Source

Adapted from my own handwritten and hand-typed notes from Year 1 to Year 3 of my EEE degree

Conditional probability is the probability of given that has already happened. Knowing shrinks the sample space down to .

Definition

and in the other direction . In general .

Multiplication rule

Rearranging the definition gives the probability that both events happen:

For more events, keep conditioning on everything that came before:

This is the natural tool for draws without replacement.

Worked example (club members)

75% of members play golf (G), 65% play tennis (T), 50% play both.

Solution:

Worked example (multiplication rule)

A lot of 1000 bulbs has 20 defective. Draw 2 without replacement. Let = first is defective, = second is defective.

After the first defective bulb is removed, only 19 defective bulbs remain among 999. When the conditioning event is split into several cases, this rule extends to Total Probability and Bayes Theorem.