How Sportsbooks Make Money: The Math Behind Every Dollar They Keep
How sportsbooks make money on 6 revenue streams, from 4.5% vig on spreads to 30%+ on SGPs. Worked examples show book revenue on $1M in total handle.
The Vig Is the Business Model
Every sportsbook on the planet makes money the same way. They build a margin into the odds. That margin is called the vig (short for vigorish, also called juice). It is not a hidden fee. It is not a mystery. It is a mathematical structure that guarantees the book earns revenue on every dollar wagered, regardless of which side wins.
How sportsbooks make money comes down to one principle: the implied probabilities of all outcomes sum to more than 100%. The excess is the house's cut. On a standard -110/-110 NFL spread, both sides imply 52.38% probability. That adds up to 104.76%. The extra 4.76% is the overround, and it translates to a hold percentage of roughly 4.5%. For every $100 wagered across both sides, the sportsbook keeps $4.54 in expected revenue.
The hold calculator computes this instantly for any market. Plug in odds for both sides and it shows you exactly how much the book is charging.
This article breaks down six revenue streams sportsbooks use to generate profit, with worked examples showing what a book earns on $1 million in handle.
Hold Percentage: The Core Revenue Engine
The hold percentage is the fraction of total handle (money wagered) the sportsbook expects to keep. It is the single most important number in the sportsbook business model.
The formula:
Hold = 1 - (1 / book sum)
Where book sum = the sum of implied probabilities for all outcomes.
Worked example: $1M in handle on NFL spreads
An NFL spread is priced at -110 / -110.
Book sum = 52.38% + 52.38% = 104.76%
Hold = 1 - (1 / 1.0476) = 4.54%
On $1,000,000 in handle: Expected revenue = $45,400
That is the theoretical revenue assuming perfectly balanced action. In practice, books do not always achieve balanced action, but $45,400 per million is the baseline expectation from the vig structure alone. Strip the vig from any line with the de-vig calculator to see the true odds underneath. For the full breakdown of how vig works at every level, read what is vig.
Hold varies dramatically by market type:
| Market Type | Typical Hold | Revenue per $1M Handle |
|---|---|---|
| NFL/NBA spreads | 4.5% | $45,000 |
| Moneylines (lopsided) | 5-7% | $50,000-$70,000 |
| Player props | 8-15% | $80,000-$150,000 |
| 3-leg parlay | ~15% | ~$150,000 |
| Same-game parlay | 20-40% | $200,000-$400,000 |
| 6-leg parlay | ~32% | ~$320,000 |
The revenue difference is massive. A dollar wagered on a same-game parlay generates 5 to 9 times more revenue than a dollar wagered on an NFL spread.
Balanced Books vs. Liability Management
The textbook explanation says sportsbooks want equal action on both sides. If $500,000 comes in on Team A and $500,000 on Team B, the book collects $1M and pays out roughly $909,000 to the winners (at -110). The remaining $91,000 is risk-free revenue.
This model is real, but it is incomplete. Modern sportsbooks use two distinct approaches.
Market-making books (Pinnacle, Circa) do aim for balanced action. They post tight lines, accept sharp bettors, and profit from volume. Their margins are thin, typically 2-3% on major markets, but they make it up by processing enormous handle. These books rarely move their lines based on which side is getting more money. They move lines based on what sharp money tells them about true probability.
Retail books (DraftKings, FanDuel, BetMGM) operate more like asset managers taking positions. They shade their lines to maximize revenue, limit sharp bettors, and often carry one-sided liability intentionally. If 70% of the public is on the Chiefs, a retail book might shade the line to extract more vig from that popular side rather than moving the line to attract money to the other side.
The distinction matters for bettors. If you are looking for the sharpest lines to build your probability estimates, use market-making books. If you are looking for the softest lines to bet against, look at retail books where public money has moved the line away from true probability. Use the EV calculator to compare odds across books and find where the price is wrong.
How Sportsbooks Shade Lines Toward Public Money
Line shading is the deliberate practice of adjusting odds to exploit public betting patterns. Sportsbooks know which teams, players, and outcomes the public overvalues, and they price accordingly.
The mechanics are straightforward. The public consistently overbets favorites, overbets overs, and overbets popular teams. Sportsbooks respond by making favorites slightly more expensive and underdogs slightly cheaper. This is not balancing the book. It is maximizing the vig collected from the heavier side.
Worked example: public-side shading
True probability (from a sharp line): Home favorite 62%, Away underdog 38%.
A balanced line would set the favorite at approximately -163 and the underdog at +163 (with standard vig applied symmetrically).
A shaded line might set the favorite at -175 and the underdog at +150. The overround is now loaded more heavily onto the favorite. If 65% of handle comes in on the favorite at -175, the book collects more vig per dollar from the popular side.
On $1M in handle split 65/35:
- Revenue from favorite side (worse odds): $650,000 x effective vig of ~3.3% = $21,450
- Revenue from underdog side (better odds): $350,000 x effective vig of ~1.8% = $6,300
- Total expected revenue: $27,750 (2.78% blended hold)
Compare that to a symmetric -163/+163 line with balanced 50/50 action:
- Revenue: $1,000,000 x 2.17% hold = $21,700
The shaded line earns more despite the same total vig because it loads the margin onto the side receiving more money. This is why closing line value matters. If you are consistently beating the closing line, you are likely on the opposite side of public money, where the vig load is lighter.
Player Props: The Highest-Margin Product
Player props are the highest-margin product most sportsbooks offer. The hold percentage on player props routinely runs 8-15%, compared to 4-5% on mainline spreads. Some exotic props push above 20%.
Three factors drive the higher margin.
Less efficient pricing. Mainline spreads are set by models informed by sharp money, injuries, weather, and decades of historical data. Player props are softer. Fewer professional bettors model individual player outcomes, which means less sharp money correcting the lines.
Higher volume per game. A single NFL game might have 10-15 player prop markets per team (passing yards, rushing yards, receptions, touchdowns, etc.). Each one carries higher vig than the spread. From the sportsbook's perspective, props multiply the revenue opportunity per event without requiring a new game.
Public bias. Casual bettors love props, especially on star players. They bet Mahomes overs, Curry threes, and Ohtani strikeouts based on narratives rather than numbers. This demand imbalance gives sportsbooks room to shade the lines further, because the money keeps flowing regardless of the margin.
Worked example: props vs. spreads on a single NFL game
One NFL game generates approximately:
| Market | Handle | Hold | Revenue |
|---|---|---|---|
| Spread (both sides) | $500,000 | 4.5% | $22,500 |
| Moneyline | $200,000 | 6.0% | $12,000 |
| Over/Under (game total) | $200,000 | 5.0% | $10,000 |
| Player props (all combined) | $300,000 | 12.0% | $36,000 |
| Total | $1,200,000 | $80,500 |
Player props generate 45% of the revenue on 25% of the handle. This is why every sportsbook app puts props front and center. Read what does over/under mean for how totals markets work, and sports betting math for the complete formula set.
Parlay Revenue: How Compound Vig Prints Money
Parlays are the sportsbook's most profitable product per dollar wagered. The reason is compound vig. Each leg of a parlay carries its own margin. When you multiply vigged odds together, the vig compounds geometrically.
A single -110 bet has a 4.5% hold. A 3-leg parlay at -110 per leg has a 14.9% hold. A 6-leg parlay reaches 31.6%. By the time a bettor builds a 10-leg parlay, the book is keeping more than half the expected value.
| Legs | Hold | Revenue per $1M Handle |
|---|---|---|
| 1 (straight bet) | 4.5% | $45,000 |
| 2 | 9.7% | $97,000 |
| 3 | 14.9% | $149,000 |
| 4 | 20.3% | $203,000 |
| 5 | 25.9% | $259,000 |
| 6 | 31.6% | $316,000 |
| 10 | ~47% | ~$470,000 |
Same-game parlays are even more profitable. Because the legs are correlated, the sportsbook uses proprietary models to set the combined odds. The bettor cannot independently verify the pricing. SGP hold routinely runs 25-40%, and complex constructions can exceed 50%.
This is why DraftKings, FanDuel, and BetMGM all aggressively promote parlays and SGPs. The marketing spend is justified because the margin is so wide that even after parlay boosts and insurance promotions, the book still comes out ahead. The parlay math breakdown walks through the full compounding mechanics with five worked examples.
How Prediction Markets Differ: Explicit Fees vs. Hidden Vig
Sportsbooks hide their margin in the odds. Prediction markets charge explicit fees. The structural difference changes the math significantly.
On Kalshi, you buy and sell contracts at market prices set by an order book. Kalshi charges a taker fee of 7% x p x (1-p) on each trade, where p is the contract price. On a $0.50 contract (the equivalent of a 50/50 bet), that fee is $0.0175 per contract, or 1.75% of notional. On a lopsided market at $0.85, the fee drops to $0.00945 per contract.
On Polymarket, there is no explicit trading fee. Polymarket charges 2% on net profits from winning trades. If you buy at $0.50 and the contract resolves to $1.00, your gross profit is $0.50 and Polymarket takes $0.01 (2% of $0.50).
Compare these to the sportsbook model:
| Scenario | Sportsbook Cost | Kalshi Cost | Polymarket Cost |
|---|---|---|---|
| 50/50 market (-110 both sides) | 4.5% hold | 1.75% fee | ~1.0% fee |
| Lopsided market (-300/+250) | 5-7% hold | 0.9% fee | ~0.6% fee |
| 3-leg equivalent | 15% hold | ~3.5% cumulative | ~2.0% cumulative |
The cost difference is 40-80% depending on the market structure. Over hundreds of bets, this gap determines whether a small edge is profitable. Sportsbooks compensate by offering more sports, more markets, faster settlement, and promotional incentives. Prediction markets compensate by being cheaper to trade.
For the complete cost comparison, read prediction market no-vig advantage. The prediction market fee calculator shows exact costs for any contract price on any platform.
What Sportsbooks Earn on $1M in Total Handle
Putting it all together, here is what a typical sportsbook earns on $1,000,000 in total handle across all product types. These numbers reflect the blended revenue across a realistic product mix.
Revenue breakdown by product
| Product | % of Handle | Handle | Avg Hold | Revenue |
|---|---|---|---|---|
| Straight bets (spreads/totals) | 35% | $350,000 | 4.5% | $15,750 |
| Moneylines | 15% | $150,000 | 6.0% | $9,000 |
| Player props | 20% | $200,000 | 12.0% | $24,000 |
| Parlays (standard) | 15% | $150,000 | 18.0% | $27,000 |
| Same-game parlays | 10% | $100,000 | 30.0% | $30,000 |
| Other (futures, live, teasers) | 5% | $50,000 | 8.0% | $4,000 |
| Total | 100% | $1,000,000 | 10.98% | $109,750 |
The blended hold is approximately 11%. That means for every $1 million in handle, the sportsbook expects to keep roughly $110,000.
Notice where the money comes from. Player props (20% of handle) and SGPs (10% of handle) together generate $54,000, which is 49% of total revenue from 30% of the handle. Straight bets on spreads and totals, the markets most bettors think of as "sports betting," generate only 14% of revenue despite being 35% of handle.
This revenue structure explains everything about how modern sportsbooks operate. It explains why they promote parlays and props. It explains why they limit sharp bettors who focus on low-margin spreads. And it explains why the industry is profitable despite spending billions on customer acquisition. The margin on high-hold products is wide enough to fund the entire operation.
Run the hold calculator on your most common bet types. If you are betting primarily player props and parlays, you are paying the maximum possible price. If you are focused on mainline spreads at the sharpest available odds, you are minimizing what the house takes.
Frequently asked questions
- What percentage does a sportsbook keep?
- It depends on the product. NFL spreads carry about 4.5% hold. Player props run 8-15%. Parlays reach 15-32% depending on the number of legs. The blended hold across all products at a typical sportsbook is roughly 10-11%.
- How do sportsbooks make money if they pay out winners?
- The vig built into the odds ensures the sportsbook collects more in total wagers than it pays in total winnings. On a -110/-110 market, the book collects $110 from each side's losers but only pays $100 to each side's winners. The $10 gap per $110 wagered is the profit margin.
- Why do sportsbooks push parlays and same-game parlays?
- Because the hold percentage on parlays is 3 to 7 times higher than straight bets. A 4-leg parlay at -110 carries 20.3% hold compared to 4.5% on a single straight bet. Same-game parlays are even more profitable at 25-40% hold because the sportsbook controls the correlation pricing.
- Do prediction markets charge vig like sportsbooks?
- Prediction markets charge explicit fees instead of hidden vig. Kalshi charges 7% x p x (1-p) per trade. Polymarket charges 2% on net profits. These explicit fees are typically 40-80% cheaper than sportsbook vig on equivalent markets.
- How much does a sportsbook make on $1 million in bets?
- Approximately $110,000 at a blended hold of 11%. This varies by product mix. A book with more parlay and prop handle earns more per dollar wagered. A sharp-focused market maker with thin spreads earns closer to $30,000-$40,000 per million.