KEY TAKEAWAYS
CONTENTS
Sportsbook hold is the actual percentage of total money wagered — the “handle” — that a sportsbook keeps as revenue after paying out winners, measured across a market, a game, or a reporting period. Vig (also called juice or margin) is a different number: the theoretical edge built into the posted odds on a single market before a single bet is placed, calculated purely from the prices themselves. The two terms get used interchangeably in casual conversation, and in a perfectly balanced market they can land close to each other. In practice they usually don’t match, because real money is never split perfectly evenly across the available sides of a bet. This article separates the two concepts precisely: how vig is computed from odds alone, how hold is computed from actual handle and payouts, why the two numbers diverge once real bettors get involved, and what the hold percentages regulators publish in official revenue reports actually represent.
What Is Sportsbook Hold?
Hold is a revenue metric, not a pricing formula. It answers a business question — how much of the money that came through the window did a sportsbook actually keep — rather than a math question about a single set of odds. The formula is simple: hold = (total handle − total amount paid to winning bettors) ÷ total handle, expressed as a percentage. Handle is every dollar wagered on a market, game, or period, regardless of which side it landed on or who eventually won.
Because hold is computed from what actually happened — which side people bet, how much they staked, and which side the event landed on — it can only be calculated after the fact, once a market or period has closed and every winning ticket has been paid. That is a fundamentally different starting point from converting a posted price into implied probability, which can be done the instant odds go up, before a single dollar has been risked. Hold describes what a sportsbook actually earned; on its own it says nothing about how fair or generous the original odds were.
How Hold and Vig Are Actually Calculated
Calculating Vig (Theoretical, Pre-Wager)
Vig comes entirely from the posted odds, with no reference to actual betting activity. Convert each side of a market to implied probability, add the two figures together, and whatever the total exceeds 100% is the vig. A standard two-way market priced at -110 on both sides implies roughly 52.38% per side (110 ÷ (110 + 100)); the two sides together sum to about 104.76%, so the vig on that market is 4.76%, fixed the moment the price is posted — the full derivation, including how to back out no-vig fair odds, is covered separately in how vig is calculated market by market. This article doesn’t repeat that derivation; it starts from the 4.76% figure and asks what happens to it once real money is involved.
Calculating Hold (Realized, Post-Wager)
Hold requires two inputs vig doesn’t need: how the actual handle split between the two sides, and which side the event actually landed on. A sportsbook collects stakes from everyone who bets, pays winning bettors their stake back plus profit, and keeps whatever handle is left over as revenue for that market. If money happened to split exactly the way the odds imply — proportional to each side’s implied probability — realized hold would land close to the theoretical vig. Real handle almost never splits that precisely, because bettors follow public teams, star players, recent narratives, and news, not a sportsbook’s probability model. That gap between how the odds assumed money would split and how money actually split is the reason hold and vig diverge.
| Question | Vig (Theoretical) | Hold (Realized) |
|---|---|---|
| What it measures | Margin priced into the odds | Actual revenue kept vs. total handle |
| When it’s known | Before any bet is placed | Only after a market or period closes |
| Depends on the outcome? | No | Yes |
| Depends on how money splits? | No | Yes |
| Typical use | Calculating fair, no-vig odds | Financial and regulatory reporting |
The table isolates the one variable that actually explains the gap between these two figures: vig is fixed by the price alone, while hold moves with real outcomes and real money — the same distinction the worked example below makes with numbers.
A Worked Example: Theoretical Vig vs. Realized Hold
Suppose a hypothetical sportsbook offers a two-way market at -110 on both Team A and Team B. As shown above, the theoretical vig on this market is 4.76%, regardless of what anyone ends up betting. Now suppose $100,000 in total handle comes in, split unevenly: $65,000 on Team A and $35,000 on Team B — a realistic pattern when one side is the more popular public team.
At -110, a winning stake returns itself plus profit of stake × (100 ÷ 110); a $100 stake would return $190.91 in total. If Team A wins, the book pays Team A’s bettors $65,000 × (210 ÷ 110) ≈ $124,090.91 — more than the $100,000 handle it collected, producing a realized loss of about $24,090.91, or a hold of roughly -24.09% on that market. If Team B wins instead, the book pays Team B’s bettors $35,000 × (210 ÷ 110) ≈ $66,818.18, keeping $33,181.82 of the $100,000 handle — a realized hold of roughly 33.18%. Same market, same 4.76% vig either way; two very different realized hold outcomes, because the actual result and the money split — not the odds — drove the number. All figures here are hypothetical and illustrate the mechanism, not any real sportsbook’s actual numbers.
How to Interpret a Sportsbook’s Hold Percentage
A single-market hold number, like the example above, is close to meaningless on its own — it’s driven heavily by which side happened to win, not by how the book is run. Hold becomes a meaningful signal only in aggregate, across many markets, many games, and enough handle for the wins and losses on each side of the ledger to smooth each other out. That’s why the hold percentages sportsbooks and regulators actually publish — monthly, by sport, or statewide — are always aggregated figures, never a single game’s result.
When aggregate hold sits well above a market type’s typical vig, it usually means the money leaned toward the side that lost, or the book priced a market type with room for extra margin (parlays, props, and futures structurally carry higher vig than a straight point spread or total). When aggregate hold sits below typical vig — or turns negative for a stretch — it usually means bettors as a group leaned toward outcomes that actually happened, the same imbalance working in the other direction. Neither outcome means the sportsbook’s pricing was wrong; it means results and handle distribution, not the posted odds, drove that period’s number.
Common Mistakes When Talking About Hold and Vig
The most common mistake is treating hold and vig as perfectly interchangeable numbers instead of related-but-different concepts. Using a single market’s realized hold to judge whether the odds were fair is a related error — as the worked example shows, a standard -110 market can realize a hold anywhere from deeply negative to well above 30% depending purely on which side won, without the posted price ever changing.
Another mistake is assuming a reported hold percentage describes an individual bettor’s odds of winning; it doesn’t — hold describes a sportsbook’s revenue outcome across every bettor on a market, not any single wager’s expected value. It’s also a mistake to assume a lower aggregate hold automatically means a market was “easier to beat” — a low hold in a given period can reflect an unusually strong run of results for bettors just as easily as it can reflect generous pricing, and the reverse holds for a high-hold period. Neither figure by itself measures skill, luck, or value on a single bet.
Where Hold Shows Up in the Real Sportsbook Business
Hold is the number that actually appears on a sportsbook’s financial results and in state gaming reports — not vig, which never shows up as a line item anywhere because it’s a pricing assumption, not a cash flow. Licensed sportsbooks are required to report handle and revenue to state regulators, and hold — revenue divided by handle — is the resulting figure regulators and the industry track over time, by state, by sport, and by month.
For a bettor, this distinction has a practical use: since hold is a business-level revenue outcome and vig is what’s priced into any one specific bet, comparing the same bet’s price across multiple sportsbooks targets the number a bettor can actually influence — the vig baked into the exact price taken — rather than a sportsbook’s aggregate hold, which no individual wager meaningfully moves.
Related Concepts to Learn Next
Hold and vig sit on top of two ideas worth understanding first: how a single posted price converts into implied probability, and how a market’s two-sided prices combine into a calculable margin. From there, the natural next step is bettor-facing rather than sportsbook-facing: taking the lower-vig price whenever one is available is the closest a single bettor comes to influencing the margin a sportsbook actually realizes over time.
