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.
In-game motion capture data is coming to the NBA. For teams, this means a lot of questions- the biggest one being “what are we going to do with all this data”?
A too common answer, and a wrong answer, is nothing.
Just ask any MLB club that dismissed ball flight or biomechanics over the past decade.
Or, ask your predecessors who have only recently caught up to the three point revolution.
The truth is, a new data source should be viewed the way the tech world looks at any innovation- even if we do not know how the story will end (the early internet in the 2000s, AI today), we have to stay ahead of the competition and dive in.
What Not to Do
With any new technology, people routinely take one of two approaches: 1) jump to the end or 2) wait and see. True innovators do neither.
Those that jump to the end require perfection today. Whether it be demanding in-stadium systems capture flawless, lab quality data, or assuming they should find groundbreaking takeaways immediately, they let perfect be the enemy of progress.
Others simply wait. Sure, this has its advantages- investing in biomechanics in a few years will likely come at a lower financial cost with quicker turnaround times- but you’ll always be behind, learning 2024 lessons in 2026.
The Better Path
Teams are like any other business. They need to execute today, but they also need to innovate for tomorrow.
To properly balance two competing priorities, franchises need to identify single threaded leaders. The same way teams have a head coach, who is singularly focused on this year’s championship, they need a leader for tomorrow.
Your innovation leader- likely a Director of R&D reporting to an analytics focused AGM- is not stressing today’s wins and losses. On the contrary, they are in charge of discovering tomorrow’s advantages.
Goals and Hypotheses
The innovation leader will have a scope beyond leveraging motion capture data. But in today’s (and tomorrow’s) world, it must play a factor.
To get the right plan in place, the leader must define their goal, develop a few hypotheses, and bring in the right team and system to execute.
Goals may vary slightly between organizations, but they’ll usually be centered around player development, scouting, or on-court strategy.
Hypotheses, however, will differ widely. The following are some of the things this new data source may unlock:
For player development, there is an inherent tradeoff between launch angle and velocity. All players may benefit from having a minimum launch angle, and each player may have an ideal shot trajectory based on how they miss.
For scouting, there may be better ways to judge on-court athletic ability. Specifically, we may be able to measure how quickly and powerfully someone jumps when rebounding. And that may be a better indicator of success than “combine metrics”.
For on court strategy, different defensive closeouts may affect different shooters in various ways. And different defenders may be more susceptible to drives in a certain direction.
These hypotheses- as well as the broader goals they’re based on- are dynamic. However, having a V1 prior to taking action is vital. It aligns the front office, coalescing everything around a common vision with an empowered leader.
Then it is about execution.
Execution
Motion capture does not live in a vacuum. To properly make use of it, the innovation leader needs to balance 1) movement data intricacies with 2) ensuring the information gathered can be brought into the organization’s pre-existing workflow.
Movement Data Intricacies
Let’s start with what actionable movement data isn’t.
Actionable movement data is not 29 key points of all ten players tracked 60 times per second continuously for 48 minutes.
Rather, actionable movement data is what can be derived from those 50 million plus data points. It is joint angles, shot trajectory, energy flow, balance, workload, and more.
But getting from key points to usable metrics is not easy.
To do so, teams need to have a Data Processing Platform they can rely on. Whether done internally by adding a team of biomechanists, developers, and data engineers, or using an external third party, a reliable data processing platform is a non-negotiable for analyzing movement at scale.
And before determining how to bring a data processing platform to your team, you need to understand what a world class pipeline entails:
Data Collection and Ingestion. Given the league wide deal between Hawk-Eye and the NBA, this is mostly taken care of.
However, teams may want to supplement this data with single camera motion capture (scouting, weight room, etc.) or their own marker-less systems in practice facilities.Data Transformation. Once teams have motion capture data, they need to transform key point data into stuff they can actually use.
This means doing inverse kinematics to calculate joint angles, then going further to actionable metrics such as momentum, torque, and energy flow.
And it means doing it at all at scale.Data Integration. Movement metrics are great, but they can’t live in a vacuum. Any output needs to easily pair with the existing work already being done.
Research. Once an organization ingests motion capture data, transforms it, and pairs it with the rest of their data and workflows, the fun begins. They can now use internal or external tools for 1) biomechanics research 2) model building and 3) report generation.
Implementation. Whether it be player development reports, valuation models, or game prep, research isn’t valuable unless it can be packaged and delivered to coaches and other decision makers.
Side note: Build vs Buy
A lot goes into turning raw motion capture data into innovation and insights. To build a world class R&D department, the innovation leader needs to decide what they can do, who they need to hire, and possible third parties to bring in.
What the current team can do vs who they need to hire is an org specific question that is based on their current makeup and their ability to bring in great talent.
Build vs buy, however, should follow a more formulaic approach.
Teams should build things that allow the franchise to economically create a competitive advantage. This is generally proprietary modeling, reporting, athlete programming, and coaching.
In other words: your organization has likely hired great coaches and great analysts. They should spend their time coaching and innovating.
Everything else- everything that isn’t the secret sauce of the people in the building- should be taken off their plate- and more often than not third parties can help.
For movement data specifically, the delineation occurs somewhere in step 4 above. Data collection, transformation, and integration are hard technical problems. Building the solution takes time and money, and constant improvement is needed. But they are not a pro sports team’s secret sauce.
Empowering Your Innovators
While value is accrued throughout the data processing pipeline, it is only realized when it is implemented by the experts in the buildings.
Therefore, as teams explore movement data, they need to ensure their solution is seamlessly integrated into what the rest of the organization is doing.
This means having:
A strong link between movement data and the core software and tools their analysts already use to test hypotheses, build models, and bring their takeaways to the relevant parties.
A reporting system that gets buy-in from coaches and the S&C staff, as the key translators for the athlete.
Conclusion: Setting Yourself Up for the Future
When thinking long-term, organizations should focus on the things that don’t change.
There will always be new
datainformation available as technology improves. Organizations that are open to exploration will always have an edge.
The best organizations will always prioritize great coaches, analysts, and leaders. They know these are the people that impact winning. And they know setting them up for success will always benefit the organization.
Outsourcing the non-essentials is always a great way to drive focus. Coaches want to coach. Analysts want to innovate. It will always be the organization’s job to give them the tools to do so.
At the end of the day, motion capture data is no different than any other new piece of information. No one knows exactly how it will be used in 2, 5, and 10 years. (And if they tell you they do- run.)
But the spoils will go to those who know we can learn from having x-ray vision into how athletes move. And it will go to those who know the race to innovate starts today.