Abstract:
|
Broadly speaking, statistics can be categorized into description, prediction, and (causal) inference. Sports analytics are often focused on description (eg. What happened in the game last night? Who is the greatest player of all time?) and prediction (eg. How good will this newly drafted player be? How many games will a team win this season?). However most questions of interest at the team level are about causal mechanisms (eg. How effective is this play call? How do we win more games?). We will review several papers addressing causal problems in various sports and discuss a roadmap for better causal research in the field.
|