Independence product
The displayed combined estimate multiplies the selected-win probability for each leg as if those events were independent.
Parlay analysis
The headline product multiplies leg probabilities under independence. Contender places that assumption beside dependence bounds, linked-leg flags, and recorded outcomes.
P(A∩B)≠certainty
Historical parlay replay
Historical example — replayed as of noon UTC on game day; not a live pick.
3 of 3 selected
Shai Gilgeous-Alexander
OKC vs MIN · Mar 15, 2026
Over 30.5 PTS
P 86.6%
missNikola Jokić
DEN @ LAL · Mar 14, 2026
Over 12.5 REB
P 52.1%
hitStephen Curry
GSW vs DET · Jan 30, 2026
Over 4.5 3PM
P 41.2%
miss
Historical example — replayed as of noon UTC on game day; not a live pick.
Dependence range
Fréchet bounds communicate how wide the mathematically possible range becomes when joint behavior is unknown.
Failed-leg semantics
A recorded parlay misses when any required leg misses. Each historical leg keeps its own graded outcome beside the aggregate.