A market with two possible outcomes shows 1.91 on each side. If that were truly fifty-fifty, the fair odd would be 2.00 on both. The difference is the house margin.
How the margin works
Convert each odd into implied probability — one divided by the odd. At 1.91, that's roughly 52.4%. On both sides, it adds up to 104.8%. That percentage above one hundred is the margin: what the house keeps, on average, regardless of the outcome.
A sharp book works with very thin margins, sometimes below 2%. A retail book can go to 7% or higher. This means the same estimate produces visibly different odds depending on who publishes it.
Why this matters
Because comparing raw odds across different books is comparing things on different scales. An odd of 2.00 on a book with 2% margin represents a completely different estimate than an odd of 2.00 on a book with 7%.
Removing the margin returns the fair price — the probability estimate without the commission. Only after that do the numbers become comparable.
The part that goes wrong
The margin doesn't distribute equally across both sides. Favorites typically carry proportionally more margin than underdogs, because that's where most public money goes.
So dividing the margin equally produces wrong fair prices, especially in imbalanced markets. Methods that weight the distribution get closer to reality.
The most common mistake
Comparing a raw price with a fair price. They're different scales, and the gap between them is exactly the margin — typically four or five percentage points. Making that comparison produces a result that looks like analysis but is really just the house commission dressed up as a number.