- Foundation
- About 17 min
- 5 practice questions
The axiomatic approach to probability
State the three axioms of probability, prove P(φ) = 0, check whether an assignment of probabilities is valid, find the probability of an event by adding its outcomes, and derive P(E) = n(E)/n(S) for equally likely outcomes.
If something looks wrong, please tell us.
In earlier classes probability was defined as favourable outcomes divided by total outcomes. That works only when outcomes are equally likely. A biased coin or a loaded die needs a more general foundation. The axiomatic approach gives a few rules that every probability must obey, and everything else follows from them.
The axioms
Let S be the sample space of a random experiment. A probability P is a rule that assigns to each event E a real number P(E) in [0, 1] such that:
- P(E) ≥ 0 for every event E.
- P(S) = 1.
- If E and F are mutually exclusive, P(E ∪ F) = P(E) + P(F).
Consequence: P(φ) = 0. E and φ are mutually exclusive (they share nothing) and E ∪ φ = E. By axiom 3, P(E) = P(E ∪ φ) = P(E) + P(φ), so P(φ) = 0.
Probabilities of outcomes
If S = {ω₁, ω₂, …, ωₙ}, the axioms require:
- 0 ≤ P(ωᵢ) ≤ 1 for each outcome;
- P(ω₁) + P(ω₂) + … + P(ωₙ) = 1;
- for any event A, P(A) is the sum of P(ωᵢ) over the outcomes ωᵢ in A.
A valid assignment gives each outcome a number between 0 and 1, with all the numbers adding up to 1. The probability of an event is the sum of the probabilities of its outcomes.
Many assignments can be valid for the same sample space. For a coin, P(H) = P(T) = 1/2 is valid, and so is P(H) = 1/4, P(T) = 3/4 (a biased coin). Which one fits a real coin is a question about the coin, not about the axioms.
Worked example 1: which assignments are valid?
S = {ω₁, ω₂, ω₃, ω₄}. Check each row.
| Assignment | ω₁ | ω₂ | ω₃ | ω₄ | Valid? |
|---|---|---|---|---|---|
| (a) | 0.1 | 0.2 | 0.3 | 0.4 | Yes: all in [0, 1], sum 1 |
| (b) | 0.5 | 0.5 | 0 | 0 | Yes: zeros are allowed, sum 1 |
| (c) | 0.6 | 0.3 | 0.2 | −0.1 | No: a negative value |
| (d) | 0.3 | 0.3 | 0.3 | 0.3 | No: sum is 1.2 |
Worked example 2: a biased die
A die is weighted so that P(6) = 0.25 and each of the other faces has probability 0.15 (made-up figures). Check: 0.25 + 5 × 0.15 = 1, so the assignment is valid. Then P(even) = P(2) + P(4) + P(6) = 0.15 + 0.15 + 0.25 = 0.55, and P(a number less than 3) = 0.15 + 0.15 = 0.3.
Equally likely outcomes
Suppose all n outcomes are equally likely, each with probability p. Then p + p + … + p (n times) = 1, so p = 1/n. If an event E contains m of the outcomes, adding gives P(E) = m × (1/n).
For equally likely outcomes: P(E) = n(E)/n(S) = (number of outcomes favourable to E)/(total number of outcomes).
Finding a probability with equally likely outcomes
- List or count S n(S) outcomes, all equally likely
- Identify E Write the event as a subset of S
- Count E n(E) favourable outcomes
- Divide P(E) = n(E)/n(S)
Worked example 3: two dice
Two fair dice are rolled: n(S) = 36. 'Sum is 7' has 6 outcomes, so P = 6/36 = 1/6. 'At least one six': 6 pairs have a six on the first die and 6 on the second, but (6, 6) is counted twice, so there are 11 outcomes and P = 11/36. 'Sum at least 10': sums 10, 11, 12 occur 3, 2 and 1 times, so P = 6/36 = 1/6.
Worked example 4: three coins
Three fair coins are tossed: 8 equally likely outcomes. P(exactly two heads) = 3/8. P(at least one head) = 7/8 (every outcome except TTT). P(all three alike) = P({HHH, TTT}) = 2/8 = 1/4.
Key takeaways
- Axioms: P(E) ≥ 0, P(S) = 1, additivity for mutually exclusive events.
- P(φ) = 0 follows from the axioms.
- Valid assignment: each value in [0, 1], total 1.
- Equally likely outcomes: P(E) = n(E)/n(S).
Check your understanding
-
1. For S = {a, b, c}, which assignment of probabilities is valid?
- P(a) = 0.5, P(b) = 0.5, P(c) = 0.5
- P(a) = 0.7, P(b) = 0.4, P(c) = −0.1
- P(a) = 0.2, P(b) = 0.3, P(c) = 0.5
- P(a) = 1.2, P(b) = 0, P(c) = −0.2
Show answer
Answer: C. Only 0.2, 0.3, 0.5 has every value in [0, 1] and a total of 1. The others have a total of 1.5, a negative value, or a value above 1.
-
2. A biased coin has P(H) = 3P(T). What is P(H)?
- 3/4
- 1/4
- 1/3
- 2/3
Show answer
Answer: A. P(H) + P(T) = 1 and P(H) = 3P(T), so 4P(T) = 1, P(T) = 1/4 and P(H) = 3/4.
-
3. Two fair dice are rolled. What is the probability that the two numbers are equal?
- 1/36
- 1/12
- 1/3
- 1/6
Show answer
Answer: D. The doublets (1, 1), …, (6, 6) are 6 of the 36 outcomes, so the probability is 6/36 = 1/6.
-
4. A die is weighted so that the probabilities of 1, 2, 3, 4, 5, 6 are 0.1, 0.1, 0.1, 0.2, 0.2, 0.3. What is P(a number greater than 3)?
- 0.5
- 0.7
- 0.3
- 0.6
Show answer
Answer: B. Add the probabilities of 4, 5 and 6: 0.2 + 0.2 + 0.3 = 0.7.
-
5. Assertion (A): P(φ) = 0 for every probability assignment. Reason (R): E and φ are mutually exclusive with E ∪ φ = E, so P(E) = P(E) + P(φ).
- Both A and R are true, and R is the correct explanation of A
- Both A and R are true, but R is not the correct explanation of A
- A is true, but R is false
- A is false, but R is true
Show answer
Answer: A. Both are true; R is the proof of A from axiom 3.
Sign in to keep track of the lessons you finish.