Implied Probability: What the Odds Think Will Happen

Odds are easier to understand when you turn them back into percentages.

Decimal odds of 2.00 imply a 50% chance. Odds of 4.00 imply 25%. Odds of 1.25 imply 80%.

The formula is simple:

Implied probability = 1 ÷ decimal odds × 100

For odds of 2.50:

1 ÷ 2.50 × 100 = 40%

The bookmaker is pricing the outcome as though it has roughly a 40% chance before adjusting for the margin across the market.

Why “before adjusting” matters

In a fair two-outcome market, the implied probabilities should add to 100%.

Suppose a tennis match has:

• Player A at 1.80
• Player B at 2.10

Player A:

1 ÷ 1.80 = 55.56%

Player B:

1 ÷ 2.10 = 47.62%

Total:

103.18%

The extra 3.18 percentage points are the overround. It is one way of seeing the bookmaker’s built-in price.

The individual percentages are therefore not pure forecasts. They include the shop’s markup.

Removing the margin

A simple method is to divide each implied probability by the total.

Player A:

55.56 ÷ 103.18 = 53.85%

Player B:

47.62 ÷ 103.18 = 46.15%

Those adjusted figures add to 100% and provide a rough estimate of the market’s margin-free probabilities.

More advanced methods distribute the margin differently because bookmakers may not apply it evenly. For everyday comparison, proportional removal is a useful first look.

Fractional odds to probability

For fractional odds A/B:

Probability = B ÷ (A + B)

Odds of 3/1 imply:

1 ÷ (3 + 1) = 25%

American odds to probability

For positive American odds:

100 ÷ (odds + 100)

+300 implies:

100 ÷ 400 = 25%

For negative American odds:

absolute odds ÷ (absolute odds + 100)

-200 implies:

200 ÷ 300 = 66.67%

The percentage is not the truth

A market price is a crowd estimate filtered through a bookmaker’s model and margin.

If odds of 2.50 imply 40%, the event does not become 40% true. The price says that a bettor needs the real chance to be above 40% for the wager to have positive expected value before other complications.

This is where betting changes from prediction to pricing.

Saying “I think Arsenal will win” is not enough. The useful statement is “I think Arsenal wins 48% of the time, and the odds are priced as though the chance is 40%.”

The gap is the argument.

A worked value example

You estimate an outcome has a 45% chance.

The bookmaker offers decimal odds of 2.40, implying:

1 ÷ 2.40 = 41.67%

Your estimate is 3.33 percentage points higher.

Expected value on a €10 bet can be written as:

(chance of winning × profit) - (chance of losing × stake)

At 2.40, the profit on €10 is €14.

(0.45 × €14) - (0.55 × €10) = €0.80

On your assumptions, the expected value is €0.80 per €10 bet.

The dangerous part is obvious: your probability estimate might be wrong. Mathematics cannot rescue a bad input. A polished spreadsheet can still be an expensive opinion.

Why longshots deserve extra care

Research across betting markets has repeatedly found a favourite-longshot bias, where long-priced outcomes tend to offer worse average returns than favourites. [1] One explanation is that bettors disagree more about unlikely events and are attracted to lottery-like payouts. [2]

That does not mean every favourite is good value or every outsider is bad. It means the exciting price may contain a larger hidden tax.

My verdict

Implied probability is the translator between odds and judgment.

Convert the price to a percentage. Remove the market margin. Then ask whether you have a credible reason to disagree.

Most people begin with a winner and look for odds. A more disciplined bettor begins with the odds and asks what must be true for the price to be wrong.