Do Mid-Range Attempts Hurt Offensive Efficiency?

January 31st, 2014 | by Andy Larsen
Photo by Fernando Medina/NBAE via Getty Images

Photo by Fernando Medina/NBAE via Getty Images

The midrange taboo

A whole legion of sports fans with fancy spreadsheets have waged a war on the long two. Many have gone as far as to refer to shot charts with few to no long twos as “moneyball” shot charts. I wonder if those people have actually read Moneyball.1

More importantly, I wonder if they understand spacing. Effective offenses have to possess at least a threat of midrange ability in order to preserve spacing. Jeff Hornacek, Bryon Russell, Matt Harpring and various other Jazz alumni will tell you that the elbow jumper was the way they punished defenses for overplaying screens or crowding Karl Malone, and it opened the weak-side baseline for cutters. Around the league today, smart teams are selectively using 18 and 20-footers to keep defenses honest.

Broadcaster David Locke mentioned this a recent podcast. He cited the league average for PPP on non-transition plays2, and said something like, “Maybe we’ve been too critical of the midrange jumper.”

We have.

There’s almost no correlation between high mid-range attempts and low offensive efficiency. The league’s best offenses are just as likely to be at the top of the midrange list (Blazers, Clippers) as they are at the bottom (Heat, Rockets) or the middle (Spurs, Mavs). Same is true of the least efficient offenses.

Where's the "moneyball" correlation between midrange twos and poor offense?
Where’s the “moneyball” correlation between long twos and poor offense?

In fact, the league’s best offense also takes more 16-24′ shots than anybody else. The difference between Portland and, say, the Wizards — who take virtually the same number of long twos per game but have the 21st overall offense — is that the Blazers make theirs. They execute their way to the right kind of midrange shots, and thus are above average in percentage from four of the five shooting zones at that range.

At the other end you have Houston, often used as the example for the argument against long twos. The Rockets take fewer than half the 16-24′ shots of the next stingiest team and less than 1/3 of the league average, so with an offensive ranking in 6th, the narrative works, right? But coexistence ≠ correlation ≠ causation. I’m not convinced their strategy makes their offense better. First, they’re actually an above-average team from 16-24′. Also, their aversion to these shots makes them pass up decent looks for, in some cases, worse shots. Their 36% mark on 8-15′ shots (one of the lowest) tells me they’d rather take a bad close jumper than an open midrange one.

With new tools, we’re confirming what scouting and intuition have told us all along: a good midrange look isn’t a cardinal sin, and “good shot” is completely situational. Look at league-wide percentages on off-the-dribble vs. catch-and-shoot jumpers, or assisted vs. unassisted, open vs. guarded, off-balance vs. squared up. Getting obsessed with the fact that a shot was two feet too far in any direction seems arbitrary given all of the other factors that impact the likelihood of it going in.

So let’s stop acting like an open 18-footer is an affront to the stats movement. It’s not. Smart teams have balanced offenses and keep their opponent honest by making good situational calls all over the basketball floor.

10 Comments