Foundations · 7 min read

Positive expected value (+EV) in plain English

Fair odds versus market odds. A worked example. Why a bet can lose and still have been the right bet.

The one-sentence definition

A bet has positive expected value (+EV) when, on average, you would make money taking it again and again at the same price. Negative expected value (-EV) is the opposite. It is not about whether this specific bet wins — it is about what would happen if you could replay the situation a thousand times.

The arithmetic, slowly

Expected value is just probability times payoff, summed across outcomes:

EV = (probability of winning × profit if you win) − (probability of losing × amount you lose)

Suppose a coin lands heads 55% of the time. You bet $100 on heads at decimal odds of 2.00, which means a $100 win pays you $100 profit. The math:

• Probability of winning: 0.55. Profit if you win: $100.
• Probability of losing: 0.45. Amount you lose: $100.
• EV = (0.55 × $100) − (0.45 × $100) = $55 − $45 = +$10.

That bet is +EV by $10 per $100 staked. If you could repeat it endlessly, you would make about $10 per bet on average. Any single bet is a coin flip; the expected value is a statement about long-run averages.

Where the probability comes from

Most punters skip this step, because it is the hard one. To know whether a bet is +EV you need a probability estimate that is more accurate than the bookmaker's implied probability. There are essentially three sources for that estimate:

1. A model. A statistical or machine-learning model that takes inputs (form, schedule, lineups, weather, etc.) and outputs a probability. This is what a serious analyst, a syndicate, or a research desk like StatLine builds.
2. Information the market doesn't have yet. A late lineup change, a piece of injury news, a venue oddity. Edges from information are real but tiny — the market reprices in seconds.
3. Comparing prices across books. If five bookmakers say 1.85 and one says 2.05, the 2.05 is probably mispriced. The “true” probability is implied by the consensus, not the outlier. (See line shopping.)

Implied probability and the “fair” price

Every decimal-odds quote can be converted to an implied probability:

Implied probability = 1 / decimal odds

• 2.00 → 1/2.00 = 50%.
• 1.91 → 1/1.91 ≈ 52.4%.
• 3.50 → 1/3.50 ≈ 28.6%.

If you genuinely believe an outcome will happen 60% of the time, then any price longer than 1/0.60 = 1.667 is +EV. A 1.91 (52.4% implied) is generously +EV. A 1.65 (60.6% implied) is roughly fair. Anything shorter is -EV by definition, regardless of how confident you feel.

A worked example with a real-feeling market

Say the AFL Round 5 line on the Tigers vs Demons is hanging at:

• Tigers -10.5 @ 1.91
• Demons +10.5 @ 1.91

The book is implying ~52.4% on each side, summing to 104.8%. The 4.8% over 100 is the bookmaker's margin (see why most punters lose).

Suppose your model gives the Tigers an honest 56% chance of covering -10.5. The expected value of a $100 bet on the Tigers:

• Profit if win: $91 (decimal 1.91 = $100 stake → $191 return → $91 profit).
• Loss if lose: $100.
• EV = (0.56 × $91) − (0.44 × $100) = $50.96 − $44.00 = +$6.96.

The bet is +EV by ~$6.96 per $100 staked, or about 7% of stake. That is a meaningful edge. Notice it does not guarantee the bet wins — it loses 44 times in 100. It only guarantees that if your 56% probability is right, you would make money on average across many such bets.

Where punters get +EV wrong

Overestimating their own probability

The single most common error is assuming your probability estimate is more accurate than the market's. The market price is the consensus of every dollar that has been bet, often by people with better models than you. If your model says 60% and the market is priced at 52%, the prior should be that you are wrong unless you have a well-tested reason to think otherwise.

The honest test of an edge is closing-line value over hundreds of bets, not how confident this week's pick feels. See closing-line value.

Confusing “value” with “long shot”

A 4.00 outsider is not automatically +EV. It is +EV only if its true probability is better than 25% (1/4.00). Most outsiders are outsiders for a reason. Equally, a short favourite at 1.30 can be +EV if its true probability is better than 76.9%. Price tells you nothing about value on its own.

Treating one bet as if it were the average

+EV is a long-run statement. A +EV bet can lose. Ten +EV bets can all lose if you are unlucky. The way to make +EV pay off is volume: many bets, small stakes relative to bankroll, taken consistently. That is also why bankroll discipline matters — you need to stay solvent long enough for the law of large numbers to do its job. (See bankroll management.)

How StatLine uses +EV

Every published pick has an internal probability estimate. We compare it to the bookmaker's implied probability at the moment of publish, and we will only ship a pick where the gap clears a minimum edge threshold and where the closing-line evidence in similar past picks has been positive. We do not publish “value” as a feeling. The number is computed, and the gap is the price's answer to whether the bet was good — independent of whether it eventually won or lost.

Educational content only — not personal financial advice. Sports are uncertain and any bet can lose. Past results do not predict future results. 18+. Gamble responsibly. Responsible gambling resources.