Las Vegas Point Spread Values and Quantifying the Value of an NBA Player

Adam Hoffer and Jared A. Pincin

This paper uses Las Vegas Sportsbooks individual point spread values (PSVs) to estimate the first marginal product estimates of NBA players. Starting with the individual PSVs, we predict PSVs from performance statistics and use the predicted values to estimate a player’s marginal product. We then compare the NBA estimations with existing, better-established measures of player performance, including player efficiency rating, win shares, and value over replacement player. The results show that the estimates using statistical performance to predict PSVs are in-line with other estimates of player performance. We conclude that PSVs are a comparable measure of marginal product to existing NBA performance composites.

JEL codes: L83 and Z20

https://doi.org/10.32731/IJSF/153.082020.03