Dissertations and Theses @ UNI


Open Access Thesis


Football--Iowa--Cedar Falls--Statistical methods; Iowa--Cedar Falls; University of Northern Iowa--Football--Statistical methods; University of Northern Iowa; Academic theses;


While the popularity of American Football has grown over the last century, so has the complexity with which the game is played. The hours spent by coaches, players, fans, and experts dissecting every aspect of the game is astonishing. While some people analyze football as a hobby or to report the game, coaches and player's careers are drastically affected by what they can draw from the data research. Statistical analysis strategies in football have become more complex and more widespread over the last 25 years, particularly since the use of computers and computer software has become so common place. Any small advantage in knowledge could mean the difference between winning and losing in the game of football today.

The purpose of this study was to investigate the relationship between selected team performance measures and game outcome for a college football program. Six team performance measures were included: turnover margin, explosive play differential #1, explosive play differential #2, rush yardage differential, red-zone efficiency, and first down efficiency. The University of Northern Iowa football team and their opponent's statistics were analyzed for this study. Game data from 2001-2004 was utilized. It was hypothesized that there would be a significant correlation between the selected team performance measures and game outcome. As a basis for this study, it was essential to determine which of the team performance variables yielded a significant correlation to game outcome, as well as determining which collection of variables provided the highest predictability of game outcome.

In order to determine that there is a significant correlation between the selected team performance measures and game outcome, correlational statistics were computed. Since game outcome, winning or losing, represented a dichotomous, nominal scale variable ( coded: loss = 0, win = 1 ), a point-biserial correlation coefficient was utilized. The correlation coefficients between the six team performance variables and game outcome for all games played, as well as conference and non-conference games, were found to be statistically significant. These results support the hypothesis that there is a significant correlation between the performance measures chosen for this study and winning football games.

Year of Submission


Degree Name

Master of Arts


School of Health, Physical Education, and Leisure Services

First Advisor

Jennifer Waldron

Second Advisor

Mick Mack

Third Advisor

Trey Leech


If you are the rightful copyright holder of this thesis and wish to have it removed from the Open Access Collection, please submit a request to scholarworks@uni.edu and include clear identification of the work, preferably with URL.

Date Original


Object Description

1 PDF file (59 leaves)



File Format