Tuesday, October 13, 2015

Analytics in Today’s Sports Landscape



Do you remember when winning or gaining an advantage in sports was achieved by a coach’s gut decisions – or when the value of a player was attributed to the “eye test?”  Analytics has spread like wildfire in sports and the days of “pure” sports are long gone. 


The Spark that Ignited the Fire


Billy Beane, the GM for the Oakland Athletics, is notorious for his use of analytics in baseball. Chronicled by the Michael Lewis book “Moneyball”, Beane used analytics to better understand the “secret sauce” of winning players. Beane looked past the traditional way of evaluating players (i.e. scouting services) and looked at the variables that led to more wins. It turned out that a team with a higher “On-Base Percentage (OBP)” would likely score more runs, and ultimately win more games. Beane revamped his entire roster with players who fit those specs…. and the rest was history. 


Analytics in Today’s Sports Landscape


Fast forward over a decade after Billy Beane introduced analytics to baseball and the landscape has completely changed. Now-a-days, if you don’t have an analytics team on staff as a professional sports organization, you’re at a disadvantage. 

The National Basketball Association (NBA) has been on the forefront of the analytics movement. NBA teams are now using a form of technology called “Player Tracking,” which evaluates the efficiency of a team and players by their movement. NBA arenas now feature 6 cameras that track player and basketball movement 25 times per second. No longer are NBA players evaluated by the “eye test” or basic statistics like Points-Per-Game (PPG) or Rebounds-Per-Game (RPG). Teams are now looking at the speed of a player, how many rebounding and scoring opportunities he had, how far a player traveled during the game (i.e. 2.8 miles), etc. 

In 2014, Benjamin Morris wrote an article on FiveThirtyEight.com about the hidden value of steals in the NBA. To illustrate this, Morris created a regression using players’ box score stats (points, rebounds, assists, blocks, steals, and turnovers) to predict how much teams would suffer when someone couldn’t play. Shockingly, Morris discovered that a “steal” is worth as much as 9 points. Put another way, Morris said, “A marginal steal is weighted nine times more heavily when predicting a player’s impact than a marginal point.”

In 2014, Kirk Goldsberry of Grantland wrote a fascinating article on how he used this new available player data to estimate the value of a possession – moment by moment. As an example he used the last 9 seconds of a game where Spurs PG Tony Parker maneuvered through the lane to find an open player for the winning shot.  The article features an interactive visual of how the estimated possession value (EPV) changed after each second. 

The articles by Goldsberry and Morris are examples of how the advancement of analytics in sports has led to a new appreciation for the “little things” of the game – things that were previously overlooked or viewed as unimportant. 


What’s The Future?


Wearable tech is the trend now. We’ve already seen the NBA experiment with materials used for uniforms and basketballs. Perhaps the next step is to bring wearable tech to the players? For example, major league baseball players often wear sunglasses when at bat. Maybe if their sunglasses featured technology similar to Google Glass and the Apple Watch, then we could measure players’ emotional and mental health throughout the game. 


Closing Thoughts


Is advanced analytics in sports a good thing or a bad thing? I’m not convinced one way or the other but I don’t believe it’s been all “good.” Just like in Market Research, there will always be value in the consultative side of things. I’m not against Do-It-Yourself research tools like SurveyMonkey but there’s a reason why Fortune 500 companies don’t assign interns to conduct their market research. We’ll always need people who can structure studies appropriately and make since of the data afterward. Analytics just for the sake of analytics can only go so far in sports. There must always be direction (a focus on things that make sense in context of the game/sport) and people in place who can make sense of the raw data. 

What do you think? Is advanced analytics in sports a good thing?  What do you think is next?

***

Isaiah Adams is the Manager of Social Media Development at Optimization Group, a marketing research and analytics firm that uses cutting edge technology to help clients make fact-based decisions. Follow Optimization Group on Twitter @optimizationgrp

Image Credit: lonely11 / 123RF Stock Photo

No comments: