Build home run, prop, and same-game parlays where the legs actually carry an edge — each one priced from the model's win probability, with correlation built in for same-game slips. These are model probabilities, not guaranteed payouts. Pick a starting point:
It multiplies each leg's model win probability — and for same-game parlays, adjusts for the correlation between legs from the same game — to estimate the parlay's true probability and fair value, then compares that to the offered payout. These are model probabilities, not guaranteed payouts.
A same-game parlay combines multiple bets from a single game — for example a pitcher's strikeouts plus a hitter's total bases. Because those outcomes are correlated, sportsbooks price SGPs differently from independent legs, and modeling that correlation is the key to fair pricing.
Home runs are high-variance, so HR parlays are long shots by nature — small stakes, big payouts. The edge comes from stacking genuinely favorable matchups rather than popular names; BallBet's HR DNA board is built to find those spots, but no parlay is a lock.