Law of Total Probability
What you'll learn: How to break down complex probability questions into simpler, non-overlapping scenarios that together cover all possibilities.
The Core Idea
Imagine you need to find the probability of an event (like "user clicks the ad"), but the situation is complicated. The Law of Total Probability lets you split the entire sample space into separate, non-overlapping scenarios (called a partition), calculate the probability within each scenario, then add them up.
Think of it like organizing your closet: every piece of clothing goes into exactly one drawer—shirts, pants, or socks. If you want to know "what's the chance I grab something blue?" you check each drawer separately, then combine the results.
The Formula
If events B₁, B₂, ..., Bₙ partition the sample space (they're mutually exclusive and cover everything), then for any event A:
P(A) = P(A | B₁)·P(B₁) + P(A | B₂)·P(B₂) + ... + P(A | Bₙ)·P(Bₙ)
Each term uses the multiplication rule for conditional probability you've already learned.
Example in Plain English
Suppose 60% of website visitors use mobile, 40% use desktop. On mobile, 5% click an ad; on desktop, 10% click. What's the overall click probability?
- Partition: {Mobile, Desktop} (mutually exclusive, cover all users)
- P(Click) = P(Click | Mobile)·P(Mobile) + P(Click | Desktop)·P(Desktop)
- P(Click) = (0.05)(0.60) + (0.10)(0.40) = 0.03 + 0.04 = 0.07 or 7%
You broke a complex question into two simpler conditional probabilities.
Key Takeaway: The Law of Total Probability lets you compute probabilities by dividing the sample space into mutually exclusive, exhaustive scenarios and combining their weighted contributions.