Audience often comes into question when considering the Sloan Sports Analytics conference. For whom, exactly, are these presentations intended?
The answer to this has changed overs. Whereas in the past, the conference was more of a gigantic brainstorm, the purpose has somewhat shifted away from an academic mindset to that of a business conference. It still focuses on the advancement and place of analytics in sports, but not necessarily enhancing our understanding of the game.
Yesterday’s paper presentation, “Automatically Recognizing On-Ball Screens” encapsulated this shift in focus.
The process, thought and work behind the paper and presentation was beyond admirable. However, the presentation did nothing to advance how we comprehend basketball. One could almost hear the specter of Stan Van Gundy hovering in the back, yelling at the presenters that he doesn’t need a computer to tell him what was and wasn’t an on-ball screen.
At first, the presentation was a little off-putting, if not insulting. This is an analytics conference, but it’s also a sports conference, one co-founded by the general manager of an NBA team. Recognizing an on-ball screen should be second nature to most in attendance. That teams set screens differently is an obvious fact, and not some sort of revelation. Why was this being presented at an elite, forward-thinking conference?
Herein re-enters the consideration of audience. The presentation wasn’t meant for writers, or for those wanting a new dimension of basketball revealed to them. There was no value in it for them. The value of such a program lies with teams and data loggers. Imagine the Synergy logger or team video coordinator whose job is made that much easier by not having to log every single screen. It doesn’t completely revolutionize those established programs, but it makes the process at least a little easier.
Still, even with this consideration in mind, this presentation was an example of the, for lack of a better word, “dangers” of the Sloan conference. The problem isn’t that everything can’t be quantified — we still don’t know the answer to that — it’s more some things may not need to be broken down at the most microscopic level.