Whether Wearables Make Athletes Better Is Still Unclear — But They Seem to Be Deciding Pro Athletes' Fates

Leagues used to treating athletes as merchandise have found a new way to put numbers on their trades.

"Curry is healthy, he's only 26, and he's not playing!" said former NBA player and TV commentator Mychal Thompson on a Warriors broadcast. "What is going on here? How did this game suddenly turn into this?"

He said it on March 17, 2015, as the Golden State Warriors played the Denver Nuggets. The lineup Warriors head coach Steve Kerr sent out that night left fans staring: four starters on the bench, including Stephen Curry, a superstar paid more than $12 million a year.

With the team's starters off the floor, Kerr became the target of fan anger — Twitter filled with complaints and criticism, and some fans wrote in to curse him out.

The club, though, had no problem with Kerr's decision. After the game, Kerr wrote back to say his call had a scientific basis: data gathered before the game from the players' smart wristbands and questionnaires had been used to gauge fatigue, and the conclusion was that Curry was too tired to play.

Professional athletes have been handing their body data to clubs, and sometimes to the public, for years. In the NBA, the data a wristband collects no longer just decides whether a player takes the court, or how many games he plays in a season. It can now decide how much he can be "sold" for, and when.

Two years later, the use of that data has raised questions about fairness.

This January the NBA put limits on player wearable data. Clubs may no longer use monitored performance data as a bargaining chip when buying or selling players.

The rule is not alarmism. The NBA players' union says it has already received at least two cases of scouts abusing players' body data to negotiate trades.

Wearables have failed to win over consumers. Surveys suggest at least a third of smartwatches sit in a drawer gathering dust after purchase, and wristbands do worse. The Apple Watch, Fitbit and the rest are all stuck on the same problem: how to get more people to put them on.

In professional sport, the same devices have become hard to do without.

Someone Got Caught Cheating by His Own Smartwatch Data

Monitoring athletes' body data is now routine, and the data can be read backwards to work out what an athlete actually did and what shape she was in.

Last week a Garmin 325 running watch became the evidence that exposed a marathon runner's cheating.

Jane Seo, 24, is an amateur marathon runner who writes about food and sport for The Huffington Post.

On February 20 she ran the Fort Lauderdale A1A Half Marathon in Florida and finished second in 1 hour 21 minutes — a very good result for an amateur, given that the world record is 1 hour 5 minutes, only 15 minutes faster.

The day after the race she posted the data from the Garmin watch she had worn on the athlete community Strava: a GPS map, heart rate and cadence.

That data caught the eye of independent sports blogger Derek Murphy. From the length of the straight segments in the cadence trace, Murphy concluded this was cycling data, not running. He also pulled up running data Seo had posted long before and found that her average heart rate on race day was off.

Murphy then found a photo of Seo taken after she crossed the finish line and blew up her Garmin watch face in the shot. The numbers did not match what she had uploaded to social media: at the very least, the distance on the watch was several kilometers short.

The evidence was conclusive, and Jane Seo admitted cheating. She said she had run out of energy during the race and cut the course instead of following the official route, then faked the data afterwards by riding the same marathon route on a bike.

The More Commercial the Sport, the Closer It Watches Body Data

Marathons and other track-and-field events are no longer the biggest category for data collection. In basketball, baseball, American football and soccer — sports worth far more commercially — athletes have far more recorded about them.

Take the NBA. Besides the six SportVU cameras hanging from the arena ceiling, players in daily training wear a smart vest while warming up, with a GPS unit hidden inside that tracks acceleration and distance run. In shooting drills they add a chest strap that measures jump height, heart rate and breathing rate. Even asleep, a player may be asked to wear a smart wristband so the coach can monitor how well he sleeps.

The NBA's Toronto Raptors once made public the devices they use. In training, players carry a device called OptimEye, which holds a gyroscope, an accelerometer and other sensors and can detect exactly how an athlete moves: how he accelerates, decelerates, balances, how far he jumps, and how hard each of those movements is.

OptimEye picks up a player's movement habits, down to details the naked eye may miss.

How fine does it get? It can tell that a player pushes harder on one side of his body than the other, that he habitually holds one foot above the other when he jumps, that one foot lands before the other.

The data sounds dull, but players have to keep moving hard for long stretches. A player whose two sides work unevenly is more likely to strain one spot; a player with a particular jumping habit may find that jump training everyone else treats as moderate is, for his feet, heavy — and more likely to injure him.

A device like OptimEye collects that data and warns the coach during training, so the workload can be tuned to each player and fewer injuries happen off the court.

Catapult, the Australian company behind the device, helped the Raptors become the NBA's least-injured team in the 2013-2014 season. Orders from American sports clubs then jumped 70%, and today NBA and NFL teams, plus soccer and basketball teams under U.S. sports federations, bring the company $20 million a year.

Finland's Omegawave also makes small monitoring devices that catch signals the eye cannot see: whether a player is about to run out of gas, when the coach should substitute, how to stop overtraining before a match. A full hardware-and-software package costs $200,000 a year, but after Omegawave caught on with European soccer clubs it became standard kit for American football and basketball teams too. Catapult is cheaper, and still charges $100,000 for devices, software and upkeep for a 15-man basketball roster.

Plenty of tech companies want a piece of the game. Microsoft and Apple have both paid to get their tablets onto the sidelines. In 2015 Microsoft spent $400 million to promote its Surface tablet, buying the right to be the NFL's official sideline tablet, in the hope that coaches would use it to go over data and talk tactics with players.

Coaches have not been enthusiastic:

Neither have players:

Commercial monitoring packages are not cheap, and teams buy them anyway.

Take soccer. A club's money comes from four places: tickets, broadcast contracts with TV networks, commercial income tied to players such as endorsements and transfer fees, and league prize money.

For clubs that earn little from the commercial side, whether TV networks will pay more for their matches, and how much sponsors will put up, comes down mostly to where they finish in the league.

Leicester City, the mid-table club that won the Premier League last year, took in £128.7 million last season, which vaulted it into the world's 20 highest-earning soccer clubs — five times what it earned in 2013/14.

By earlier estimates, Major League Baseball teams pay $665 million a year in salaries to injured players, and NBA teams $358 million to keep injured players on the books, with the Los Angeles Lakers alone spending around $44 million. In the NFL, the average player salary is $2 million. The loss is not only the wages paid to injured men on the bench; it also shows up in what it costs to win games.

For a club with revenue in the hundreds of millions, spending a few hundred thousand more on monitoring devices and data — probably less than a team's annual travel bill — to cut injuries by a few percent, which is worth several million, looks like a decent trade.

For all the talk, though, it is hard to say whether monitoring really reduces injuries. The odds of getting hurt depend on an athlete's body, his state of mind, the weather, the condition of the field, and, of course, the coach's tactics and mood. Sometimes you know a player is fragile and there is simply no one else to put in.

Where Sport and Data Met: Buying the Right Player for the Least Money

Sport and data first crossed paths in the 1970s.

The idea goes back to Bill James and Charles Reep. James began publishing The Bill James Baseball Abstract in 1977, picking apart problems in baseball statistics; Reep had been recording and analyzing soccer matches since 1950, more than 2,200 of them in the end.

In 1982, Sports Illustrated writer Dan Okrent interviewed James and turned it into a story, which made The Baseball Abstract a national bestseller instead of a cult item. But the small group James was trying to reach — the owners of MLB clubs — still ignored his theories about data.

James did give rise to another company. From the 1960s on, Dick Cramer had spent his spare time using computers to analyze baseball data, and in 1981 he founded the data company STATS. "The company was founded to collect as much baseball data as possible."

Cramer's attempts to sell what he had collected to MLB teams went badly. No club was interested, and STATS was left in a bind.

TV networks and news organizations wanted the data instead. ESPN and USA Today were STATS clients from the start, and in 1999 Fox Broadcasting bought the company for $45 million. STATS went on to buy more companies, including Bloomberg's sports analytics arm and the well-known data firm Prozone, whose clients included the Premier League.

James and his theories did not really carry weight until 2002, when Michael Lewis's Moneyball hit the bestseller lists.

The book was built on a real case: the Oakland Athletics, a lower-middle-tier Major League Baseball team, used player performance data to match the results of the richest clubs while spending the least money. The A's were short of cash to begin with; their $40 million payroll was third from the bottom, a third of the league-leading Yankees'.

No money for good players or good coaches, and no way to improve results: that was the trap weak baseball teams could never climb out of.

Team manager Billy Beane brought in Paul DePodesta, a Harvard economics graduate, as his assistant. Working from Paul's statistics on how league players had performed in the past, Billy put together a team nobody rated, including a pitcher with a bad elbow and an aging hitter.

In the season's first games the team was unremarkable and stayed near the bottom of the standings. Things changed between mid-June and the end of July, when Billy started trading: selling players and buying cheap the ones he liked, the ones other teams had undervalued. The starting lineup got a transfusion; the team swapped out half its roster and a lot of pitchers.

After that, with Billy and Paul using the numbers to guide players in changing how they hit and pitched, the A's climbed off the bottom, won 103 games in the 2002 regular season and made the playoffs.

The A's experience in Moneyball changed how baseball thought about weak teams and player data: an underfunded club that spends little and takes data seriously may not win a title, but it can save itself, and can even do well.

2002 is also considered an important year for sport. After it, using data to succeed on the cheap became something professional coaches studied.

Baseball Was the First Sport Mass Data Changed

Moneyball is a story about a manager using data from the bottom up to change a team's results. MLB as a league used data from the top down to change how professional baseball developed.

"Remember, baseball was the first sports league on radio, then TV, then cable, and now on digital media," said Bob Bowman, CEO of MLB's subsidiary MLBAM. "Think about the 1950s — the three biggest sports leagues were baseball, boxing and horse racing. Two of them are gone. Baseball isn't."

MLBAM was founded single-handedly by former MLB commissioner Bud Selig. Born in Wisconsin in 1934, this old-fashioned American kept no computer on his desk — he usually had someone print out the news from MLB.com for him to read — carried a plain feature phone, and ate a hot dog with a Diet Coke every day at noon, year after year.

Selig also worried that baseball's audience was aging along with him. So how do you get young people to keep watching?

In 2000, not long after the internet became widespread, Chicago White Sox owner Jerry Reinsdorf put an idea to Selig: to stop rich teams from using the new internet to widen the gap between rich and poor clubs, the league should set up one company to run every team's website and share the revenue.

Many teams doubted it would bring in much, but the league's 30 clubs still put up $77 million to found it. For the first few years the company really did struggle and lose money, though it tried charging subscriptions early on. That turned out to work: 11 years later it had 2.2 million paying subscribers, and MLB.com was drawing 10 million visits a day.

MLBAM now does far more than run MLB's website. It is in the technology business: cameras inside the ballparks record a pitcher's motion and speed and deliver the game to broadcasters like ESPN within 25 seconds. Other companies have even used the technology to stream other sports; in 2011 alone it carried 270 million broadcast streams.

Since MLB brought in the StatCast camera system in 2014, it has gathered much more data and given fans better games and better analysis. In a typical game, the instant the batter is about to make contact, StatCast can put on screen the pitch's speed and spin rate, the speed of the ball off the bat, how long the catcher took to glove it, and more.

Last year BAMTech, a video-streaming technology company spun out of MLBAM, drew a $1 billion investment from Disney at a $3 billion valuation and became the partner for ESPN's online streaming service.

After Wearables Arrived, Sports Data Collection Grew Exponentially

The 1990s were the golden years for sports data outfits like Opta and Prozone, which began working with commercially valuable leagues — the NBA, the Premier League, baseball — supplying the most basic match, score and scoring statistics. Opta, a player-data company founded in 1996, got its first job from Sky Sports, which used Opta's numbers to measure what players contributed in a match; STATS followed a similar path, with its earliest orders from ESPN.

After 2013, with wearables and smart on-field monitoring equipment, sports data stopped being limited to outside statistics like scores and points.

"Since I got into this business 15 years ago, analytics has been a big part of what we do," Dallas Mavericks owner Mark Cuban told ESPN in 2015. "What's changed is that every team now uses every bit of data it can get to help make decisions, so the whole market has become more efficient."

MIT's Sloan conference is another example. It started with two teachers showing a group of students how to analyze games; 11 years later Sloan is where tech products, professional teams and data companies come together, drawing 4,000 attendees last year, with clubs from all four of America's most popular leagues — baseball, basketball, hockey and football — taking part.

Today's capture and counting leave nothing out. A single soccer match can be broken into more than 350 kinds of data — a player's running, tackles, position, time on the ball — and each match yields 1,600 to 2,000 such data points. Arsène Wenger, manager of Arsenal, gets a 60-page data report after every match, including an analysis of how each player moved.

Most of this collection happens in daily training, mainly because few of today's most popular leagues — the Premier League, the NFL, MLB — let players wear such devices during actual games. Even those leagues only changed the rule in the past two years, and allow only one or two approved products on the field. Other leagues ban them outright.

The Gambling Industry Is a Big Buyer of Data

Databases that cost hundreds of thousands a year, monitoring gear with radar built in, commercial packages that put a dozen cameras in one stadium — they have a market because someone is willing to pay. Besides the clubs, sports betting companies are the biggest buyers of data.

In August 2016 the NBA struck a deal directly with the data analytics and gambling company Sportradar: six years, $250 million, allowing NBA game data to be sold to betting firms. All four of America's big leagues have data-sale contracts. Sportradar had already signed the NBA, the NHL and the NFL; STATS holds the MLB contract.

Sportradar's data deal with the International Tennis Federation is not cheap either: $70 million over five years, more than $10 million a year.

Betting companies badly need live sports data, because they have to keep adjusting their odds.

Bet365, Britain's biggest online sports betting site, lets customers bet while a match is in progress. The numbers that matter most to a betting company are the ones it uses to set the odds on each team and guarantee its own profit, and Bet365's odds move with live data from both sides, down to whether a player is hurt or has been sent off.

That flood of constantly changing raw data does not come from Bet365's own staff. It comes from the sports database company RunningBall, which the British sports data group Perform Group bought for $135 million back in 2012.

Behind the prices betting companies pay for data are the profits sports gambling throws off. In 2015 Bet365's revenue passed £1.5 billion and its profit reached £400 million. For comparison: Manchester United, the world's highest-earning soccer club in 2016, had revenue — not profit — of £500 million.

If betting firms are this keen on mere match statistics, think what would change if players' body data, which bears directly on the result, could also be gathered and sold in bulk. Bettor or bookmaker, who wouldn't want to know the odds of Wayne Rooney getting hurt in his next match? Who wouldn't want to know how Stephen Curry's heart rate moves during a game?

Professional Athletes: How Much a Pound?

Analyzing an athlete's body data cuts both ways. It can lengthen a career by heading off overtraining and injury; it also means the athlete's own weaknesses can no longer be hidden from the club.

If the data shows a high risk of injury, will a club still buy him? Will his next contract pay less? The NBA has banned using players' body data in contract talks, but the loopholes are wide: the club holds the data and is also the party negotiating the new deal. Will it ever admit it used the data to push the price down? There are always plenty of reasons to cut a salary or sell a man cheap.

Brian Bulcke is one athlete who has spoken against wearables.

From 2010 he spent six years as a professional in the Canadian Football League, playing defensive line for the Calgary Stampeders, the Hamilton Tiger-Cats and others.

2015 was his turning point. At 28, Bulcke suspected his playing days were running out, so during the off-season he took a side job at Ryerson University's startup incubator in Toronto, advising more than a dozen sports data companies on their products.

That was when he saw the problem with being turned into data. Technology moves fast, and the performance data his team collected on him could be used to improve a company's commercial products as easily as to improve the team's results. He came to worry that the data served commercial ends more than it prevented injuries.

In the locker room, Bulcke and his teammates worried that data collection was blurring the line between private life and professional competition, but it was only last August, after his playing career ended, that he began telling reporters he opposed collecting athletes' data.

"When it comes to athlete data and technology, I feel like a lab rat. I think we're people too. We're professional athletes, and there should be a line that respects our privacy, one that guarantees the security and the limits of how our data is used. Athletes, and the products used on the front line, are defining this industry. We're the early users, which makes us the mouthpiece for every other athlete in the market."

Zu Bell, agent for NBA player Steve Blake, also suspects body data will end up as one more tool for driving prices down:

"What are clubs doing with this data? What are their rights? In the end, big data could become the pivot, the thing league teams use to bargain players down."

Kirk Lacob, son of Warriors owner Joe Lacob and the team's assistant general manager, thinks players are team employees and the team is doing nothing wrong. "From my perspective, we pay these players a lot of money, so we should be able to know what we don't know. This is negotiable. It can be negotiated, and it will be negotiated."

Clubs want to use players' body data to maximize their results and their income. In theory the NBA lets a player decide for himself whether to wear a wristband. In practice, the side paying the money always has the upper hand.

MLB, which took up data earlier, lets players use wearables voluntarily, but this collection is not covered by HIPAA, the Health Insurance Portability and Accountability Act, written to protect patients' medical records, which requires that patient information be kept private and held for six years.

In the NFL, where nothing limits how the data may be used, an athlete's body data could even be shared with commentators, with sports video games, with betting companies.

With body data this detailed, trading a professional athlete looks more like weighing two goods on a precise scale. A is $5 million cheaper but 5% more likely to get hurt than B; B's muscles are 15% more explosive than A's. Which one do you buy? However you decide, turning a person into data is itself a way of removing the person.

Relying on data can also lead to the wrong call.

People tend to treat what they know as what matters. It is like losing your keys in a parking lot and starting the search under the street lamp — not because the keys are more likely to be there, but because that is where it is easy to look.

Set against the complexity of the human body, the data now being collected is thin. And this partial data is becoming a key measure of an athlete.

Professional athletes, already bought and sold by clubs like goods, now look more like goods than ever.

Cover image from: Pixabay