Next big thing in NBA analytics might be moving from the external to the internal

During a panel discussion at the 2013 MIT Sloan Sports Analytics Conference on how football coaches do (or don't) use analytical information when making in-game coaching decisions, former New York Jets and Kansas City Chiefs head coach Herman Edwards presented a good point: The health of a team's roster — not which names are and aren't on the injury report, but the actual physical well-being of the human people on the field and the sideline — really isn't something fans can't know just watching the team from TV. Interestingly enough, though, it's become increasingly evident that it's not always something you can know in the locker room or at practice, either.

Want to know how serious Mark Cuban is about improving the Dallas Mavericks' visibility into their players' health so that they can not only get injured players back faster, but also prevent health from deteriorating to the point where a stint on the injured list is necessary?

"We're locking up our medical staff to longer-term deals than our players," he said during the conference's opening panel on Friday morning.

(Actually, come to think of it, given the Mavs' "get under the cap and avoid the luxury tax" salary structure and the once-and-future openness of the Bank of Cuban, maybe that's not that big a statement.)

Yes, we've come a long way from the mid-1990s, when — as Stan Conte, the former director of medical services and head athletic trainer for the Los Angeles Dodgers, said during the conference's injury analytics panel — it took 2 1/2 years to figure out how many total player games a Major League Baseball team had lost to injury because there wasn't yet such a thing as a disabled list log. But while NBA teams continue to develop better analytical tools and collect more information about what happens on the court, there's still quite a bit of work to be done on understanding what goes on within player's bodies and inside their heads.

Advancements might be coming there, too, however.

STATS vice president Brian Kopp, whose company makes and markets the vaunted SportVU optical tracking technology that's got NBA teams and statistically inclined types on tilt, said during the conference that the special six-camera system can pay major dividends for teams' medical departments thanks to the motion tracking information it collects on how far and how fast players run during a game.

That info, according to Kopp, can offer new insight into how quickly players get fatigued, how being tired affects their productivity and more, which could help coaches better manage rotations and substitution patterns to both maximize their players' effectiveness and (theoretically) reduce the likelihood that players suffer the kinds of injuries that some researchers believe to be associated with fatigue, including joint and ligament injuries. Using SportVU data for health/injury analysis was also one of the primary next steps forwarded by NYU-Polytechnic Institute professor Dr. Philip Z. Maymin in his research paper on how often individual players speed up and slow down while playing — with more data and time, he'd like to study whether players who accelerate more frequently tend to get injured more often and whether changes in a player's frequency or level of acceleration contribute in any way to the onset of injuries.

When you combine that with STATS' partnership with Apollo Medical Information Systems — which extends the tracking and data collection on players to practices and workouts by measuring distance traveled, speed, acceleration, heart rate and other things — you start to see the framework for more informed, intentionally designed training programs that could give teams' training staffs an opportunity to keep players in better working order for longer.

That's obviously an appealing prospect for teams reliant on older stars, like Cuban's Dallas Mavericks with Dirk Nowitzki and the San Antonio Spurs, whose general manager, R.C. Buford, emphasized the importance of getting as much information as possible on health, rest, recovery and when an injured player is legitimately ready to resume activity.

"I think we're trying to capture more data than we've ever captured," Buford said during Saturday afternoon's basketball analytics panel. "I think that the two fastest-growing areas [of study] are medicine and technology, and the more you put those two together, the better return-to-play criteria you'll have."

Of course, as fellow panelist Kevin Pritchard of the Indiana Pacers noted, investing in injury prevention analysis seems right up the alley of a Spurs franchise that hasn't exactly shied away from using, shall we say, unique approaches to managing rest and minutes in pursuit of long-term returns.

"It's cost us a little money," Buford responded with a slight smile.

As is the case with in-game performance, because only 15 of 30 NBA teams have purchased SportVU for their home arenas, that data set isn't yet comprehensive enough to draw significant conclusions on player rest, health and physiology. And for one executive of a camera team that recently lost three players to season-ending injuries in the space of three weeks ruefully noted at the basketball analytics panel, the day when there's enough information to find some real answers can't come soon enough.

"I wish we were doing some more injury analytics," Boston Celtics assistant general manager Mike Zarren said.

(For a good read on another element of injury and performance analysis, check out Ryan Weisert's chat with endurance trainer Dr. Philip Skiba at Suns blog Valley of the Suns.)

The drive to get more and better information extends to psychological profiling of players, too. The idea of having players work with sports psychologists is nothing new, but given how much of performance seems related to focus and mental acuity, it's not quite as prevalent as you might think. Cuban's Mavs not only employ a full-time sports psychologist, Dr. Don Kalkstein, but bring him out on the road with the team and have him sit behind the bench during games to be able to work directly with Dallas players who might be struggling in a particular instance.

While such interactions could have a positive effect on players, though, the psychological aspect of the game remains one of the most difficult things to effectively measure, as Houston Rockets general manager/Sloan co-founder Daryl Morey noted after Cuban referenced Kalkstein during the conference's opening panel.

To best-selling author and moderator Michael Lewis' question about what's hard to measure with stats, Morey immediately replied, "Team psychologists — Mark has no idea if his team psychologist is good," eliciting some laughs from the crowd. (Morey and Cuban also exchanged some playful barbs over the situation surrounding rookie Royce White, who seemed to be the unspoken impetus for some of the discussion of psychological profiling and training — Morey charged that the Mavs "were going to draft the same guy we did," with Cuban retorting, "Yeah, but we would have dealt with it differently.")

As you might expect given the massive amount we still don't know, opinions vary on the value of psychological profiling and testing, even among the league's smartest and highest-performing teams.

"I think if you spend time doing your intel work, talking to people who are coaching [a player] and playing with him, getting to know the person, the things you learn from there are going to be more important than any psychiatric metric to date," the Spurs' Buford said. "Everybody feels different on this."

On the topics of "everybody," "feels" and "interior stuff that's hard to measure," one area of study that could provide a massive amount of insight for NBA organizations, coaches and fans is team chemistry, which Daniel McCaffrey and Kevin Bickart of SyncStrength Analytics sought to measure. Their presentation applied approaches from their neuroscientific, psychological and physiological backgrounds to the tracking of heart-rate data for members of basketball and soccer teams, with the results mapped to motion charts showing how closely players' heart rates matched up during particular plays.

When the players involved in making a play happen — those who passed to one another, or were responsible for helping create space for a scoring opportunity to develop — did their jobs successfully, you saw the nodes representing them bunched tightly together in the center of the chart, showing that these players were operating at very similar heart rates, which has been shown in past studies of married couples to indicate lower levels of conflict, increased abilities to solve problems and higher likelihood of staying together longer.

There are millions of questions about how predictive measurements like these can be and just how much they can tell us about which groups of players might be better suited to share the floor than others; I unfortunately didn't catch the entire presentation, so I'd like to point you toward posts on it written by Andrew Lynch and Jordan White at Hardwood Paroxysm for more details. But even early-stage studies like this, and the increased attention to training programs, injury prevention, psychological testing and all the rest of it, indicate that NBA teams and the analytics industry that serves them are placing at least as high a premium on learning more about the kind of stuff we can't see as the kind of stuff we can. Before too long, a lot that we've always considered to be "intangible" might be anything but.

More from the 2013 MIT Sloan Sports Analytics Conference:

Needles, haystacks and failures to communicate: The challenges of advanced stats
Optical tracking data has us asking better questions, but answers remain elusive
Sloan wrap-up: Loving Larry Sanders, Stan Van Gundy's secretly an analytics guy and more