The Durant-LeBron Arms Race Is Real—And Statistically Unprecedented

We mentioned this back in November, but LeBron James and Kevin Durant—the two best basketball players in this galactic quadrant—are outdoing themselves in 2013-14. Here are some per game lines:

  • James: 26.2 PPG, 6.5 AST, .578 FG%, .656 TS%, 29.8 USG%
  • Durant: 31.1 PPG, 5.3 AST, .512 FG%, .645 TS%, 32.0 USG%

There have been some other impressive numbers flying around—Durant put up a 37.1 PER in January, for example—but those last two figures, true shooting and usage percent, are the most remarkable; no other player this season has this combination of offensive centrality and scoring efficiency. In fact, they might be outpacing every season of every player in history. The chart below (which we'll explain in full in a bit) shows how Durant and LeBron have progressed in these key stats over the last four years—compared to some of the NBA's top seasons—with USG% on the x-axis and TS% on the y-axis:

That diagonal white boundary line represents a theoretical limit on combined volume and shooting efficiency, a concept developed by Evan Zamir back in 2012. It's a mindbending thing when you realize that, for the past 30-odd years, this relationship between efficiency and usage has toed a frontier that KD and LeBron haven't just surpassed, but tomahawked clear over. Since LeBron's first year in Miami, the two stars have moved in lockstep to break past this boundary, making particularly massive gains in true shooting percentage.

To understand just how remarkable this growth has been, we need to take a step back. Below is a plot of usage vs. true shooting for all 4,105 player seasons in the three-point era, where the player qualified for the FG percentage title. It's generally accepted that more usage leads to lower efficiency, but in this sort of broad analysis there doesn't appear to be much of a direct relationship:

11 percent of the seasons fall into that red box in the top right, where the player was in the top third in both usage rate (24 percent and higher) and true shooting percentage (.558 and higher). For obvious reasons, this square contains some pretty goddamn great offense.*

Let's zoom in on the 448 player seasons in that corner:

This is basically the chart that Zamir made back in 2012, when he found that—in this subset of already outstanding seasons—there was a linear relationship among the furthest outlier seasons, which constitute some of the NBA's best.

This line was defined as TS= 0.9 - 0.89*USG, and it implied that there was a limit on the combined usage and efficiency at the highest level of basketball play. You could have a ridiculous .665 true shooting season like Barkley in '87-88 (26.7% usage), or you can have a ridiculous 38.7% usage season like Kobe in '05-'06 (.559 true shooting), but it looked like you couldn't have it both ways—although 1989-90 Malone got close.

Of course, that boundary line was calculated almost two years ago. As you can see in the chart at the top, LeBron James and Kevin Durant finished just south of the boundary line for 2011-12, which, as Zamir put it at the time, "gives you some idea how great those other seasons were." Both are now far beyond Zamir's boundary, in a space uncharted by even the very greatest scorers, putting together seasons with a combination of usage and efficiency that are unprecedented in modern NBA history. If these stat lines hold, or if they even just regress a bit, 2013-14 could have not one, but two of the most remarkable offensive performances the league has ever seen.

*For rule-minded readers out there, we ran the numbers, and—at the highest level of play—the usage/true shooting relationship didn't seem to be significantly affected by the introduction of the hand-check penalty in 2004-2005.