Add up to 4 legs with their American prices, click Price it, and see the independent-baseline parlay price. This demo computes the product of marginal implied probabilities, which is the price the legs would parlay to if they were uncorrelated. Real SGP prices differ by 5 to 40% depending on per-leg correlation. The full /v1/sgp/price endpoint applies literature-derived correlation coefficients.
Build a parlay
Independent-baseline price
American
+0
Decimal
0.00
Implied prob
0.00%
Legs
0
Per-leg breakdown
// Press "Price it" to compute
This is not a real SGP price. Books apply correlation adjustments per-pair: positively-correlated legs (e.g. "Mahomes 300+ pass yds" + "Chiefs win") push the price longer than the independent product; negatively-correlated legs (e.g. "Yankees ML" + "Yankees lose by 2+") collapse the parlay to zero correlation discount. The full /v1/sgp/price endpoint applies per-pair literature-derived correlation coefficients and shows you which paper sourced each.
How real SGP pricing differs
Three coefficients drive the difference between independent baseline and real SGP:
Marginal-marginal correlation: if leg A's win is correlated with leg B's win, the joint probability is higher than the product. Books shorten the parlay price to capture this.
Same-player residual: two props on the same player (e.g. "rush yds 80+" and "anytime TD") share a base process. Literature gives Pearson coefficients in the 0.25 to 0.55 range depending on the prop pair.
Team-game outcome residual: "team wins" and "team total goes over" share scoring distribution. Coefficients typically 0.15 to 0.30.
The full /v1/sgp/price endpoint returns the independent baseline alongside the correlation-adjusted price and names the literature source for each pair. Sign up free at /signup for access.
Rate-limited to 60 req/h per IP, max 4 legs per request. No auth required. The full endpoint at /v1/sgp/price supports up to 20 legs and applies correlation modeling.