A Statistical Analysis of Landmark Conference Women’s Soccer

Author/Creator

Author/Creator ORCID

Date

2014-05

Department

Program

Masters of Education

Citation of Original Publication

Rights

Collection may be protected under Title 17 of the U.S. Copyright Law. To obtain information or permission to publish or reproduce, please contact the Goucher Special Collections & Archives at 410-337-6347 or email archives@goucher.edu.

Abstract

The purpose of this study was to determine if a statistically significant relationship existed between box score statistics and success in winning Landmark Conference women’s soccer games. The researcher used measures in the form of box score statistics from previously contested Landmark Conference women’s soccer games. An independent samples median test was used for each of the following factors: number of shots taken, shots on goal, shots on goal percentage, corner kicks, and the number of fouls committed. A stepwise regression analysis, in particular a discriminant analysis, was used to determine whether variables entered after the best predictor contributed anything additional to the predictive power of the model. Initially, both shots taken and shots on goal were identified as significant predictors. However, once the best predictor, shots taken, was taken into account, none of the others added any additional information to improve the prediction. Shots taken and shots on goal are correlated and predictive of winning versus losing. The other two actions—fouls and percent shots on goal—are not. Therefore, the null hypothesis was retained for fouls and percent shots on goal and rejected for the other two actions. Despite the demonstrated importance of taking more shots, a team may not put all of its effort into shooting, as many other aspects of a match will affect shooting rates. This points to the need for future research on factors such as passing accuracy and team possession that affect the volume of shooting opportunities generated within a match.