How Gambling Odds, Probability and Payouts Fit Together

Odds do two jobs at once. They describe the return attached to a winning outcome, and they imply something about how likely that outcome is.

Those ideas are connected, but they are not interchangeable.

A bet can offer a large payout because the outcome is unlikely. A short-priced outcome can be much more likely but return relatively little. Neither tells you on its own whether the price is fair, and neither should be confused with a game's RTP.

Probability starts with the chance of something happening

Probability expresses how likely an outcome is.

A probability of 50% means the outcome would be expected to occur about half the time across a sufficiently large number of comparable trials. It does not mean results will neatly alternate between win and loss.

A fair six-sided die gives a simple example. The probability of rolling a six is:

1 ÷ 6 = 0.1667

Expressed as a percentage, that is about 16.67%.

There are six possible equally likely results and only one of them is a six. If three of the six results counted as a win, the probability would instead be 50%.

Real gambling products can have much more complicated probability models, but the underlying idea is the same. Each possible result has a probability attached to it.

Digital games use random inputs to produce those results. Our guide to random number generators explains how regulated games generate and map those inputs.

Decimal odds are the price offered for an outcome

In Great Britain, betting prices are commonly shown as decimal odds.

Decimal odds tell you the total amount returned if the bet wins, including the original stake.

At decimal odds of 2.00, a £10 stake returns £20 in total. The profit is £10.

At odds of 3.00, the same £10 stake returns £30 in total. The profit is £20.

At odds of 1.50, it returns £15. The profit is £5.

The calculation is:

stake × decimal odds = total return

Profit is then:

total return - original stake = profit

This distinction matters because a £30 payout from a £10 stake is not a £30 profit. £10 of that return is simply the original stake coming back.

Turning decimal odds into implied probability

Decimal odds can also be expressed as an implied probability.

The basic calculation is:

1 ÷ decimal odds × 100

So odds of 2.00 imply 50%.

Odds of 4.00 imply 25%.

Odds of 1.25 imply 80%.

The word "implied" matters. This is the probability represented by the price. It is not proof that the event genuinely has that exact chance of happening.

A bookmaker can price an event differently from your own estimate of its probability. Prices can move as information or the market changes. The bookmaker's margin can also affect the relationship between the quoted prices and a perfectly fair set of probabilities.

It is therefore safer to read odds of 2.00 as representing a 50% break-even probability before other factors, rather than saying the event definitely has a 50% chance.

Fair odds and offered odds are not necessarily the same

Imagine an event with only two possible outcomes.

If both outcomes genuinely had a 50% probability and there were no margin built into the prices, fair decimal odds for each would be 2.00.

Now imagine both outcomes are instead offered at 1.91.

Each price gives an implied probability of about 52.36%.

Add the two together:

52.36% + 52.36% = 104.72%

Both outcomes obviously cannot each have a true probability of 52.36% in a two-outcome event.

The amount above 100% reflects margin within the set of prices. This is commonly called the overround.

That does not mean the bookmaker will make exactly 4.72% from every customer or event. Real trading results depend on the bets actually taken, the prices at which they were taken and the result.

It does show why converting one quoted price into a percentage should not automatically be treated as a neutral estimate of the real-world probability.

Bigger payouts usually come with lower probabilities

Suppose two hypothetical bets are fairly priced.

Bet A has a 50% chance of winning. Its fair decimal odds would be 2.00.

Bet B has a 10% chance of winning. Its fair decimal odds would be 10.00.

A £10 winning bet would return £20 on Bet A and £100 on Bet B.

The £100 return looks much larger in isolation, but it comes with a much lower probability of receiving it.

That trade-off sits underneath fixed-odds gambling. All else being equal, a larger potential return is attached to a less likely outcome.

The size of a payout therefore tells you very little unless you also understand the probability attached to receiving it.

Probability does not tell you when an outcome will happen

A 10% chance does not mean an outcome must occur once in every ten attempts.

Across ten attempts it might happen once. It might happen three times. It might not happen at all.

The 10% describes the probability attached to each comparable trial, assuming the underlying conditions remain the same.

This becomes particularly important where outcomes are independent. If a random game gives an outcome a fixed probability, previous results do not create a debt that the next result has to repay.

Five losses do not automatically make the sixth attempt more likely to win.

Long-run proportions can move towards their theoretical probabilities without results arriving in an orderly pattern. Our guide to volatility and variance explains why short sequences can look very different from the long-run average.

Odds and RTP answer different questions

Odds usually describe a particular outcome. RTP describes the theoretical proportion of total wagering returned as prizes across a very large amount of play.

Those are different measurements.

A game can contain many possible winning outcomes, each with its own probability and payout. The combined effect of all of those outcomes contributes to the game's overall theoretical RTP.

Take a deliberately simplified £1 game:

• 50% chance of receiving £1.80
• 50% chance of receiving nothing

Its expected return is:

0.50 × £1.80 = £0.90

The theoretical return is therefore 90p for every £1 wagered, equivalent to a 90% RTP in this simplified model.

The winning outcome occurs half the time in the probability model, but the game still has a negative expected value because the £1.80 return does not fully compensate for the losing outcomes.

This is why win probability and RTP should not be treated as synonyms.

A game can produce frequent winning outcomes while still returning less than is staked overall. It can also produce infrequent large prizes that account for a substantial part of its theoretical return.

Our RTP guide covers the long-run return calculation in more detail.

A winning bet can still be badly priced

Probability and payout also allow expected value to be calculated.

Suppose an outcome has a genuine 50% probability of occurring, and a £1 bet returns £1.80 in total if it wins.

The possible net results are an 80p profit on a win and a £1 loss if it loses.

Expected value is:

(0.50 × £0.80) + (0.50 × -£1.00)

That gives:

£0.40 - £0.50 = -£0.10

The expected loss is 10p for every £1 wagered.

The fact that the bet wins half the time does not make it a fair bet. The payout matters as much as the probability.

The same logic explains why a very unlikely event can still be poorly priced despite offering a large headline return. Our house-edge guide develops the expected-value calculation from the player's and operator's perspective.

Read the figures together

Probability answers one question: how likely is the outcome?

Odds answer another: what return is being offered if it happens?

Expected value asks whether that return adequately compensates for the probability and the amount at risk.

RTP and house edge describe the longer-run mathematical effect when probabilities and payouts are combined across repeated gambling.

None of those figures predicts what will happen on the next bet. They describe the structure underneath it.