The addition rule combines probabilities across alternative events. If events A and B are mutually exclusive, they cannot occur together, so P(A ∪ B) = P(A) + P(B). The union A ∪ B means that at least one event occurs; mutual exclusivity prevents double counting. section 2.3.1
Rolling a die
For a fair six-sided die, rolling a 2 and rolling a 5 are mutually exclusive. The probability of either outcome is 1/6 + 1/6 = 1/3. section 2.3.1
Mutual exclusivity does not by itself make probabilities sum to 1. They sum to 1 when the events are also collectively exhaustive: together they account for every possible outcome. For example, conditional probabilities of 0.3 for spam and 0.7 for not spam sum to 1 because the two classes partition the possibilities. section 2.3.1
When events can overlap, simply adding their probabilities counts the overlap twice. The general rule corrects this: P(A ∪ B) = P(A) + P(B) − P(A ∩ B), where A ∩ B is the event that both occur. section 2.3.1
The rule supports probability distributions across competing outcomes in clinical diagnosis and classification. In logistic regression and classification random forests, mutually exclusive and exhaustive classes have probabilities that sum to 1. Alongside the multiplication and complement rules, the addition rule also underpins Bayesian inference. section 2.3.1 section 2.4.2 section 2.2.3

