Picture a pick and roll in your mind’s eye. What do you see?
If you’re of a certain age, you probably see grainy images of John Stockton working his way around a screen from Karl Malone, sizing up the defense and waiting until exactly the right moment before delivering a pinpoint pass to The Mailman. If you’re a bit younger, maybe you see Steve Nash threading the needle between multiple defenders to find a rolling Amar’e Stoudemire, who powerfully slams the ball through the rim; or perhaps Tony Parker veering away from his defender and ducking behind a Tim Duncan screen, wrong-footing his man as he pulls up for a midrange jumper. If your basketball memory doesn’t extend far beyond the past few years, maybe you see Stephen Curry luring a double-team out to half-court while Draymond Green rolls to the nail, catches a pass and coolly navigates a four-on-three opportunity.
If you’ve been playing close attention in recent seasons, however, what you’ve been seeing more of looks something like this.
In each of those clips, Clippers forward Kawhi Leonard receives a screen from one of Patrick Beverley, Lou Williams or Landry Shamet. Aside from being Leonard’s Clipper teammates, those players have an interesting thing in common: They’re all guards.
Through Nov. 10, 9.3 percent of all plays involving on-ball screens during the 2019-20 NBA season have been set by guards, according to Second Spectrum tracking data. While that number might not sound enormous, it’s the continuation of a trend that has seen the share of plays involving ball screens set by guards approximately double over the past six years.
Even a cursory glance at slightly more granular data offers an explanation as to why this is happening: Pick and rolls with a guard as the screener have been more efficient than those with a forward or center as the screener in five of the past six seasons.1 The difference in per-possession efficiency between pick and rolls working off of guard-set screens (1.14 points per possession) and those using forwards or centers as the screener (1.05 points per possession) has also never been wider over this period than it is right now.
Throughout NBA history, pick and rolls have almost always seen a forward or center screening for a guard, so forwards and centers, in turn, have far more experience defending screeners. Using a guard in that role can therefore put opposing guards in unfamiliar and often uncomfortable situations. Defensive players put in unfamiliar and uncomfortable situations are more likely to make mistakes, which may present the offense with better scoring opportunities.
NBA defenses have been far more willing to switch defenders on pick and rolls involving a screen set by a guard than they have those with a screen set by a forward or center.2 The switch rate on plays involving guard-set screens has ranged between 32 and 42 percent over the past six seasons, while the high-water mark for switching on plays with forward- or center-set screens during the same time period is just under 15 percent. Overall, over the past six seasons, pick and rolls where a guard sets the screen have been switched nearly 3.5 times as often (39.1 percent vs. 11.5 percent) as those that used a forward or center as the screener.
Despite that increased willingness to switch, offenses have still managed a greater level of per-possession efficiency on guard-screened pick and rolls than the more traditional kind. Why might that be the case? At first glance, nothing really stands out as the reason such plays have been more efficient. But there’s an obvious through-line here. Pick and rolls using guard-set screens over the past six seasons have resulted in turnovers slightly less often than those with forwards or centers doing the screening. Guard-screened pick and rolls have also yielded shooting fouls slightly more often than the forward- or center-screened plays. And the leaguewide shooting percentages on both 2-point and 3-point shots have been slightly better on plays where a guard sets the screen, while those plays have also yielded 3-point looks slightly more often than those where a forward or center did the screening.
|As the screener in a pick and roll|
|Rate of drawing a shooting foul||6.4||5.7|
|Shooting percentage on 2-pointers||47.9||46.8|
|Shooting percentage on 3-pointers||35.3||35.1|
|3-point attempt rate||35.6||32.2|
None of those gaps alone can explain the observed difference in efficiency. But a bunch of small advantages stacked on top of each other will do the trick — and that appears to be what’s happening here for the guards involved in such plays.
The frequency of these plays also warrants a deeper dive into which teams and players are taking advantage of them most often. On the team level, it’s clear that the Houston Rockets are at the top of the league. No team has run more pick and rolls with a guard as the screener over the past six seasons than Houston. Of the 19 teams that have run at least 1,500 of them during that time period, none has been more efficient on a per-possession basis than the Rockets, who have scored 1.12 points per possession.
That the Rockets are at the forefront of this movement shouldn’t be surprising to anyone who closely follows the league; Daryl Morey, Mike D’Antoni and Co. orient their offense around The Math in many ways, and this is just another. Predictably, given the Rockets’ heavy usage of these plays, James Harden is among the league leaders in being screened for by guards. This season, Harden has been screened for by a guard in the pick and roll 112 times. That’s nearly as many as the next two closest players (Damian Lillard and Kyrie Irving) combined. His 12.44 guard-screened pick and rolls per game this year is on track to be the most any player has run in the past six seasons, as it would dwarf the 8.43 per game LeBron James averaged during the 2016-17 campaign.
Harden has been incredibly efficient on those plays this season — producing 1.28 points per possession, which would be the sixth-best mark among the 158 individual seasons in which a player has received at least 100 guard-set screens in the pick and roll over the past six years. Zooming out over the full sample, however, reveals who the real masters of these types of plays are. The table below shows the number of players who have run a given amount of guard-screened pick and rolls over the past six seasons as well as the points per possession leader among that group of players. There’s a familiar name among the high-volume orchestrators.
|No. of screens received||No. of Players||Points per possession leader at that level|
|500||21||Stephen Curry, LeBron James|
LeBron James sitting atop this list seems quite appropriate. It could be argued that the Cavaliers’ relentless targeting of Curry during the 2016 NBA Finals — with J.R. Smith, Iman Shumpert and Matthew Dellavedova repeatedly screening for James in order to involve Curry in the action — led to the subsequent spike we’ve seen in these types of plays. We saw the Rockets pursue the same strategy during their playoff series against the Warriors last season, and we saw Golden State’s opponents target Curry with seven screens in 3.5 games this season before he broke his wrist.
Given how well these plays seem to work, we shouldn’t expect that teams will stop running them anytime soon. If anything, we should get ready to see more and more guards setting screens in the pick and roll, until opposing defenses figure out a better way to stop it.
Check out our latest NBA predictions.
CORRECTION (Nov. 12, 2019, 4:38 p.m.): A previous version of this article incorrectly described the recent injury to Golden State Warriors guard Stephen Curry as a broken wrist. Curry broke his hand.