Underdog spread + Under
Lower-variance games make it easier for an underdog to stay inside the number.
Every empty possession matters more in a low-total game, compressing the final margin.
Questions
Are Underdog spread and Under correlated?
Lower-variance games make it easier for an underdog to stay inside the number. Our working same-game coefficient is about 0.19 (a seed estimate, not yet measured).
Why can't I just multiply the two prices?
Multiplying assumes independence. With a correlation of about 0.19, the true joint probability differs from the naive product, which is why the Autopsy simulates the pair 20,000 times instead.
Does the correlation hold across different games?
No. This relationship is driven by shared game state, so across separate games we treat the legs as independent.
Price this pair in the Autopsy →
https://blacklightparlays.app/correlation/underdog-spread-and-under · coefficient is an estimate, not a measured result