|
Activity Number:
|
453
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Sports
|
| Abstract - #301599 |
|
Title:
|
Determination and Analysis of Factors Determining the Outcomes of National Football League Games
|
|
Author(s):
|
Christopher Cohea and Mark E. Payton*+
|
|
Companies:
|
Oklahoma State University
|
|
Address:
|
Department of Statistics, Stillwater, OK, ,
|
|
Keywords:
|
Football ; Sports ; NFL ; Logistic ; Regression ; Factor Analysis
|
|
Abstract:
|
The outcome of a National Football League game depends on numerous factors related to events that occur during the game. Over 40 variables were collected on a sample of 534 NFL games from the 2004 and 2005 seasons. Data from only one team (selected at random) will be utilized for each game. Logistic regression is employed to assist in understanding the relative importance of the collected variables. Due to multicollinearity among the predictor variables, factor analysis was used to determine if a set of underlying characteristics exists that explains a significant portion of the measured variability. The factors are named according to their perceived characteristics and are used as predictor variables in a logistic regression model. Game outcome predictions are also explored with both the logistic model with factors and the logistic model with the original variables.
|
