Betting 101

How to convert betting odds to implied probability (properly)

Published 2026-08-19

Every set of betting odds is a probability statement with a fee hidden inside it. Learning to convert betting odds to implied probability — and then to strip out the bookmaker’s margin — is the single most useful piece of arithmetic in football betting. It’s how you find out what the market actually believes, and it’s the first thing our pipeline does to every odds feed before comparing the market’s probabilities with our model’s. The whole calculation fits on a napkin. Here it is, with the traps labelled.

Step one: convert the odds to implied probability

For decimal odds, the raw implied probability is simply 1 divided by the odds:

• Odds of 2.00 → 1 / 2.00 = 50%
• Odds of 3.40 → 1 / 3.40 = 29.4%
• Odds of 1.25 → 1 / 1.25 = 80%

The intuition: odds of 2.00 pay out double your stake, which is a fair deal exactly when the outcome is a coin flip. Shorter odds mean the bookmaker thinks the outcome is more likely; the inverse converts that price back into a belief.

If you bet in fractional or American odds, convert to decimal first (fractional 5/2 → 3.50 decimal; American +150 → 2.50, −200 → 1.50) and proceed the same way.

Step two: notice the sum is more than 100%

Take a plausible 1X2 market — home 2.10, draw 3.40, away 3.80 — and invert all three:

One of these three outcomes will happen, so true probabilities must sum to exactly 100%. The extra 3.3 points are the overround (also called the margin, the vig or the juice): the bookmaker has shaded every price slightly below its fair value, and that shading is their structural profit. Whoever you back in this market, you’re paying a share of those 3.3 points.

Margins vary widely — sharp bookmakers run 1X2 markets at 2–3%, softer books at 6–8%, and margins tend to be fatter still in lower divisions and obscure markets where the book fears sharp customers more than it wants turnover. That’s worth internalising: the margin is a price, and it’s not the same price everywhere.

Step three: remove the margin

The standard fix — the basic or multiplicative method — is to divide each raw probability by the sum, so the set is rescaled to 100%:

Those de-vigged numbers are the market’s actual opinion: this home team wins about 46 times in 100, not the 47.6 the raw inversion suggested. The “fair odds” column (1 divided by the de-vigged probability) shows what the prices would be with no margin — and the gap between 2.10 and 2.17 is what the bet costs you beyond its risk.

Two-way markets work identically. The classic Over/Under 2.5 line at 1.91/1.91 inverts to 52.4% + 52.4% = 104.7%; de-vigged, it’s a dead-even 50/50 and the fair odds are 2.00 on both sides. Any time you see 1.91/1.91, you’re being quoted a coin flip with a 4.7% fee.

The trap: margin isn’t spread evenly

The basic method assumes the bookmaker shades every outcome proportionally. Decades of betting-market research say they don’t: a disproportionate share of the margin sits on the longshot, a pattern known as the favourite–longshot bias. Odds of 15.00 on an away upset typically overstate the true probability by much more (relatively) than odds of 1.30 on the favourite understate theirs — partly deliberate pricing, partly the market catering to bettors who love big payouts.

The practical consequence: after a basic de-vig, longshot probabilities are usually still too high. More careful methods — the power method, Shin’s method, or weighting the margin in proportion to each outcome’s odds — push more of the correction onto the outsider. For a typical 1X2 market with a 3–5% margin the differences are fractions of a percentage point; for a 15.00 longshot they can be material. If you only remember one thing here: be most sceptical of implied probabilities at long odds.

Why this arithmetic matters

Three immediate uses:

1. Comparing bookmakers honestly. The best price on your outcome matters more than the headline margin, but a book that consistently runs 7% overrounds is charging you more than one running 3% — compounded over every bet you ever place.
2. Reading any probability claim. When a tipster says a team is “60% to win” at odds of 2.00, you now know the market says ~48–50% after the margin comes out. One of them is wrong, and it’s usually not the market.
3. Judging predictions properly. Probabilities are the only honest format for football predictions — and once everything is a probability, models and markets can be scored on the same scale, which is exactly what we do in public on our track record.

This is also why every FootInsights prediction page shows the market’s de-vigged probabilities next to our model’s raw ones: same units, no margin, disagreements visible. We think anyone quoting odds-derived “percentages” without removing the margin first is — knowingly or not — inflating every number by several points. The fix, as you’ve now seen, is one division.

Want the arithmetic done for you? The odds converter turns any price — decimal, fractional or American — into its implied probability instantly, right in your browser.