Understanding Professional Baseball Contracts: An Empirical Study Using Predictive Statistics

Abstract

This thesis aims to provide an empirical study to answer the research question: “How do professional baseball contracts account for future productivity in determining compensation?” These players in past literature have been compared as an asset to a company. If this analysis were true, it would be clear to see that future output of that player should be the only metric to value him. To begin the narrative around these two concepts it is important to analyze Major League Baseball's free agency market. This is a complex system where players are allowed free movement to whatever organization they desire after a certain amount of experience in the league. The free agency market has allowed for the explosion of larger, over the top, contracts in the past few years. And I wish to study how players are valued during this process. In order to do this, I have enlisted the help of Sabermetric statistics. These are advanced statistical analysis of a player's game to better understand their worth on the field. The statistic I utilize most in this study is called Wins Above Replacement Player (WARP). WARP is a statistic which takes into account many aspects of the game including, hitting, baserunning, and defending. Here data will be collected from 2008-2024 regarding a player contract information, like Average Annual Value (AAV), and number of years the contract is for. I will then collect the players average WARP for the past 3 seasons. As well as the players predictive WARP for the season following when the contract was signed. Utilizing a matching system in order to pair players up with someone of similar qualities. The matching system will be based on a player's past WARP then filtered by age. Now that the matching is complete STATA is used to run the regression relating the difference in a match’s predictive WARP to the difference between a match’s AAV. While we see a significant positive correlation between the two variables, it appears that there are more external variables that will have an effect on the difference between a match’s AAV.