Nylon Questions: How can an analytic approach be applied to rest and injury prevention?

Sports analytics is a constantly evolving field and keeping up can be a challenge, especially with so much work being divided between the public and private spheres. As we head into the 2018-19 season, Nylon Calculus wanted to take stock of where we are and of what’s coming next.

This project is a throwback to work Keith Woolner did for Baseball Prospectus nearly two decades ago, and an update to Kevin Pelton’s basketball-specific version from five years ago. Our staff compiled 10 questions whose answering will likely guide the next few years of public analytic work. Not knowing what has been accomplished in private by NBA teams and consulting firms, we focused on questions that could be worked on in the public sphere, wouldn’t have to be answered with existing datasets (we don’t want to imagine the data we have now is all we’ll ever have) and things that would theoretically have an effect on how teams operate on and off the court.

Hopefully, these questions will help spark, refocus, and recalibrate conversations and lead to collaborative progress here at Nylon and everywhere else sports analytic work is being done.

6. How can an analytic approach be applied to rest and injury prevention?

One improper cut. One mistimed jump. One off-balance dive. That’s all it takes to hear a dreaded “pop” or to pull up limping. Human bodies are marvels of engineering, but when subjected to the rigors of professional athletics at the highest level, it’s a wonder they don’t give in more often. Of course, athletes are acutely self-aware of that risk. That’s why players like LeBron James invest so heavily into techniques, routines, and technologies that optimize their physical fitness. The name of the game is minimizing risk.

Unfortunately, nature still finds a way. Tom Haberstroh documented earlier this year how, despite all the advancements in physical training, injuries were still up significantly in the most recent NBA season. In February 2018, Haberstroh noted that “there were 3,798 games missed due to injury, up 42 percent from the same portion of games last season.” In a strongest-link sport that is only as great and as popular as its best players, not having those players be on the floor night in, night out is a head-turning challenge.

There are two different phases to this: part one, which I will attempt to focus more on today, is about minimizing the risk of injury. Part two is what happens after injury. As I alluded to earlier, the human body is a machine. That’s certainly as good a place to start as any because it means we can apply mathematical and engineering models to understanding the biomechanics of every action!

For example, let’s take ACL (anterior cruciate ligament) tears, which were responsible for felling the Large Latvian Son Kristaps Porzingis last season. Research shows that neuromuscular imbalances are a primary mechanism for causing ACL injuries. Specifically, when an athlete allows his or her ligament to absorb most of the force from movements as opposed to distributing that force along the constituent muscles in the area. Look at this free body diagram of forces acting on the ACL from that same paper:

V indicates valgus loading, representing an external force that causes the knee to move horizontally from the center, a common occurrence that can arise from any sort of dynamic cut or landing from a jump. Typically, the quadriceps force (Q) and medial hamstrings (MH) are responsible for helping stabilize the ACL. However, like Meyer, Ford, and Hewitt write any reduction in those muscle forces under an external valgus load shifts more stress onto the ACL. Keep accumulating load, and boom goes the dynamite (simplistic depiction, to be sure, but illustrative nonetheless). ACL tears may seem like freak injuries, but in fact, there are biomechanical models to help us understand how those ligaments get stressed.

Muscular imbalance, improper mechanics, overloading joints — these are all things that can be measured and trained against, at least to an extent. Scientific research has enabled us to come a long way in this regard, such as with this sample study from Quatman et al. using cadavers to simulate landings after a jump under various conditions and evaluating the resulting effects on ligament stress in the knees.

So we know we can calculate how our body moves, and should only expect the biomechanics modeling to get more and more sophisticated with more research. That’s great, but how can we actually take advantage of that ability? Well, we need to be able to collect the data of course!

In an analytic approach, the ability to collect data and how we observe the world around us is intertwined with how we utilize the data. Sports are obviously not akin to a highly controlled laboratory study. A live game is a frenetic mess of bodies and collisions and high-speed actions and reactions. That means in order to be able to collect the data required to better train and monitor athletes’ bodies, we will need to continue utilizing wearable technologies, a path that the NBA has already started down.

NBA teams already utilize products from companies like Catapult during practices in order to collect data on factors like heart rate, and sometimes go as far as monitoring sleep patterns. As part of the most recent collective bargaining agreement, the use of technologies such as Catapult in practice became allowed. Additionally, an advisory committee on wearable tech was created that will, as Haberstroh wrote, “examine which devices, if any, will be permitted in the future and how the data will be monitored, protected, and potentially monetized.”

Unlike in the MLB, which allows the Motus mThrow sleeves for pitchers to be worn in a game, no wearables are approved yet for use in NBA games. As detailed in this Atlantic article by Jeremy Venook, the legal implications are massive, and the league is still struggling with how to properly integrate the use of wearables in a controlled manner. It’s easy to envision how this data becomes far more advantageous for the teams than for the players and begins to contentiously creep into things like contract negotiations.

One should never expect athletes to give up total control over their bodies’ data, but with the way that professional sports are heading and the immense potential benefits of biomechanical tracking, it only seems a matter of when, not if, wearable technology will become omnipresent in the NBA, both in and out of games. And as Tom Haberstroh alluded to with the use of the word “monetized,” the possibility certainly exists that this data, or at least a limited subset of such, will become available for folks outside of teams to get their hands on.

Imagine being able to parse through data from a simple 9-axis inertial motion unit (comprising an accelerometer, gyroscope, and magnetometer) inserted into a player’s socks from a mid-season game, juxtapose that against play-by-play and video, and analyze how the mechanics of different possessions (a drive, a fast-break, a hard off-ball cut) affect the accumulated load on the player’s legs, then using that to potentially predict how that player might play on the second leg of the back-to-back the next night. And that’s just a fraction of what is possible (not to mention a tiny part of what teams and players might privately have access to) with the advent and integration of wearable tech.

The road forward is fraught and the challenges are obvious, but if there’s even one or two less Kristaps Porzingis cases a year, the league, its fans, and everyone in the orbit of the NBA will be better off for it.