What does "machine learning" really mean? Step inside for a closer look.
In a relatively short period of time, big data has become a big business in professional sports. The market for sports analytics is expected to reach almost $4 billion by 2022, and teams around the world are racing to find a competitive advantage. The way we play may soon never be the same.
Let’s take a closer look at how four sports have been radically changed by big data – and how forward-thinking teams leveraged new tools to reach greater heights of success.
Anyone who has read or seen Moneyball knows that baseball and the Oakland Athletics were at the forefront of the sports analytics movement – using data to identify players whose statistical contributions were being undervalued by traditional metrics – but the league has come a long way since then.
When the Houston Astros won the World Series in 2017 – the first championship in their 56-year history – they credited a fulsome commitment to analytics and data-driven decision making that began when general manager Jeff Luhnow was hired in 2011. Luhnow has acknowledged that, at the time, the team was among the least sophisticated in baseball when it came to analytics, and it took years to get everyone necessary on board.
Eventually, he made believers out of the rest of the organization, and the Astros began using analytics to determine everything from game-day lineups to defensive configurations to scouting. They’ve also managed to significantly boost the performance of their pitching staff by using data to change pitch selection, location, and even the height of the pitchers’ release points.
“Big data combined with artificial intelligence is the next big wave in baseball, and I think we’re just starting to scratch the surface,” Luhnow said. “There’s so much being captured. There’s radar and video at every facility in baseball now … we know what every person is doing on the field at all times. We know what the bat and the ball are doing on the field at all times. We now have information we didn’t dream we’d have a few years back.”
If Houston’s data-driven success seems like an anomaly, consider this: following their World Series Win, Astros manager Alex Cora left for a job with the Boston Red Sox. As part of the interview process, Cora peppered Boston management with questions about their analytics approach, ultimately arguing that they needed to be doing a lot more with data.
Management bought in, and the Boston Red Sox won the World Series the next season.
In 2017, venerable English soccer side Lincoln City F.C. made a historic and unlikely run to the quarter-finals of the FA Cup for the first time since the Victorian era, becoming the first non-league side to reach that stage in more than 100 years. And big data played a big role.
At the root of their analytic strategy was a video review tool called Hudl, which is used by 150,000 teams in more than 35 sports worldwide. Lincoln recorded all their matches and uploaded the footage to Hudl, which automatically generated statistical reports. The game could be broken down into playlists focusing on specific play types – penalty kicks, for instance – and coaches could tag those clips with comments or illustrations. Hudl’s game reports provided both raw and aggregated data, linked directly to video.
Players could access Hudl’s reports and video wherever they were, and management could even track how much time players are spending on the platform.
“I would say that the video analysis we conduct using Hudl has certainly played an important part in our cup run, as well as our position in the National League,” the team’s performance analyst, Glenn Skingsley, told the Independent.
“Look, it may only be 1 percent of the reason behind our recent success – or it may be a lot more – but we really believe that the analysis we’ve performed before matches has helped the team an awful lot.”
When the Philadelphia Eagles won the Super Bowl in 2018, it wasn’t just a win for a long-suffering fanbase – it was a win for the big-data analytics community.
As the NFL began to embrace the big-data revolution, it was the forward-thinking Eagles who were at the head of the pack. The team, which has its own analytics department, hired EdjSports to provide predictive and prescriptive tools for game planning, play calling, and post-game analysis. Using more than 20 years of play-by-play data, EdjSports’ EdjFootball division assesses a team’s Game Winning Chance at any moment of the game and helps coaches make decisions accordingly.
Perhaps the biggest impact this approach has had on strategy can be seen in fourth-down play-calling. Where conventional, risk-averse wisdom would compel most teams to punt the ball on fourth downs, the Eagles’ analytics team instead found that in many of those instances, they would have had a better chance at winning the game if they had gone for a first down or a touchdown.
Though it could seem random when the Eagles decided to strike, in fact, those situations were often decided before the start of the game or even the start of the season. Two members of the Eagles’ analytics team even communicated with coach Doug Pederson during the games.
The result? The Eagles went for it on fourth down 26 times during that regular season – second only to the Green Bay Packers – and converted a league-high 17 of those attempts. According to ESPN, the Eagles scored a touchdown or field goal on 13 of the 18 drives in which they converted a fourth down, and when they went for it and didn’t convert, the opposing teams didn’t score a point on the next drive.
Ultimately, the decisions still lay with Pederson, but it’s hard to argue with the results.
“He can do whatever he chooses to do, but when you have the resource of data, why not?” said Eagles owner Jeffrey Lurie. “It’s kind of a no-brainer.”
Big data has affected the NFL in plenty of other ways. As Los Angeles Rams executive Kevin Demoff explained at the Sloan Sports Analytics Conference, data is practically the only tool used now in deeper stages of the draft.
“We really let our analytics group lead us in the sixth and seventh round – and especially undrafted free agency, they run it all,” he said. “Your scouts may have seen these guys one time, but increasingly teams are looking to analytics to fill the gaps.”
And the most recent 2018-19 season was the first in which teams were able to see not only their own player tracking data but the full league’s worth of information. Some predicted this change to lead to an all-out analytics arms race in the NFL
“Make no mistake, it is going to be a separator in terms of your competitiveness, both in personnel and on Sundays,” one NFL executive told ESPN. “My belief is it will drastically help teams compete if they can embrace it and integrate it. I think it will be more of a separator early.”
While it’s true that the addition of Kevin Durant turned the Golden State Warriors into an unstoppable dynasty, their golden run is rooted in something less showy: numbers.
The Warriors’ front office was an early adopter in the world of basketball analytics, which has since taken the entire league by storm. They were one of the first six teams – along with Dallas, Houston, San Antonio, Boston, and Oklahoma City – to install SportVU cameras at their home arena after the technology was introduced in 2009.
The camera system collects data at a rate of 25 times per second for every player on the court, allowing teams to track speed, distance, and possession data, which STATS then leverages to present performance metrics to players.
“We really didn’t know what we were going to do with that information, but we were going to figure it out,” owner Joe Lacob said. “It took us a while to understand the value, but we knew that was going to happen.”
Now, the team’s analytics guru, Sammy Gelfand, sits behind the bench during home games as the Warriors’ coaches rely on him for his in-game video skills and statistical insights. After every game, Gelfand spends two hours assembling a report that features passing numbers, situational results (for instance, how the Warriors defended pick-and-rolls) and overall offensive and defensive evaluations. He also compiles preview reports on the team’s next opponent.
Golden State is unique in the visibility of their analytics specialist, but not in how heavily they rely on the numbers.
The Houston Rockets are another famous case. Led by General Manager Daryl Morey, the Rockets perhaps more than any other team, transformed themselves around the notion that three-point attempts from the corner of the court – where the distance to the basket is smaller compared to the top of the key – were the most valuable shots in basketball. Conversely, long two-pointers were considered the least efficient shot.
The Rockets have now led the NBA in three-point attempts for six straight seasons. In each of the past three years, they’ve broken their own record for the most three-point field goals in a season.
But it’s not just the Rockets. The game has been transformed by the three-pointer, and that owes to data analytics.
“Now you look at a stat sheet after a game and the first thing you look at is the threes,” San Antonio Spurs coach Gregg Popovich said. “If you made threes and the other team didn’t, you win. You don’t even look at the rebounds or the turnovers or how much transition D was involved. You don’t even care. That’s how much an impact the three-point shot has and it’s evidenced by how everybody plays.”
NBA teams continue to look to data to gain an advantage. Israeli startup RSPCT uses a sensor attached to the backboard to analyze and track every shot, including the arc of the ball and the exact location that the ball hits the basket. Meanwhile, 14 of the NBA’s 30 teams now use KINEXON, wearable tech that tracks players’ movements on the court and allows coaches, players, and executives to monitor player performance and health through data. Analytics also play a role in when teams decide to rest players – Toronto Raptors fans will be familiar with the “load management” strategy used to manage star Kawhi Leonard’s minutes this past season.
Taken together, you can see that data has had a profound impact on the sport. NBA commissioner Adam Silver summed it up this way: “Analytics are part and parcel of virtually everything we do now.”
Looking to improve your data skills? Find out more about BrainStation’s data science and data analytics courses.