Expected Value: Why a 10% Chance Is Not 10% of a Win

A Super Mega Bowl guide

You have a 10 percent chance to win a prize. So do you get a tenth of the prize? Of course not. You either win the whole thing or you walk away with nothing. That gap between the odds and what actually lands in your hands is one of the most useful ideas in probability, and one of the most misread.

Outcomes are whole, odds are not

Most random events are all or nothing. You win the raffle or you do not. The bet pays or it does not. The coin is heads or tails. There is no partial win, no 10 percent of a prize, no half a heads. Each trial delivers one whole outcome from a small set of possibilities.

Probability describes how those possibilities are distributed. It is not an amount you receive. So a 10 percent chance does not hand you 10 percent of anything. It means that if the same situation played out many times, you would win about one time in ten and lose the other nine.

What expected value actually is

Expected value is the bridge between the odds and the outcomes. It is the average result you would get per trial if you repeated the bet a huge number of times. You find it by multiplying each outcome by its probability and adding the pieces together.

Say a ticket wins 10 dollars with probability 0.10 and wins nothing with probability 0.90. The expected value is 0.10 times 10 dollars plus 0.90 times 0 dollars, which is 1 dollar. On any single ticket you get either 10 dollars or nothing. You never get exactly 1 dollar. Yet 1 dollar is the honest long-run value of the ticket, and it is exactly what you should compare against its price.

The average is often impossible on a single trial

Here is the part that trips people up. The expected value is frequently a number that can never actually occur on one try.

Roll a fair six-sided die and the expected value is 3.5, but you will never roll a 3.5. It is the average of 1 through 6, not a face on the die.
A single lottery ticket has an expected value of a few cents, yet it pays either a jackpot or nothing, never a few cents.

Expected value is a long-run average, not a forecast of the next result. Treating it as what will happen on a single try is the core mistake.

Why the average still matters

If you never actually get the expected value on a single trial, why care about it? Because of the law of large numbers. Over many trials the average of your real, whole outcomes converges toward the expected value. One roll is unpredictable. Ten thousand rolls average almost exactly 3.5. Expected value is not what happens once. It is what the pile of results averages to, and that average becomes reliable as the pile grows.

Where it decides winners and losers

Casino games. Every bet has a negative expected value for the player, which is the house edge. You still win or lose each bet as a whole, but the average per bet is a small loss, and over a night the law of large numbers turns that average into a near-certain result. This is why the house wins if it keeps you playing. See roulette and craps for the numbers.
Lottery tickets. The expected value sits far below the price. That does not make buying one for fun irrational, only unprofitable on average.
Insurance. You pay a premium with a negative expected value to you on purpose, trading a small certain cost for protection against a rare, large loss.

The two mistakes to avoid

Treating a probability as a fractional payout. A 10 percent chance is not 10 percent of a prize in hand. It is a whole prize one time in ten.
Treating expected value as a guarantee. A positive expected value does not promise a win, and a negative one does not promise a loss today. It tells you which way results lean over the long run, not what the next trial holds.

See it for yourself → Roll a die many times and watch the running average. Every roll is a whole number from 1 to 6, yet the average creeps toward 3.5, a value no single roll can ever produce.

The takeaway

Odds and outcomes are two different things. Any one trial is all or nothing, a whole win or a whole loss. Expected value is the long-run average those whole results settle into, often a number you can never get on a single try. Use it to judge whether a bet is worth making, and never mistake it for what will happen next time.