Park effects

by Leber, Kyle.

Jim Albert, Advisor The way a baseball stadium is designed can have an affect on the events, hitting, defense, etc., of the game. The affect is called a park effect. Many baseball analysts in the past have tried to account for park effects when studying the events of a baseball game. This study also tries to account for park effects using three specific events. They are the home-game homerun percentage for a team, the number of doubles and triples divided by the number of balls put in play for a team at their home park vs. all the other parks in which they played, and finally, the number of hits not including homeruns divided by the number of balls put in play for a team at their home park vs. all other parks in which they played. Given some of the ballpark’s measurements, average distance to the wall, area of fair territory, area of foul territory, and average wall height, along with the ballpark’s elevation, if the park has turf or grass, and if the park has roof or not, it is possible to build regression equations to that will give the analyst a way to estimate the three events described before. This study compares regression equations (employing best subsets and stepwise regression) estimate of variance and ability to predict the three events described above. The conclusion was that the predictor variables area of fair territory and elevation predicted the first event the best, the variables average wall height, elevation and roof or not predicted the second event the best, and the variables average wall height and elevation predicted the third event the best. iii The analyst now has a fast effective way to account for some of the affects that a ballpark will have on certain events of the game. iv