Part of an extensive longread by Blake Murphy
In the long-run, there's evidence that teams have far less control over three-point percentage than three-point volume, and so it's an ability to chase teams off the line that's worth examining more than three-point percentage from each spot. In other words, it's maybe not a big deal that opponents are shooting a bottom-10 percentage on both corner and above-the-break threes against the Raptors, but it's instructive that they're not taking all that many – the Raptors rank fourth in terms of opponent corner-three frequency, per Cleaning the Glass (which eliminates garbage-time), and they rank third in limiting above-the-break threes. Only Brooklyn allows a smaller percentage of opponent shots to come from beyond the arc.
This is a dramatic turn around from last year, when a nearly historic three-point differential in the sweep at the hands of the Cavaliers highlighted a very shaky approach to defending the arc. The Raptors ranked 22nd in opponent three-point frequency last year and 16th the year prior; they've managed to buck league-wide trends this year, cutting opponent three-point frequency by 3.1 percentage points as league-wide three-point frequency has risen 2.1 percentage points. As a result, the Raptors rank fourth in opponent effective field-goal percentage, with their opponents unable to fire up high-efficiency threes at a rapid rate.
Shots have to come from somewhere, though, and the Raptors haven't been able to just force teams a step within the three-point line. In fact, the Raptors are completely average in terms of opponent “long mid-range” (twos taken 14 feet and out) frequency. What that means, then, is that the Raptors have been willing to give up shots in the paint to cut down on threes. Only two teams are allowing a higher frequency of opponent shots at the rim, and only five are allowing a higher frequency of “short mid-range” (four-to-14 feet) shots. Still, the Raptors are succeeding, allowing very low percentages on these looks. It doesn't seem quite like an anomaly, but a strategy producing a high volume of shots in close seems difficult at first blush.
A big part of limiting threes, especially in the corner, has been to limit the frequency with which the team sends a help defender to tag the roll-man in the pick-and-roll. Many of the league's top defences have grown more conservative in sending help off of shooters in this way, deeming the trade-off too extreme – it is perhaps better to guard a pick-and-roll 2-on-2 without help rather than collapsing and trusting rotations or scrambling back out to shooters.
For the Raptors, this has long been at least a moderate concern, as Jonas Valanciunas has not been the most adept pick-and-roll defender. Not only has he been susceptible against shooting bigs who are a threat to pop into the mid-range or beyond the arc, he's lacked the quickness to stay with point guards coming downhill at him. Valanciunas has been much better this year, with the Raptors asking him to hedge or high-wall along the sideline far less often. Instead, to paint with reductive strokes, the Raptors' pick-and-roll strategy sees Valanciunas (or whoever the big is) drop back to contain a drive while the ball-handler's man fights over top of the screen. The hope is that the big can prevent the guard from getting to the rim long enough for his primary defender to recover and the centre can stay close enough to his man that only modest help is required.
It's still a work in progress, but the results so far are encouraging. The Raptors are giving up 0.82 points per possession to pick-and-roll ball-handlers, eighth-best in the league (up from 21st last year). They're even stingier against roll-men, in relative terms, surrendering one point per possession, second only to Philadelphia (up from 19th). And because the Raptors are defending those actions better and sending less help, they've been able to do a better job on spot-up shooters than any team in the league at 0.93 points per possession (they were second last year, too).
Helping matters fairly often is that the Raptors are also switching a lot moreon and off the ball this year, which theoretically cuts down on open above-the-break opportunities or chances to get space without a defender nearby when attacking. They're still figuring out the proper frequency with which to switch – there are still instances of over-switching, especially between Serge Ibaka and Valanciunas off the ball – but their willingness to trade off a potential match-up advantage to prevent a total breakdown is working so far.
There's still the matter of all of the shots coming in close. Opponents are shooting “just” 58.7 per cent at the rim against the Raptors (second in the NBA), and even if that's a good mark in relative terms, it's still a lot of shots coming in a high-yield area. The Raptors have to defend them well to survive like this, and they are – the Raptors contest the fifth-most shots at the rim per-game (and fourth-most twos overall) and allow the second-lowest field-goal percentage on contested shots at the rim.
This has been a team-wide effort. Among 221 players with 100 or more shots contested within six feet, the Raptors' centres rank fourth (Lucas Nogueira, 50 per cent), eighth (Poeltl, 50.9 per cent), and 42nd (Valanciunas, 55.4 per cent). Pascal Siakam (54.1 per cent) has been great, too. Even Ibaka (58.3 per cent), worst among the team's bigs, still has a mark better than 27 teams in the NBA. These individual marks in isolation are somewhat noisy; as a team-wide trend that applies to every frontcourt regular, it may be more revealing.
There have been two primary problems with the Raptors' defence so far this year, and they're both tied directly to the scheme, trade-offs the Raptors aren't necessarily willing to accept but understand are a work in progress.
The first is that the Raptors foul. A lot. They send opponents to the line at the fifth-highest rate in the league, which is a problem on its own and introduces another issue – the Raptors have spent the third most time in the league in the penalty, per research from Matt Femrite. The Raptors are the league's best defence when not in the penalty but fall to 20th when they are, which is a problem late in quarters (and may explain some of their crunch-time woes). There aren't a lot of prescriptions here. They want to contest everything, and that aggression is going to lead to some fouls.
The other is that the Raptors are only an average defensive rebounding team, a bit of a surprise since they've stayed traditional at the frontcourt positions more often than maybe anticipated. (That they have taken these defensive strides without downsizing and while leaning heavily on Valanciunas is encouraging, and probably worth a follow-up article. “Hopefully that continues,” Casey offered about the Raptors being able to defend well with a true centre on the floor.) The Raptors rebound at a borderline top-five rate when Ibaka shares the floor with Poeltl or Valanciunas but rank just 17th overall, and the tie-in to the scheme once again makes sense – bigs contesting the shot aren't in great position to also grab the rebound, and the Raptors are still working out the kinks of a strategy that requires guards and wings to crash, especially if the power forward has been lifted out of the paint against a stretchier opponent.
“I still think we can be better at getting there (to contest in help), and when that happens, we've gotta make sure we cover them on the backside and crack-down when the big goes to help. I don't think we're doing a good job of that right now,” Casey said.