Abstract:

The National Hockey League (NHL) is the fifth wealthiest professional sports league, worth USD$5.09 billion, but it has received little attention compared with other sports such as football and basketball in the data analytics. Subsequently, there are few research works related to ice hockey game outcomes using statistical and data analytics techniques. The outcomes of sporting matches are of great interest to many stakeholders because of the money involved—club managers, sport analysts, bookmakers, sport fans, and potential bidders are interested in approximating the odds of a game in advance in an objective manner. Given the vast amounts of data available and generated during games, this research explores features related to real ice hockey games that may contribute to the outcome of a match, and identifies features that are correlated with the outcome to assist stakeholders in decision making.