When Is a Parlay Positive EV? Calculations and Correlation

A parlay has positive expected value when its true probability of winning, multiplied by its offered decimal odds, exceeds 1. For independent legs, you can calculate that probability by multiplying the individual win probabilities; for correlated legs, you need the joint probability instead.

What makes a parlay positive EV?

A positive EV parlay offers more than fair compensation for the chance that every selection wins. Its payout alone tells you nothing about value: decimal odds of 10.00 are attractive at a true win probability of 12%, but unattractive at 8%.

OpenStax explains expected value as a weighted mean: multiply each possible return by its probability, then add the results. For a simple all-or-nothing parlay, there are two outcomes: the slip wins or the stake is lost.

The net expected return per unit staked is EV = q × D − 1, where q is the probability that every leg wins and D is the offered decimal odds, including returned stake. The break-even probability is 1 ÷ D. These calculations assume no additional costs and no pushes, voids or partial settlements.

How do you calculate a positive EV parlay?

Consider this hypothetical two-leg football accumulator across separate matches. The probabilities are illustrative estimates, not live odds or predictions, and we assume the outcomes are independent.

• Leg A: Match A home win, estimated probability 55%, offered decimal odds 2.00.
• Leg B: Match B over 2.5 goals, estimated probability 50%, offered decimal odds 2.10.
• Combined win probability: 0.55 × 0.50 = 0.275, or 27.5%.
• Combined decimal odds: 2.00 × 2.10 = 4.20.
• Combined EV: 0.275 × 4.20 − 1 = 0.155, or +15.5%.

At a $10 stake, the winning return is $42, including the stake, so the winning profit is $32. Expected profit is therefore 0.275 × $32 − 0.725 × $10 = $1.55. That is a probability-weighted average, not the profit you receive on an individual ticket: this example still loses 72.5% of the time.

Now contrast a hypothetical three-leg parlay with each leg priced at 2.00 but having only a 45% true win probability. Assuming independence, its odds are 8.00 and its win probability is 0.45³ = 9.1125%. Its EV is 0.091125 × 8 − 1 = −27.1%. The larger payout conceals worse value: the ticket needs a 12.5% win probability merely to break even.

Can you combine the EV percentages of individual legs?

Yes, under the independence and multiplicative-pricing assumptions, but you multiply return factors rather than add EV percentages. In the two-leg example, Leg A has EV of 0.55 × 2.00 − 1 = +10%, while Leg B has EV of 0.50 × 2.10 − 1 = +5%.

Combined EV is (1 + EV₁) × (1 + EV₂) × … − 1, with each EV written as a decimal. Here, 1.10 × 1.05 − 1 = 0.155, or +15.5%, rather than the +15% obtained by simple addition. Each factor equals that leg’s estimated probability multiplied by its decimal odds.

This shortcut works only when the joint probability is the product of the individual probabilities and the offered parlay odds equal the product of the leg odds. If the sportsbook quotes a different total price, use that actual price in q × D − 1. Positive EV legs do not remove the need to check the final proposition.

How does correlation change parlay EV?

ProbabilityCourse defines independent events by the equality P(A and B) = P(A) × P(B). Without independence, the general two-leg calculation is P(A and B) = P(A) × P(B given A). You need to estimate how the first outcome changes the chance of the second.

For a hypothetical same-match football parlay, suppose the home win has a 60% probability and over 2.5 goals has a 50% probability. The sportsbook offers combined decimal odds of 3.20. Multiplying the standalone probabilities gives 30%, but that is only valid if the outcomes are independent.

• Independence assumption: joint probability 30%; EV = 0.30 × 3.20 − 1 = −4%.
• Positive-dependence scenario: joint probability 36%; EV = 0.36 × 3.20 − 1 = +15.2%.
• Negative-dependence scenario: joint probability 24%; EV = 0.24 × 3.20 − 1 = −23.2%.

These are alternative illustrative probability models, not measurements of a particular match. They show why correlation can reverse a value judgment. Positive correlation does not automatically create value, because the offered payout still matters. Likewise, negative correlation does not automatically rule value out if the payout is sufficiently generous, as the examples in Wizard of Odds’ correlation explainer demonstrate.

For three or more legs, checking pairs is not enough: ProbabilityCourse’s definition also requires the relevant higher-order intersections to satisfy the product rule. Separate matches are a useful starting point for screening dependencies, not proof of independence. If you cannot defend the joint-probability estimate, treat the calculated parlay EV as conditional on an unverified assumption.

How does OddsFantasy build positive EV accumulators?

The OddsFantasy accumulator builder builds two- to seven-leg slips from live pre-match value bets at one bookmaker you choose. The qualifying prices beat Pinnacle’s no-vig odds. That benchmark provides a probability reference for identifying value; it is not a guarantee that the estimated probabilities are correct.

1. Choose one bookmaker. The builder assembles the slip at that bookmaker rather than combining selections that would need to be placed at different sportsbooks.
2. Set the leg filters. Choose minimum and maximum odds per leg, minimum EV per leg, sport and how far ahead the match is, from one to seven days.
3. Review the generated slips. Each uses one leg per match and displays total odds and combined EV, plus every leg’s event, market, selection, odds and bookmaker link.
4. Sort and inspect. Sort slips by last update, match date, EV or odds, then examine the selections and any potential dependencies before placing a bet.

The builder calculates total odds by multiplying leg odds and combined EV using the return-factor formula above. One leg per match avoids combining multiple selections from the same fixture, but it does not establish independence across every slip. The accumulator builder is included in the Classic plan, and slips can be added to the OddsFantasy bet tracker.

Sportsbook placement remains manual: OddsFantasy finds and alerts value bets, and you place them yourself. Automatic value-bet placement is available only on Polymarket through the positive EV tool; that is separate from placing a sportsbook accumulator.

What should you check before placing the slip?

1. Verify the final price. Recalculate EV from the odds actually available when placing the bet, not an earlier displayed price.
2. Check market definitions and settlement. Confirm the selections and how pushes or voids would affect the ticket; the simple win-or-lose formula does not model those additional outcomes.
3. Separate payout arithmetic from probability estimation. OddsFantasy’s free calculator handles 2–20 legs in decimal or American odds and shows total odds, payout, profit and the odds-implied chance of all legs winning. That implied chance is a break-even threshold, not an independent estimate of the true chance.
4. Allow for estimation error. A small change in several leg probabilities can materially change combined EV. Reject the slip if its apparent edge depends on assumptions you cannot support.

Does positive EV justify a larger stake?

Not by itself. Stake size must account for the probability of losing, uncertainty in your estimates and exposure shared with other bets. Our example’s +15.5% EV still comes with a 72.5% loss probability. A positive EV parlay is not arbitrage and does not guarantee a profitable session.

Use the Kelly criterion calculator to explore how probability and price affect sizing, without treating uncertain estimates as precise inputs. For longer-term evaluation, see how to judge positive EV results over time. The practical decision is whether the available payout compensates for the joint risk—not whether the slip offers an impressive headline return.