Seeing the Future
How Kirk Goldsberry's "Biomechanics of Buckets" talk was 2 years ahead of being 10 years ahead
Every other week I write an email discussing what I learn launching and growing Reboot Motion. If you would like to receive it directly in your inbox, subscribe below.
Kirk Goldsberry is no stranger to innovation.
While many know him as the author of the 2019 book "SprawlBall”, or for his popularization of visual and spatial analysis during his time at ESPN, his journey started a decade earlier- and included stops at Michigan State, the Spurs, and USA basketball.
So, when I came across his 2022 Sloan Analytics Conference talk targeting biomechanics as basketball’s next revolution, I immediately got excited.
Seeing the Future
One of Reboot Motion’s large initiatives is bringing movement analysis to the NBA. With the arrival of 3d pose data, we strongly believe the next decade will mirror what occurred in MLB over the previous one:
A shift from understanding why a thing happened to how it happened.
The implementation of better data, both in-game and in labs, for player development, strategy, scouting, and more.
An increase in coaches’ and athletes’ thirst for information.
While we like to think we’re innovators, Goldsberry was already seeing this trend two years ago, over a year before Hawk-Eye’s, and therefore 3d pose data’s, entrance into the NBA.
Finding a Starting Point
As Goldsberry thought about biomechanics in basketball, he did the logical thing and looked to baseball, which was similarly ahead in the rise of advanced analytics 20 years ago.
In MLB, biomechanics found its beachhead with pitching. This wasn’t because it was the only use case- hitting is catching up, and we’ve seen teams start to look at catchers and other fielders. Rather, it was because it was the simplest, highest leverage starting point.
For basketball, we should start with shooting for the same reasons.
When diving into shooting, Goldsberry notes the majority of missed shots miss short or long, as opposed to left or right, deducing that most errors are, first and foremost, velocity errors.
This led to a hypothesis that the best shooters are the ones that best control velocity, which they likely do via consistent, repeatable kinematics.
In order to test this belief, Goldsberry got even more specific and targeted free throws, which strip away the variability (location, defense, etc.) inherent with most shots.
And in his early, admittedly limited, research, that’s exactly what he found: better shooters had more consistent joint angles, joint locations, and timing.
Looking Forward
A talk on basketball biomechanics in 2022 is by definition just a starting point- and Goldsberry himself would tell you he has more questions than answers.
Is consistency the only thing that matters, or are there other commonalities?
Can these takeaways be taken to the court by a coach?
What about all other shots, that have far more variability?
Maybe the most interesting part of Goldsberry’s talk was how much has changed in our ability to tackle those questions in just two years.
For the study above, he stated, “the only way to get accurate biomechanical data [was] to use compression suit markers…the NBA 2k suit”. This meant he, and his research partner Kait Jackson, were working with one hand tied behind their backs- as asking athletes to “marker up” meant 1) limited access to data and 2) a natural deviation from one’s normal environment.
However, they both knew this was just the beginning, with Jackson looking forward to a time when advancements in computer vision allowed us to “light these markers on fire”.
Thankfully, those advancements have arrived.
Due to in-game motion capture, teams have access to 50 million movement specific data points every time they take the court- data that can be transformed to impactful metrics and 1) incorporated into existing models and analyses or 2) used as the backbone of reports for coaches, training staffs and front offices.
Still, even with the seismic shift in data availability since Goldsberry’s Sloan talk , we’re only in the first inning quarter of biomechanics adoption, which will allow us to continually innovate as we move backwards on the data timeline.
Goldsberry ended his talk with a vision for the future- and it’s one I agree with. One where:
Front offices have in-house biomechanists involved in modeling and player development...just like baseball
Teams build labs to systematically help their athletes improve... just like baseball.
The franchises that embrace innovation will reap the rewards…just like baseball