What is the rake in sports betting? (also called the vig)

The rake - you may also see it called the vig, or the juice - is the sportsbook's built-in margin on a bet, and it is why both sides of a coin-flip market are priced a little worse than even money. To measure it, add up the chances the two prices imply: they come to more than 100%, and that excess is the rake. How much it costs depends entirely on the market. The average two-sided hold on an MLB game line is about 3.9%. Anytime home run props, which are almost always one-sided, graded out around -23% expected value across 18,468 props in our historical data set. So, yes, some props are worse than others.

Written by Jesse, NegativeEV. Last updated 12 August 2026.

The rake is the sportsbook's built-in margin. It is why a coin-flip market is priced worse than even money on both sides.
Measuring it takes both sides. Add up the chances the two prices imply: they come to more than 100%, and the excess is the rake.
Game lines carry the least. The average two-sided hold on an MLB game line is about 3.9%, and shopping the best price on each side strips a good part of it out.
One-sided markets hide it. Most player props post a price on only one side, so there is no second price to add up and the cost surfaces as expected value instead.
Expected value is the related number, not the same one. It sets the real chance of an outcome against the chance the price implies. Anytime home run props grade out around -23% expected value across 18,468 graded props, far outside any noise floor.
Graded against real results, the average moneyline came back around -2% per bet over 1,111 games. That is a different quantity from the hold, and it does not separate from zero on that sample.
The rake is baked into every expected value on the site, which is a large part of why so few prices come back positive.

How does the rake work?

On a market with two roughly equal sides, a true coin flip would pay +100 on each. A sportsbook instead posts something like -110 on both. That gap is the rake. Bet both sides and you would still lose a little, and that little is the book's guaranteed cut. The book's goal is to set the line so money comes in evenly on both sides, then collect the margin no matter who wins.

How much does the rake cost you?

It depends on the bet. A moneyline, which takes heavy two-way action, is priced tight: the average two-sided hold on a game line is about 3.9%, and that action splits so evenly that shopping the best price on each side strips a good part of it out. Graded against real results, the average moneyline came back around -2% per bet over 1,111 games - a separate quantity from the hold, and one that does not separate from zero on that sample. A player prop takes far less action, and mostly on one side, so the book can pad the price. Anytime home run props grade out around -23% expected value across 18,468 graded props, far outside any noise floor. Same game, wildly different rake.

How do you avoid paying too much rake?

Know the true probability, and shop the price. The rake hides inside the odds. On a two-sided market you can see it directly by adding up what both prices imply, and on a one-sided market it only surfaces once the price is measured against the real chance of the outcome. Markets that take balanced action, like game lines, carry the least. Longshot props carry the most.

How can the rake on a specific bet be measured?

Add up what both prices imply and take the excess over 100%. That works on a market with two priced sides, like a moneyline. Most player props price only one side, so there is no second price to work from, and the honest measurement there is expected value: the real chance of the outcome set against the chance the price implies. NegativeEV simulates each game thousands of times, across MLB, WNBA, PGA, and ATP, and reports both the true probability and the expected value for any bet. Checking is free: 10 checks a day with no account, and unlimited signed in. The rake is already baked into that expected value, so a negative number means the price is worse than the outcome deserves. Adding up what two prices imply starts with how odds convert to probability.

More on the number: https://negativeev.com/about/expected-value