The Vig Tax: A Quantitative Survey of Sportsbook Margins

Every sportsbook bet carries a built-in cost. It goes by several names — the vigorish, the juice, the overround — but the mechanics are identical: the book sets its odds so that the implied probabilities of all outcomes sum to more than 100%, and the excess is the house's margin. If you bet both sides of a -110/-110 line, you put up $220 to win $210. The $10 difference is the vig. You paid it whether you won or not.

The concept is well understood. What is less well understood, because nobody publishes the data systematically, is the magnitude of the variation from book to book. We set out to measure it.

We assembled 1,016,310 closing moneyline odds from 40 sportsbooks across four sports (NBA, NHL, MLB, NCAAB) and two full seasons. For each two-way market, we summed the implied probabilities and subtracted one. The remainder is the overround — a direct measure of what the bettor pays for the privilege of having an opinion.

The cheapest and most expensive NBA moneyline differ by a factor of seven. Matchbook charges 0.93 cents on the dollar; Fliff charges 6.82. On any individual wager this is a rounding error. Over a season of regular betting it is the difference between a profitable year and a losing one, all else being equal. A bettor placing 500 bets at $50 each pays roughly $233 in annual vig at Matchbook and $1,705 at Fliff. Same bets, same outcomes, $1,472 difference in friction cost.

Three things are worth noting.

First, the exchanges (Matchbook, Betfair) and Pinnacle cluster at the low end, as you would expect. Exchanges charge commission on net winnings rather than embedding margin in the price, so their overrounds are structurally lower. Pinnacle operates with thin margins because its business model depends on volume from sharp bettors who would not tolerate a 5% tax. This is not new information, but the magnitude of the gap may be: Pinnacle charges roughly 25% less vig than FanDuel on the same NBA games.

Second, and this is the observation that surprised us most in the data, the vig is not constant across sports for the same book.

Pinnacle charges twice the vig on college basketball moneylines (4.08%) as it does on MLB (2.04%). This is rational: MLB moneylines are among the deepest and most efficiently arbitraged markets in sports betting. NCAAB has 5,700 games per season across hundreds of programs, with thinner markets, less sophisticated price discovery, and more asymmetric information. Higher uncertainty justifies higher margin. But the practical consequence is that a bettor who uses Pinnacle for everything is getting a very different deal depending on which sport they are betting. And if they are using a higher-vig retail book, the sport-specific dispersion is even wider.

Third, the rank ordering of books by vig is not identical across sports. A book offering competitive NBA margins may charge substantially more on MLB, or vice versa. This means that if you bet multiple sports through a single book, you are almost certainly overpaying on at least one of them. The solution is the oldest and most reliable edge in all of gambling: look at more than one number before placing the bet.

We should be careful not to overstate the implication. Low vig is not synonymous with "best sportsbook." It does not account for sign-up bonuses, ongoing promotions, user experience, withdrawal speed, or the critical question of whether the book will limit or ban you for winning. There are sharp bettors for whom a 4% vig book that does not restrict them is worth more than a 2% vig book that will close their account after three winning weeks. But the vig is the one cost that is perfectly measurable, perfectly persistent, and perfectly within the bettor's control to minimize. It seems worth knowing what it actually is.