A working probability deck for SOA Exams

Probability is a useful place to practise actuarial flashcards because a small change in wording can change the calculation. A question may ask for a conditional probability, a tail probability, or an expected value; recognising which quantity is wanted comes before substituting numbers. The deck below focuses on foundational probability, not a claim about the syllabus of any particular SOA exam. Check the Society of Actuaries’ official exam page and syllabus for the material required for your sitting.

Use the deck as a drill, not as a formula sheet to skim. Read the question, say or write the answer from memory, then turn the card. Grade the attempt according to how readily you recalled the correct fact: a forgotten or incorrect answer needs another look soon; a correct but effortful answer needs review; an immediate, accurate answer can wait longer. When a card includes a formula, state what its symbols mean and any condition needed to use it.

MySummaries turns notes on probability into a deck like this:

Cards — Probability foundations18 due

For an event A, what is P(Aᶜ)?

1 − P(A).

All 18 cards
For an event A, what is P(Aᶜ)?1 − P(A).
Define P(A | B), where P(B) > 0.P(A ∩ B) / P(B).
What condition makes A and B independent?P(A ∩ B) = P(A)P(B).
State the law of total probability for a partition B₁,…,Bₙ, with P(Bᵢ) > 0.P(A) = Σᵢ P(A | Bᵢ)P(Bᵢ).
State Bayes’ theorem for P(B) > 0.P(A | B) = P(B | A)P(A) / P(B).
How do you calculate E[X] for a discrete random variable with values x and probabilities p(x)?E[X] = Σₓ x p(x), when the sum exists.
Give the variance identity in terms of the first and second moments.Var(X) = E[X²] − (E[X])², when the moments exist.
Define covariance using expectations.Cov(X,Y) = E[XY] − E[X]E[Y], when the expectations exist.
Give the probability mass function of X ~ Binomial(n,p).P(X = k) = C(n,k)pᵏ(1 − p)ⁿ⁻ᵏ for k = 0,…,n.
For X ~ Binomial(n,p), what are E[X] and Var(X)?E[X] = np; Var(X) = np(1 − p).
Give the probability mass function of X ~ Poisson(λ), λ > 0.P(X = k) = e⁻λ λᵏ / k! for k = 0,1,2,… .
For X ~ Poisson(λ), what are its mean and variance?E[X] = Var(X) = λ.
For T ~ Exponential(λ), λ > 0, give the survival probability P(T > t), t ≥ 0.P(T > t) = e⁻λᵗ.
If Z is standard normal, what is P(Z ≤ z) written using the standard normal CDF?Φ(z), where Φ is the CDF of N(0,1).
How do you standardise X ~ N(μ,σ²), with σ > 0?Z = (X − μ)/σ, so Z ~ N(0,1).
For independent variables X and Y with finite variances, what is Var(X + Y)?Var(X + Y) = Var(X) + Var(Y).
For independent variables X₁,…,Xₙ with common mean μ and variance σ², what are the mean and variance of their sample mean?For X̄ = (ΣᵢXᵢ)/n, E[X̄] = μ and Var(X̄) = σ²/n.
For independent Bernoulli trials with success probability p, what is the expected number of trials until the first success, counting the successful trial?For the geometric distribution on 1,2,…, E[T] = 1/p.
An 18-card probability deck with the first card open and four recall grades.

That is a MySummaries deck, filled with SOA Exams material. Yours is written from your own notes. Start free

How to work the cards

Take one card at a time. Before revealing the answer, identify the object in the question: is it an event, a random variable, or a sample mean? Then recall the relationship and its conditions. For example, the conditional probability card requires P(B) > 0; the variance-of-a-sum card requires independence. Leaving out a condition can turn a memorised formula into a wrong answer.

After revealing the answer, grade what you actually retrieved, not what looks familiar now. If you remembered the binomial mean but not its variance, that card is not fully secure. If you gave the right formula but used the wrong support for k, review it again. On calculation questions, practise adding one check: probabilities must be between zero and one, a variance cannot be negative, and a reported expected count should make sense for the setup.

Must not miss core

A short core list is useful before timed practice. It should hold the relationships you repeatedly need to retrieve, plus the conditions that stop you applying them carelessly. The deck’s core can be kept beside your worked problems as a final recall check.

The probability board’s Must not miss core is:

Key pointsProbability foundations
Conditional probability: P(A | B) = P(A ∩ B)/P(B), provided P(B) > 0.
Independence: P(A ∩ B) = P(A)P(B); for independent variables with finite variances, variances add.
For X ~ Binomial(n,p): E[X] = np and Var(X) = np(1 − p); support is k = 0,…,n.
For X ~ Poisson(λ): P(X = k) = e⁻λ λᵏ/k! and E[X] = Var(X) = λ.
The four probability relationships selected as the deck’s Must not miss core.

Turn missed marks into another recall

A flashcard is most useful when it responds to a specific mistake. If you lose a mark because you treated dependent variables as independent, the follow-up should test the missing condition, not merely repeat a long derivation. If you misread “given that”, make yourself identify the conditioning event before calculating.

For every missed question, record the precise failure: definition, setup, algebra, arithmetic, or interpretation. Then make a short card that asks for the correction. Keep separate cards for separate facts; a card that asks for a theorem, its proof, three exceptions and an application is difficult to grade honestly. Return to the original question later and solve it again, including a check that the answer has the correct sign, scale and probability range.

When the same conditional-probability error costs a mark twice, the remediation tray might surface this prompt:

Remediation tray

You lost this mark twice: if a question asks for P(A | B), what is the denominator, and what condition must hold?

Add cardDismiss
A targeted reminder for a repeated error in conditional probability.

Use the deck alongside timed problems

Recall alone does not prove that you can set up a calculation under time pressure. After a short card session, use a representative problem from your own study material. First underline the requested quantity and list the givens. Choose the method before calculating: for example, a partition suggests total probability, while a specified binomial model points to its mass function. Write the assumptions you rely on, substitute carefully, and check the result.

For a probability, check the bounds. For a variance, check non-negativity. For an expected value, check whether it is on a plausible scale. If an answer is a tail probability, make sure you have not returned the complementary event. Those checks catch errors that formula recall alone will not.

Keep the deck tied to the material for your actual exam. SOA exams are typically computer-based, proctored assessments with time-pressured objective questions; some exams also include written-answer components. The exact content and requirements vary, so verify the official syllabus for your exam rather than assuming this probability set is complete. Add cards from your own notes and worked solutions when they cover a required definition, condition or calculation you have not practised.

How MySummaries helps

MySummaries builds a revision board from your own material and can turn its sections into spaced-repetition flashcards. For SOA study, that gives you a way to keep formula conditions, worked examples and error checks together, then practise recalling them between timed problems. Open MySummaries.