Probability vs Odds: How to Work Out the Chance of Something

Probability and Odds Are Different Numbers

The two words are used interchangeably in conversation and mean different things in arithmetic, which is the source of most confusion about chance.

Probability compares the favourable outcomes to all outcomes: favourable ÷ total.
Odds compare the favourable outcomes to the unfavourable ones: favourable : unfavourable.

Drawing one particular suit from a shuffled pack is a probability of 1 in 4 — 0.25, or 25%. The odds against it are 3:1, because there are three losing suits for every winning one.

The conversions both ways are short:

Odds against, from probability: (1 − p) ÷ p
Probability, from odds against of a:b: b ÷ (a + b)

So 3:1 against gives 1 ÷ (3 + 1) = 0.25. A probability of one half is evens, or 1:1 — which is why "50:50" is strictly a statement of odds rather than of probability, even though everyone means the probability by it.

Three Ways to Write the Same Chance

Probability always sits between 0 and 1, where 0 is impossible and 1 is certain, and it appears in three interchangeable forms:

Fraction: 1/6 for one face of a die
Decimal: 0.1667
Percentage: 16.7%

Any figure above 1, or above 100%, means an error has crept in — usually probabilities added together that should have been multiplied.

Betting Odds, and the Number the Bookmaker Adds

UK betting uses two notations for the same thing, and converting either to a percentage is the only way to compare them honestly.

Fractional odds such as 5/1 state profit against stake: a winning £1 returns £5 profit plus the £1 back. Decimal odds state the total return per unit staked, so the same bet is 6.0.

Implied probability = 1 ÷ decimal odds. For 5/1 that is 1 ÷ 6 = 16.7%.

Now the part worth knowing. Apply that formula to every outcome in a market and the total should be 100%. It never is.

A two-way market priced at 10/11 on both sides converts to decimal odds of 1.909 each, an implied probability of 52.4% each — a total of 104.8%.

That surplus 4.8% is the overround, the bookmaker's margin, and it is the reason a run of correctly judged bets can still lose money. Summing the implied probabilities of any market tells you what the margin is before you place anything.

Combining Events: Multiply for AND

When independent events must all happen, multiply their probabilities.

Two coins both landing heads: ½ × ½ = ¼, or 25%.

Independence matters. Two coins do not influence each other, so multiplying is valid. Drawing two aces from a pack without replacement is not independent, because the first draw changes what is left: 4/52 × 3/51, not 4/52 × 4/52.

The "At Least One" Shortcut

Questions of the form what is the chance of at least one are painful head-on and trivial backwards, because the opposite of "at least one" is simply "none":

P(at least one) = 1 − P(none)

Rolling a die four times, the chance of no six at all is (5/6)⁴ = 0.482, so the chance of at least one six is 1 − 0.482 = 51.8%.

Just over an even chance, which is a slimmer margin than most people expect from four attempts. This exact problem, posed by a seventeenth-century gambler, is one of the questions that started probability theory.

Combining Events: Add for OR, Then Subtract the Overlap

When either of two events will do, add — but only when they cannot both happen at once. Rolling a 1 or a 2 on one die is 1/6 + 1/6 = 1/3, and those outcomes are mutually exclusive.

When both can happen together, plain addition double-counts the overlap:

P(A or B) = P(A) + P(B) − P(A and B)

Drawing a heart or a face card: 13/52 + 12/52 would give 25/52, but the three face cards that are hearts have been counted twice. Subtract them: (13 + 12 − 3) ÷ 52 = 22/52 = 42.3%.

The Birthday Problem

The most famous demonstration that intuition is unreliable here, and it follows directly from the two rules above.

In a room of 23 people, the chance that two share a birthday is 50.7% — better than even.

It sounds wrong because people picture themselves and compare against 22 others. The question is not about you: it asks about every pair in the room, and 23 people form 253 pairs. Working it backwards, the chance that all 23 birthdays are different is 365/365 × 364/365 × 363/365 and so on, which falls below one half at the 23rd person.

The lesson generalises well beyond birthdays: coincidences that feel improbable for a specified individual become likely once any match anywhere will do.

Combinations and Permutations

Counting the possible outcomes is usually the hard half of a probability question, and it turns on one distinction: does order matter?

Permutation — order matters. A four-digit PIN, a podium finish, a password.
Combination — order is irrelevant. A lottery draw, a hand of cards, a committee.

The names are the wrong way round in everyday speech: a "combination lock" needs its digits in order, so it is really a permutation lock.

Choosing 6 numbers from 59, the number of possible combinations is C(59,6) = 45,057,474 — the familiar odds of a jackpot. If the order of the balls mattered, the count would be P(59,6) = 32,441,381,280, which is 720 times larger, because each set of six can be arranged in 6! = 720 ways.

That factor of 720 is the whole distinction between the two, and it is why identifying which one a question needs matters more than the arithmetic that follows.

Where Intuition Goes Wrong

The gambler's fallacy. After eight reds on a roulette wheel, black is not "due". The wheel has no memory, and the next spin is what it always was. The probability of eight reds in advance is genuinely small; the probability of red on the ninth spin, given eight already, is unchanged.

Confusing a sequence with an outcome. Any specific sequence of ten coin flips has a probability of 1 in 1,024, including ten heads. Ten heads feels special only because we group all the mixed results together and compare one sequence against a large class of them.

Adding when you should multiply. Three independent 50% chances do not make 150%, or a certainty. All three happening is 12.5%; at least one happening is 87.5%.

Conclusion

Almost every probability question reduces to four moves: count the favourable outcomes, count the total, multiply for events that must all happen, and add for alternatives while subtracting any overlap. When a question asks about at least one, turn it round and work out the chance of none.

Odds are the same information in a different frame, and converting betting prices to implied percentages is the only way to see what is actually being offered. Use a Probability Calculator for the combinations and permutations, which is where the counting gets heavy, and keep the four rules above for the rest.

Try our Probability Calculator

Try our Statistics Calculator

Related Calculators: Factorial Calculator