Nylon Notebook: Nylon Calculus at the 2016 MIT Sloan Sports Analytics Conference, Day 2

Nick Turchiaro-USA TODAY Sports
Nick Turchiaro-USA TODAY Sports /
facebooktwitterreddit
MIT Sloan Sports Analytics Conference
Nick Turchiaro-USA TODAY Sports /

Nylon Calculus is well-represented at this year’s MIT Sloan Sports Analytics Conference and we’ll be bringing you our thoughts and reactions throughout the conference. Check back in here during the day for updates.

What’s left to learn?

At today’s Sloan Conference panel, Basketball Analytics: Hack-a-Stat, the guiding questions was, what’s next? With all our advances in technology, data, and the understanding of optimal on-court strategies, what problems are left for analytics to solve. Moderator Zach Lowe started by asking the panel, Mike Zarren, Brian Scalabrine, Brian Kopp, Tom Thibodeau, “So, what do you do, what are you doing right now?”

Zarren answered that all the problems and questions are the same, saying about his organization’s focus, “It hasn’t changed we’ve just got better data.”

The conversation meandered through strategy, roster building, biometrics and training, but the focus was always in the same place. Basketball analytics has provided rich new data sets and work with that data has yielded tremendous insights. The focus now is figuring out how to implement those insights, turning the information into something useful that helps players reach peak performance, that helps teams play better together. The new glut of biometric data is the most obvious example, figuring out how to manage minutes and train for specific minute ranges and situations.

Distance traveled is an element of the public SportVU data set, and one as often maligned as useless. But Zarren talked about just how important that sort of information was.

“Should you train to run three miles or five miles? Those are really different training regimes, right? We didn’t know until we installed the cameras during the playoffs in 2010, how far you guys were running during the games.”

That’s a potential development, derived from analytic data, acted upon by athletic trainers using their own continually evolving techniques and technology, and one that you might never see reflected in strategy or style of play. Brian Kopp, formerly with SportVU and now with Catapult Sports, reinforced this idea that the frontier of analytics is putting new information into many of the existing organizational structures.

“When we started doing the SportVU stuff, I remember talking to Mike [Zarren] about this in the early days, people assumed that when you had new data it’s going to flip everything on its head. And that was never the point. There are a lot of coaches out there that are great at their job. It’s really just furthering and advancing what you already know in a different way. Maybe challenging you on a couple things, but mostly reinforcing what you already know.”

Implementation is the piece that coaches, front offices, and training staffs have always worked on. Their role has not changed and the problems they are working on haven’t either. Analytics has simply given them more information to work with.

— Ian Levy (@HickoryHigh)

Who Is the Target Market for Sloan?

Over the previous few years, the Sloan Conference alternated between the smaller, more intimate, centrally located Hynes Convention Center and the sprawling new development of the Seaport District’s Boston Exhibition & Convention Center. Attendance usually ranged between 1,500-2,500 or so. That’s massive for a sports conference, undoubtedly. But it was limited out at Hynes and practically unlimited in Seaport.

This year, Sloan has a record attendance or 3,900-plus, as Jessica Gelman shared at the intro panel on Friday morning. Which begs the question yet again: Who is the target market for the world’s leading sports analytics forum? The oft-stated description is that it is an ESPN-sponsored trade show. Attendees often skew young, with more and more high school and college students showing up each year.

But are we — the writers and consumers of a niche basketball analytics website called Nylon Calculus — the target market? We, as that segment of possible consumers, have complained more and more over the years about the lack of hard-hitting analytical research and insight in Boston. Or is the target market more skewed toward the traditional ESPN fan, who is very thrilled to be in the presence of former players and hush-hush sports executives? To them, it might just be good enough to keep throwing out the same panels and the same analytics v. gut instincts questions every year.

It’s true: This year does have more panels, more workshops, and I did just enjoy my time in a crowded-as-heck Introduction to R talk. Maybe having 4,000-and-counting attendees can enable Sloan to stretch out the programming to tinker with more education-based and more hard-hitting discourse. But in the end, I often do leave frustrated each year. And I probably should just admit that Sloan isn’t really for me. It’s good for networking. It’s not really about the content.

— Jacob Rosen (@JacobLRosen)

The Softer Side of Analytics

This afternoon, the Sloan Sports conference hosted another basketball panel, entitled Modern NBA Coaching: Balancing Team and Talent. The panel featured some of Basketball Twitter’s favorite coaching punching bags — Vinny Del Negro, Mike Brown, and Scott Brooks. It was a curious mix since all three of those coaches have been derided at times for embracing “old school” styles and strategies at the expense of more progressive strategies. None of the three really did much to change perceptions of themselves as analytics-savvy.

Brooks, in particular, struggled to articulate a consistent framework for how analytics had influenced his own coaching, or how it should influence NBA coaching in a generalized way.

A few of his comments on biometrics and training data were particularly troubling. He talked about the physicality of Russell Westbrook, acknowledging none of the nuance of training data is meant to provide.

“He can probably play back-to-back in the same day, he has that much energy. With the biometrics and the wearables, there are certain things that you don’t need to know. With Russell [Westbrook], you know that he’s going to bring it every night. You don’t need to know he’s going be fatigued with four games in five nights.”

No one doubts that Westbrook will play hard every minute that you let him on the floor, regardless of the situation. The question is whether those kinds of minutes increase injury risk and whether that is really in the best interests of the player and the team. A little later in the panel, Brooks was dismissive of some parts of popular data sets.

“You want real data. You want something that’s going to help yourself, your coaches, and your team. You know the fun stuff that’s out there…you know, he runs 4.2 miles during the game, Russell Westbrook dribbled 16 times on a possession…those are all great things to look at but it doesn’t really tell anything that you need to know.”

The fact that he derided distance traveled as something you didn’t really need to know, stood in pretty stark contrast to the basketball analytics panel this morning where Mike Zarren quoted that exact statistic as something that had helped the Boston Celtics begin to develop individualized training regimens.

There were some interesting comments — not related to analytics — early in the panel about the importance of fostering a positive culture for players and generating buy-in from the entire roster. However, the ideas seemed increasingly off-tune as the discussion moved through using data and implementing analytic strategies — things that are increasingly becoming part of the culture of many teams, things that this group doesn’t seem to have bought in on.

— Ian Levy (@HickoryHigh)