User:SportsAnalyticsInstitute/sandbox

Hockey Analytics is the process of using empirical analysis of hockey and is primarily used to determine player and team performance in terms of winning. Hockey analytics remains in its infancy due to perceived challenges resulting from a different underlying production function of winning than other sports that have made sport analytics popular (e.g., baseball and basketball). Hockey analytics will become a fundamental management tool when teams incorporate these production differences into their analytics methodology. Currently, derived statistics remain the prominent methodology utilized.

Perceived challenges

 * The continuous flow of the game with intermediate stoppages and frequent in-play player (i.e., changes on-the-fly).
 * The nature of the game forces a high degree of substitution rate between goals scored and goals allowed, or the likelihood of a net positive goal scored.
 * Varying benefits of goals scored and costs of goals allowed across score margin games states makes the high degree of substitution an important factor.
 * For example, the benefit of scoring a goal when leading by two goals scored is less than the benefit of scoring a goal when trailing by two goals. Therefore, with the exception of tied games, leading and trailing competing teams have different incentive to play relative amounts of offense and defense.
 * A high degree of teammate and opposing team interactions are included in player specific derived statistics.
 * Most player actions are not recorded and unobservable for analysis purposes.