What vig actually is
A sportsbook is not a casino — it does not own the dice or the deck. It is a broker that prices a future event and takes a commission for hosting the trade. That commission is vig. Unlike a stock brokerage that charges a visible $4.95 ticket fee, the sportsbook embeds its commission directly into the odds. There is no line item. You pay vig every time you place a bet, whether you win or lose, by accepting odds that are tilted away from the book's true probability estimate.
The mechanism is simple. A fair coin flip should pay even money (+100 / +100). A sportsbook offering a 50-50 proposition will instead post -110 / -110. The 10-cent gap on each side is the book's revenue. Multiplied across millions of bets, it is the entire industry's profit engine.
The -110/-110 math
Convert the standard NFL spread price to its real meaning:
# American odds → decimal → implied probability -110 → decimal 1.909 → implied 52.38% -110 → decimal 1.909 → implied 52.38% # Overround 52.38% + 52.38% = 104.76% # Vig as % of total handle (104.76 − 100) / 104.76 = 4.55%
That 4.55% is the bookmaker's average revenue per dollar of handle, assuming balanced action. Multiplied across an NFL season's tens of billions of handle, it explains why DraftKings, FanDuel, BetMGM, and Caesars build glass towers in Las Vegas, London, and Macau.
Vig by sport — the 2026 landscape
| Market | Overround | Vig % | Notes |
|---|---|---|---|
| NFL spreads / totals (Pinnacle) | 102.5% | 2.44% | Lowest in major US book ecosystem |
| NFL spreads / totals (DraftKings) | 104.5% | 4.31% | US retail standard |
| NBA moneyline (Bet365) | 103.0% | 2.91% | Two-way moneyline market |
| MLB moneyline (FanDuel) | 106.5% | 6.10% | Run line and totals often lower |
| EPL match result (1X2) | 104-108% | 3.8-7.4% | Three-way market, vig compounds across legs |
| NHL player props | 108-112% | 7.4-10.7% | Books exploit weaker player-prop models |
| UFC live in-play | 110-115% | 9.1-13.0% | Wide because of latency and model uncertainty |
| NFL same-game parlay (3 legs) | ~120-135% | 17-26% | Stacked vig + correlation manipulation |
The pattern is consistent across operators: the more illiquid the market, the wider the vig. Books charge a premium where their probability estimate is least confident, partly to compensate for model risk, partly because casual demand is highest and price-sensitive sharp money is lowest.
Three ways to remove vig
To recover the book's probability estimate, the bettor must strip vig. There are three accepted methods and they often disagree, sometimes by 1-2 percentage points on the same market.
1. Multiplicative (proportional) method
Divide each side's implied probability by the overround. Simple, fast, and the industry default for two-way markets.
# Lakers +130 / Celtics -150 implied_LAL = 1/2.30 = 0.4348 implied_BOS = 1/1.667 = 0.6000 overround = 1.0348 novig_LAL = 0.4348 / 1.0348 = 42.02% novig_BOS = 0.6000 / 1.0348 = 57.98%
2. Additive (Shin) method
Subtract an equal slice of the overround from each side. Accounts better for favorite-longshot bias on lopsided lines. Underestimates vig on heavy favorites and overestimates on heavy dogs relative to the multiplicative.
3. Power (logit) method
Apply a power exponent to each implied probability so the result sums to 1. Best fit for soccer 1X2 and other three-way markets where favorite-longshot bias is strong. Computationally heavier but the academic gold standard (Štrumbelj 2014, "On determining probability forecasts from betting odds").
Practical implication: multiplicative for two-way, power for three-way, Shin for futures with extreme longshots. Differences typically matter only at the 1-2 cent level — meaningful to sharps, irrelevant to recreational bettors.
The 52.38% break-even rule
Every -110 bet requires a 52.38% win rate to break even on a flat-stake basis. The intuition: out of 100 bets at $110 risk to win $100 each, you stake $11,000. To break even you must win enough $100 payouts to cover the $11,000 stake plus the losing bets. Solve: 0.5238 × 100 = 52.38 wins, returning 52.38 × $210 = $10,999.80 ≈ $11,000.
This is the most important number in sports betting. Recreational bettors hit ≈49% historically, paying 3-4 percentage points to vig over a long sample. Sharps clear 52.38% by 1-3 points. The thin margin is why bankroll management and Kelly sizing matter — variance can swamp a 1% edge for hundreds of bets.
Where vig hides — parlays and same-game parlays
The casual bettor's worst leak. A two-leg parlay at -110/-110 advertised "true" odds of +264 actually pays +260 — the 4-point gap is compounded vig on each leg. A four-leg parlay stacks vig multiplicatively: theoretical hold ≈ 17-20%. A same-game parlay (SGP) compounds further because the book also adjusts for correlation, often unfavorably. Books love SGPs because the embedded vig is invisible — the bettor sees "+1200 payout!" without realizing the no-vig fair price would be +1500.
Industry data (NJ DGE 2024 sports report): parlay handle ≈ 22% of total, but parlay revenue ≈ 41% of net win. The casual bettor subsidizes the entire sportsbook P&L by playing parlays.
Pinnacle — the low-vig benchmark
Pinnacle Sports (Curaçao-licensed, not legal in the US but available offshore) is the industry's reference book. Standard NFL/NBA/UFC markets routinely post at 102.5-103% overround vs. 104.5-106% at US operators. Pinnacle's model: high limits, low margins, welcomes sharp action. Most professional bettors hold Pinnacle accounts because the line is closest to fair, and CLV against Pinnacle is the gold standard for measuring sharpness.
The flip side: Pinnacle accepts heavy action and rarely limits accounts, but it doesn't offer the boosts, free bets, and promotions that bloat US books' effective vig downward for new customers.
Why vig is sticky — the regulator angle
Regulators do not cap vig in any major jurisdiction. The argument: competition will discipline it. In practice, US state-licensed monopoly/oligopoly markets (six operators in most states) have seen vig widen since legalization (May 2018 PASPA repeal). NJ DGE, PA Gaming, and NY Gaming reports all show hold percentages climbing from 6% to 7%+ across 2022-2024.
The structural reason: state tax rates (NY at 51% of GGR, PA at 36%) force operators to recover margin via wider lines. NY DraftKings posts NFL spreads at 105% overround while NJ DraftKings posts 104%. The customer pays the tax through vig.
Line shopping — the only edge most bettors can earn
If you can't model better than the book, you can at least pay less for the same outcome. A bettor with five US books (DraftKings, FanDuel, BetMGM, Caesars, Bet365) who takes the best available price on every bet has an effective vig of 1.5-2.5%, not 4.5%. The improvement compounds across volume.
Modern line-shopping tools (OddsJam, DonBest, our own OddsCipher feed) surface the best price across 20+ books in real time. A bettor placing 500 NFL bets per season at a 2.5-point line-shop improvement saves roughly $400-700 in vig versus a single-book bettor of equal volume. That is not edge — that is just paying retail vs. paying wholesale.
The professional bettor's checklist
- Calculate vig before every bet — if > 6%, walk away unless your model edge is documented at > 8%.
- Line shop ruthlessly — paying -108 vs -110 is a 0.9% EV improvement, the same magnitude as a small modeling edge.
- Avoid parlays unless you have correlated-leg insight the book has under-priced (rare, exists in same-game parlays with sharp props).
- Track CLV — beating the closing line by 1-2 cents is the only durable evidence of sharpness.
- Book mix — Pinnacle/Bet365/Circa for low-vig majors; FanDuel/DraftKings/BetMGM for promotions and boosts; exchanges (Betfair, Prophet) for niche markets.
Sources & further reading
- Levitt, Steven D. "Why are gambling markets organised so differently from financial markets?" Economic Journal, 2004 — foundational analysis of bookmaker margins vs. exchanges.
- Štrumbelj, Erik. "On determining probability forecasts from betting odds." International Journal of Forecasting, 2014 — power method for vig removal in three-way markets.
- Pinnacle Betting Resources — "Margin calculator and the importance of vig" (low-vig book's own public documentation).
- New Jersey Division of Gaming Enforcement — Sports Wagering Monthly Reports 2022-2025 (hold percentage and parlay handle data).
- Paul, Rodney J. & Weinbach, Andrew P. "Bettor preferences and market efficiency in football totals markets." Journal of Economics & Finance, 2011.