David Griffin on analytics

Home for all your discussion of basketball statistical analysis.
Post Reply
Posts: 2137
Joined: Tue Sep 10, 2013 3:58 pm

David Griffin on analytics

Post by rlee » Sat May 25, 2019 4:48 pm

“I think you can (gain an advantage) with the analytics side. Not necessarily relative to saying we have the greatest ‘Nerd Cave’ in the NBA, which is what we called our group in Cleveland. It’s not about that. It’s about, do we have the best ability to translate what the data reveals in a meaningful way? Because data is not an answer in and of itself.

“It’s only as good as quality of questions you ask it to solve. That’s what analytics needs to be able to do. That requires not just the right people crunching numbers, but the right people communicating them.”

Posts: 6250
Joined: Thu Apr 14, 2011 11:10 pm

Re: David Griffin on analytics

Post by Crow » Sat May 25, 2019 5:49 pm

Lots of people can ask questions. Players, coaches, GMs, owners, the first line quants, interns, people good at identifying research questions.

Between the number "crunchers" and the "communicator" there might be 1-2 links. The analytics director or somebody below that that studies the produced analytics and sifts out the relevant stuff found or identifies some outputs for use as new inputs to get further processing. If the analytics director has enough time for ALL of that, fine. But the volume of mid-stream review and reaction should be huge and may not be possible for the analytics director to handle. The analytic process probably needs mid-stream analytic USERS / reviewers. Do teams have enough? Probably not in terms of doing the max. You may not need to be a code expert to look at the outputs intelligently. It might be good to back away from the coding and focus fully on the quality review and utilization of the produced data. I would think most teams could use a few thousand more hours of review / use of the produced analytics.

If the frontline quants are spending most of their time producing analytics, they may nit be spending enough time reviewing it. The analytic director or somebody probably has what should be a near full time job trying to communicate the findings and checking / fighting for proper implementation. So hire more middlemen to carry stuff from output pile A to ready to be discussed pile B.

Post Reply