Random variables and expected value

A random variable is a number attached to a random outcome — the total on two dice, tomorrow's rainfall, the payout on a bet. Once outcomes have numbers, you can average them, and that average is the most useful number in probability.

Expected value: the long-run average

Line up every value a random variable X can take, weight each by its probability, and add:

$E[X] = \sum_i x_i \, P(x_i)$

The expected value is the balance point of the distribution — where it would tip if the bars were weights on a ruler. Notice it needn't be a value you can actually land on: a fair die has E[X] = 3.5, though you'll never roll a 3.5.

"Expected" means average, not likely

E[X] is the average over many repetitions, not the most probable single outcome. A lottery ticket's expected value might be −$0.50 — you don't lose exactly 50 cents, but across millions of tickets that's the average loss per ticket. Read "expected" as "long-run average," and the idea stops being confusing.

Fairness and decisions

Expected value is the yardstick for a fair game and a rational bet:

• A game is fair when its expected value is 0 — neither side profits long-run.
• A bet is worth taking (on average) when the expected payoff beats the cost.

The house edge in every casino game is just a small negative expected value for the player, applied a few million times.

Why this matters

Expected value turns uncertainty into a single number you can compare and decide on. Insurance premiums, option prices, and A/B-test payoffs are all expected-value calculations. It's also the bridge to the next idea: the distributions whose expected values and spreads we can write down in a formula.