There has been much attention paid to predicting the winner of the NCAA Men`s Basketball Tournament each year. In this research, we concentrated on deriving a method to predict the winner of the NCAA Women`s Basketball Tournament. Various models were developed to predict winners in each round using data from the 2011 and 2012 tournaments to develop these models. The models selected were verified using data from the 2013 tournament, and then used consecutively to predict the results from the 2014 tournament before any games were played. The variables found to be significant in developing the models included the differences in the following seasonal variables: field goal percentage; average number of assists per game; average number of steals per game; average number of points scored per game; average percentage of 3 point goals made; average percentage of free throws made; average number of blocks per game; and seed numbers. The point spread models, when used continuously from rounds 1 through 6, had a correct prediction percentage of 76% in 2014, and correctly predicted all four teams making it to the final four, the two teams in the championship game, and the winner.
© Copyright 2014 International Journal of Sports Science. Scientific & Academic Publishing. Alle Rechte vorbehalten.