, or take any extra risk. It multiplied the same four prices it was already offering.
| Legs at −110 | Book pays | Fair price | Book's hold |
|---|---|---|---|
| 1 | −110 | +100 | 4.55% |
| 2 | +264 | +300 | 8.88% |
| 3 | +596 | +700 | 13.03% |
| 4 | +1228 | +1500 | 16.98% |
| 6 | +4741 | +6300 | 24.36% |
| 8 | +17545 | +25500 | 31.08% |
The formula is one line. If each leg holds h, a slip of n legs holds:
That is why the parlay is the most promoted product on every sportsbook app. It is the same inventory at four to seven times the margin. Our parlay calculator guide walks through the payout side of the same arithmetic.
Here is the part almost every page on this topic misses. You might assume a parlay is expensive because it is a longshot, and longshots carry more margin. Both halves of that are true, and they are still not the same mechanism.
SmartStake's own closing line pipeline has to price a leg when only one side of a market is quoted. It cannot normalize the overround without the other side, so it assumes one, and the margin it assumes rises with the odds. The model is 1 + 0.04 × (2d / (1 + d)) on decimal odds d, an odds-scaled approximation of favorite-longshot bias. It floors at a 4 percent overround on an extreme favorite and ceilings at 8 percent on an extreme longshot.
Run a 4-leg parlay price through it. At +1228, that model assumes a hold of 6.92 percent on a single bet. The actual 4-leg parlay at the same price holds 16.98 percent, roughly two and a half times as much. At eight legs the gap is wider still: 7.37 percent assumed on a single bet at +17545, against 31.08 percent on the slip.
A long price built from many legs is not the same product as a long price on one outcome, even when the number on the screen is identical. The price tells you how many times the margin was applied, not how unlikely the bet is.
One honest caveat: the 6.92 percent is a modelling assumption, not a measured market average. Treat it as the shape of the effect rather than a precise quote. The parlay side of the comparison is exact arithmetic on the stated prices, and the direction and rough size of the gap hold up either way.
The same curve explains something you can see on any book's screen. Margins are thinner on main lines between two well known teams and thicker on longshots, alternate lines and deep player props.
| Single price | Assumed overround | Assumed hold |
|---|---|---|
| −500 | 4.36% | 4.18% |
| −200 | 4.80% | 4.58% |
| −110 | 5.25% | 4.99% |
| +200 | 6.00% | 5.66% |
| +400 | 6.67% | 6.25% |
| +1000 | 7.33% | 6.83% |
Two forces drive this. Bettors systematically overpay for longshots, so a book can charge more there and still take the action. And a longshot is harder to price accurately, so the book widens the margin to cover its own uncertainty.
The practical read: the further you get from a heavily traded main line, the more you are paying for the privilege. That is also where line shopping turns up the widest price differences, because books disagree most where they are least sure.
Knowing a market holds 4.76 percent does not tell you how that charge is split between the two sides. That question has no single answer, and the devigging method you use is a claim about it.
Take a lopsided market at −400 and +300. Under the multiplicative method, both sides pay the same 4.76 percent and fair value is −320 against +320. Under the additive method, the favorite pays only 3.13 percent while the longshot pays 10.00 percent, putting fair value at −344 against +344.
Same market, same posted prices, and the two methods disagree by more than threefold about who is being charged. Run it yourself:
The devigging calculator does the same job on any market you paste in, and the full devig guide covers when the choice of method actually changes a decision.
The textbook says a sportsbook wants equal money on both sides. Balanced action means it pays the winners with the losers' stakes and keeps the margin regardless of the result.
In practice, money almost never arrives evenly. Public money concentrates on favorites, overs, popular teams and parlays, so a book holding a real position on most games is the normal state, not a failure. Trading desks manage that with price moves, stake limits and hedges rather than by chasing a perfect balance.
The margin is the dependable part of the business. The position is the variable part, and over a full season the margin is what dominates the result.
Because each book manages its own position, two books often land on noticeably different prices for the same market. That disagreement is an opening: pair the best price at one book with the best price at another and the combined hold can fall to almost nothing, which is what low hold betting and the low hold calculator are for.
A $200 free bet is a marketing cost with a known conversion rate, not a gift. A free bet returns only the winnings, not the stake, so its cash value is well under face value. In the worked example in our free bet converter guide, a $100 free bet hedged at the prices shown there leaves about $67 in cash. That figure is an illustration at those specific prices rather than a typical result, it moves with every price you can actually find, and no conversion is ever without risk: odds shift, stakes get limited, legs get voided.
Books price these knowingly. The offer buys a funded account, and the play-through terms that usually come with it push turnover through markets that hold 4.55 percent or more. Sign-up offers are almost always new customer only, and the terms are where the real cost sits.
That is not an argument against taking them. Sign-up promotions are among the few genuinely favorable prices a retail bettor gets offered, and the free bet calculator exists to recover as much of the face value as the hedge allows. It is an argument for knowing what the book is buying.
The last piece of the model is who a book keeps. A sportsbook with a 4.55 percent hold does not need you to lose. It needs volume it can price.
Accounts that consistently beat the closing line are the ones books restrict, because closing line value is the strongest available signal that a bettor's prices are better than the book's. Stake limits, delayed acceptance and account closures are all the same risk tool applied to that signal. Our guide on managing sportsbook limits covers the bettor's side of it.
Everything above points at the same conclusion. You are not trying to beat a coin flip, you are trying to beat a coin flip plus a fee, and the fee is negotiable in four ways.
No approach removes the margin. Beating it requires finding prices better than fair value often enough to cover it, and any individual bet can still win or lose.
How do sportsbooks make money?
Sportsbooks make money from a margin priced into both sides of every market. Two sides at −110 imply 104.76 percent of probability between them, and the extra 4.76 percent is the book's cut. Spread over every dollar wagered it works out to 4.55 percent, so a book with balanced action keeps about 4.55 cents of every dollar no matter which side wins.
What is the difference between vig and hold?
Vig, juice and overround all name the same thing, the amount the two prices add up to over 100 percent. Hold is that same margin measured against every dollar wagered. Two sides at −110 carry a 4.76 percent overround and a 4.55 percent hold. The overround is the bigger number because it is measured against a smaller base.
Why do parlays have a higher hold than single bets?
A parlay multiplies the leg prices together, so it multiplies the margin too. Four legs at −110 each carry a 4.55 percent hold, and combined they pay +1228 where a fair price would be +1500. The hold on that slip is 16.98 percent, close to four times the hold on one leg, and no price had to be shaded to get there.
Do sportsbooks always balance both sides of a bet?
No. Balanced action is the textbook model, not daily practice. Books run open positions on most games because the money rarely arrives evenly, and they manage the risk by moving prices, capping stakes, and leaning on the fact that recreational money concentrates on favorites, overs and parlays.
Does the sportsbook want me to lose?
A book with balanced action is indifferent to the result, because it collects its margin either way. What it does care about is which bettors it keeps. Books restrict accounts that consistently beat closing prices, and they spend heavily on promotions to recruit new customers.
A sportsbook's revenue is a fee charged on both sides of every market. At −110 that fee is 4.76 percent of the payout and 4.55 percent of the handle, and it is collected whichever side wins.
The fee is not constant. It grows with the odds, and it grows much faster with the number of legs, because a parlay reapplies the same margin once per leg. A four leg slip at +1228 costs about two and a half times what a single bet at that price would.
Reading a price for what the book is charging is the whole skill. Start with the devigging calculator on any market you are about to bet, and check what the fair number underneath actually is.
Practical betting guides, new tools, and product updates from the SmartStake team. Unsubscribe in one click, any time.
By providing your email, you are consenting to receive communications from SmartStake. Visit our Privacy Policy for more info
No Vig Fair Odds: What Fair Value Really Means in Betting
No vig fair odds show what a price would be if the book took no margin. That number is the market's honest estimate of an outcome, and it is the single benchmark every serious betting strategy measures against.
Jul 26, 2026 · 9 min read

Every sportsbook price hides a margin called the vig. A no vig calculator removes it to show the fair odds and true probability behind a line, the number you need before you can measure value.
Jul 7, 2026 · 8 min read

Player props are bets on a single player's statistics rather than the game result. They carry roughly double the vig of a side, and they are the one market where the method you use to strip that vig out can flip a bet from positive to negative.
Aug 1, 2026 · 11 min read