Why the shown price is never fair
Every priced market sums to more than 100%, and that surplus is the overround. It is spread across all the outcomes, which means no single price you see is the market's true estimate; each one has been shaded short to make room for the margin. If you want to know what the market actually believes an outcome's chance is, you have to take the margin back out first.
This matters because a fair price is your yardstick. Without it you cannot tell whether 2.75 was a good number or a bad one, because you have nothing honest to compare it against. Strip the margin and you get the no-vig price, the price that would exist if the bookmaker took no cut at all. That is the number your bet has to beat.
The proportional method
The clean way to remove the margin is proportional. Convert every outcome to its implied probability, add them to get the book total, then divide each individual implied probability by that total. This rescales the whole market back down to 100% while keeping each outcome share of the book intact. Turn each rescaled probability back into a price with 1 divided by the probability, and you have the fair odds.
Work an example. A three-way market is priced 2.10 / 3.50 / 3.80. The implied probabilities are 47.6%, 28.6% and 26.3%, which sum to 102.5%. Divide each by 1.025 and you get 46.5%, 27.9% and 25.7%, now summing to 100%. Convert back: the fair prices are 2.15, 3.59 and 3.90. Every fair price is slightly longer than the shown one, which is exactly what removing a margin should do.
| Outcome | Shown price | Implied % | Fair % | Fair price |
|---|---|---|---|---|
| Home | 2.10 | 47.6% | 46.5% | 2.15 |
| Draw | 3.50 | 28.6% | 27.9% | 3.59 |
| Away | 3.80 | 26.3% | 25.7% | 3.90 |
| Book | — | 102.5% | 100.0% | — |
Turning fair price into edge
Once you have the fair price you can measure your edge. If your own estimate of an outcome is stronger than the fair probability the market implies, you have a positive edge; if it is weaker, you do not. The size of the edge is simply your probability minus the fair implied probability. Betting only when that number is positive, and ignoring the shaded shown price, is the whole discipline in one sentence.
Be honest about the direction of the comparison. Most punters compare their fancy against the shown price and talk themselves into a bet because the shown price looks generous. But the shown price is shaded in the bookmaker's favour, so it flatters every selection. Compare against the fair price instead and a lot of tempting bets quietly stop qualifying, which is the point.
The closing line is the honest referee
The best available fair price to compare against is often the one the market settles on at the off, the closing line, stripped of its margin. Markets get sharper as money comes in and information arrives, so the closing price is the crowd's best collective estimate. If you consistently took prices that beat the no-vig close, your process is finding value even on the nights the result goes against you.
This is why serious bettors track the price they took versus the price at the close, not just their profit and loss. Profit is noisy and needs hundreds of bets to mean anything. Beating the fair closing line is a signal you can read far sooner, because it shows up across dozens of bets rather than hundreds.
A repeatable routine
Make it mechanical. Add the book up. Rescale every implied probability by the book total to get fair probabilities. Convert those back to fair prices. Compare the fair price to both your own estimate and the price you can actually take. Bet only where your number beats the fair one, and keep a record of how your taken price compares to the fair close. Do that for long enough and the results stop surprising you, because you already knew whether the process was sound.