As Golden State Warriors power forward Draymond Green elevates for a three pointer, he releases a shot that he sinks less than one-third of the time, a success rate that is relatively low amongst a cast of long range marksman.
The Warriors, however, are all too happy to accommodate Green’s misses, choosing to instead play the numbers game. They are a team that is heavily reliant on the three-point efficiency of its sharp shooting personnel, a group built with an analytics mindset rather than a traditional basketball framework.
The result is a team that divides public basketball opinion at times, offending the senses of those who value traditional basketball skillsets while delighting those who love high scoring and dramatic comebacks.
The truth is, however, they are a result of data driven thinking rather than a symptom of basketball skills disappearing. The use of basketball analytics has been building over the course of the last decade and has led to increased efficiency in NBA scoring rates, as teams analyze more than just how often a player makes the shots he takes.
It’s also no accident that the team, long a laughingstock of the league, built the sport’s top squad after being bought by Silicon Valley venture capitalist Joe Lacob.
Golden State Pioneers
Lacob and Warriors management immediately started a turnaround that has taken them from perennial losers to three straight appearances in the NBA Finals and two championships since 2015.
Erik Malinowski describes the transformation in the book “Betaball,” released this fall. One of the first moves by the Warriors was recognizing the potential uses of SportVU, a system that captures every movement on the court and gives teams the opportunity to analyze data in a number of areas.
The Warriors are not alone. At an analytics conference earlier this year, NBA commissioner Adam Silver said technology has progressed to the point where teams have players wear sensors not only during games and practices, but even when they sleep. They collect data on performance, the effects of fatigue and sleep patterns.
“Analytics are part and parcel of virtually everything we do now,” Silver said.
The Use of Basketball Analytics
Basically, every movement, personnel package and on court decision is analyzed with scrutiny. Does a player have to dribble to his left in order shoot a decent percentage from beyond 8 feet? How many points does a team average when they get off a shot with more than 10 seconds left on the clock? Is there a necessary number of offensive rebounds and points in the paint that help guarantee victory? The data doesn’t lie and the exploration of it through basketball analytics is a necessary skill for everyone from scouts to coaching assistants and journalists.
Enormous amounts of data are now analyzed before teams will even consider drafting a player. As Silver noted at the conference, a team can set itself back years picking the wrong player in the draft. While any other business would simply fire a high-profile employee who didn’t work out and find a replacement, NBA teams can’t typically fill holes left by bad draft choices with free agents.
Teams now analyze data from college and high school. They develop metrics for each player on performance in specific matchups or game situations. Silver called the amount of data collected and analyzed “incredible.”
No in-game strategy offers better proof that teams use data analytics more than the three-point shot. The Warriors use it repeatedly, and in an effort to keep up with higher scoring, others have followed. Teams averaged about 18 three-point attempts per game in 2012. Now, it’s 27.
Analysis of thousands of games led teams to a conclusion that math and logic already indicated. The more three-point shots you take, the more a team scores, even if the team goes through cold periods of the game where everyone misses.
Matchups have always driven much of the strategy in sports. Quarterbacks look for receivers matched up against less talented defenders in the NFL. Managers make pitching changes to have the person with the best chance to succeed on the mound against a specific hitter in baseball.
Basketball, like soccer, is a much more fluid sport. Players make split-second decisions in the flow of the game. Part of the preparation for players now comes in seeing analysis of the opposing team’s success rate against certain offensive attacks.
Huge datasets on a player’s ability to defend jump shots, three-pointers and drives to the lane, separated by distance and location of the shot are analyzed. The key is then to get the ball into the hands of the player in the best position that suits his skillset and in areas the opposing player has shown an inability to defend well.
Data collection on players’ diet and sleeping patterns – as well a close monitoring of fatigue levels – has helped teams to keep players healthy. Some teams now go as far as testing saliva, which contains indicators of fatigue, according to Silver.
Players tend to get injured more often when fatigued, and many teams will now rest a player if fatigue levels are high. Also, it makes more sense strategically to rest a player during a meaningless game and have them in top fitness for a playoff run.
Using data doesn’t always make for a better game, however. Hacking is a prime example of this as team’s take advantage of poor free throw shooting statistics and target players to send to the foul line so they can keep the game close.
Data analysis shows that the arc of a shot from a taller player has less chance of going in. Coaches have taken advantage of this fact, and it has led to a lot of fans watching big men clang balls off the rim from the foul line. However, NBA officials are considering rule changes to end this practice.
Of course, all the analysis in the world means nothing without player performance. That includes not only physical ability, but also the quick thinking needed to read a defense and develop the best attack strategy on offense (and the reverse on defense).
With big data now able to analyze a player’s ability on the court in dozens of ways, on-court decision-making is now part of what is evaluated.
Teams such as the Warriors had an advantage by adopting much of what Big Data has to offer before other clubs. Now, every NBA team has a data analytics department. The question is, when will they catch up to the Warriors analytic fueled efficiency.