Picking apart the pick-and-roll: A deep dive into the Toronto Raptors’ two-man game.

26 mins read

The pick-and-roll is the most important quality in modern NBA offense, and the fact of the matter is last year’s Toronto Raptors were not very good at it. Before the season began, I wrote for Sportsnet about five lineups that could define Toronto’s season. I wrote this about Toronto’s pick-and-roll struggles last year:

“Their average points per possession when the roller finished such plays was 1.02, second-lowest in the league. When the ball-handler finished pick-and-rolls plays, Toronto scored closer to league average, at 12th, but handlers for Toronto finished pick-and-rolls with the fourth-lowest frequency in the league.”

Neither ball-handler not screen-setter was an efficient scorer for Toronto in the league’s most important offensive play. That one weakness went a long way towards explaining why Toronto last year had the 15th-ranked half-court offense and the second-ranked transition offense. It’s a simplification, but you don’t run pick-and-roll in transition. Shooting and athleticism become more important, as does the ability of the defense to fuel the offense. But when an opposing defense is able to set up, have its chosen matchups, call out the plays and dig into whatever Toronto does — that’s when the Raptors struggled. It mattered, so much so that Toronto’s meek scoring in the half-court was a huge reason why they lost to the Boston Celtics in the playoffs.

Then in the off-season Toronto lost 53.4 combined minutes of center play in Marc Gasol and Serge Ibaka (higher than 48 minutes because the two often were injured at different times) and had to rebuild the position from scratch. Incoming Aron Baynes was selected to start at center. He’s a bone-crunching screen setter, a surprisingly competent shooter, and a solid finisher when others create for him. Then the team retained Chris Boucher, who is a pogo-stick leaper and can finish above the rim. Between the two, it was clear how Toronto’s centers could contribute in the pick-and-roll. Through two games, they’ve been solid.

I think we’ve got to be extremely pleased with what we’re getting out of that [center] position in our first two games. I think both those guys have been outstanding,” said Nurse after the team’s loss to the San Antonio Spurs.

In fact, it’s possible that Toronto’s pick-and-roll attack will be better this year than last. Ibaka and Gasol both loved to short roll, or drift to the middle of the floor rather than diving all the way to the rim. Ibaka did it to fire mid-range jumpers and Gasol to sling passes across the floor. Both are incredible skills on their own, but Toronto lacked a consistent rim-runner who gave handlers Fred VanVleet and Kyle Lowry easy passes for layups. Baynes and Boucher should help there. Both were more efficient than either Gasol or Baynes as pick-and-roll finishers, with Boucher topping out as the best of the four at 1.15 points per possession, good for 62nd percentile.

Furthermore, Pascal Siakam and OG Anunoby are inching their way towards larger shares of the screening duties. Their skills as strong wings, who can make contact and also make plays in space, offers Toronto skills from the screener that aren’t common to centers. Between those four, Toronto has plenty of options, none of whom fully duplicate the skills of the other three.

Last season, I looked at a wide variety of pick-and-rolls in a monster breakdown of the Lowry-Siakam two-man game. The piece encompassed 45 pick-and-rolls. I thought I’d outdo myself this year. Two games into the 2020-21 season, and Toronto has run 81 pick-and-rolls. I watched and charted and studied them all.

So, two games in, how has it been?

Of those 81, 65 have ended in an event, either a shot or turnover. This piece will focus only on those pick-and-rolls that ended in an event, rather than those that were reset, either by choice or compulsion via the whistle, with Toronto keeping possession. And of those 65 pick-and-rolls, Toronto has scored 1.03 points per possession. Because I’m counting all of these numbers by hand, they don’t compare directly to nba dot com’s advanced stats numbers, which only count pick-and-rolls that are finished by the roller or finisher. But I can do the math on Toronto’s pick-and-rolls that have ended in those events. Of Toronto’s 22 pick-and-rolls finished by the handler, they’ve scored 0.95 points per possession. Last year, they managed 0.87. Of the 15 pick-and-rolls this year finished by the screener, they’ve scored 1.50 points per possession. Last year, they managed 1.02. S0, you know, there’s definitely some noise here due both to the small sample sizes and the huge gaps. The most likely explanation is some differences in definitions between myself and the professional stat keepers, but the point is that Toronto has had an effective pick-and-roll game so far this year. It’s hard to compare it to last year, but my hunch looking at the numbers is that it’s been better.

(If you compare current stats using nba dot com’s playstyle numbers, Toronto’s statistics are much lower for both handler and roller categories than I calculated, although the roller efficiency is much higher than it was last year, so even if there are different definitions involved, the point probably remains that Toronto’s pick-and-roll has been relatively effective this year compared to last.)

The task, then, is to analyze how Toronto has performed in the pick-and-roll through two games. Let’s start by diving into Toronto’s options as handlers.

Lowry, VanVleet, and Siakam eat the lion’s portion of pick-and-roll possessions. But there is a difference in what the three have brought to the table. For one, Lowry and Siakam have been much more efficient scorers when handling than VanVleet. Correspondingly, the team has reached elite point per possession marks with either of the two handling, while the mark with VanVleet in charge pales in comparison. Much of that has to do with the individual skills of the three players (or at least the skills shown through two games). Lowry has hit 3-of-8 pullup triples, while VanVleet has only connected on 1-of-11. While Siakam’s pullup shooting from distance has been poor, he’s connected on 4-of-4 two-point pullups. He can get to his shot against switches in isolation in a way that isn’t as easy for the smaller guards.

VanVleet has missed plenty of pullups and layup attempts as the handler, which could be chalked up to poor luck. But he’s been an average pullup three-point shooter and below-average rim finisher for his entire career. Siakam and Lowry are more established at creating their own shots, and the numbers have been consistent with those trends this season. Thus Siakam and Lowry have threatened defenses more on-ball with their own scoring, leading to more efficient team scoring when they run pick-and-rolls. It won’t necessarily remain true, but that the two games have been predictive with trends from past seasons means it’s more likely to be predictive than not; the Raptors want VanVleet off the ball and Lowry or Siakam in control of the pick-and-roll.

While all three sport relatively similar assist-to-possession ratios, Lowry has been unique in the way he’s collected those assists. VanVleet and Siakam have dished all their assists either to popping screeners or orbiting shooters, but Lowry alone has been able to find the roller diving to the rim for points. That’s not a knock on VanVleet or Siakam; both have been quite good at collecting assists as handlers. VanVleet in particular is wonderful at hitting shooters after reaching the paint. But Lowry offers an element unique on the Raptors. He’s the only player comfortable throwing bounce passes into space that his roller can collect while gaining a head of steam.

It’s good that Toronto has three options who can capably run five or more pick-and-rolls per game. VanVleet should remain an initiator, of course. But it’s possible that if some of VanVleet’s attempts are shifted to Siakam, the team will be more successful. VanVleet is a brilliant catch-and-shoot shooter and an average pullup shooter. That on its own helps explain why Toronto wants him orbiting a few more pick-and-rolls and initiating a few fewer.

While the Raptors have only three players they trust to run primary pick-and-rolls, Toronto’s offerings at the screener position have been more varied. Boucher and Baynes have screened more than any other players, but Toronto has plenty of off-speed pitches in the bag. And they’ve used more screeners in multiple situations than they have handlers.

Let’s start with the traditional screeners. Boucher and Baynes have used a relatively equal number of possessions as the screener in the pick-and-roll. Their own numbers as shooters, and their handlers’ numbers are shooters, are similar, though tilting towards Boucher. But the team has experienced polar opposite levels of success, scoring 1.16 points per possession with Boucher as the screener and only 0.62 with Baynes. What’s to explain that vast gap?

Likely, it’s been luck. This is true of everything in this piece, of course, as two games is far from a predictive sample size. To support that, when you actually dive into the tape, Toronto has created excellent looks with Baynes as the screener, even if they haven’t fallen. His gravitational pull when he dives all the way to the rim has been enormous.

Baynes has already busted out the Gortat screen, wherein he turns and screens in the paint a second time, rather than trying to make himself available for a pass. This opens up a wider driving lane for the handler. Siakam blew the layup here, but the rim was poorly protected because of Baynes’ seal. Those tricks will help Toronto’s handlers convert efficiently around the rim with Baynes screening.

Boucher, on the other hand, has likely been the beneficiary of some positive luck. He’s shot extremely well from beyond the arc through two games, at 3-of-5, with two of those makes coming in the pick-and-roll. His finishing has been great, but given his career three-point percentage of 32.9, he probably won’t space the floor to the extent he has. (Man, do those swishes look good, though. His shots just hit different because of the moonball arc.) However, some elements of Boucher’s success have seemed replicable, namely his finishing inside the arc. He’s always been great at diving hard to the rim. But he’s shown some craft on the short roll, too.

Remember, this is a player who in 2018-19 in the G League was so inexperienced when it came to team concepts that he would frequently roll into his own ball-handler with the Raptors 905. He’s come a long way. When I asked Boucher how he improved as a short roller, he only spoke in generalities; confidence in your own shot and etcetera. His guards put him in positions to succeed. Well, Boucher has never lacked confidence. Such a wide gap between Toronto’s success with Boucher screening and Baynes screening surely won’t continue, but there are signs that Boucher will remain useful in a variety of ways, at the very least.

Perhaps most exciting, when it comes to Toronto’s screening options, is the wing tandem of Siakam and Anunoby. Neither has been used as a high-volume screener in the past, but both have been excellent thus far. The team has scored well, of course, in those setups. But even more encouraging has been the tape of how it’s happened.

Both are effective scorers, particularly when others create for them. But both have also dished an assist after setting a screen. Here Siakam popped and used the momentum of the closing defender to drive and create for Anunoby. Anunoby, for his part, showed more balance and fluidity in this finish than he has in past seasons.

Whereas Anunoby did his best Gasol impersonation as he surveyed the floor and found an open shooter drifting to the corner.

Neither Baynes nor Boucher are above-average passers from the big spot, so it’s incredibly useful to be able to turn to options who can create for others after setting a screen. Flexibility is key, and Toronto is flexible at the handler spot or screener spot. Siakam, of course, with his abilities in both situations, is the biggest reason why.

Another fun reason why Toronto can wreak havoc in small spurts is because of the screening ability of its shooters. Lowry, of course, has always been a deadly screener. The Spain pick-and-roll has been in Toronto’s playbook ever since Lowry joined the Raptors. And they’ve already dusted off the cover to get Lowry an open shot after he set a pair (!) of vicious back-screens.

But Toronto has revealed a newer variation of the shooter-as-screener setup, with Matt Thomas slipping a screen before flying into a flare screen set for him. The two times Toronto has run such plays (only once with Thomas as the triggerman), the team has earned excellent looks. It’s those type of plays that new assistant coach and offensive guru Chris Finch alluded to before the season when he mentioned that in a good offense, the system creates shots for players rather than asking players to do all the work themselves.

Those are the skills of individual players, both handlers and screeners. But if you dive even deeper (deeper?) into the pick-and-roll, there are even more revealing trends. Namely, how two-player combinations reveal which handlers succeed with which screeners.

There’s been a real divide in terms of success when it comes to pairing handlers and screeners. For one, Lowry has been much better alongside Boucher. That makes some intuitive sense — Lowry has always been at his best alongside electric energy bench players, and Boucher is the Charizard evolution of the Lucas Nogueira player type. VanVleet, though, has been best alongside Siakam. That, too, is logical. VanVleet’s poor numbers around the rim mean that he needs a screening partner who occupies maximal attention. Siakam offers that. And though he sometimes whiffs on screens when he knows he’s going to get the ball after the action, he can make solid contact when it’s required.

VanVleet has shot well from the mid-range so far through two games, hitting 3-of-6 in that range. He’s needed to add mid-range scoring to his game in order to create more space at the rim and punish opponents who run him off the line. He’s done that, leading the team in scoring at 18.0 points per game. He hasn’t been Toronto’s most accomplished individual scorer in the pick-and-roll, but him adding ways to become better there can only help the team.

That brings us to the Lowry-Siakam two-man game. Last year, it was rarely used, and I predicted that a play that achieved 1.58 points per possession would become a staple in the playoffs. That didn’t really happen. But Siakam and Lowry have remained underused together, though effective. The team has only finished possessions with four pick-and-rolls between the two, resulting in two triples. But boy oh boy can the result look good. Here Siakam’s threat as an off-ball finisher drew multiple defenders, and Lowry’s wizardry and VanVleet’s shooting took care of the rest.

(Fun note: of Toronto’s 16 pick-and-rolls that didn’t change ownership of a possession, a full four have come on Lowry-Siakam pick-and-rolls. That’s a larger portion than should be true given how few pick-and-rolls the team runs between the two, but I’ve thought long and hard for an explanation, and I don’t think there is one. Sample size silliness.) Lowry and Siakam together should remain Toronto’s most effective offensive weapon. With Siakam’s ostensibly improved jumper and certainly improved passing, he will be much more effective as a handler and screener this year. And Lowry remains above everything.

There’s plenty on which to improve. For one, most bizarrely, Toronto hasn’t drawn a single free throw out of the pick-and-roll in 96 minutes of gameplay. That’s bizarre. To be fair, the Raptors have only attempted 13.5 free throws per game, by far the lowest in the league. But it’s still outrageous to have drawn zero in the single most used offensive play in NBA basketball. And the Raptors have attacked the rim plenty. That’s probably what Nick Nurse has meant when he said outright that “we’re gonna definitely need to get to the line more. It’s certainly a high priority, we’re going to need to need to shore that up a little bit.” 

Beyond the need to earn more free throws, Toronto handlers outside of Lowry need to find ways to create for the rollers when they dive to the rim. Threes are important, but layups remain the best shot in the game. And VanVleet and Siakam could be better at creating layups for teammates in the pick-and-roll. VanVleet could improve as a lob passer, and Boucher’s wide catch radius there will help. Siakam is making much faster reads this year, but his passing on the move could still be faster. Improvement is always necessary, but overall Toronto has been fairly effective in the pick-and-roll.

Thus far, the pick-and-roll hasn’t really helped Toronto. Their offensive rating of 101.9 ranks a putrid 27th in the league. That isn’t fully to be explained by half-court struggles; in fact, those were to be expected. Toronto simply hasn’t found its bread and butter baskets. Last year, Toronto was the second-best transition offense in the league. This year they’re 21st. And they were the second-best defense in the league. This year, they’re 18th.

Such sickly numbers are probably to be explained by an early-season lack of chemistry, more than anything else. Early games have wackadoo results, particularly ones without training camps. And short off-seasons. Just look at the Los Angeles Clippers losing by 50 points, or the loss by the Brooklyn Nets to the Charlotte Hornets, or the Sixers to the Cavs, or the Bucks to the Knicks. Or the fact that all four of those upsets happened on the same day! Shooting variance is more likely to decide the results of these games. Besides, the Raptors haven’t played defense like themselves. They’ve made mistakes, haven’t communicated, haven’t taken the fight to opponents. And without defense, they haven’t been able to score on the fastbreak. The point is that there are plenty of reasons why the Raptors have struggled. But those things should come. Any team that employs Lowry, VanVleet, Anunoby, and Siakam will be elite defensively and in transition. It will take a lot more than two poor showings to disprove that.

If the Raptors are also to be improved in the pick-and-roll, then this team will have a bright future in the playoffs. The pick-and-roll is the most important quality in a good NBA offense. That isn’t changing. But Lowry remains one of the best pick-and-roll operators in the league. VanVleet is improving, but he’s already fantastic at finding shooters orbiting the play. It’s surprising, but the unheralded Baynes and Boucher could be a better pick-and-roll center rotation than the Raptors fielded last year. And to top it all off, Siakam is emerging as a force as both handler and screener. It’s early days, of course, and two games can’t prove anything. But If the Raptors are going to improve on last year’s playoff loss, then their performance in the pick-and-roll demands improvement. At 0-2 it’s crazy to say, but so far, so good.