^^thanks for comment. I'm not an analytics person but pretty sure your closing statements are why hockey would look at shot attempts or shot attempts on goal or shots on goal and treats goals themselves as random events.
Correct. Hockey is also extra-screwed vs the other continuous games (soccer and basketball) because they’re the last sport without player tracking.
One of the reasons xG or expected points per shot has been more integrated into mainstream analysis elsewhere, is because you can get a readout of exactly how a team generated their statistical profile over hundreds of games. For example, NBA player tracking highlighted how the Rockets moved away from the pick and roll while maintaining their offensive efficiency, which led to the Clint Capela trade because they didn’t need a dunker.
With hockey, that same insight still requires hours of grinding tape, or access to hand-tracked data. So rather than “this player’s shot share is bad because they’re a puck-on-stick guy being asked to dump-and-chase,” you have to wade through a lot of “this player is bad because their shot share is bad.”
