At The Launch Angle, we have developed a highly accurate model that predicts salary for impending free agent players in Major League Baseball. This model, entitled the Player Prediction Model, or PPM, uses a wide range of variables to determine projected salary for a given upcoming Major League season. Using the last five free agency periods, from 2014 to 2018, we used historical data to show how accurate the model truly is.
Figure One is a distribution of the salary residuals of the free agent outfielders from 2014 to 2018. As shown in Figure One, there were only eight outlier salaries that were generated by the PPM, out of a possible sixty-eight free agent players. This is a margin of 88% accuracy in projecting free agent salaries for outfielders.
Let's transition over to the next group in the game: infielders. Using the same criteria as above, we reference Figure Two below. Of the ninety four qualified free agent infielders, there were only seven salaries that were considered to be outliers, representing an accuracy margin of nearly 93%.
For the grand finale, we collectively focus on two unsung position groups: catchers and designated hitters (DH's). Using the same variables as the two groups above, Figure Three shows that of the fifty-three qualified players, only one salary residual that was considered to be an outlier, representing an accuracy margin of 98%:
So let's review: we have created a highly accurate, tested salary prediction model that generates pinpoint contract value estimates across all offensive position player groups. That's all there is to it, the results speak for themselves.
Comments