|
Activity Number:
|
391
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics in Sports
|
| Abstract - #307670 |
|
Title:
|
Prediction of Hitting Performance in Baseball
|
|
Author(s):
|
Shane T. Jensen*+
|
|
Companies:
|
University of Pennsylvania
|
|
Address:
|
The Wharton School, 463 Huntsman Hall, 3730 Walnut Street, Philadelphia, PA, 19104,
|
|
Keywords:
|
Baseball ; Bayesian modeling ; Prediction
|
|
Abstract:
|
There is substantial public and private interest in the projection of future performance in sports. We have developed a statistical model for the prediction of hitting performance for major league baseball players. The Bayesian paradigm provides a principled means for balancing past performance with crucial covariates, such as player age and position. We compare the performance of our model to current sabermetric methods on the 2005 season, and discuss both successes and limitations. We also discuss our procedure in the context of past investigations into shrinkage estimation of baseball performance.
|
