Suppose a factory buys parts from three suppliers, each with a different defect rate. How do you find the overall probability that a randomly picked part is defective, without knowing which supplier it came from? This is exactly what the theorem of total probability answers — and it’s the essential stepping stone to Bayes’ theorem, which comes right after. Understanding partitions and total probability thoroughly makes Bayes’ theorem feel like a natural next step rather than an intimidating new formula.

The Theorem of Total Probability

Let {E₁, E₂, …, Eₙ} be a partition of the sample space S, where each Ei has nonzero probability. Let A be any event associated with S. Then:

P(A) = P(E₁)·P(A|E₁) + P(E₂)·P(A|E₂) + … + P(Eₙ)·P(A|Eₙ) = Σ P(Ej)·P(A|Ej)

In words: to find the overall probability of A, break it down by every possible “cause” or “scenario” Ei, multiply each scenario’s probability by the probability of A occurring within that scenario, and add everything up.

Proof

Since {E₁, E₂, …, Eₙ} partitions S, we have S = E₁ ∪ E₂ ∪ … ∪ Eₙ with all Ei pairwise disjoint. For any event A:

A = A ∩ S = A ∩ (E₁ ∪ E₂ ∪ … ∪ Eₙ) = (A ∩ E₁) ∪ (A ∩ E₂) ∪ … ∪ (A ∩ Eₙ)

Because the Ei’s are pairwise disjoint, the sets (A ∩ Ei) are too. By the addition rule for disjoint events:

P(A) = P(A ∩ E₁) + P(A ∩ E₂) + … + P(A ∩ Eₙ)

Applying the multiplication theorem to each term, P(A ∩ Ei) = P(Ei) · P(A|Ei), giving:

P(A) = P(E₁)·P(A|E₁) + P(E₂)·P(A|E₂) + … + P(Eₙ)·P(A|Eₙ)

Solved Examples

Example 1 — Construction Job and Strike Probability

A person has taken up a construction job. The probability of a strike is 0.65. Without a strike, the job finishes on time with probability 0.80; with a strike, it still finishes on time with probability 0.32. Find the overall probability that the job finishes on time.

Let A: “job completed on time,” B: “there is a strike.”

P(B) = 0.65, so P(B′) = 0.35. P(A|B) = 0.32, P(A|B′) = 0.80.

Since {B, B′} forms a partition of S:

P(A) = P(B)·P(A|B) + P(B′)·P(A|B′) = (0.65 × 0.32) + (0.35 × 0.80) = 0.208 + 0.28 = 0.488

So there’s about a 48.8% chance the job finishes on time, averaged over both scenarios.

Example 2 — Bolts From Three Machines

A factory has three machines A, B, and C producing 25%, 35%, and 40% of all bolts, with defect rates of 5%, 4%, and 2% respectively. Find the overall probability that a randomly selected bolt is defective.

Let B₁, B₂, B₃ denote “bolt made by A/B/C” and E denote “the bolt is defective.”

P(B₁) = 0.25, P(B₂) = 0.35, P(B₃) = 0.40; P(E|B₁) = 0.05, P(E|B₂) = 0.04, P(E|B₃) = 0.02

P(E) = (0.25 × 0.05) + (0.35 × 0.04) + (0.40 × 0.02) = 0.0125 + 0.014 + 0.008 = 0.0345

So roughly 3.45% of all bolts across the factory are defective.

Example 3 — Drawing From Two Bags

Bag I has 3 red and 4 black balls; Bag II has 5 red and 6 black balls. A bag is chosen at random, and one ball is drawn from it. Find the probability that the ball drawn is red.

Let E₁: “Bag I chosen,” E₂: “Bag II chosen,” A: “ball drawn is red.” P(E₁) = P(E₂) = 1/2, P(A|E₁) = 3/7, P(A|E₂) = 5/11.

P(A) = (1/2)(3/7) + (1/2)(5/11) = 3/14 + 5/22 = 33/154 + 35/154 = 68/154 = 34/77

Why This Theorem Matters Beyond the Exam

The theorem of total probability formalises something we do intuitively — weighing different scenarios by how likely each is, then combining outcomes. Insurance companies use it to price policies across risk categories; quality control teams use it to estimate defect rates across production lines; and it’s precisely the missing piece needed to “reverse” a conditional probability question, which is what Bayes’ theorem does next.

Common Mistakes Students Make

  1. Forgetting to check that the events truly partition S. If the Ei’s overlap or don’t cover the whole sample space, the formula doesn’t apply.

  2. Missing a scenario. In multi-supplier or multi-machine problems, make sure every probability adds up to 1 across all partition events.

  3. Confusing P(Ei) with P(A|Ei). P(Ei) is how likely the scenario is; P(A|Ei) is how likely A is within that scenario — mixing these up flips the whole calculation.

Ready for the Next Step — Bayes’ Theorem?

Total probability is the launchpad for Bayes’ theorem, one of the highest-weightage topics in CBSE boards, CUET, and IPMAT. Practice AI-generated questions on partitions and total probability with instant solutions with PrepCoin packs from ₹59 on ChampionsPrep — built for CBSE and Maharashtra Board Commerce students.

Theorem of Total Probability Class 12: Partition, Formula and Examples podcast artwork
Audio LessonClass 12

Theorem of Total Probability Class 12: Partition, Formula and Examples

ChampionsPrep MathematicsEpisode 1356:45

Subscribe in any podcast app:

Frequently Asked Questions

What is a partition of a sample space? +

A set of events that are pairwise disjoint, together cover the entire sample space, and each have nonzero probability.

What does the theorem of total probability state? +

For a partition of the sample space and any event A, P(A) equals the sum over all partition events of P(Ej) times P(A given Ej), meaning the overall probability of A is the weighted sum of its probability within each scenario.

How is the theorem of total probability proved? +

By writing A as the disjoint union of A intersected with each partition event, applying the addition rule for disjoint events, and then applying the multiplication theorem to each term.

Why must the partition events be pairwise disjoint? +

So that the intersection terms with A do not overlap, allowing the simple addition rule to be applied without subtracting any intersections when summing up the total probability.

How is the theorem of total probability related to Bayes' theorem? +

The denominator of Bayes' theorem is calculated using the theorem of total probability, since it is literally the overall probability of the event expanded across the same partition used in Bayes' formula.