The purpose of this study was to present the statistical model to predict the winning of ice hockey game and determine the contributing factors for win in the world ice hockey championship. In order to find the optimal regression model for ice hockey, we compared two regression models (logistic and linear model) with the database of all games and the seperate databases of top/bottom teams. The logistic regression model using the seperate database was most accuratly predicted the actual outcome of games. This model and database further revealed that goalkeeping and scoring efficiencies and the number of shots on goal were significantly contributing factors to win. In addition, the results for prediction analysis of winning rate for each team indicated that offensive skills were more inportant factors than defense power to increase winning rate for teams.
© Copyright 2018 ISBS Proceedings Archive (Michigan). Northern Michigan University. Published by International Society of Biomechanics in Sports. All rights reserved.