Probability is Topic 3 of Paper 3's Statistics section, and splits cleanly across the two years. At AS it's mutually exclusive and independent events, worked through Venn and tree diagrams; conditional probability and modelling with probability are held back to year two, with no conditional probability tested at AS. Across the 300 marks it's a smaller strand, typically around 10 marks depending on the series, though it also underpins the binomial and Normal work in Topic 4.
Almost every formula you need — P(A′) = 1 − P(A), the addition and multiplication rules, and the full conditional probability identity — is printed in the booklet. What's actually tested is judgement: which rule fits, whether a diagram needs a reduced sample space, and whether a model's assumptions hold. Probability questions reward careful reading as much as calculation.
The specification statements this topic covers. AS = Year-1 content, also assessed in the standalone AS course (8MA0); A2 = full A level only. Typical share of a 300-mark series: ≈10 marks — our estimate from the 2018–2025 papers, not an official weighting.
| Ref | Spec statement | Level |
|---|---|---|
| 3.1 | Mutually exclusive and independent events | AS |
| 3.1 | Probability and distributions | AS |
| 3.2 | Conditional probability | A2 |
| 3.3 | Modelling with probability | A2 |
Examiners repeatedly report that many candidates don't spot when a question has already restricted the sample space — a phrase like 'given that' changes what you're dividing by. In one widely cited series only the strongest tenth read the conditioning correctly. Underline the 'given' clause first, then decide whether you need P(A∩B) or P(A|B).
The two ideas pull in different directions: exclusive events can't both happen; independent events don't affect each other's chances. Testing independence needs P(A∩B) = P(A) × P(B), not the addition rule, and a Venn diagram for exclusive events should never show overlapping regions.
When modelling shifts into justifying a binomial model, examiners repeatedly report conditions listed in generic textbook language — 'trials are independent' — rather than tied to context, such as 'each throw is independent of the last'. Naming the actual objects and actions in the question is what earns the mark.
Questions asking you to critique or refine a probability model — is the die really fair, is the bag's replacement realistic — reward a precise limitation and its effect on the answer, not a general doubt. 'The probability might not be exactly ⅙ every time' scores; 'the model could be wrong' doesn't.
Using P(A∩B) = P(A) + P(B) to test independence, or assuming independent events can't overlap, mixes up two separate properties. Independence is checked only by comparing P(A∩B) with P(A) × P(B); exclusivity is about whether the events can occur together at all.
Conditional probability answers sometimes skip the division by P(B) altogether, giving P(A∩B) instead of P(A|B), or answer P(B|A) when the question asked the other way round. On a tree diagram, branch probabilities after the 'given' event should already reflect the reduced space — check the letter order in the notation before calculating.
Writing that 'trials are independent and probability is constant' without saying what that means for this die, bag or survey doesn't score — conditions have to be pinned to the nouns and actions in the question. It's also worth checking whether a condition genuinely fails, such as sampling without replacement changing the probability each time.
We haven’t published checked questions for this topic yet — a worked sample appears here only once a question has passed every check. In the meantime you can practise in the app.
Because the formulae are supplied, there's relatively little to memorise for probability itself — P(A′) = 1 − P(A), the addition and multiplication rules, and the full conditional probability formula are all in the booklet, at AS and A2 respectively. What's worth memorising instead is the pattern-matching: recognising a 'given that' clause, knowing when a Venn or tree diagram is the faster route, and testing independence or exclusivity without reaching for notes.
Practise by drawing the diagram first, every time, even for questions that look short enough to do in your head — a labelled Venn or tree diagram catches conditioning and double-counting mistakes before they reach the arithmetic. Keep probabilities as exact fractions unless the question allows a decimal, and sanity-check that every answer sits between 0 and 1.
Probability also sets up the statistical distributions that follow, so time spent here pays off in Topic 4 — a shaky grip on independence or conditioning makes binomial modelling harder than it needs to be. When a question asks you to critique a model, practise naming one specific assumption and saying, in context, what would change if it didn't hold.
All 19 topics: Edexcel A level Maths topic guides. Reference: formula booklet vs memorise and grade boundaries.
Ready to practise this topic properly? Start free — no card needed.
Original questions written for the Pearson Edexcel A Level Mathematics (9MA0) specification. Not affiliated with or endorsed by Pearson.