,

DIY or Die: the Toronto Raptors 2016 Offensive System

Season to date the Toronto Raptors have one of the most effective offenses in the NBA, but all season long there have been questions about how sustainable it is or how effective it can be come playoff time. One of the main reasons for these questions is the heavy reliance on free throws and one…

Season to date the Toronto Raptors have one of the most effective offenses in the NBA, but all season long there have been questions about how sustainable it is or how effective it can be come playoff time. One of the main reasons for these questions is the heavy reliance on free throws and one on one play, as the Raptors are among the leaders in points per possession and free throws made but among the worst teams in assisted baskets. This was an issue for the team last season before the team famously self-destructed in the first round and it has become even more of an issue this year with the Raptors dropping from being mediocre at generating assists in 2015 to being among the worst in the NBA at it in 2016. Do the Raptors not have enough offensive talent, are they incapable of playing together or is this an issue with the core principles of the Raptors offense?

It can be tempting to chalk the low assists up to a personnel issue but the Raptors do have some capable passers on the roster, they’re just not getting the assist opportunities that they’re used to. Luis Scola is currently averaging a career low in assists per 36 minutes and seeing a 44% drop from last season while Patrick Patterson’s has dropped 45%, the injured DeMarre Carroll’s has dropped 47% and the enigmatic Terrence Ross has seen his drop 30%. If there are cutters to hit the Raptors have players who have proven capable of hitting them in the past so the issue has to be one of strategy, not personnel.

If you watch any of the Raptors home broadcasts you will inevitably hear the broadcast team talk about the importance of ball movement, but the volume of passes is not the issue – the Raptors rank 13th in the NBA in passes thrown per game – it’s the quality of shots those passes give them, if they get a shot off the pass at all. Despite being in the upper half of the league in passes thrown they’re 29th in assists, 22nd in secondary assists and 30th in assist chances so the issue has less to do with how often they pass than with the action that comes before those passes. The team doesn’t use motion and passing to create open shots for the team, they use it to shift the defense to make it easier for one of Kyle Lowry, DeMar DeRozan or Cory Joseph to get inside the defense and any assisted baskets happen after that penetration.

Seth Partnow of Nylon Calculus has made some very handy matrices showing who assists on whose baskets for each team in the NBA, and when you view the Raptors matrix this becomes painfully obvious:

TORasts
None of the other top offensive teams are so reliant on three players to generate looks for teammates, not even the top-heavy and iso-centric Cleveland Cavaliers. The Raptors appear to be getting more looks off of the individual play of a few players than any other team in the league and that’s a problem for a few reasons.

First, there is the danger that having to work a few players so hard for every open look will wear the team down. DeRozan and Lowry are already 3rd and 7th in the league in minutes per game while ranking 12th and 19th in total distance covered. Even with DeRozan’s work ethic and Lowry’s skinnier physique this year wearing them down is a very real possibility. When they’re resting so little and running so much having an offensive system that can get them easier baskets and delegate responsibility should be a priority so they don’t have to add the constant physical punishment necessary for their slashing style of play for the team be effective. It would be nice if every now and then all those picks that DeRozan uses off the ball would lead to something immediately productive instead of another post up or pick and roll opportunity or if Lowry could take a few more possessions off as a deadly spot up shooter, especially considering Lowry’s eFG% as a spot up shooter is an amazing 67% and good for 9th in the NBA.

Second, the offense becomes predictable. By now opposing defenses know to pay little mind to off-ball actions because the real action will occur when one of those three has the ball and is able to attack a closing out defender, come off a ball screen or try to exploit a perceived mismatch. The Raptors are talented enough to overcome this when the opponent is an average or worse defensive team and things will click enough times against top defensive teams to keep their overall offensive numbers respectable, but what happens when they have to face a top defensive team every night in a playoff series? You can’t expect the Raptors to suddenly change up everything they do overnight and even if that was a reasonable expectation there are serious questions as to whether Dwane Casey is a coach capable of recognizing the need for or implementing that change.

Sometimes it can be tough to see whether this is the result of poor play design and strategy or poor execution but it seems to be more of the former, though I’m sure there is plenty of the latter. Here we have one possession against the Cavs and it’s unclear what the point of any of this movement really is:

There’s some off-ball action for Lowry but Joseph is ignoring it so the defense does as well – that may just be poor execution by Joseph. The screens are also really weak, there is zero chance that they could spring Lowry for an open look unless the defense really falls asleep. Even if those screens weren’t so weak and Joseph didn’t ignore the off-ball action, Lowry’s only option is to come out too high for a shot to be a reasonable possibility or curl because they’re too close to the sideline for him to be able to fade for an open jumper if his man goes under the screens and shoots the gap – that seems like poor design. It looks like that off-ball action is supposed to be a distraction while Scola and Joseph run a two man game, but it doesn’t work on any level. The Raptors don’t sell it so the Cavaliers ignore it and most of the shot clock is wasted.

To see how another team runs a similar action, consider this Atlanta Hawks set from their recent loss to the NY Knicks:

They don’t spend 10 seconds of the shot clock playing hot potato with the ball to set up the two man game, they just go into it and it keeps Arron Afflalo sufficiently distracted that he doesn’t realize what is happening until it’s far too late for him to stunt at Hawks point guard Jeff Teague. If he had, the screens set for Kent Bazemore were more solid and could have had led to an open look. If Teague doesn’t feel he has the layup and Bazemore isn’t available for the pass Teague may be able to thread a pass to Thabo Sefalosha in the corner, with Walter Tavares in position for a potential backscreen. It’s also worth noting that their entire play was done in less time than it took for the Raptors to set up their primary action.

Here’s another token Raptors offensive set:

The Raptors spend 10 seconds setting up this pick and roll with DeRozan, and when he doesn’t get a shot there’s nothing else happening so there’s nothing to do but throw it back outside and try a one on one play or another pick and roll. This takes up the whole shot clock.
When a team like the Hawks use up an entire shot clock it looks more like this:

Unlike a typical Raptors set, there are 4 actions there that could have produced a shot: Shelvin Mack off the Kyle Korver backscreen after his initial pass, the Korver run to the corner and two quick ball screens at the top near the end of the possession. Because the Raptors don’t have multiple options built into their offensive sets like that there’s very little chance of anybody but DeRozan, Lowry or Joseph generating any assists so almost all of the Raptors shots come off their penetration or the few post up touches that Scola or Jonas Valanciunas manage to get.

This isn’t really ideal, and it’s why the Raptors are among the worst at generating assists. Assists are generated when the offense is able to hit open shooters or cutters, both of which become less likely when attacking a set defense, as the Raptors usually do. It seems pretty simple: the less a defense has to do the more likely they are to be successful and the Raptors barely make them work. While looking into how the Minnesota Timberwolves offensive system functions(or rather, doesn’t function) Partnow posted this chart:

defmovement

The Raptors are among the worst at making a defense move, which means they attack a set defense more often. Cutters like Carroll and James Johnson are at their best when they have seams in the defense to attack or when a defender is trying to read what is happening elsewhere and loses track of his own man and that is not going to happen when the defense is stationary and there is nothing noteworthy to distract individual defenders. As anyone who watched the Raptors play last year can attest a lot of open looks happen off of blown rotations, which don’t happen as often if the defense doesn’t need to make rotations constantly. A defense can be expected to make mistakes on a certain percentage of rotations or shifts in position so the more you make them do it the more mistakes you get to take advantage of. By not making teams move much the Raptors are giving them a break on that front.

I’m not sure exactly what the solution is but I know this is something that everybody wishing for some long-overdue postseason success for the Raptors should be concerned about. To their credit the Raptors seem to be trying to implement some new things into their offense and have taken out or reduced some things that weren’t working earlier in the year – most notably sets designed to spring DeRozan for catch and shoot jumpers off of screens – but with practice time being scarce during the season it can be really tough to revamp the offense on the fly. I’m hoping that all-star weekend being a local event this year will allow the Raptors staff to think of some tweaks and get in some repetitions with the players over the extended break even with two of their players likely being occupied with the festivities. That’s a best-case scenario but even if that is possible I’m somewhat skeptical that Casey is capable of making the necessary changes because he’s been in charge as the team has moved in this direction over the last few years – the teams assist percentage has actually dropped every year since he took over in 2012, starting at 60.9 in 2012 and dropping to a low of 50.9 this year. For better or worse this really seems to be the direction Casey has wanted to go – we can’t really blame the assistant coaches anymore, they’ve been overhauled already – making what happens the rest of the year something of a referendum on him as a coach. After the Wizards game I wrote that it’s becoming more common to see some solid game plans and in-game decision making from him and I definitely stand by that but that’s all for naught if the team’s offense or defense is not built on a solid foundation.