The Biggest Misconception about Basketball Analytics
In the past decade, Daryl Morey helped revolutionize the NBA with his mathematical approach to basketball known as Moreyball. The concept of Moreyball advocates taking as many layups, free throws, and open 3s as possible: the shot types that yield the most points per shot (distribution pictured on right). By this metric, taking 3s and layups should be prioritized before any other type of shot. Consequently, the Rockets have almost entirely removed the midrange shot from their offense and many other NBA teams have followed suit. Despite being trendsetters in this analytics-savvy style of play, the Rockets haven’t even made it to the NBA Finals in the past few years.
Kirk Goldsberry/FiveThirtyEight
The Rockets lack of playoff success has led a lot of people to question the efficacy of analytics, particularly the Rockets heavy reliance on 3s. Moreyball has appropriately convinced NBA offenses that 2-point jumpshots within one or two feet of the 3-point line are objectively the worst shot in a normal halfcourt setting. However, it also created the biggest misconception about basketball analytics: thinking that all midrange shots are always the worst shots a team can take.
The philosophy of Moreyball doesn’t explicitly state that midrange shots are bad, but with the Rockets reputation of passing up open midrange shots for contested 3s, it isn’t difficult for many to connect the dots. There is empirical evidence that shooting a respectable amount of midrange shots can be good for an offense.
There is no obvious correlation between midrange frequency and points per play. Credit: Thinking Basketball
Noah Graham/Getty Images
In my opinion, the Rockets have focused on Moreyball too much in the past few seasons - or at least on a shallow understanding of it; I’ll try to keep my reasoning short and simple. Chris Paul, one of the top midrange shooters in NBA history took half as many long midrange shots on the Rockets than when he was on the Clippers. Surprisingly, over the course of his two seasons with the Rockets, Paul’s midrange shots (16-3P) had the exact same amount of points per shot as his 3 point shots at 1.1 points per shot. If Paul was getting 1.1 points per shot from his midrange, why was he still encouraged to take fewer midrange shots than when he was on the Clippers? It begs the question: is Moreyball more about taking the shot that yields the highest points per shot, or does it strictly mean taking as many 3s, layups, and free throws as possible? I’d argue that it’s the former, but it seems like the general basketball public seems to think it’s the latter.
Chris Paul’s shooting splits over the course of his career. Source: basketball-reference
Regardless of how you interpret Moreyball, you can still find problems in its underlying philosophy. Shooting 67% from the 3 point line yields a higher points per shot than a 100% 2 point shot. So in a vacuum, taking the 67% 3 is a better shot. But is that actually the case in reality? This hyperbolical example highlights a flaw of the points per shot metric. In the 4th quarter of a close game, do you go for a 50% midrange shot or take your chances with a 33% 3 point shot? On the other hand, if you completely remove the midrange shot from your offense, opposing defenses can create schemes to closely guard the paint and the 3 point line - leaving the midrange open. Does avoiding the “least efficient” shot in basketball end up causing more problems than it solves?
Just like other playstyles, Moreyball has its own advantages and disadvantages - there’s no definitive best way to play basketball. A NBA team doesn’t need a lot of great 3 point shooters to be successful, but a team can still be successful with a lot great 3 point shooters.
Moreyball is A way to play, not THE way to play.
If you’re interested in the basketball theory and strategy behind the midrange shot, Thinking Basketball made a very informative and entertaining youtube video that I highly recommend watching.
