Most people compute one margin for a match. The margin is one number, but its distribution across home, draw and away is not uniform — and that distribution is where the interesting part is.
Three prices: 2.10 / 3.40 / 3.60.
The standard assumption is that the book applies the same proportional loading to each outcome. Divide each implied probability by the sum:
Invert those and you get fair prices: 2.20 / 3.56 / 3.77. Compare with what you were offered: 2.10 / 3.40 / 3.60. Every outcome is short by roughly the same proportion.
Books do not have to load every outcome equally. Some price the favourite close to fair and put most of the margin on the draw and the outsider, because that is where the money is least price-sensitive. Others do the opposite.
You can see this when two books show the same favourite price but very different margins overall: the difference is being carried by the other two outcomes.
For a given match, compute the proportional fair price for each outcome and compare it to the offered price as a percentage:
shortfall = offered / fair − 1For the example above: home −4.5%, draw −4.5%, away −4.5%. If one outcome comes out at −2% and another at −7%, the book is not loading proportionally, and the expensive side is the one to avoid — not the one with the shortest price.
It is not a prediction method. A fair price is still just the market's opinion with the cut removed. What it does give you is a like-for-like comparison, which is the only way to shop for price without being misled by the headline number.