Add up the chances implied by a bookmaker's odds on a football match and you do not get 100%. You get something like 105.
Take a match priced at 2.50 the home win, 3.40 the draw, 3.00 the away win. Turn each into an implied chance by dividing one by the odds:
That totals 102.7%. But exactly one of those three things is going to happen, so the true chances have to total 100%. The extra 2.7 points is the margin — sometimes called the overround or the vig. It is the bookmaker's cut, priced into every selection on the board.
Football 1X2 markets typically run a few points over. Some markets, and some books, run a good deal wider than that.
The margin is not a fee you pay once. It is applied to every bet you place, forever. It is the reason most bettors lose over time — not bad luck, not bad discipline, just arithmetic running quietly against them on every single wager.
To be profitable you do not need to be better than the bookmaker. You need to be better than the bookmaker plus the margin, which is a meaningfully higher bar and the one almost everyone fails to clear.
When someone says the model rates a selection at 63% while the price implies 55%, that eight-point gap sounds enormous.
It is also partly fictional, because the 55% has the margin baked into it. Strip the margin out first — spread it across the selections in proportion, which is the standard approach and the one we use — and the honest comparison might be 63% against 57%. Still a gap, but a smaller one, and a good number of gaps disappear entirely once you do this properly.
We de-vig every market before comparing anything to it. Most of what looks like value in the wild is the margin wearing a disguise, and a model that skips this step will generate hundreds of exciting selections that are worth nothing at all.