PrizePicks combo lines — PRA (Pts+Reb+Ast), Pts+Reb, Pts+Ast, Reb+Ast and Double-Double — priced with a correlation-aware joint simulation. We draw each player's points, rebounds and assists together using their own stat-to-stat correlations (Gaussian copula over negative-binomial margins, 20k sims), then read the probability of the combo clearing the PrizePicks line. Why this matters: pts/reb/ast move together within a game, so the naive independent-sum that most tools use mis-prices the tails. The Correlation edge column shows our correlated fair price side-by-side with the naive-independent price every other tool uses — and the gap between them. No competitor surfaces this. PrizePicks is a pick'em-payout product, so edge = our probability on the favored side minus the 50% break-even.
How to read Correlation effect:
0.512 correlated fair · 0.521 naive-independent · +0.9pp = the correlation moves the favored side +0.9pp vs assuming the stats are independent.
r = the player's own pts/reb/ast correlation that drives the shift.
| Player | Tm |
Combo | PP Line |
Proj | Pick | P |
Correlation effect ▾ |
No PrizePicks combo props posted for this slate yet — lines post near game time. Check back closer to tip-off.
Joint-sim, correlation-aware. Marginals come from the same projection model the board uses (calibrated where a single-stat line is posted, else from the season game log); correlations are each player's own pts/reb/ast Pearson correlations, shrunk toward the league average for small samples. Confidence is Low when a marginal falls back to the game log, the correlation sample is thin (<8 games), or the combo line is small (role-player noise). Not financial advice.