
Probability and Outcomes
Probability is careful counting dressed up as prediction. Every question in this topic reduces to comparing two numbers: the theoretical fraction you calculate before you roll, spin or draw, and the experimental fraction you calculate after. Get comfortable listing outcomes without missing any — the rest (trees, grids, Venn diagrams) is just bookkeeping around that one skill.
Overview
This chapter has exactly two families of question: those asking you to predict a probability from counting outcomes (theoretical), and those asking you to measure one from recorded data (experimental). Everything else — dice, coins, spinners, trees, Venn diagrams, sample space grids — is just the setting the counting happens in.
- The probability scale (0 to 1) gives you the vocabulary: impossible, unlikely, even chance, likely, certain.
- Single events (one die roll) become combined events (two dice, spinner + die) once you're tracking more than one action at once.
- Listing outcomes systematically — as a set, a grid, or a tree — is the skill every later calculation depends on.
The shape of the chapter
Command terms this topic actually uses
| Command term | What it demands | AO | Mark-earning move |
|---|---|---|---|
| Calculate | Obtain a numeric answer, showing the working/formula used. | Knowing and applying | Mark is for the correct method AND the final value — a bare answer with no formula shown can lose the method mark. |
| Explain | Give a reason connected to the maths, not just restate the situation. | Communicating | Vague answers ('it's rare', 'it's not fair') earn nothing without a specific mathematical or physical reason. |
| Construct | Build the actual diagram or table completely and correctly. | Applying maths | Marked on structure/completeness — one missing cell or branch loses marks even if later arithmetic is fine. |
| Compare | Put two values in the same format and state which is bigger and by roughly how much. | Communicating | Comparing raw counts instead of probabilities (e.g. '6 blue > 3 blue') scores zero — must normalise first. |
| Justify | Give a clear verdict (correct/incorrect, true/false) backed by a calculated value. | Reasoning | A calculation with no stated verdict, or a verdict with no calculation, only earns partial credit. |
| Apply | Use a given formula on new numbers. | Knowing and applying | Marks for correct substitution into the stated formula — using a different (unstated) method can cost marks even if numerically correct. |
| State | Give a fact directly, no working required. | Knowledge | No explanation expected or rewarded — but the value must still be exact (e.g. 1/6, not 0.17). |
Key point
Theoretical vs Experimental Probability
- Theoretical probability is calculated before anything happens, purely by counting the sample space — no die needs rolling.
- Experimental (empirical) probability is measured after actually running trials: frequency of the event divided by total trials performed.
- The Law of Large Numbers says experimental probability tends towards theoretical probability as the number of trials increases — but it never has to match exactly, and small samples can look wildly 'unfair' even from a perfectly fair object.
- A single unusual run (three 6s in a row, or 8 heads out of 10 tosses) is evidence, not proof — you need many trials before concluding anything about fairness or bias.
- Theoretical probability
- P(event) = number of favourable outcomes ÷ total number of possible outcomes, assuming all outcomes are equally likely.
- Experimental probability
- P(event) = number of times the event occurred ÷ total number of trials performed.
- Bias
- A systematic tendency for an outcome to occur more or less often than the theoretical value predicts, usually caused by a physical flaw in the object generating the outcomes.
Theoretical probability
- — number of favourable outcomes
- — total number of possible outcomes
Experimental probability
Predicted count of an event, not what you'll necessarily observe
Worked example
Q1 — is the die actually unfair?
Worked example
Q5 — does the coin get 'more fair' with more tosses?
Common mistake
Treating one small experimental result as the 'true' probability and ignoring the theoretical calculation entirely — e.g. stating P(rolling a 3) = 7/50 from a bar chart instead of the theoretical 1/6.
Consequence: Loses the mark on 'state the theoretical probability' questions (like Q8a), because examiners want the exact fraction 1/6 stated, not the observed frequency.
Fix: Underline the keyword before calculating: 'theoretical' → use the formula on the sample space; 'experimental'/'from the data'/'recorded' → use the table.
Examiner tip
Frequently confused
| Concept A | Concept B | The distinction that matters |
|---|---|---|
| Theoretical probability | Experimental probability | Theoretical is reasoned out before any trial from equally likely outcomes; experimental is measured from real data after trials — they only have to be equal in the limit of infinitely many trials, never in a small sample. |
Key takeaways
- Theoretical: P(E) = n(E)/n(S), calculated with no experiment needed.
- Experimental: P(E) = frequency/trials, calculated only from real recorded data.
- More trials → experimental probability drifts closer to theoretical (Law of Large Numbers), but a small sample proves nothing about fairness.
- Expected frequency = theoretical probability × number of trials — a prediction, not a guarantee.
Single and Combined Events
- A single event involves one action with one outcome recorded — rolling one die, tossing one coin, drawing one marble.
- A combined event involves two or more actions (or one action checked against two conditions) considered together — rolling two dice and adding the sum, or a spinner AND a die outcome checked at once.
- For combined events built from independent actions, multiply the individual probabilities: P(A and B) = P(A) × P(B) — this is the multiplication rule.
- When 'either/or' connects two outcomes that can't both happen at once, add the probabilities instead: P(A or B) = P(A) + P(B).
- Independent events
- The outcome of one event has no effect on the probability of the other, e.g. two separate die rolls.
- Mutually exclusive events
- Events that cannot both happen at the same time, e.g. rolling a 2 and rolling a 5 on the same single die roll.
- Combined event
- An outcome built from two or more single events considered together as one unit, e.g. 'spinner shows even AND die shows a factor of 6'.
Multiplication rule — valid only for independent events
Addition rule — valid only for mutually exclusive events
Worked example
Q6 — marbles: single event and comparing bags
Frequently confused
| Concept A | Concept B | The distinction that matters |
|---|---|---|
| Independent events | Mutually exclusive events | Independent = one outcome doesn't affect the other's probability, and they CAN happen together (use ×); mutually exclusive = the two outcomes physically cannot both occur at once (use +) — mixing these up means adding when you should multiply, or vice versa. |
Mnemonic
AND means multiply (×), OR means add (+)
For combined events, decode the connecting word first: 'AND' chains two conditions that both must happen (multiply their probabilities); 'OR' offers alternatives (add them, provided they can't overlap).
Common mistake
Adding P(A) and P(B) when the question says 'AND' (or multiplying when it says 'OR' between mutually exclusive outcomes).
Consequence: The arithmetic can be perfectly correct but answer the wrong question entirely — full method marks are lost because the operation itself is what's being assessed.
Fix: Circle the connecting word before doing any arithmetic. AND → ×. OR (mutually exclusive) → +.
Key takeaways
- Single event = one action, one outcome recorded.
- Combined event = two or more actions treated as one unit — decode AND/OR before choosing an operation.
- P(A and B) = P(A)×P(B) only for independent events; P(A or B) = P(A)+P(B) only for mutually exclusive events.
- Comparing probabilities always needs a common format (matching decimals or fractions) — never compare raw favourable counts.
Probability with Dice, Coins and Spinners
- These are the 'equally likely outcomes' machines of the course: a fair die has 6 equally likely faces, a fair coin has 2 equally likely sides, a spinner with equal sectors has as many equally likely landing zones as it has sectors.
- 'Fair' and 'equal sectors' are doing real mathematical work in the wording — that assumption is exactly what lets you use n(favourable)/n(total) at all.
- Once wear, weighting, or unequal sector sizes are involved, theoretical probability calculated this way stops being trustworthy — recorded experimental frequencies become the more reliable guide to what's really happening.
- Spinners with unequal-sized sectors are NOT '1 divided by the number of sectors' — you need each sector's angle (or area) as a fraction of the whole circle (360°).
Dice, coin and spinner — the three standard probability generators
Standard fair generators: a six-sided die, a two-sided coin, and an 8-sector spinner with labelled sectors.
Only needed when a spinner's sectors are NOT equal in size
For equal sectors this simplifies back to 1 ÷ number of sectors.
Worked example
Q2 — a worn spinner
Real-world application
Key takeaways
- Dice: 6 equally likely faces; coin: 2 equally likely sides; spinner: equally likely only if sectors are equal in size/angle.
- Unequal-sector spinners need angle/360° (or area/whole), not 1 ÷ number of sectors.
- Wear, weighting or damage break the 'fair' assumption — theoretical probability then stops matching reality, and experimental data becomes the more reliable evidence.
Using Tree Diagrams and Venn Diagrams (Introductory)
- A tree diagram maps combined/sequential events branch by branch: each branch is labelled with its probability, and every complete path from start to end represents one possible combined outcome.
- Rule for reading a tree: multiply the probabilities along a single path (AND — narrowing to one exact sequence) to get that outcome's probability; add the probabilities of different complete paths (OR — separate routes) that all satisfy what you want.
- A Venn diagram shows how two events overlap inside a rectangle representing the universal set (everything possible); circles represent each event, and the overlapping region is the intersection (both happening at once).
- At this level, Venn diagrams are mostly given with actual counts placed in each region; probability is then that region's count divided by the total in the universal set.
- Tree diagram
- A branching diagram showing all possible sequences of outcomes for combined events, with probabilities labelled on each branch.
- Venn diagram
- A diagram using overlapping circles inside a rectangle (universal set) to show which outcomes belong to one event, both events, or neither.
- Intersection (A ∩ B)
- Outcomes belonging to both event A and event B at once — represented by the overlapping region of a Venn diagram.
Tree diagram for two coin tosses
First toss branches into H (1/2) and T (1/2); each of those branches again into H (1/2) and T (1/2), giving four paths each with probability 1/4.
Venn diagram — two overlapping events
Universal set U contains two overlapping events, Set A and Set B, with counts shown in each region.
Choosing the right diagram
| Diagram | Best used when… | What you read off it |
|---|---|---|
| Sample space grid | Two single events combined together, both with a small number of equally likely outcomes (e.g. spinner + die) | Every ordered pair of outcomes at a glance — count favourable cells over total cells |
| Tree diagram | Sequential events (one after another), especially with different numbers of stages or unequal branch probabilities | Multiply along one path for AND; add across separate complete paths for OR |
| Venn diagram | Two events with real overlap in membership (people/items belonging to more than one group) | Count in the intersection, union, or 'neither' region directly, divided by the universal set total |
Worked example
Two coins, tree diagram — 'exactly one head'
Common mistake
On a tree diagram, adding branch probabilities along a single path instead of multiplying (e.g. writing 1/2 + 1/2 for two branches in a row).
Consequence: Produces probabilities that don't shrink as you go deeper into the tree — sometimes even exceeding 1, an immediate red flag, and full method marks are lost.
Fix: Along ONE path = multiply (AND, narrowing down to one exact sequence). Across DIFFERENT complete paths = add (OR, separate routes to the wanted result).
Examiner tip
Key takeaways
- Tree diagrams: multiply along one path (AND), add across separate complete paths giving the wanted result (OR).
- All the path probabilities in a complete tree diagram must sum to 1 — a quick self-check for arithmetic errors.
- Venn diagrams show overlap: intersection = both, union = either/both, complement = neither — always confirm region counts sum to the universal set total.
- Choosing tree vs Venn vs sample space grid depends on whether events are sequential, overlapping-membership, or simply two combined single events.
Introduction to Probability and the Probability Scale
- Probability is a number between 0 and 1 (equivalently 0% to 100%) measuring how likely an event is — 0 means impossible, 1 means certain, 0.5 means an even chance.
- Every probability value can be written as a fraction, decimal or percentage — any format is fine unless the question specifies one, but fractions should always be simplified fully.
- An outcome is one possible result of a trial (e.g. 'rolling a 4'); an event is the outcome or collection of outcomes you're actually interested in (e.g. 'rolling an even number' covers outcomes 2, 4 and 6).
- The basic formula for equally likely outcomes only works because every outcome in the sample space is assumed equally likely — that assumption (a fair die, coin or spinner) is the whole reason the formula is valid.
- Outcome
- A single possible result of a trial.
- Event
- An outcome, or set of outcomes, that satisfies some stated condition.
- Sample space (S)
- The complete list of all possible outcomes of a trial; n(S) is the total count of outcomes in it.
- Probability scale
- A number line from 0 (impossible) to 1 (certain) used to represent and compare the likelihood of events.
The probability scale
0 = impossible, 0.5 = even chance, 1 = certain, with example events positioned along the scale.
Basic probability formula for equally likely outcomes
Only valid when the object generating outcomes is fair/unbiased — every outcome must be equally likely.
Worked example
Q7 — why does 'fair' matter?
Frequently confused
| Concept A | Concept B | The distinction that matters |
|---|---|---|
| Outcome | Event | An outcome is one single specific result (e.g. 'rolling a 4'); an event is often a broader condition satisfied by several outcomes (e.g. 'rolling an even number' = outcomes 2, 4 or 6) — always check whether the question wants one specific result or a whole category. |
Key takeaways
- Probability scale runs 0 (impossible) to 1 (certain); 0.5 is an even chance — 'unlikely'/'likely' sit either side of 0.5.
- P(event) = n(event)/n(S) — valid ONLY when every outcome is equally likely (a fair object).
- Outcome = one specific result; event = the outcome or group of outcomes satisfying the condition asked about.
- Fractions, decimals and percentages are interchangeable formats for the same probability — simplify fractions fully.
Listing Outcomes and Sample Space Diagrams
- Listing outcomes systematically is the foundation skill under everything else: for one die, S = {1, 2, 3, 4, 5, 6} — miss one out and every probability calculated afterwards is wrong.
- For combined events (two dice, spinner + die, coin + die), the sample space grows multiplicatively: n(S) = n(first event) × n(second event) — a 4-sector spinner with a 6-face die gives 4×6 = 24 total equally likely combined outcomes, never 4+6 = 10.
- A sample space diagram (two-way table/grid) is the clearest way to list every combined outcome without missing or repeating any: one event's outcomes go along the top, the other's down the side, and each cell is one ordered pair.
- Once the grid is built, any probability question becomes pure counting: count the cells satisfying the condition, divide by the total number of cells n(S).
Sample space grid: spinner (1–4) × die (1–6)
Spinner outcomes 1–4 across the top, die outcomes 1–6 down the side; each of the 24 cells is one ordered pair (spinner, die).
Worked example
Q3 — spinner and die sample space
Sample space grid: spinner (1–4) × die (1–6) — answer to Q3(a)
| Spinner \ Die | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| 1 | (1,1) | (1,2) | (1,3) | (1,4) | (1,5) | (1,6) |
| 2 | (2,1) | (2,2) | (2,3) | (2,4) | (2,5) | (2,6) |
| 3 | (3,1) | (3,2) | (3,3) | (3,4) | (3,5) | (3,6) |
| 4 | (4,1) | (4,2) | (4,3) | (4,4) | (4,5) | (4,6) |
Common mistake
Adding the number of outcomes from each event instead of multiplying to find the total sample space size (e.g. 4+6=10 instead of 4×6=24 for a spinner and a die).
Consequence: Every subsequent probability calculated from a wrong-sized sample space is wrong, even if the favourable-outcome counting is otherwise perfect — this is a total-method error, not a small slip.
Fix: Total combined outcomes = n(first event) × n(second event), always. Only use addition when combining separate favourable counts WITHIN an already correctly sized sample space.
Examiner tip
Key takeaways
- List systematically: for one die, {1, 2, 3, 4, 5, 6} — missing an outcome breaks every probability calculated afterwards.
- Combined sample space size = n(first) × n(second), never n(first) + n(second).
- A sample space grid turns any combined-event probability question into counting cells over total cells.
- 'Justify' answers need the correct calculated value AND an explicit verdict on whether the given claim is right or wrong.
Quick Reference
Probability and Outcomes — everything on one page
Formulas
- Theoretical probability — equally likely outcomes only
- Experimental probability from recorded data
- A prediction, not guaranteed to match observed data
- Independent combined events only
- Mutually exclusive events only
- Size of a combined sample space — never add the two counts
Key facts
- Probability of an impossible event
- 0
- Probability of a certain event
- 1
- Even chance
- 0.5 = 1/2 = 50%
- Sum of all probabilities in a sample space
- 1
Mnemonics
- AND = ×, OR = +Decode the connecting word before choosing to multiply or add for combined events.
Don't confuse
- Theoretical probability vs Experimental probability: Calculated by reasoning (before trials) vs measured from data (after trials) — they converge only over many trials.
- Independent events vs Mutually exclusive events: Can happen together, don't affect each other (×) vs cannot both happen at once (+).
- Outcome vs Event: One specific result vs a category that can include several outcomes.
Command terms
- Calculate:
- Show the formula, substitute values, give a numeric final answer — usually a simplified fraction or decimal.
- Explain:
- Give a reason tied to the maths or physical situation, not a restated fact.
- Construct:
- Build the actual diagram/table fully and correctly — marked on completeness, not just the final number.
- Justify:
- State a clear verdict (correct/incorrect) AND back it with the correctly calculated value.
- Compare:
- Convert both values to the same format and explicitly state which is bigger/smaller and by how much.
Exam checklist
- Have I identified whether this needs theoretical (formula) or experimental (data) probability?
- Is the sample space size found by multiplying, not adding, event outcome counts?
- Have I decoded AND (×) vs OR (+) correctly for combined events?
- Is my final probability between 0 and 1, and fully simplified if a fraction was requested?
- For 'justify' questions, have I stated a clear verdict, not just a number?