Is betting the over a good bet? What MLB totals actually pay

The over is the popular side of a totals bet. Measured against 1,067 games of real final scores, it grades worse than the under - but neither side separates from fair.

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

A totals bet takes the combined final score of a game, the sportsbook posts a line like 8.5 runs, and both sides pay close to even money. Over or under, the price looks symmetrical. What the slip does not show is that the two sides have not been paying anywhere near the same.

What do overs and unders actually pay?

Across 1,067 games, betting every over returned about -4.4% per bet and betting every under about -0.7%. On the smaller July-only sample the over looked clearly expensive, but the full archive window pulls both sides back inside their noise floors. The direction survives - overs still grade worse - but neither side of a total separates from fair on the season measured so far.

Why is the over priced worse?

A sportsbook sets each line to balance its own action and bank its margin. On totals, the action leans heavily to the over: runs are what casual money roots for, and an under ticket asks its holder to cheer for nothing to happen. When one side draws most of the money, the book can shade that side's price and still keep its book balanced. The over pays for its popularity.

Does that mean the under is always the better bet?

No. The -0.7% on unders sits inside its own noise floor, so it is not evidence of a cheap side, and a blanket rule is exactly the kind of shortcut the numbers warn against. What this window does support is the direction - the over graded worse than the under - not a size, so an over price deserves more scrutiny than its even-money look suggests rather than a standing rule against it. Whether tonight's total is priced fairly is a question about tonight's game.

What does this page not establish?

Which side of tonight's total to take. Every figure here is an average across 1,067 games, and an average does not sort one game from the next.

Neither side clears its own noise floor, so what this window supports is the direction, that the over graded worse than the under, and not the size of either number or the gap between them. The decimals are estimates rather than measured costs, which is also why the smaller July-only pass read so differently.

The reason offered here for why the over is shaded, that the action leans to it and a book can price a popular side worse and stay balanced, is an explanation for the result rather than something measured on this page.

Both sides were graded at the best price posted across books, at the total most books were quoting. That describes what the market charged at its best number, not what one account at one book would have returned, and not what the same total was worth earlier in the day.

And this is MLB game totals over one archived window. Nothing here covers team totals, player totals, totals in another sport, or anything priced after the window closed.

How can a totals bet be checked?

NegativeEV simulates each game thousands of times, play by play, and counts how often the combined score actually clears the line. That rate is compared with what the posted price implies, over or under, and the bet grades out as an expected value. Paste a bet slip or type the bet at negativeev.com and it comes back graded, free.

More on the number: https://negativeev.com/about/expected-value The measurement behind this page: https://negativeev.com/research/mlb-totals