Predicting MLB Win-Loss records by Runs Scored vs Runs Allowed by Brandt Mandia '23

Predicting MLB Win-Loss records by Runs Scored vs Runs Allowed by Brandt Mandia ’23, Mathematics Colloquium, Wednesday, February 22, 1 – 1:45 pm, North Science Building 113, Wachenheim.

Abstract:  Bill James empirically discovered a good predictor of a team’s winning percentage is RS^2 / (RS^2 + RA^2), where RS is the average number of runs scored per game, and RA the average number allowed.  We discuss a derivation of this result using techniques from probability and statistics, under the assumption that runs scored and allowed are independent random variables drawn from a three parameter Weibull.  If time permits we discuss how to estimate these values from seasonal data.