Expected Value in Sports Betting: The Only Math That Matters

If you spend any time in sports betting forums, you will hear the term "expected value" thrown around constantly. Experienced bettors treat it like a religion. And for good reason: expected value (EV) is the single most important concept separating long-term winners from long-term losers in sports betting.

Forget about hot streaks, gut feelings, and "lock of the week" picks. If you want to understand whether your betting strategy will make or lose money over time, you need to understand exactly one thing: expected value.

This post breaks down what EV is, how to calculate it, and why positive expected value betting is the only mathematically sustainable path to profit.

What Is Expected Value?

Expected value is a concept from probability theory that tells you the average outcome of a bet if you could repeat it an infinite number of times. It answers a deceptively simple question: "Is this bet worth making?"

Every bet you place has an expected value. It is either:

• Positive (+EV): You expect to profit over time
• Negative (-EV): You expect to lose over time
• Zero (neutral EV): You expect to break even over time

The vast majority of bets offered by sportsbooks are negative EV for the bettor. That is how sportsbooks make money. Your job as a sharp bettor is to find the rare situations where the odds offered are better than they should be, creating a positive expected value opportunity.

The Expected Value Formula

The basic EV formula for a simple win/lose bet is:

$$ EV = P(\text{win}) \times \text{Profit if win} - P(\text{loss}) \times \text{Amount lost if loss} $$

Or equivalently:

$$ EV = (P_w \times W) - (P_l \times L) $$

Where:

• $P_w$ = Probability of winning
• $W$ = Net profit if you win
• $P_l$ = Probability of losing (which is $1 - P_w$)
• $L$ = Amount you lose (typically your stake)

If EV is positive, the bet is profitable in the long run. If EV is negative, you will lose money over time.

Worked Example 1: A Simple Coin Flip Bet

Let's start with the most basic example possible.

Scenario: Your friend offers you a bet on a fair coin flip. Heads, he pays you $110. Tails, you pay him $100.

• $P_w = 0.50$ (fair coin)
• $W = \$110$ (your profit on heads)
• $P_l = 0.50$
• $L = \$100$ (your loss on tails)

$$ EV = (0.50 \times 110) - (0.50 \times 100) = 55 - 50 = +\$5.00 $$

This is a +EV bet. On average, you make $5 every time you take this bet. You will lose plenty of individual flips, but over hundreds or thousands of flips, you will come out ahead by roughly $5 per flip.

Now flip the terms: if your friend pays you only $90 on heads and you pay $100 on tails:

$$ EV = (0.50 \times 90) - (0.50 \times 100) = 45 - 50 = -\$5.00 $$

That's a -EV bet. You lose $5 on average per flip. This is essentially the position most recreational bettors are in at sportsbooks.

Worked Example 2: NFL Moneyline Bet

Now let's apply this to a real sports betting scenario.

Scenario: The Kansas City Chiefs are playing the Denver Broncos. The sportsbook offers the Broncos at +200 (American odds). You believe the Broncos have a 38% chance of winning.

First, convert the odds to understand the payout. At +200, a $100 bet returns $200 in profit if the Broncos win.

• $P_w = 0.38$
• $W = \$200$ (profit on a $100 bet)
• $P_l = 0.62$
• $L = \$100$ (your stake)

$$ EV = (0.38 \times 200) - (0.62 \times 100) = 76 - 62 = +\$14.00 $$

This is a strong +EV bet. For every $100 wagered, you expect to earn $14 in the long run.

But here is the critical question: what probability does the sportsbook's line imply?

At +200 odds, the implied probability is:

$$ P_{\text{implied}} = \frac{100}{200 + 100} = \frac{100}{300} = 0.333 = 33.3\% $$

You believe the Broncos win 38% of the time, but the sportsbook is pricing them as if they win only 33.3% of the time. That gap — your estimated probability being higher than the implied probability — is where the +EV lives.

Key insight: You do not need to be right about who wins. You need to be right that the probability is higher than what the odds imply.

Worked Example 3: NBA Point Spread

Scenario: The Milwaukee Bucks are -6.5 against the Charlotte Hornets at -110 odds on both sides. You have built a model that projects the Bucks to win by 8.5 points. Based on your model, the Bucks cover -6.5 about 58% of the time.

At -110 odds, you risk $110 to win $100.

• $P_w = 0.58$
• $W = \$100$ (profit on a $110 bet)
• $P_l = 0.42$
• $L = \$110$ (your stake)

$$ EV = (0.58 \times 100) - (0.42 \times 110) = 58 - 46.2 = +\$11.80 $$

Per $110 wagered, you expect to profit $11.80 over time. Expressed as a percentage of your stake, that is a 10.7% edge — an excellent EV margin.

Now, what if the line were -110 and you only estimated a 52% chance of covering?

$$ EV = (0.52 \times 100) - (0.48 \times 110) = 52 - 52.8 = -\$0.80 $$

At 52%, the bet is actually slightly -EV against -110 odds. You need better than a 52.38% win rate to break even at -110, because:

$$ P_{\text{breakeven}} = \frac{110}{100 + 110} = \frac{110}{210} = 0.5238 = 52.38\% $$

This is the sportsbook's margin (the "vig" or "juice"), and it is why most bettors lose.

The Break-Even Table

Understanding what win rate you need to break even at common odds is essential:

American Odds	Decimal Odds	Implied Probability	Break-Even Win Rate
-200	1.50	66.7%	66.7%
-150	1.67	60.0%	60.0%
-110	1.91	52.4%	52.4%
+100	2.00	50.0%	50.0%
+110	2.10	47.6%	47.6%
+150	2.50	40.0%	40.0%
+200	3.00	33.3%	33.3%
+300	4.00	25.0%	25.0%
+500	6.00	16.7%	16.7%

To be +EV, your estimated probability of winning must be higher than the implied probability for that price.

Why +EV Is the Only Path to Long-Term Profit

This is where the law of large numbers enters the picture. The law of large numbers states that as you repeat a random experiment more and more times, the average result will converge to the expected value.

What this means for betting:

• A single +EV bet might lose. In fact, many will lose.
• But over hundreds and thousands of +EV bets, your results will converge toward your expected profit.
• Conversely, -EV bets will grind your bankroll down to zero over time, no matter how clever your staking plan is.

Think of it like a casino. A casino does not win every hand of blackjack. It does not need to. The house edge means every hand is +EV for the casino, and over millions of hands, the casino's profit approaches the expected value with near-certainty.

When you bet +EV, you are the casino. When you bet -EV, you are the gambler. There is no staking system, no parlay trick, and no "feel for the game" that can overcome a negative expected value in the long run.

The law of large numbers is not a guarantee on any single bet. It is a guarantee on the process. If you consistently find and bet +EV situations, the math will reward you.

How to Identify +EV Bets

Finding +EV opportunities requires estimating probabilities better than the sportsbook (or at least better than the market). Here are the primary approaches:

1. Build Your Own Models

The most robust approach is to build statistical models that estimate game probabilities independently of the betting market. This might include:

• Regression models using team and player statistics
• Elo rating systems that track team strength over time
• Machine learning models trained on historical data
• Simulation-based approaches (e.g., Monte Carlo simulations)

If your model consistently gives different probabilities than the market, and those differences are accurate, you have a systematic source of +EV.

2. Shop for the Best Lines

Different sportsbooks offer different odds on the same event. If one book offers the Chiefs at -150 and another at -130, taking -130 is clearly better. Line shopping does not guarantee +EV, but it improves your EV on every bet you make.

3. Exploit Market Overreactions

Markets tend to overreact to recent events:

• Recency bias: A team that just lost a blowout may be undervalued
• Public bias: Popular teams (Cowboys, Lakers, Yankees) tend to have inflated lines because recreational bettors load up on them
• Injury overreaction: The market sometimes adjusts too much for a single player's absence

4. Use Closing Line Value (CLV) as a Proxy

Many sharp bettors track whether the line moves in their favor after they place a bet. If you consistently bet a line before it moves in the direction of your bet, you are likely finding +EV. For example:

• You bet Team A at +150 on Monday
• By game time on Sunday, Team A is at +130
• The market moved toward your position, suggesting you had an edge

Closing line value is the best single indicator of long-term betting skill.

Common Mistakes About Expected Value

Mistake 1: Confusing +EV with a Guaranteed Win

A +EV bet can and will lose. Expected value is a long-term average, not a prediction for a single event. If you bet on a 60% chance and it loses, the bet was still correct if your probability estimate was accurate. The key question is never "did I win?" but "was the bet +EV?"

Mistake 2: Ignoring the Vig

Many bettors look at a game and think "I like this side" without considering whether the price is right. Liking a team to win is not the same as having a +EV bet on that team. At -200, you need that team to win more than 66.7% of the time. Your analysis has to be better than "I think they'll win."

Mistake 3: Overestimating Your Edge

The most dangerous mistake in EV betting is overconfidence in your probability estimates. If your model says a team wins 60% of the time and the true probability is 54%, you might think you have a huge edge when you actually have none (or a negative one, depending on the odds).

Good EV bettors:

• Track their results rigorously
• Compare predictions to actual outcomes over large samples
• Calibrate their models regularly
• Accept that their edge, if it exists, is small (1-5% on most bets)

Mistake 4: Thinking Parlays Are a Shortcut

Parlays multiply your odds but also multiply the sportsbook's edge. Each leg of a parlay adds vig. A three-leg parlay at standard -110 odds on each leg has significantly more vig built in than three individual bets. Parlays are almost always -EV.

The exception is correlated parlays where the outcomes are not independent, which some sportsbooks misprice. But these are specialized opportunities, not a general strategy.

Mistake 5: Thinking You Need to Win Most Bets

At +200 odds, you only need to win more than 33.3% of the time to be +EV. You could lose two-thirds of your bets and still be profitable. EV betting is about the relationship between price and probability, not about your win percentage in isolation.

EV and Bankroll Management

Even with a +EV strategy, variance can destroy you if you bet too much of your bankroll on each wager. This is where concepts like the Kelly Criterion come in.

The Kelly Criterion tells you the optimal bet size based on your edge:

$$ f^* = \frac{p(b + 1) - 1}{b} $$

Where:

• $f^*$ = fraction of bankroll to wager
• $p$ = probability of winning
• $b$ = decimal odds minus 1 (net payout per unit wagered)

For example, if you have a 55% chance on an even-money bet ($b = 1$):

$$ f^* = \frac{0.55(1 + 1) - 1}{1} = \frac{1.10 - 1}{1} = 0.10 = 10\% $$

Most sharp bettors use fractional Kelly (betting 25-50% of the Kelly-recommended amount) to reduce variance while still capturing the edge.

How Sportsbooks Set Odds and Create -EV

Understanding the "enemy" helps. Sportsbooks set lines through a combination of:

1. Opening lines based on power ratings and statistical models
2. Market adjustments based on where money flows
3. Sharp action tracking — they move lines when known winning bettors take a side
4. Vig inclusion — the margin built into every line

A "fair" line on a 50/50 game would be +100 on each side. Instead, books offer -110 on each side, collecting $110 from both sides and paying $100 to the winner. That extra $10 per side is the vig, and it makes both sides -EV for the bettor.

Your job is to find spots where the book's probability estimate is wrong enough that even after the vig, the bet is still +EV for you.

Putting It All Together

Here is the mental framework for every bet you consider:

1. Estimate the true probability of the outcome (your model, your research)
2. Calculate the implied probability from the odds offered
3. Compare: Is your estimated probability meaningfully higher than the implied probability?
4. If yes, calculate the EV to confirm a positive number
5. Size appropriately using Kelly or fractional Kelly
6. Track everything — you need hundreds of bets to know if your edge is real

Expected value is not a magic formula that tells you who will win tonight's game. It is a disciplined framework that, applied consistently over thousands of bets, separates profitable bettors from everyone else.

The math does not care about your feelings, your favorite team, or your "read" on a situation. It only cares about one thing: is the expected value positive? If yes, bet. If no, pass. Repeat for the rest of your betting life.

That is the only math that matters.