RA sproj 01231019081015100608022205190602221321

This candidate applies the baseball won-loss records analysis to basketball. The small roster of starting players, the long season, and the relatively low stochasticity of basketball suggest that a regression equation with few past years may be optimal. She finds that a single past year is optimal.

We post this student project to show the difference between baseball and basketball in the number of past years for the optimal regression equation. The same difference in the opposite direction occurs for football.

The NEAS faculty comments point out a few errors in the student project. The comments here focus on the intuition; they do not cover the statistics in the Excel file.