Expected Value Explained

By 4Football Staff · November 1, 2026 · 9 min read

What expected value means, how to calculate it from odds and a probability, why a winning bet can still be a bad one, and where the same idea shows up in football strategy.

Suppose someone offers you a deal: flip a coin, and if it lands heads you win $120, and if it lands tails you lose $100. You will only get to flip once. Do you take it? Most people say yes, and they are right to, even though half the time they will lose. The deal is good because the average result, if you could repeat it again and again, is a profit.

That average has a name: expected value, usually shortened to EV. It is the single most useful idea for thinking clearly about any uncertain choice, from a coach deciding whether to go for it on fourth down to a fan trying to understand why sportsbooks make money. This article explains what expected value is, how to calculate it, how it connects to betting odds, where it misleads, and how the same logic runs through football strategy. It is educational only. It does not recommend any wager, and sports betting is legal only in some places and only for adults of legal age.

What expected value means

Expected value is a weighted average. Take every possible outcome of a decision, multiply each by how likely it is, and add the results. It tells you what you would earn per attempt, on average, if you could make the same decision a very large number of times under identical conditions.

The formula for a simple win-or-lose bet is short:

EV = (chance of winning × amount won) − (chance of losing × amount risked)

Using the coin deal above, the chance of heads is one half. So EV = (0.5 × 120) − (0.5 × 100) = 60 − 50 = +10. On average you gain ten dollars per flip. That does not mean any one flip pays ten dollars; each flip pays either +120 or −100. The +10 is the long-run average.

Three ideas hang on that definition.

• EV is not a prediction of the next result. It describes the average over many repetitions, not what will happen this time.
• EV can be positive, zero or negative. Positive means the decision tends to pay, zero means it breaks even, and negative means it tends to lose.
• EV depends entirely on the probability you plug in. If your estimate of the chance is wrong, the EV you calculate is wrong, however carefully you do the arithmetic.

Turning betting odds into a calculation

A sportsbook price tells you two things: how much you win if the bet succeeds and how much you must risk. The three common formats (American, decimal and fractional) all carry the same information. Here we only need the basic arithmetic, using American odds.

A standard point-spread bet

Spread and total bets are commonly priced at -110, meaning you risk 110 to win 100. Suppose, hypothetically, the true chance that your side covers is exactly 50 percent. Then:

EV = (0.50 × 100) − (0.50 × 110) = 50 − 55 = −5

You lose five units on average for every bet, which is about 4.5 percent of the 110 you risk. A coin-flip bet at this price loses money in the long run. That is not a trick. It is the sportsbook's margin, the vig or juice, built into the price. Our article on understanding the vig explains it further.

The break-even point

You can use the same formula in reverse. Ask what chance of winning makes the EV exactly zero at -110. Setting (p × 100) − ((1 − p) × 110) = 0 gives p = 110 / 210, or about 52.4 percent. A bettor must win more than roughly 52.4 percent of such bets to come out ahead.

The break-even chance for any price is simply the amount risked divided by the total of amount risked plus amount won. That is also what the price implies as a probability, a topic our article on implied probability explores.

A table of common prices

Read the table as the hurdle your estimate has to clear. If you believe a +300 outcome has a better than 25 percent chance, the bet has positive expected value by your estimate. If you think it is below 25 percent, it does not, no matter how exciting the payout looks.

A worked example, step by step

All numbers here are made up for illustration.

Imagine a game where a sportsbook offers a team at +150 to win outright. Before looking at anything else, calculate the break-even chance: 100 divided by 250 is 40 percent.

Now suppose you study the matchup and estimate the team wins 45 percent of the time. Then:

EV = (0.45 × 150) − (0.55 × 100) = 67.5 − 55 = +12.5

That is 12.5 units of profit per 100 risked, on average, if your 45 percent estimate is right. But if the true chance is 38 percent, the same bet has EV = (0.38 × 150) − (0.62 × 100) = 57 − 62 = −5. A seven-point error in your probability flips a good bet into a bad one.

This illustrates a central difficulty. The arithmetic is easy. The estimate is hard. Football outcomes are noisy, and the book's price already contains many people's best estimate. To have positive expected value you must know something the price does not reflect, and be able to quantify it.

Why a good bet can lose and a bad bet can win

Expected value is about the average, so any single result can go the other way.

Return to the +12.5 example. With a 45 percent chance of winning, that bet loses 55 percent of the time. A bettor who made this good wager and lost has not made a mistake. A bettor who took a bad wager, such as a -5 EV bet, and won also has not shown skill. Judging a decision only by its result is called results-oriented thinking, and it is the most common error in sports and in life.

The way out is to separate decision quality from outcome. A decision should be judged by the information available when it was made, not by the dice afterward. Our article on sample size and small edges shows how many repetitions it takes before outcomes start to reveal decision quality, and the answer is far more than a season or two.

Where expected value misleads

EV is a powerful tool, but it has limits that careful readers should know.

It assumes you can repeat the choice

The average only appears if you can make the choice many times. A decision you can make once, with stakes that matter to you, deserves a different kind of care. A bet that is positive EV but would cost you rent money if it lost is not a good decision for you, however favorable the average.

It ignores how much you risk

A small positive EV on a tiny bet gains little. The same EV on a large bet comes with a large chance of an uncomfortable loss before the average arrives. Choosing the size of a wager is a separate skill, which our guide to bankroll management covers.

It depends on an uncertain probability

In casino games the odds can be calculated exactly. In football the probability is an estimate, and estimates of the same game vary from person to person. A positive EV from your own estimate is a hypothesis, not a fact. People tend to overestimate their own edge, particularly when they have a favorite team or a recent success.

It treats all money as equal

Expected value in dollars assumes losing a dollar hurts exactly as much as winning a dollar helps. For most people, a loss of a large sum hurts more than a gain of the same amount helps. This is why insurance, which has negative expected value in dollars for the buyer, can still be sensible. The same logic is why a bet that is fair in dollar terms may still be a poor idea when the downside matters a lot to you.

Expected value across a parlay

A parlay combines several bets into one, and the payout is larger because every leg must win. Parlays are a good test of the concept.

Imagine two legs, each with a genuine 50 percent chance, and suppose each is priced fairly with no margin. The chance both win is 0.5 × 0.5 = 25 percent. A fair payout on a 25 percent chance is 3 to 1, since the break-even calculation is 100 / (100 + 300) = 25 percent. If the book instead pays 2.6 to 1, the EV per 100 risked is:

EV = (0.25 × 260) − (0.75 × 100) = 65 − 75 = −10

The shortfall comes from the margin being applied to a combined bet, so the cost of the margin grows as legs are added. The lesson here is simple: the bigger and more exciting the payout, the more carefully you should look at what the true chance is.

Expected value outside of betting

EV is not just a betting concept. It is how modern football strategy thinks about decisions, and it is a good way to see why the idea matters.

Fourth-down choices

A coach facing fourth and short can punt, kick a field goal or go for it. Each option has a set of outcomes with different probabilities and different results for points and field position. Analysts translate those into an expected number of points or a chance of winning, then compare the options. The choice with the highest expected value is not guaranteed to work. It is simply the one that tends to do best across many similar situations. Coaches also weigh things the model does not capture, such as the state of their defense or the weather.

Two-point conversions

Here is a clean hypothetical. A team can kick an extra point, which succeeds with probability k, for one point, or try a two-point conversion, which succeeds with probability p, for two points. In raw points, going for two is better when 2p is greater than k. If the kick succeeds nearly every time, the two-point try needs to succeed nearly half the time to match it. The rules about where the ball is placed and how often each succeeds vary by level, so check the current rules for the league you follow.

Fantasy football

Every start-or-sit decision is a small expected-value problem. A player with a high ceiling and low floor may have the same average projection as a steady one. Which you start depends on whether you need the average or the chance of a big week, a topic our fantasy section touches on.

Common mistakes with expected value

1. Confusing EV with odds. A bet with long odds is not necessarily positive EV. It is positive only if the true chance beats the break-even chance.
2. Treating a projection as a probability. An estimate of how many points a team will score is not the same as the chance it wins.
3. Ignoring the margin. Doing the math without including the vig makes every bet look better than it is.
4. Rounding the probability favorably. Moving an estimate from 51 to 53 percent, a small and self-serving change, can turn a loser into a winner on paper.
5. Forgetting correlation. Two events that tend to happen together, such as a team scoring a lot and the game going over the total, are not independent, so multiplying their probabilities gives the wrong answer.
6. Mistaking one win for a vindication. A single result carries very little information about the quality of the decision behind it.

How to think about it as a fan

You do not need to bet to benefit from this idea. When you read a price on the odds page, you can convert it into the chance it implies and ask whether that matches your sense of the game. When you watch a coach make a surprising call, you can ask what the alternative would have cost on average. And when you see a record on the standings page, you can remember that results are a noisy reflection of decisions and strength.

If you do choose to wager, treat it as entertainment spending with a fixed limit that you can afford to lose. Do not chase losses, and step back if betting stops being fun. Anyone worried about their own gambling or a loved one's can contact the National Problem Gambling Helpline at 1-800-GAMBLER for free, confidential help.

The short version

• Expected value is the probability-weighted average result of a decision, found by multiplying each outcome by its chance and adding them.
• For a win-or-lose bet, EV = (chance of winning × amount won) − (chance of losing × amount risked).
• At -110, the break-even win chance is about 52.4 percent, and a true coin flip loses roughly 4.5 percent of the amount risked on average.
• The arithmetic is easy and the probability estimate is hard; a few points of error can reverse the sign of the result.
• A positive-EV bet still loses often, and a negative-EV bet still wins sometimes, so judge decisions by process and not by single results.
• EV ignores bet size, repeatability and how much a loss hurts you, which is why bankroll and personal limits matter.
• The same logic drives fourth-down and two-point decisions in football strategy.