What is Expected Value?
Expected value is a mathematical estimate of the average profit or loss a bet would produce if the same type of opportunity could be repeated many times under the same probabilities and prices.
Expected value is not a prediction of what will happen to one individual bet. A bet with positive expected value can lose. A bet with negative expected value can win. The concept describes the long-run mathematical quality of the price under a stated probability estimate.
How expected value is calculated
For a simple two-outcome bet, the basic formula is:
EV = probability of winning × net profit if the bet wins + probability of losing × loss if the bet loses
The loss is written as a negative number.
If a 10 MDL bet at decimal odds of 2.20 wins, the total return is 22 MDL and the net profit is 12 MDL. If it loses, the net result is -10 MDL.
Examples of expected value
The same stake can have positive, zero or negative expected value depending on the price and probability estimate.
| Odds | Win probability estimate | EV per 10 MDL | Interpretation |
|---|---|---|---|
| 2.20 | 50% | +1 MDL | Positive EV |
| 2.00 | 50% | 0 MDL | Break-even |
| 1.80 | 50% | -1 MDL | Negative EV |
The table uses a 10 MDL stake and a 50% estimated chance of winning for every example.
Positive expected value example
Suppose a bettor estimates that Team A has a 50% chance of winning and the sportsbook offers decimal odds of 2.20.
- Stake: 10 MDL
- Net profit if Team A wins: 12 MDL
- Loss if Team A loses: -10 MDL
- Estimated win probability: 50%
- Estimated loss probability: 50%
EV calculation:
0.50 × 12 MDL + 0.50 × -10 MDL = 6 MDL - 5 MDL = +1 MDL
The expected value is +1 MDL per 10 MDL staked under that probability estimate.
Expected return on stake:
1 MDL / 10 MDL × 100% = +10%
This does not mean every 10 MDL bet earns 1 MDL. It means that if the probability estimate is accurate and the same value can be repeated many times, the mathematical average is +1 MDL per 10 MDL stake before other costs or rule differences.
Negative expected value example
Now suppose the same 50% estimated chance is offered at decimal odds of 1.80.
- Stake: 10 MDL
- Net profit if the bet wins: 8 MDL
- Loss if it loses: -10 MDL
EV:
0.50 × 8 MDL + 0.50 × -10 MDL = 4 MDL - 5 MDL = -1 MDL
The expected value is -1 MDL per 10 MDL stake, or -10% of the stake.
The bet can still win today. Negative EV means the offered return is too low for a 50% win probability in the long-run calculation.
Break-even probability
The break-even probability is the win probability at which the expected value of a standard bet is zero before other costs.
For decimal odds:
Break-even probability = 1 / decimal odds
Examples:
- 2.00 = 50.00%
- 2.20 = 45.45%
- 1.80 = 55.56%
- 1.50 = 66.67%
At odds of 2.20, an estimated probability above 45.45% produces positive expected value. An estimate below 45.45% produces negative expected value. Exactly 45.45% is the break-even point under the simplified calculation.
Expected value and implied probability
Implied probability is the probability represented by the offered odds.
At 2.20:
1 / 2.20 × 100% = 45.45%
If the bettor estimates the chance at 50%, the bettor's estimate is higher than the price-implied probability. That difference creates positive expected value in the model.
The important difficulty is that the bettor does not know the future probability with certainty. The expected value is only as reliable as the probability estimate and the settlement assumptions used in the calculation.
Expected value is not a guarantee
Positive expected value does not guarantee profit on one bet, ten bets or any specific short sequence.
Consider a bet with a 55% estimated chance of winning. Even if that estimate is correct, losses can occur several times in a row because each event still has uncertainty.
Expected value describes the average mathematical result over repeated opportunities. Actual short-term results can be far above or below that average.
Expected value and variance
Variance describes how widely actual results can move around the expected average.
Two betting strategies can have the same positive expected value but very different short-term behavior. One can produce many small wins and losses. Another can depend on rare large wins and therefore show much larger swings.
Expected value answers the question “What is the average mathematical result under these probabilities?” Variance answers “How widely can actual results move around that average?”
Expected value and value betting
A value bet is a bet whose offered price is considered higher than the bettor's estimate of the fair price.
If a bettor estimates a 50% chance, fair decimal odds under that estimate are:
1 / 0.50 = 2.00
If the sportsbook offers 2.20, the price is above the bettor's fair-price estimate and the bet has positive expected value under that model.
The phrase “value bet” therefore describes the opportunity. Expected value measures the mathematical size and direction of the advantage or disadvantage under the estimate.
Expected value and bookmaker margin
Bookmaker prices normally include a margin. In a complete market, the implied probabilities of all outcomes can add to more than 100%.
For example, a two-way market at 1.91 and 1.91 implies about 52.36% on each side, for about 104.71% in total.
This does not mean both outcomes truly have a 52.36% chance. The total above 100% reflects the pricing load. A bettor evaluating expected value therefore needs an independent probability estimate rather than simply treating the displayed implied probability as the exact fair probability.
Expected value with refunds or pushes
If a market can return the stake, the refund outcome has a net result of 0 MDL and should be included in the formula.
For example:
EV = P(win) × winning profit + P(loss) × negative stake + P(refund) × 0 MDL
The refund does not add profit or loss, but its probability changes the weights of the winning and losing outcomes.
Why the probability estimate matters
Expected value is not created by the formula alone. The formula only processes the assumptions entered into it.
If the bettor estimates a team at 60% when its realistic chance is much lower, the calculated EV can appear positive even though the underlying estimate is poor.
Useful probability estimates require a consistent method, relevant information and awareness of uncertainty. A precise-looking percentage is not automatically an accurate probability.
Common beginner mistakes
- Treating positive EV as a guaranteed winning bet.
- Using total return instead of net profit in the EV formula.
- Forgetting to write losses as negative values.
- Assuming the sportsbook's implied probability is automatically the true probability.
- Believing one losing result proves a positive-EV calculation was wrong.
- Believing one winning result proves a negative-EV bet was good.
- Ignoring refunds, commissions or special settlement rules.
- Confusing expected value with potential payout.
- Assuming higher odds automatically mean higher expected value.
- Using an unsupported probability estimate and treating the result as precise.
What to remember
Expected value is the average mathematical profit or loss of a bet under a stated probability estimate and price.
Positive EV means the model expects an average gain over repeated opportunities. Negative EV means the model expects an average loss. Neither result guarantees what happens on the next bet.
For a standard decimal bet, the break-even probability is 1 divided by the decimal odds. Expected value becomes positive only when the bettor's estimated chance is high enough relative to the offered price.
Related terms
- Odds
- Implied Probability
- Value Bet
- Bookmaker Margin
- Stake
- Payout
- Variance
- Probability
- Bet
