Expected Value Explained: the Math Behind Smart Bets and Big Decisions

Most people make decisions under uncertainty by gut feel. They take the bet that feels lucky, buy the warranty because losing the phone would hurt, or chase the startup with the biggest possible payday. Expected value gives you a single number that cuts through the feeling and tells you what a choice is worth on average. It is one of the most useful ideas in probability, and you can compute it in your head once the pattern clicks.

What Expected Value Actually Means

Expected value is the probability-weighted average of every outcome a random event can produce. You list each possible result, multiply it by how likely it is, and add up those products. The formula is short:

EV = Σ (probability × value)

Read that as "sum, over all outcomes, of probability times value." A coin flip that pays you $10 on heads and $0 on tails has an expected value of (0.5 × $10) + (0.5 × $0) = $5. You will never actually win $5 on a single flip — you get $10 or nothing — but $5 is the long-run average per flip if you played forever.

That gap between the number and any single result trips people up. Expected value is not a prediction of what happens next. It is the center of gravity of the whole distribution, the figure your average converges to as you repeat the experiment again and again.

A Worked Dice Example

Let's make it concrete with a game I can play on the back of a napkin. You roll one fair six-sided die. If you roll a 6 you win $20. If you roll a 1, you lose $10. Anything else (2, 3, 4, or 5) and nothing changes hands. Should you pay $2 to play?

Each face has probability 1/6. The outcomes in dollars are:

Roll a 6: +$20, probability 1/6
Roll a 1: −$10, probability 1/6
Roll 2–5: $0, probability 4/6

Plug into the formula:

EV = (1/6 × 20) + (1/6 × −10) + (4/6 × 0)
   = 3.333 − 1.667 + 0
   = +$1.67 per roll

The game is worth $1.67 before you pay to play. Since the entry fee is $2, your net expected value is $1.67 − $2 = −$0.33. On average you lose 33 cents every time you play, so over many rounds you should walk away. Drop the fee to $1 and the math flips positive — now you should play all day. You can recreate this exact table in the expected value calculator, and if you want to sanity-check the per-face odds first, the dice probability calculator lays them out cleanly.

Why a Positive EV Bet Can Still Lose

Here is the part that separates people who understand probability from people who just memorized the formula: a positive expected value does not guarantee a profit in the short run. It only promises that the average wins out eventually.

Take the $1-entry version of the dice game above, where each roll is worth a positive $0.67 to you. Sit down for ten rolls and you might never see a 6. You could roll three 1s and lose $33 against your $10 in fees, and walk away down badly — even though every single roll was a "good" bet. The expected value was real; variance just hadn't averaged out yet.

This is why casinos love volume. Each spin of American roulette has a tiny negative expected value for the player, about −$0.053 on a $1 bet, because two green zeros tilt the wheel toward the house. Any individual gambler can win big on a lucky night. But the casino isn't playing one night — it's playing millions of spins, and across that many trials the law of large numbers drags the average down to that −5.3% edge with brutal reliability. Positive EV is your edge; sample size is what lets the edge show up.

The lesson runs both ways. A positive-EV bet you can only place once, with money you can't afford to lose, is genuinely risky even though the math favors you. Expected value tells you the direction; the variance and the standard deviation tell you how bumpy the ride is. You need both.

Decisions Beyond the Casino: Insurance and Business

The same arithmetic prices insurance and steers business calls. Insurers are professional expected-value machines. Suppose an extended warranty pays out $4,800 with probability 0.03 and collects a $200 premium the other 97% of the time. From the insurer's seat the net per policy is:

EV = (0.97 × +200) + (0.03 × −4,800)
   = 194 − 144
   = +$50 per policy

That $50 is the margin baked into the price — enough to cover overhead and turn a profit across thousands of policies. As the customer, you're paying a $200 premium to dodge a 3% chance of a $144 expected loss, which is a negative-EV trade for you. You take it anyway when the rare loss would be ruinous, which is exactly the point of insurance: it isn't supposed to be a good bet on average, it's supposed to cap your downside.

Business decisions work the same way. Project A might return $50,000 with probability 0.25 and nothing otherwise, giving an expected value of $12,500. Project B is a sure $10,000. On expected value alone, A wins. But A is far more volatile — most of the time it returns zero — so a founder who needs reliable cash this quarter might rationally take B. Expected value ranks the options; your tolerance for variance breaks the tie.

Putting It to Work

When I size up any choice with an uncertain payoff, I run the same three-step routine. First, list every distinct outcome and tag it with a probability — and force those probabilities to sum to 1, because if they only reach 0.9 I've forgotten a case. Second, compute the expected value to see which way the math leans. Third, glance at the spread: a high EV wrapped in enormous variance is a different animal from a modest, steady one, and only one of them belongs in a single high-stakes shot.

You don't need to do the multiplication by hand. Lay your outcomes and probabilities into the expected value calculator and it returns EV, variance, and standard deviation together, with fraction support so 1/6 stays exact. The number won't make the decision for you — that still takes judgment about how much risk you can stomach — but it replaces a vague feeling with a figure you can defend.

Expected value is the closest thing decision-making has to a north star under uncertainty. It rewards patience, punishes wishful thinking, and quietly explains why the house always wins and why insurance always sells. Learn to compute it, respect the variance around it, and you'll make sharper bets at the table, in the market, and in the boardroom.

Made by Toolora · Updated 2026-06-13