That Microsoft is (secretly, as of now) deep in the throes of crafting a successor to its wildly successful Xbox 360 console should come as little surprise. These development cycles are predictable like the seasons. But the specific information leaked to Kotaku about the system upgrade is quite encouraging: the Kinect camera system is getting a hefty improvement. Better resolution, more powerful tracking, and deeper integration with the new console mean the burgeoning Kinect hacker movement will have access to a new, more powerful gaggle of goodies with which to analyze…well, anything. And the implications for sports could trigger a revolution in not just how we play games but the way teams, agents, and specialists approach injuries.
To this point, the forefront of athletic-injury studies are specialty outlets like the American Sports Medicine Institute, founded 26 years ago by the esteemed Dr. James Andrews. Over the past decade or so, researchers down in Birmingham, Alabama, have studied more than 2,000 pitchers using basically the same 3-D motion-tracking camera system. Much like actors on a soundstage performing special F/X movements, athletes are dotted up with reflective markers that feed an approximation of biomechanical movements into the mainframe. Ten years ago, the cameras filmed at roughly 500 frames a second. To this day, the same camera speeds remain.
Enter Kinect’s souped-up infrared system, which doesn’t rely on reflective markers, but rather a camera that can collect many more data points, determining depth and forms by calculating the distance light is traveling from the sensor to objects standing in its range (read: people) and then back again. Every specification that made Kinect so addictive for sports games is purportedly receiving an sizable upgrade. According to the Kotaku report, here’s how the new camera will interpret movements, as compared with the original Kinect.
Now, this looks like a motion-tracking camera setup, with important joints and body positioning indicated by an unspectacular dot formation. But look closer, specifically at the hands and fingers, and that’s where the real breakthrough becomes clear. By further fleshing out the thumbs and top-hands, Kinect is going to make it possible to indicate more specific and minute movements within the hands and lower arms. Adding a spine joint, likewise, can help measure curvature of the back during sports motion, and an extra skeletal joint near the neck also means being able to integrate more biomechanical data concerning head movements and how they relate to the way we compete. Does the angle of Colin Kaepernick’s head change on screen passes? How about Tom Brady’s thumb position on Hail Marys? From now on, we can know these things.
Still, this sort of technology is painfully immature. Though Kinect was launched in the fall of 2010 and has sold more than 20 million units since, the integration with mainstream gaming, let alone sports bioinformatics, has come at a maddeningly slow pace. Madden NFL has had it for less than a year, and for all the features forthcoming on MLB 2K13, Kinect support still isn’t one of them.
Maybe a thriving Kinect hacks community will figure out a way to make that last part happen before Microsoft or 2K Sports does, and for all we know, some geeks collaborating at some university are on the verge of doing just that. (Discoveries like these have a way of staying under wraps, held instead for a splashy debut at events like the MIT Sloan Sports Analytics Conference.) Hackers (in the non-pejorative, more traditional sense of the word) are producing remarkable results from original $150 Kinect add-on. Low-cost 3-D modeling of the human body in minutes? You bet. Virtual real estate tours from afar? Sure thing. Heavy equipment training? Just move your hands around. The possibilities are, in the most realistic sense of the word, endless. It’s almost too weird to believe this sub-$200 peripheral can outperform scientific lab setups that often run six figures, but this is where we are.
So it’s entirely possible that the only reason why some jaw-dropping baseball (or football or basketball) biomechanics hack hasn’t been derived from the Kinect system yet is pure inclination. It’s not cost, it’s not time, and it’s definitely not availability. That the breakthrough hasn’t come yet, especially considering the secrecy that often pervades the analytics community, doesn’t mean it won’t happen soon.
The next-gen Kinect will be a mandatory component of every new Xbox console. When a company moves to make something so innovative (and so malleable) become even more ubiquitous, somewhere on the order of tens of millions of units sold over the next couple years, you’re asking your customers to be creative. It becomes a tacit request to please do more than what the box says you can.
In time, someone who buys a next-gen Xbox will figure out a compelling way to drill down on athletic movements in a manner that we have not yet seen before. Athletes will be evaluated on a set of terms that now exist solely in our imaginations. And when it happens, when we peg all that advancement to a low-cost camera system available to almost anyone, no one can honestly say they’ll be surprised.