Faculty Advisor

Thelge B. Peiris

Identifier

etd-043018-080853

Abstract

This project uses multiple linear regression to predict the value of Major League Baseball free agent contracts, inspired by the low volume of published research on this topic. I found one published paper that shared my research goal but its predictive power needed improvement. An in depth comparison of our models is carried out with k-fold cross validation mean square prediction error being used as the main standard. The predictor variables considered in my models were related to performance evaluation and position, and the response variable was inflation-adjusted average annual value of the contract. The result of the project is two linear regression models, one for hitters and one for pitchers.

Publisher

Worcester Polytechnic Institute

Degree Name

MS

Department

Mathematical Sciences

Project Type

Thesis

Date Accepted

2018-04-30

Accessibility

Restricted-WPI community only

Subjects

sports, linear regression, sabermetrics, statistics, moneyball, regression, diagnostics

Available for download on Tuesday, April 30, 2019

Share

COinS