Implied Probability Explained

The formula: decimal odds to implied probability

For decimal odds, the conversion is one step:

Implied probability = 1 ÷ decimal odds

Multiply by 100 to get a percentage. A few reference points South African punters see daily:

• 1.20 → 83.3%
• 1.50 → 66.7%
• 1.91 → 52.4%
• 2.00 → 50.0%
• 3.50 → 28.6%
• 5.00 → 20.0%
• 11.00 → 9.1%

Going the other way, if you rate a runner at 40% you need odds of 1 ÷ 0.40 = 2.50 just to break even. Anything above 2.50 is a positive-expectation price on that estimate; anything below is negative.

Decimal is the default display at Betway, Hollywoodbets, Sportingbet and every other locally licensed book, which makes this the only formula most South African bettors need.

Why the numbers never add up to 100%

Add the implied probabilities of every outcome in a market and you will get more than 100%. That surplus is the bookmaker margin — also called the overround or vig — and it is how the operator builds in profit.

Example: a two-way market priced at 1.91 / 1.91.

• 1 ÷ 1.91 = 52.36%
• 52.36% + 52.36% = 104.72%

The extra 4.72% is the margin. On a genuinely 50/50 event the fair price would be 2.00 / 2.00 (50% + 50% = 100%). The book has shaded both sides to 1.91.

To recover the fair (true) probability, divide each raw implied figure by the market total. For the 1.91 example: 52.36% ÷ 104.72% = 50.0% per side. That normalisation step is what separates a headline price from the model probability underneath it.

Worked example: comparing two South African prices

Say Kaizer Chiefs are priced to win a PSL match. You have two decimal prices from different books:

• Book A: 2.10
• Book B: 2.00

Step 1 — convert to implied probability:

• 2.10 → 1 ÷ 2.10 = 47.6%
• 2.00 → 1 ÷ 2.00 = 50.0%

Step 2 — apply your own estimate. Suppose your model rates Chiefs at 49% to win.

Book A implies 47.6%. Your 49% is higher than the price implies, so at 2.10 you have an edge. Break-even odds for a 49% chance are 1 ÷ 0.49 = 2.04. Book A's 2.10 clears that; Book B's 2.00 does not.

Step 3 — quantify the edge. Expected value on a R100 stake at 2.10 with a true 49% chance:

• Win: 0.49 × (2.10 − 1) × R100 = +R53.90
• Lose: 0.51 × R100 = −R51.00
• EV = +R2.90 per R100 staked

At Book B's 2.00 the same 49% estimate gives EV = 0.49 × R100 − 0.51 × R100 = −R2.00. Same match, same view, opposite decision — driven entirely by the 0.10 difference in odds. This is why line shopping matters: the 4.8% difference between 2.00 and 2.10 flips a losing bet into a profitable one on paper.

Margin is where implied probability gets expensive

The lower the margin, the closer a book's implied probabilities sit to fair value, and the more often its prices beat rivals. In BETAXIO's sampled odds data, average two-way margin varies sharply between operators:

• 10bet: 4.78% average margin (27 markets sampled)
• Betway: 6.49% average margin (238 markets sampled)
• Sportingbet: 8.63% average margin (39 markets sampled)

On the same true 50/50 event, a 4.78% margin produces a price near 1.95, while an 8.63% margin produces roughly 1.84. That is the difference between implied 51.2% and 54.3% on each side — you pay more implied probability for the same outcome at the higher-margin book.

Best-price share reflects this in practice. Betway held the top price in 90.8% of its 238 sampled markets and 10bet in 85.2% of 27, while Sportingbet led in only 28.2% of 39. Fewer sampled markets means less certainty, but the pattern is clear: lower margin correlates with more market-leading prices.

Fractional and American odds (for reference)

South African books display decimal by default, but you will meet other formats on international feeds.

Fractional (e.g. 5/2): implied probability = denominator ÷ (numerator + denominator). So 5/2 = 2 ÷ 7 = 28.6%. That is the same as decimal 3.50.

American (moneyline): for a positive line +150, implied probability = 100 ÷ (150 + 100) = 40%. For a negative line −150, implied probability = 150 ÷ (150 + 100) = 60%.

If you ever need to convert an American or fractional price, turn it into decimal first, then apply 1 ÷ odds. It keeps one mental model for every market.

Using implied probability to find value

The workflow is the same across football, rugby, cricket, tennis and MMA:

1. Convert every price in the market to implied probability. 2. Sum them and note the margin (overround). 3. Normalise to get the book's fair probability. 4. Compare against your own estimate. Bet only when your estimate exceeds the price's implied figure by more than the margin can explain.

Two practical notes for South African markets. First, margins on niche markets — lower-tier PSL, individual player props, deep MMA undercards — are typically wider than on a Premier League match-winner line, so the implied probability you pay is inflated. Second, opening versus current lines matter: if a price has drifted from 1.80 to 2.05, the market's implied probability for that outcome has dropped from 55.6% to 48.8%, which is information about where the money has gone.

BETAXIO's odds and market-movers pages exist to do the conversion and line-tracking for you, but the arithmetic above is what sits underneath every ranking.