This work is part of a retrospective collection of 179 electronic theses and dissertations (ETDs) from the VCU Libraries pilot ETD system that were designated as available only to VCU users. Please contact us at if you have questions or if you are the author of one of these and would like to release it for online public access.

Non-VCU users: Please talk to your librarian about requesting this thesis through interlibrary loan.

Defense Date


Document Type


Degree Name

Master of Science



First Advisor

Dr. Jason R. W. Merrick


The article offers a dynamic approach for predicting the outcomes of NFL games using the NFL games from 2002-2005. A logistic regression model is used to predict the probability that one team defeats another. The parameters of this model are the strengths of the teams and a home field advantage factor. Since it assumed that a team's strength is time dependent, the strength parameters were assigned a seasonal time series process. The best model was selected using all the data from 2002 through the first seven weeks of 2005. The last weeks of 2005 were used for prediction estimates.


Part of Retrospective ETD Collection, restricted to VCU only.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

June 2008