Sportsbooks price parlays assuming each leg is independent. When the legs are actually correlated, the true probability diverges from the price — and the math shifts toward the bettor.
Standard parlay math multiplies the odds of each leg together, assuming each outcome is independent. But in sports, outcomes are often correlated -- if one thing happens, another related thing becomes more or less likely.
When positively correlated legs are combined in a parlay, the true probability of both hitting is higher than the independent calculation suggests. The sportsbook prices the ticket as if the legs are independent, but they’re actually linked. That gap is the edge.
These legs are more likely to hit together:
These legs work against each other:
Honest disclosure: MyOddsy’s correlated-parlay feature does not currently apply a correlation discount. We surface four hand-picked same-game pairs per game (Favorite ML + Over, Underdog ML + Under, Underdog Spread + Under, Favorite Spread + Over) and display the combined price as the naive product of the two decimal odds — i.e., as if the legs were independent. We label the pairs “strong” or “moderate” based on intuition, not on a measured correlation coefficient.
Practically, this means: real same-game parlay prices at sportsbooks will be shorter than the combined number shown here, because books charge an SGP markup of roughly 15–30% to account for the positive correlation. The displayed product should be treated as an upper bound on the actual placeable price, not as a guaranteed price.
Building this requires (a) measured per-pair correlation coefficients from historical play-by-play data and (b) live SGP pricing from the books, not the independent-leg pricing. We don’t currently have either pipeline. We’ll update this page when we do.
Even without per-pair coefficients, there are correlation patterns books often don’t adjust for because the legs are in different games:
A correlated parlay combines bets that are more likely to hit together than independently. Sportsbooks price parlays assuming independence, so correlated parlays can have positive expected value when the true co-occurrence rate is higher than the book’s pricing implies.
Most books now offer same-game parlays that allow correlated legs, but they adjust odds to account for it. The edge lives in cross-game correlations that aren’t adjusted -- like weather affecting multiple games’ totals.
Focus on non-obvious cross-game correlations: weather affecting multiple outdoor games, division rivalry scoring patterns, pace-of-play contagion, and back-to-back scheduling effects. Avoid obvious single-game correlations that books already price in.