It started inside a MIT classroom a decade ago, with Daryl Morey and Jessica Gelman teaching a small group of students about a relatively nascent subject: sports analytics.
Fast forward to last weekend at the Boston Convention Center, where nearly 4,000 folks from all walks of the sporting world gathered for the 10th annual Sloan Sports Analytics Conference, and this much was clear: Utilizing data to win games and make more money is now a requirement — not an option — for franchises around the globe.
GeekWire traveled to Boston and attended the SSAC for the first time, hearing from players, coaches, general managers, owners, entrepreneurs, and more about how data and analytics are impacting the games we love to watch and play.
The conference, once dubbed by sportswriter Bill Simmons as “Dorkapoolza,” attracted a wide range of folks from the sports industry, with more than 120 professional teams represented. Some of the high-profile speakers included people like statistician Nate Silver, NHL Commissioner Gary Bettman, retired NBA champion Shane Battier, Los Angeles Rams owner Stanley Kroenke, and many more. It was fun watching hoards of MIT students crowd around the speakers after each fireside chat, and I enjoyed catching up with fellow Seattleites like women’s basketball legend Sue Bird and prolific ESPN football reporter John Clayton.
While technology is impacting sport from a wide variety of angles, this conference mostly focused on how teams, players, and coaches are using data and analytics for positive benefit. Panels ranged from “Sports Tracking Solutions — The Next Generation” to “Sport Science: Extending the Athlete’s Peak Performance.” There were also data science-related demos and a handful of research papers that offered a glimpse into the future of sports analytics.
“I don’t know what’s next,” Morey said as the conference kicked off. “But I do know what’s next is here.”
Here were my three main takeaways from the conference:
From data to analytics
Given how many front offices have employees focused exclusively on analytics, it’s essentially become a requirement for teams to apply some type of analytical process to their business. But while there are numerous examples that prove analytics’ worth in sports, there are still people that don’t believe in its value.
Last week, Hall of Fame relief pitcher Goose Gossage went on an “anti-nerd” rant and said baseball is “becoming a freaking joke because of the nerds who are running it.”
Bill James, the famed baseball statistician who invented sabermetrics, spoke on a “Moneyball Reunion” panel at SSAC and was asked about Gossage’s comments.
“That’s what’s changed since 2002,” James answered. “You used to have to have to pay attention to those guys. Now you can just ignore them.”
Zingers aside, the wealth of data teams now have is more important than ever to help improve decision-making and overall performance. Retired NBA champion Shane Battier noted that data was “extremely significant” and “changed the course of his career.”
But one of the main themes of the conference seemed to be defining the difference between raw data and actual analytics. Just because a team can measure something doesn’t necessarily mean it’s useful.
“Not all numbers are inherently analytics,” noted ESPN reporter Bill Barnwell. “There is this replacement of the term ‘analytics’ for any and all numbers, and that is false.”
This is why we’re seeing partnerships inked like the one announced between Intel and EXOS this week “to make wearable data more useful and actionable for the user.”
“For wearables to become an irreplaceable addition to our everyday lives, we need to answer first and foremost the question of so what,” Intel executive Steven Holmes wrote.
As repeated in years past, several panelists noted that teams can’t completely rely on analytics, but rather use it as a tool to help guide decision-making. For example, you can’t use a wearable device to measure someone’s attitude or analyze numbers to grade the chemistry on a team — at least not yet.
“Analytics are like a bikini: they show a lot, but they don’t show everything,” said Golden State Warriors General Manager Bob Myers.
For some, the conversation about how and why teams use analytics to win is already growing old. ESPN NBA writer Zach Lowe moderated the basketball analytics panel and began by noting how “we do this every year.”
“We have the basketball analytics panel every year,” he said. “We talk about the same stuff: the nerds and the jocks and who can talk to who and how do we communicate. I kind of feel like the analytics thing is over. We know how the most efficient team should play. Every team has an analytics guy now. It’s part of the process. So I guess my question is, what’s left to do? What’s left to learn?”
It seems that technology tools which can give teams answers, versus just raw data, are most valuable in today’s world. Coaches and players have a finite amount of time to prepare for games; they need answers, not just numbers. This becomes even more crucial as companies develop new wearable devices and cameras to track fitness activity and overall performance, resulting in even more data collection.
Sports science — sleep, rest, recovery
Speaking of wearables, another key theme to the conference was the ability to measure and monitor a player’s health more effectively. That information is crucial when coaches make decisions about dialing intensity levels up or down at practice, as well as with recovery and injury prevention.
We’ve seen this play out with the Golden State Warriors and its historic dominance of the NBA in recent years. The team analyzes player movement with wearable devices and cameras, and uses that data to help the coaches figure out how much rest to give their players. It’s not the only reason why the Warriors — who, by the way, took home the “Best Analytics Organization” award at this year’s conference — won the championship last season, but it’s certainly a factor.
Another example of the focus on recovery is how the Philadelphia 76ers are employing a number of methods to keep players at peak performance levels. However, former NBA head coach Jeff Van Gundy noted how poorly the 76ers are playing this season.
“The way it’s being promoted, you would have thought they were winning,” Van Gundy said last week. “The process is important, but the results you get is actually more important.”
Van Gundy also noted that it’s tough to accurately measure “rest” when you can’t track what players are doing at the club the night before a game, for example. Added Van Gundy: “You don’t get a medal at the end of the year for being ‘most rested.'”
But being more rested can clearly help players feel better. Bird, the longtime Seattle Storm point guard, noted that her team eliminated pre-game shootaround sessions in order to give players more time to rest.
“Especially on back-to-back game nights, not having shootaround changed my world,” Bird said. “I didn’t realize how lacking in sleep I was until we cut it out. With that one change with my sleep pattern, I feel so much better.”
Isaiah Kacyvenski, a retired NFL linebacker who is now an executive with MC10, said that the teams will continue to research and develop ways to improve an individual player’s ability to rest and recover.
“If you own a team, it’s crazy for you to not want make your players feel as good as they possibly can come game day,” he said. “The ability to dial that in in the best way possible should be a given across the board. It will only continue to get bigger and quite honestly, it’s a very, very powerful thing.”
As wearable technology improves — don’t be surprised if players start wearing devices during games soon — so will the quality of data. Brian Kopp, CEO of Catapult Sports, noted that we may soon be able to build new connections between how certain movements and impact affect a player’s health and performance.
“You want to know about jump shots off dribbles? How about how many jump shots you have with a certain acceleration into that shot?” Kopp said. “How about when you set not just 10 screens, but 10 really hard screens? You start to measure all that impact on your body. It’s done in sports all around the world, but being able to tie that back to what it means — right now, sport science is a very trainer, strength coach area. Forward thinking people are thinking about how to take that into the front office.”
Added Kopp: “You can be much more precise with more data.”
Looking ahead, one of the research papers presented at SSAC dove into an injury prediction model for NBA players. This topic becomes even more important as questions linger about changing the 82-game NBA season, which some say is taking a toll on players’ health.
The future of sports tech
While panel discussions featured high-profile names, some of the more innovative ideas came from the Research Paper Competition. It’s here where ideas like the NBA’s real plus-minus stat first appeared, and is a good place to look for tomorrow’s next big sports analytics trends.
FiveThirtyEight has a solid rundown of the themes with this year’s research papers. Among them include big data sets, the use of machine learning, and classifying player types. Specific topics included the “hot-hand fallacy,” MLB brand engagement, ball screen defense in the NBA, and forecasting sponsorship revenue.
The winning paper was titled The Thin End of the Wedge: Accurately Predicting Shot Outcomes in Tennis Using Style and Context Priors. The idea is to analyze player-specific movements and ball striking to determine the likelihood of winning a point during a tennis match. The researchers analyzed nearly 40,000 shots at the Australian Open over a 3-year period.
Here were the finalists who presented at SSAC, with links to the full research papers:
- The Pressing Game: Optimal Defensive Disruption in Soccer
- “The Thin Edge of the Wedge”: Accurately Predicting Shot Outcomes in Tennis using Style and Context Priors
- Using Digital Signals To Measure Audience Brand Engagement At Major Sports Events: The 2015 MLB Season
- Sponsorship Revenue Forecasting for Sport Organizations: A Survival Analysis Approach
- Recognizing and Analyzing Ball Screen Defense in the NBA
- Accounting for Complementary Skill Sets When Evaluating NBA Players’ Values to a Specific Team
- Arsenal/Zone Rating: A PitchF/X based pitcher projection system
- The Hot-Hand Fallacy: Cognitive Mistakes or Equilibrium Adjustments? Evidence from Major League Baseball
And here are 15 other papers that had posters featured at the conference:
- Tracking Pitcher Performance with Instantaneous Component ERA and Moving Averages
- zWins, an Alternative Calculation of Wins Above Replacement in Baseball
- Testing Pitch Recognition to Improve Talent Identification and Player Development
- NBA Court Realty
- Preventing in-game injuries for NBA players
- The Van Exel Effect: Adjusting for Scorekeeper Bias in NBA Box Scores
- Classifying NBA Offensive Plays Using Neural Networks
- Is the Price Right? An Analysis of the National Football League Ticket Market Relative to the Great Recession
- The Competitive Effects of PEDs: MLB in the Post-testing Era
- Competition From Sports Hurts TV Ratings: How to Shift League Calendars to Optimize Viewership
- Analyzing In-Game Movements of Soccer Players on Scale
- Draft by Numbers: Using Data and Analytics to Improve National Hockey League (NHL) Player Selection
- Esports Analytics Through Encounter Detection
- Hot heads, cool heads, and tacticians: Measuring the mental game in tennis
- Finding the Open Receiver: A Quantitative Geospatial Analysis of Quarterback Decision-Making
The sports analytics industry has grown so much in the past several years, it’s tough to cover all the bases at a weekend conference. There were a few sessions related to eSports and fantasy sports, but you can expect more on those growing areas in the coming years.
One particular topic I wanted to hear more about at the conference was ownership of data, particularly as teams collect more and more information about their players.
During the NBA panel with Battier, Van Gundy, Bird, and hoops statistician Dean Oliver, ESPN reporter Kevin Arnovitz asked who owns all the player data, whether it’s health or performance-related. His question was generally met with silence and shrugs.
“It’s a public good, I guess,” Battier said. “I don’t know.”
Beyond just data and analytics, there are plenty of other fascinating intersections between the sports and technology worlds. We’ll explore many of these topics at our inaugural GeekWire Sports Tech Summit on July 12-13 this summer, and I’m looking forward to those conversations.
