The quick version
A parlay pays more because every leg must win
Multiply the odds
Convert each leg to decimal, then multiply without rounding the intermediate result.
Multiply by stake
Combined decimal odds × stake gives total payout. Payout includes the original stake.
Check the margin
Compare the offered price with a fair-probability estimate before the bigger payout takes over the decision.
Worked example: three legs at −110
Exact −110 odds equal 1.909090… in decimal. Multiply that price three times and the parlay is 6.957926 decimal, usually displayed as 6.96 or +596. A 100-unit stake therefore returns 695.79 units, including 595.79 units of profit.
| Active legs | Est. chance all legs win | Offered payout | Estimated price cost |
|---|---|---|---|
| 1 | 50.00% | 190.91 | 4.55% |
| 2 | 25.00% | 364.46 | 8.88% |
| 3 | 12.50% | 695.79 | 13.03% |
| 4 | 6.25% | 1,328.00 | 16.98% |
Assumes each leg is independently fair at 50%. The 4.76% two-way market overround corresponds to a 4.55% expected margin on a −110 wager; those are related figures, not interchangeable labels.
Estimate first, exact when you have both sides
One price tells you the break-even probability, not the true probability. The default estimate removes an assumed market overround from each leg. If you enter the other side’s odds, the calculator proportionally removes the observed two-way overround instead. A slip with the other side on some legs is labelled mixed; one with the other side on every leg is labelled exact under that de-vig method.
“Exact” describes the arithmetic, not a promise that the market is perfectly efficient. Different de-vig models distribute margin differently, especially around longshots. Proportional de-vig is used here because it is transparent and stable; it is not the only defensible model.
Same-game parlays need a correlation model
Multiplying probabilities assumes independence. Two picks from one match often move together: a quarterback passing-yards over and a receiver yards over may be positively correlated, while opposing game-script outcomes can be negatively correlated. Marginal prices do not reveal the full joint distribution. When two legs are marked “Same match/event”, the calculator keeps the operator payout but removes the edge estimate instead of inventing precision.