Why Do Some Bets Cost More Than Others?
Every price includes the book's margin, and the margin varies by market: thinnest on main lines, wider on props and futures, and compounding on parlays.
September 27, 2026 at 5:21 PM EDT
6 min read
Every sportsbook price has a fee built into it, and the size of the fee depends on what you are betting. On a main-market point spread or total at the usual -110 each way, the book's theoretical margin is about 4.5%. On a two-way player prop it is often wider. On a parlay the margin multiplies with every leg, and on a futures market with dozens of teams it can be larger again.
That fee — called the vig, juice, margin or hold — is why two bets with the same chance of winning can have very different long-run costs. Nobody has to hide it. It is sitting in the odds, and anyone who can convert odds to probabilities can measure it.
How do you measure the margin in a price?
Convert each side to implied probability, add them up, and see how far past 100% the total goes.
For American odds: a negative price −X implies X ÷ (X + 100); a positive price +X implies 100 ÷ (X + 100).
The standard spread bet, -110 / -110:
- Each side: 110 ÷ 210 = 52.38%
- Total: 104.76%
- Hold = 1 − 1 ÷ 1.0476 = 4.55%
That hold is what the book expects to keep from each dollar wagered if it takes equal money on both sides at fair odds. It is also why a -110 bettor needs to win 52.4% to break even — Steven Levitt's figure in his 2004 study of how bookmakers price, where the standard spread bet risks 110 to win 100. Tobias Moskowitz, in a 2021 Journal of Finance study of 117,442 betting contracts, found the vig large enough that every betting strategy he tested lost money after costs, including ones that looked strong before them.
The Academy's guide to hold and vig goes further into the formulas.
Why are player props more expensive?
A book can price the NFL spread with a great deal of confidence: huge volume, a market-maker to copy, and years of data. A receiver's yardage prop has less of all three. The book has less information, less liquid reference prices, and more exposure to someone who knows the player's usage better than its model does. It protects itself with a wider margin and lower limits.
Two examples of prop prices you will see:
-120 / -120:
- Each side: 120 ÷ 220 = 54.55%
- Total: 109.09%
- Hold = 1 − 1 ÷ 1.0909 = 8.33%
-135 / +105:
- -135: 135 ÷ 235 = 57.45%
- +105: 100 ÷ 205 = 48.78%
- Total: 106.23%
- Hold = 1 − 1 ÷ 1.0623 = 5.86%
The second one looks like the better deal at a glance because one side is plus money, but what matters is the total. Both are more expensive than a -110 main line, the first nearly twice as much.
Why are parlays so expensive?
Because you pay the margin on every leg, and the payments compound.
Take legs that are each genuinely 50/50 and priced at -110 (decimal 1.909). A parlay multiplies the decimal prices:
| Legs | Pays (American) | Fair price | Book's expected margin |
|---|---|---|---|
| 1 | -110 | +100 | 4.55% |
| 2 | +264 | +300 | 8.88% |
| 3 | +596 | +700 | 13.03% |
| 4 | +1228 | +1500 | 16.98% |
| 5 | +2436 | +3100 | 20.75% |
The margin column is 1 − (1.909 × 0.5)ⁿ. For two legs: the parlay pays 3.645 times the stake and hits 25% of the time, so the expected return is 0.25 × 3.645 = 0.911 — an 8.88% cost. By five legs you are giving up about a fifth of every dollar, before any extra margin the book adds for same-game correlations. The Academy's parlays explainer covers when a parlay can make sense despite that.
Why are futures the most expensive of all?
A futures market has many outcomes, and the book puts a little margin on each one. Across 30 teams that adds up. It also ties up your money for months, and the book gets to reprice as the season goes on while your ticket is fixed.
A simplified six-team market, with illustrative prices:
| Team | Price | Implied |
|---|---|---|
| A | +250 | 28.57% |
| B | +300 | 25.00% |
| C | +400 | 20.00% |
| D | +500 | 16.67% |
| E | +800 | 11.11% |
| F | +1000 | 9.09% |
The implied probabilities add up to 110.44%, so the hold is 1 − 1 ÷ 1.1044 = 9.45%, about double a main-line spread. Real futures boards have far more entrants, and every longshot's small overpricing adds to the total.
What the state data show
States publish what books actually keep — the realised "win" or "hold", which is revenue divided by handle. That is not the same as the theoretical margin: it also reflects luck, which markets customers choose, and the fact that recreational bettors often take worse prices than the best available. But over a year it tracks the pricing closely enough to be revealing.
Nevada, from the Gaming Control Board's June 2026 revenue report, statewide figures for the twelve months from July 2025 to June 2026:
| Category | Win % |
|---|---|
| Football | 7.28% |
| Basketball | 7.31% |
| Baseball | 7.27% |
| Hockey | 8.65% |
| Parlay cards | 30.86% |
| All sports pool | 7.75% |
Nevada reports "parlay cards" as their own small category ($7.2 million of win over the year); other parlays, including those placed on mobile apps, are not broken out in the report. The category is still the cleanest regulatory evidence of what a multi-leg product costs: nearly a third of what was bet was kept.
Colorado, from the Division of Gaming's June 2026 report, which lists parlays separately for all channels. Using the state's figures (gross gaming revenue is wagers minus payments to players):
- Parlays and combinations: $166,737,128.62 wagered, $147,156,919.12 paid out → 11.74% hold. Parlays were 39.1% of all money wagered.
- Everything else combined: $259,366,634.61 wagered, $247,745,718.31 paid out → 4.48% hold.
- All bets: 7.32%.
One month is a small sample — Colorado's basketball books lost money in June 2026, for example — so these are a snapshot, not a rate you should expect. But the direction matches the arithmetic: the parlay line kept more than two and a half times as much per dollar as the rest of the board.
What this means when you bet
Three practical conclusions, none of them a reason to bet more.
The market you pick matters as much as the side. A 4.5% margin and a 13% margin on the same kind of opinion are very different bets.
Shopping prices cuts the margin you pay. If one book offers Team A at -105 and another offers Team B at -105, the two best prices imply 51.22% + 51.22% = 102.44%, a combined hold of 1 − 1 ÷ 1.0244 = 2.4% — about half of -110 each way. Books know this, and the accounts that consistently take the best number are the ones they limit.
The house margin is why most bettors lose. It is not hidden and not mysterious. Beating it requires being right more often than the price implies, by more than the margin, over hundreds of bets. Most people don't, and the state revenue reports are the record of that.
The rest of this series covers how books set odds and why lines move.