ESPN linked to this article in their Feb. 23 cover story "The Great Analytics Rankings".
Winning is the bottom line.
A soft line drive of the baseball to left field puts the player on first. The pitcher glances over to the base-runner before he throws a 97-mph fastball to home plate, but, as he does so the player on first takes off to steal second. “You’re out!” the umpire yells at second. With the sweep of the glove to the hip, the runner’s dream of crossing the bases and scoring a run is dashed. Out by an inch.
And that’s the difference. The difference between a steal and an out, a diving catch and a spectacular drop — a win and a loss.
The goal of advanced sports analytics is to account for the inches — to decide whether the runner should steal or not. It is a tracking and record system of advanced statistics used to measure performance on and off the field, with the goal of understanding how to increase performance and assign value to data. And the frenzy behind finding this new data is sweeping the nation . . . including students and professors at BYU.
The race for data
One reason for the emphasis on advanced sports analytics is clear: the more a team, player or owner knows, the better they can prepare players and win games. As a result, professional teams are hiring statisticians in hopes that they will be able to explain the game and improve odds of winning, a job that hardly existed 15 years ago, and one Shane Reese, of BYU’s statistics department, is excited about.
“I’ve always had an interest in sports,” Reese said. “I liked to play sports, and I always cared about percentages, free throw shooting and batting average. But it was something that I was interested in that I thought would never have any real professional benefit. But now there’s a real push from the collegiate ranks up to the professional ranks to get a real handle on the quantitative side of performance of athletes. It’s been great.”
While Reese, a self-proclaimed nerd of sports statistics, is seeing more and more opportunity to combine his profession with his love for sports, he’s in fact been in the business of analyzing sports data for more than a decade now, teaming up with Gilbert Fellingham, professor and associate chair of the statistics department. Together, they have worked with the U.S. Olympic Volleyball Team, consulted with the Philadelphia Eagles in the NFL, the Utah Jazz, Charlotte Bobcats and Cleveland Cavaliers in the NBA and the Colorado Rockies of Major League Baseball.
“The numbers are an additional factor that I think can be used to help teams play at a higher level, to help evaluate talent more rationally and in a better way,” Fellingham said. “Will it solve all the problems or will games cease to be games? No … and that’s what makes them fun.”
As imagined, the statistical models used to track performance and increase win percentage will not replace managers or coaches. Rather, they are in place to enhance production and provide answers to questions that can now be answered because of newly available technology and a growing interest.
Zach Bradshaw, a BYU student analyzing data for the Charlotte Bobcats, has firsthand knowledge of how his work can improve an owner’s decision-making ability but also understands the need for human judgment in evaluating players.
“I think (pro teams) are starting to see value in this realm, as analytics can reduce the uncertainty surrounding important decisions across many aspects of basketball operations,” Bradshaw said. “However, many things are very difficult to measure and assess quantitatively, so a qualitative component is still very important in the decision-making process.”
The reason for both professional and collegiate teams to focus on analytics boils down to the century-old cliche of “getting an edge” on the opponent, but only time will tell if this “new data” will be worth the fuss.
Who is involved?
The genesis of sports analytics has roots at BYU and, not surprisingly, has had a generous impact on academics and sports on campus and around the world. Hugh McCutcheon, former coach of the BYU men’s volleyball team, was a huge proponent of quantitative sports analysis and led the U.S. men’s national team to a gold medal at the 2008 Summer Olympics in Beijing.
“BYU is this funny breeding ground for all of this statistical stuff,” Reese said. “We’ve worked with the men’s basketball team by looking at every single possession one season, and we’ve done a little work with the track and swim teams. The fundamental idea is that anything that is going on in practice in improving a developing player can be tracked and used to improve coaching decisions and future performance.”
It’s also creating more work for students who, like Reese and Fellingham, have a desire to take math and analytical skills to an area that is fun and can improve the jobs of professional athletes and coaches.
“Some statisticians will say that the stats are all that matters when analyzing a player, but you can’t take the scouts out of the game,” said Kevin Williams, a student who analyzes pitching data for the Rockies. “They know what they are doing. They know what to look for in a player. They have the eye to find players that others might not be looking at because they don’t have the ideal stats. I believe that a combination of both scouts and stats is the best way to analyze players and find those who will succeed.”
Baseball analysis dates back to 2002, when Billy Beane, the general manager of the Oakland A’s, used advanced metrics to scout players who were undervalued and could help his team perform against powerhouse $250-million teams such as the New York Yankees, on only a $20-million payroll. This became referred to as “moneyball,” and teams across all markets have implemented this strategy.
“I think everyone wants to get that ‘edge’ over their competition,” Williams said. “It has been shown that this technology has helped teams and given them an advantage. Just look at the defensive shifts that teams put on certain players during a Major League Baseball game. They wouldn’t do this if the data wasn’t telling them something.”
The impact is real at BYU, where a sports analytics course has been offered since fall of 2013, and continues to gain steam as professors and students creatively analyze data and gain recognition from professional sports franchises, such as the Jazz, who employed the services of several statistics students to analyze shot charts throughout the 2013–2014 NBA season, and individuals, like Nick Martineau, who works with the Cleveland Cavaliers.
“Some people worry about sports becoming robotic or predictable, but that will never happen,” Martineau said. “There is way too much variation in sports for them to become robotic or predictable. Analytics, done correctly, can give you a competitive edge. However, that competitive edge is not huge. It may be worth just two or three extra points in a basketball game. Two or three extra points will win you a few extra games, but it won’t guarantee any championships.”
What the future holds as the stats race continues
While data and performance can be quantified, the future of it all cannot. However, since “moneyball,” a whole new wave of thought has developed that is turning the business of playing a game into a thought tank that produce some amazing results.
Major League Baseball Advanced Media announced before the start of the 2014 MLB season that it would be installing a new on-field tracking system that has the ability to track nearly every statistic imaginable, including how quickly a players swings the bat, the angle at which the ball leaves it and how quickly an outfielder reacts. To this point, three Major League stadiums have it installed, including Citi Field in New York, Target Field in Minnesota and Miller Park in Wisconsin. The plan is to have all 30 stadiums operational by 2015.
“I have seen the new on-field tracking system, which some people refer to as the ‘Holy Grail’ of baseball research,” Williams said. “It will definitely change the way that players are analyzed. I believe it will definitely help the game of baseball.”
And in 2007 a conference was created to share ideas and research on sports analytics to a mass audience. This year’s conference, which ended on March 1, discussed varying topics, from the “biases of baseball umpires” to the “three important dimensions on rebounding.” The collaborative effort is an academic approach to improving the numbers game.
But when it is all said and done, performance will still be based on the willingness and drive of an athlete, and taking a few numbers and throwing them around will not account for that concern. Some worry the push to have answers and numbers for everything will hurt competition, but in the end it is all about enhancing it.
It’s quite possible that analytics will become mainstream in collegiate athletics, as it is already taking over in the professional realm.
Time to downplay the hype?
The reality is that the hype surrounding gathering new data is big. The key to deciphering it all in a reasonable and sensible manner is to have enough people who understand the nuances of data and who understand that not all data is useful.
“I think sports analytics can have a significant impact,” Fellingham said. “One of the problems we have as people that do sports analytics is that there is a certain amount of pushback from people that have been around for a long time that don’t see the value. I think there is a tendency to oversell sports analytics, and while I’m a big believer in it, I don’t want to oversell it.”
Vital to the future of sports themselves are the players. There may be a time in the future when every piece of data that is gathered is used, but the “star” quality of an athlete will still be vital as coaches evaluate players. The numbers game in full force will help sports, but without the players, none of it would even be worth it.
