Analytics launching victories – and thousands of careers
Tom Layberger | Wednesday, June 26, 2019
Not even Brad Pitt in the 2011 movie “Moneyball” could speak in terms of launch angle, exit velocity and spin rate.
The baseball general manager the actor portrayed, Oakland’s Billy Beane, introduced a different way of analyzing a baseball player’s abilities and how it could translate into adding to the team’s victories.
The explosion of analytics across sports has provided athletes, coaching staffs and general managers data that can be applied to provide a competitive advantage for an individual and a team.
Colorado Rockies’ second baseman Daniel Murphy serves as an example of how the application of analytic data can boost performance. According to the Washington Post, when the left-handed hitter was with the New York Mets in 2015, his percentage of balls in play that were on the ground was 42.8 percent and fly balls 36.0. That season Murphy hit .281 with 14 home runs and had a slugging percentage of .449.
A year later, when Murphy was a member of the Washington Nationals, in a similar number of plate appearances, his groundball to flyball rate was virtually reversed: 36.3 to 41.9. His production soared, including an average of .347, 25 home runs and a slugging percentage of .595. Murphy finished second in the National League MVP race.
A key was the increasing of his launch angle to 16.6 degrees from 11.1.
The Dodgers’ Justin Turner serves as another example. He went from a player with 15 homers in his first 1,100 career at-bats to one who slugged 48 home runs in his roughly 1,000 at-bats during the 2016-17 seasons. In so doing, he did not sacrifice batting average, which was a career-best .322 in the latter season.
“You can’t slug by hitting balls on the ground,” he said to the Post. “You have to get the ball in the air if you want to slug, and guys who slug stick around, and the guys who don’t, don’t.”
Turner did. And his reward was a $64 million payday that offseason.
Similar to baseball’s launch angle is the angle at which a basketball travels to the hoop after being released. A 45-degree entry angle is ideal, though analysis may deviate by a degree or two.
“Different studies have shown a different optimal angle of entry from the ball,” said James Brocato, director of basketball analytics for the Dallas Mavericks. “Forty-five degrees is what you usually get, and some say 47 degrees. We know where a player’s shot is normally. If a player is missing short a lot, or missing long a lot, we can correct that. We literally track every single shot in our practice gym and that can help players a lot with their shooting.”
Analytics has also changed where on the court players are shooting.
“One thing the analytics has changed in basketball, the NBA at least, is there are more 3-pointers and fewer mid-range shots,” Brocato said. “You are now seeing players embrace the 3-point shot. As compared to a long two, players don’t shoot that different of a percentage. Getting the extra point is much more valuable.”
Brocato mentioned Milwaukee’s Brook Lopez as a prime example of the trend NBA fans have witnessed.
“He has never been a 3-pointer shooter and has always been a good mid-range shooter,” he said. “This season he started taking almost exclusively 3-pointers, had a hell of a year and the Bucks won the most games in the NBA (60) and narrowly missed out on going to the finals.”
Indeed, Lopez took 30 3-point shots in his first six non-injury marred seasons. He took more than 300 in both 2016-17 with Brooklyn and 2017-18 with the Lakers. This past season, 512 of his 786 field goal attempts (65.1 percent) were of the 3-point variety.
“That’s a pretty big success story,” Brocato said.
Accompanying the explosion of data utilized in professional sports is colleges and universities offering courses that include sports analytics or offering sports analytics as a standalone course. After all, with offices of professional sports teams adding analytical staff, or at least individuals who help in that area, the educational component is becoming critical.
“The last four years analytics has really taken off,” said Daniel McIntosh, a lecturer at Arizona State’s W.P. Carey School of Business, who during that time has consulted on analytics with the NBA. “Across basically every major sport you can find a team that is using analytics to create some kind of advantage.”
Baseball is the most stat heavy of the sports, something writer and historian Bill James brought to national attention through his many statistical abstracts and handbooks, the first of which was published in the late 1970s.
Major League Baseball, through its Statcast tracking system, has a glossary devoted to technical terms such as launch angle.
Basketball, football and hockey also incorporate analytics. While one sport may rely more on analytical data than another, each has different needs.
“I don’t think it’s fair to say any sport is ahead, or behind, another in their use of analytics,” said Tim Smith, a professor of information technology management at the University of Tampa. “I believe that properties of the sport, such as slow-paced, fast-paced, the frequency of games, the number of players involved, etc., significantly impact the analytic techniques used.”
Regardless of the sport for which they are used, data derived from analytics is not eradicating observation through the trained eyes and instincts of veteran scouts and front office types who have made a living of poring over player reports while also relying on instinct.
“Initially people thought analytics was going to be some wonder cure-all and replace emotional decision making,” McIntosh said. “Right now, (analytics) is probably used more to validate decisions or to support strategic initiatives than they are to replace human decisions.”
“All the stuff we do is supplemental,” he said. “It is more to be taken hand in hand with traditional scouting and things like that. Basketball is such a complex game with 10 working parts at all times that you have to take everything into context.”
Because a newer way of acquiring a competitive edge is bumping up against long-entrenched practices, how findings from data are communicated is critical. After all, if those making crucial playing personnel decisions, for example, do not understand what is being conveyed, then it is not going to matter how good any analytics model looks.
That’s one of the points of emphasis at Syracuse University, which offers a major in sports analytics. A model may make look pretty, but can the end user benefit from is usage?
“One of the key things that we focus on in our program is being able to communicate your findings,” said Rodney Paul, a sports economist and professor with the David B. Falk College of Sport and Human Dynamics at Syracuse. “It’s one thing to be able to crunch the numbers and be able to come up with a model and a nice visualization. It’s another thing to be able to get the point across to the parties that actually are the decision makers, whether it’s the general manager, the coach or whoever it may be.”
Another key point in the classroom is analytics is not a “cure-all” or an absolute. A top draft pick is just as prone to injury and off-field issues regardless of analytics, which can’t predict the human element part of the equation.
“Unfortunately, injuries happen, off-field things happen, random events happen,” McIntosh said. “The New Orleans Saints did everything correct. There was a bad call (that went against them late in the 2018 NFC championship game that may have prevented them from going to the Super Bowl). You cannot control that. If you have the proper mindset of what you can do with (analytics), then it can become an incredibly powerful part of an organization.”
The off-field component rose to the surface when McIntosh was consulting with an NBA team. Some red flags can be addressed by simply altering a player’s approach. However, it is not always that simple.
“Everything in our models and everything in our systems said something is not right with one of the players,” he said. “So we called (the team) and asked what’s going on with the player and the response on the other end of the phone was not to worry about it. We said, ‘Well, everything in our metrics and everything in our numbers, everything says worry about it.’”
McIntosh asked if he could receive some feedback from the team so it could be incorporated into the player’s performance data. He recalled, “The guy says, ‘I can’t tell you more than this, but he’s going through a divorce.’”
Since analytics is relatively new, uncertainty concerning acquiring an education in the discipline is understandable.
Somebody wanting to become a writer, for instance, would take journalism and/or related courses while perhaps acquiring hands-on experience by writing for a school newspaper. The paths in which to acquire education and experience in sports analytics may not seem as clear to some.
“The No. 1 thing I tell students is to just get started,” McIntosh said. “We direct them to Python and SQL (programming) conferences. Go to SABR (Society for American Baseball Research) and go to Sloan (sports analytics conferences) to be up to date and current in what is happening in that space. The most valuable skills are the technical skills.”
Getting started is what Brocato did. As a law student at Gonzaga in 2013, he spent free time researching basketball analytics and created a draft model to predict how college players would translate into the NBA. Next he thing he knew, he was presenting that model during his interview with the Mavericks, who hired him in 2014.
“The No. 1 thing I tell people whenever somebody asks me what they have to do to get into the NBA and what they have to learn, is to do something interesting with data,” he said. “Put it on a web site so that people can see it. That’s how I got in and how a pretty large percentage of guys (like me) in the NBA got in.”
A potential employee needs to know more than programming, though. Knowing the sport is key.
“You have to know about basketball because communication is pretty important in the industry, for sure,” he said.
Brocato noted the Mavericks have four full-time staffers, two programmers and three interns in the team’s analytics department. He also recommended learning Python for programming and how to do basic SQL queries.
“Those two are real important,” he said.
It is also important to speak the language.
In 2018, Paola Boivin taught a class at Arizona State that dealt with the future of sports. Within the course, a week was spent on analytics, including technical terminology that goes with the territory.
“Part of the messaging to the students was that (analytics) is part of the story you’re going to cover and you have to understand what they are talking about and become familiar with some of the terminology that goes with it,” said Boivin, who teaches sports journalism programs at ASU’s Walter Cronkite School of Journalism and Mass Communication. “Analytics will become more a part of the curriculum in colleges because you have to be well versed and well educated in what you are covering.”
Being proficient in a second language could prove beneficial as well. The major offered at Syracuse includes 12 credit hours of a foreign language.
“The ability to communicate around the world opens many more possibilities as the sports analytics revolution moves across the globe,” Paul said.
The language requirement is a prime example of the emphasis placed on the well-rounded nature of individuals who want to work for a professional sports team or a similar organization. Technology, mathematics and communication skills are among the ingredients in building the foundation.
What is being built above the foundation is growing larger and larger. As more major-league sports organizations come on board and expand their use of analytics, more opportunities are sure to arise.
“We are anticipating the spread of analytics across the college landscape,” Paul said. “(As students go through the education process,) they will find a lot of fascinating questions that people need answered, and that they can find a pretty neat job in the industry.”
In other words, what we are seeing today just might be the beginning.
“Analytics is not going away,” Boivin said. “So, I think you are going to see more colleges pay attention to analytics whether there are specific classes about it or classes that are including it.”
Indeed, analytics is not going away. As such, failure to get on board may render negative consequences. That includes those who implement the findings: the athletes.
“If it is part of the game, you might as well learn it,” said Arizona Diamondbacks pitcher Matt Andriese, who has found analytics useful with, among other things, pitch location. “The game is changing and players just need to adapt. If you get stuck in your ways you might be out of the game.”
Tom Layberger has spent more than 25 years as a writer, editor and web producer for various media outlets. Tom, who resides in Tampa, is a graduate of the University of South Florida. Follow him on Twitter @TomLay810