Independence of Events in Real-World Contexts

A fair coin has no memory. Six heads in a row don't change the next flip's odds, because each flip is a separate, independent trial. The physical process that produces heads or tails doesn't know or care what happened before. This is the intuition behind independence: past outcomes carry zero information about future ones.

The mechanism behind this is a conditional distribution that refuses to move. Take a group of people, some fraction of whom own cars. If you zoom in on just the bike owners and ask what fraction of them own cars, independence means you get the same fraction as the whole group. Conditioning on bike ownership changed nothing. If instead that fraction shifts up or down inside the bike-owning group, the two events are linked — owning a bike tells you something real about owning a car, and they are dependent.

This idea exposes a classic trap: five reds in a row on a roulette wheel feels like pressure building toward black, but each spin is independent, so the chance of black stays exactly what it always was. That gut feeling is the gambler's fallacy. On the far end of the spectrum sit events that can never happen together at all — mutually exclusive events, like rolling a 2 and a 3 on one die. Those are about as dependent as two events can get: knowing one happened guarantees the other didn't, which is the opposite of independence. Keeping these two ideas separate — unrelated versus mutually exclusive — matters when counting outcomes or building probability trees.