The future of sports is algorithms, not athletes

Spectators sit stacked one row behind another, all craning their necks to look down at the miniaturized soccer pitch below them. With bated breath, they watch as a tiny player gently bumps the ball up the pitch toward its opponent.

The goal in sight, the attacker has two options — evade its opponent by expertly moving the ball around it, or send a safer pass to a teammate outside the fray. Like any soccer star, the player chooses glory and begins to putter its feet back and forth to move the ball — when it begins to lose balance. And down it falls, like a felled tree.

This is the joy and heartbreak of RoboCup (coming to a screen near you on June 22), where robots of various sizes and skills compete against each other to take home trophies (and honor).

While the footballing escapades of these tiny bots may look pretty innocuous, they point toward a cresting tidal wave of innovation and technology that is coming from sports as a whole. This includes a report published this May that teases a VR future for up-and-coming soccer stars.

It is in fact already here in many ways, Matt Winkler, director of American University’s Masters in Sports Analytics and Management, tells Inverse, which may be changing the very heart of sports itself.

A.I. has already changed sports as we know it

Sports are often seen as the pinnacle of human ability. From grueling training sessions to grueling games, sports are an example of not just physical endurance but mental endurance as well. But while athletes may regularly perform astounding feats to score goals or land dismounts, Winkler says these accomplishments can’t be chalked up to simple human ability alone.

“Too much technology would be if there is no decision made by a human in the whole process.”

When it comes to training, Winkler says it's standard for players to be decked out in wearables (similar but more advanced than your run-of-the-mill Fitbit) that can track everything from heart rate to oxygen consumption and muscle posture.

These biometrics can be fed back into intelligent dashboards, says Winkler, that crunch the data and spit back out new recommendations to improve performance.

“[Imagine] a golfer that's having a great season,” says Winkler. “He lifted certain kinds of weights ... but the computer taught him the right swing.”

While this might not be A.I. in the robotic sense, it is an example of intelligent algorithms helping humans find connections and patterns that might go otherwise unseen.

“You drive that passion and devotion into one click.”

Combined with predictive analytics of players’ performance or carefully measured game-time variables like a baseball’s launch angle or spin rate, Winkler says that the sports industry is manufacturing data at an alarming rate — and giving it all away.

And these data are not only changing how the game is played, but how it’s watched as well, says Winkler.

“From a consumer standpoint, everybody is one click away from everybody else,” says Winkler. “So you drive that passion and devotion into one click.”

This one-click behavior drives not only sports fandoms and fantasy leagues, but also good-old-fashion betting on the games, Winkler says, which is more important than ever.

But there’s one technology that Winkler says has yet to see its heyday in sports, at least when it comes to athletes: virtual reality.

When teams began experimenting with these headsets at least five years back, Winkler says they had held promise as a personalized training program for quarterbacks to play through simulations in real-time. But headsets of the mid-2010s were not robust enough to actually achieve this in any meaningful way, and the trend fizzled out.

But the resurgence of this tech could be just on the horizon if a study published in the journal PLOS One this May is any indication.

Virtual reality and sports

It’s human nature to believe that our own actions — diving for a shot on goal or dodging an opponent — are completely unique and of our own control, but in reality, these actions actually follow very predictable patterns. In this case, Benedikt Hosp and colleagues used the minute movement of goalkeepers’ eyes (tracked using VR goggles) to estimate player ability.

“There are multiple approaches that use machine learning in sports, but using eye movements in combination with machine learning is a quite newer approach,” Hosp, the study’s first author and researcher at the University of Tübingen, tells Inverse. “These findings are especially important, as a lot of eye movements are subconscious.”

How does it work — To collect this data, the team rounded up groups of expert, intermediate, and novice players:

• For the experts, they measured the eye-tracking data of 12 youth soccer players
• For the intermediates, they measured the data of 10 regional league players,
• And for novices, they measured the data of 13 players who had no history of professional competitions in the past two years

All players were strapped into an HTC Vive headset and immersed into a gameplay scenario where the screen would cut just before they, as goalkeepers, would need to make a decision. The idea, the researchers explain, is that how the players tracked the ball with their eyes would correlate with their overall abilities in the goal. For example, how an expert watches the ball would be very different from how a novice does.

And this was exactly what they found.

“We were able to differentiate three different levels of decision-making skill by looking at the perception of the participants,” says Hosp. “The findings suppose that experts have a more fixed gaze strategy and are less distracted by the ball. Novices tend to follow the ball more often and thus, show longer smooth pursuits.”

Why do humans want this? This research is still only in its early days, but Hosp says it could be used in the future to create individualized training plans to help not only level up player’s physical skills but their cognitive ones as well.

“Young expert soccer players usually already have a lot of physical training,” explains Hosp. “In the several last years, clubs have been focusing on cognitive training too. Our system is meant to be part of a balanced perceptual-cognitive training which can also be done at home.”

And with better, younger players, says Winkler, there are more profit for sports as a whole.

The ethical implications — With the increasing role of smart technology and algorithms in sports, is there still enough room for human heart at the center of the game? It may depend on where you live, says Winkler. Sports fans in the U.S., for example, are much more willing to rely on automated replay or referee systems than sports fans across the EU are.

As for Hosp, he says that their research is not meant to replace human coaching, but simply augment it.

“I do not think that we will lose the human element in sports,” says Hosp. “We will just find new ways to train our senses. Coaches still need to think about training situations they want to address with their players, so we still need humans in that system.”

“In my opinion, too much technology would be if there is no decision made by a human in the whole process,” he adds. “So far, we are at a point where we only support the coach.”

What’s next? The future of sports technology is evolving every day to accommodate not only new technology but the increasing appetite of the fans and stakeholders themselves for more data points and more content to consume.

Will Olympians of 2050 have tiny microchips embedded in them to track their biometrics live during performances? Maybe, says Winkler.

But in the short-term at least, Hosp says that VR systems like the one they trialed are likely to become an important — if not essential — part of sports training in the next five to 10 years.