Open Like Josh Smith

May 25, 2015; Houston, TX, USA; Houston Rockets forward Josh Smith (5) dribbles during the second half against the Golden State Warriors in game four of the Western Conference Finals of the NBA Playoffs. at Toyota Center. Mandatory Credit: Thomas B. Shea-USA TODAY Sports

Josh Smith has been making three-pointers in the playoffs. Seriously—39.1 percent against a career average of 28.5 percent. It’s been bizarre to watch and one of the most striking things about the visual of him knocking down three-pointers is just how much empty space there is around him. On potentially assisted three-point attempts[1. Using Seth Partnow’s back-of-the-envelope calculation of a touch time of less than 2.5 seconds], the closest defender has been, on average, 7.9 feet away from Smith[2. Calculated using his SportVU shot log.] That’s an enormous distance, larger than Rudy Gobert is tall. For gosh sakes, it’s like four Earl Boykins’ laid out end to end.

Obviously, leaving Smith this wide open is a calculated choice by the defenses of the Dallas Mavericks, Los Angeles Clippers and Golden State Warriors. He’s made them pay to some degree but those teams have a career’s worth of evidence that supports their decision. When it comes to defensive adjustments in the plauoffs, Smith is not the only player teams have to make this kind of decision about, just one of the most obvious.

Who’s Left Open? | Create infographics

The graph above shows all players with at least 15 potentially assisted three-point attempts this season. They are graphed by their three-point percentage on those attempts on the x-axis and their average defender distance on the y-axis. The size of each bubble represents the total number of potentially assisted three-point attempts. You can see that Smith is among the least closely defended outside shooters, although not in the top spot—I must have missed the Clippers’ internal memo on deciding to let Boris Diaw beat them.

You can also see Kyle Korver the most closely defended perimeter shooter in the playoffs. Although this was a very human shooting performance for him, he still finished near the middle of the pack in efficiency despite garnering more defensive attention than anyone else.

One of the most interesting things about this data set, difficult to see in this graph, is the relationship between efficiency and defender distance. There is actually a fairly strong negative correlation (-0.267) between defender distance and three-point efficiency—that is, players with more space are shooting worse. On a bigger sample, we would expect to see the opposite relationship—more space means a shot should be more likely to be made. However, we’re looking at a somewhat constrained sample, influenced by selection bias. In the playoffs, player with poor track record as outside shooters—like Josh Smith—are given more space.

Given all this information, I would like to formally add a term to the basketball analytics lexicon. Any three-pointer with at least seven feet between the shooter and the closest defender will henceforth be characterized as “Josh Smith open.”