← Learn

Probability Studio

Intermediate → Olympiad

Probability isn't a list of formulas. It is choosing the right picture of the randomness in front of you. Each chapter builds one idea from its definition, puts it to work on problems, and then stretches it on a story.

  1. 1

    Probability as area

    Outcomes, events, and why a probability is a share of one square.

    Representation: The unit square

    Open →
  2. 2

    Counting outcomes

    Before any probability: what exactly are the outcomes?

    Representation: Tables, trees and lists

    Open →
  3. 3

    Independence and multiplication

    Why “and” multiplies, and when it doesn't.

    Representation: Area times area

    Open →
  4. 4

    Complement and “at least one”

    When counting explodes, count what didn't happen.

    Representation: The rest of the square

    Open →
  5. 5

    Exactly k

    One string's chance, times the number of ways to arrange it.

    Representation: Success–failure strings

    Open →
  6. 6

    Conditional probability

    Keep only the outcomes that fit what you saw, and count again.

    Representation: Zoom into what you know

    Open →
  7. 7

    Bayes and total probability

    From causes to evidence and back: what a positive test really tells you.

    Representation: Split by the cause

    Open →
  8. 8

    Symmetry

    Sometimes the answer needs no calculation at all.

    Representation: Random orderings

    Open →
  9. 9

    States

    The present holds everything that matters: walks, races and urns.

    Representation: State diagrams

    Open →
  10. 10

    Expectation and linearity

    Random variables, their average and their spread, and why expected totals add up even when the parts are dependent.

    Representation: Indicators

    Open →
  11. 11

    Conditional expectation

    Expected waiting times, first successes and the coupon collector.

    Representation: State diagrams with values

    Open →
  12. 12

    Chance and data

    From probability to samples: why estimates settle, and why the error shrinks like 1/√n.

    Representation: Repeated experiments

    Open →
★

Final round

Ten problems from the whole studio. Nothing tells you which picture to use.

Open →