Using Markov Models to Estimate the NHL Scoring Rate

Abstract: In hockey, the best measurement of offensive strength is goal scoring; the more goals a team scores, the higher the chances are of that team winning the game. But what factors affect the goal scoring rate? Does it depend on the current time in the game, or perhaps the time that the last goal was scored? The objective of this thesis is to develop several models for the goal scoring rate, and to determine how well the models fit the data. Some modeling techniques that will be discussed include regression, smoothing, hazard functions, and inhomogeneous Markov interval models.