Regression study in professional basketball since 1990. Does the winning team always covers the point spread?.

Abstract. In sports betting it is easier to predict the winner of a game match than the team that covers the bet. Since, a winner team might not cover the point spread.
This study focuses on the relation of the variable win to the betting variable cover the point spread. The study is performed with data from professional basketball (betting lines and scores) and tries to answer the questions: Does the favorite team always win?, Does the winning team always covers the point spread?. In order to answer these questions, a regression analysis is performed taking into account the most and less winning teams, together with their betting variables since the 1990-1991 NBA season. The regression results are inserted in the SPXS expert system revealing an indirect factor analysis that correlates betting variables with teams winning percentages.

Key words: Regression, Expert System, Sports Betting and Analysis, Prediction Algorithms, Data Mining, Machine Learning

ntroduction