At the end of my last piece on Julius Randle’s defense, I noted that the fit between he and Mitchell Robinson posed some puzzles. On the defensive side of the ball, Randle struggles in standard PnR coverages (Drop, Ice, Hedge, etc.) regardless of whether he’s involved in the action (guarding the screener or ball-handler), or whether he’s in the weak-side or strong-side help position. That, combined with his ability to defend in space, makes a switch-everything scheme ideal for him. But, while a switch-everything scheme would be ideal for Randle, it would not be for Mitchell Robinson. To be sure, Mitch can handle a switch-everything scheme. Even with sometimes questionable defensive positioning and footwork (which should improve), his rare combination of foot speed and length give him the ability to make life difficult for literally anyone on the perimeter.
Mitch can handle a switch-everything scheme... the problem is that his extraordinary skills aren’t maximized in such a scheme.
Centers are the most important defensive players on the court because they affect (or have the opportunity to affect) more shots than any other player. This is true regardless of which defensive scheme you play. But the effect that centers can have on possessions is maximized in standard schemes, and minimized in switch-everything schemes.
In standard defensive schemes, the center — unless they’re defending a stretch 5 — typically remains in the paint (2.9ing) or relatively close to the paint ready to offer help on any drives.
So centers in standard schemes not only can affect their primary assignment’s shots (and whether they shoot at all), but also most drives.
Further, centers set more screens in PnR than any other players (last year every one of the top 10 players in possessions per game used as the roll man were centers, per the NBA stats website). So, presuming there’s no crossmatch, centers will also have the opportunity to affect numerous PnRs per game in standard defensive schemes.
As a result of those two factors, bigs in standard defensive schemes have massive impacts on a team’s defense.
In a switch-everything scheme, however, both of those factors can be minimized. Regarding the first, anytime a big gets switched onto a perimeter threat they will naturally get drawn out to the perimeter. And this will prevent them from being able to 2.9 and will make it harder for them to affect offensive players other than their primary assignment.
Next consider the second factor: Since centers are the primary roll men in most offenses, having your defensive center switched away from them will result in them being involved in fewer PnR possessions. And this will straightforwardly reduce the number of actions they can affect.
(It’s also the case that switch everything schemes can cause teams to go away from PnRs and attack defenses in other ways, so things are actually a bit more complicated.)
There is one mitigating factor, however. Typically when a big gets switched out on to a perimeter player, teams are more than happy to attack that mismatch. After all, most bigs struggle to defend in space — see the below example:
As a result of this, bigs who get switched on to the perimeter will affect more actions than it may initially appear. But a lot of these will be non-ideal for the defense. This is why switch-everything teams have created schemes to help bigs in these situations (think about Tom Thibodeau’s widely-used flood-the-strong-side tactic), or they opt to go really small to mitigate this weakness (think about the 6-foot-6 PJ Tucker playing center).
While most bigs will get opportunities to affect shots when they’re switched on to the perimeter, Mitch is different than most bigs. Last year, teams did try to attack him in those situations, but with very little luck.
Why is this a problem? Teams are going to quickly learn that getting Mitch switched onto the perimeter isn’t a mismatch. As a result, if we employ a switch everything scheme, I expect we’ll start to see teams (1) use screens to get him switched on perimeter threats, (2) use those threats to pull him away from the basket and (3) run actions away from him. This will effectively remove him from plays.
At times last year, we saw teams do this. On this next play, Robinson got switched onto Harden after an offensive rebound. Recognizing this, Harden quickly cleared out. And since he’s James Harden, Mitch had to go with him. This resulted in there (essentially) being no help at the rim.
In this next clip, watch DJ Augustin. As soon as he gets the switch, he starts pulling Mitch further and further from the basket, creating room for Aaron Gordon to go to work against a weaker defender.
And, again here, the Bulls are very happy to pull Mitch away from the basket before attacking.
If the Knicks move to a switch-everything scheme, I expect this to become a common theme. With a defensive player as talented as Mitch, the Knicks should want to maximize the number of actions he can effect. For that same reason, teams will be happy to exploit the scheme to minimize the number of actions he can affect.
Towards a solution:
If Randle would be best served by a switching scheme and Robinson would be best served by a more traditional scheme, why not let them both play the scheme they are best suited for? Why not employ a hybrid scheme? Here’s what I have in mind. Players 1-4 switch most on-ball and off-ball screens, but any player involved in a screening action with Robinson’s primary assignment DOES NOT switch unless it’s late in the shot clock and a ball screen is being set by Robinson’s primary assignment.
A few quick clarifications:
First, the reason I said players 1-4 switch MOST screens, rather than all screens, is straightforward. Even teams that employ switching schemes don’t always switch. Various matchups or in-play factors can make not switching the best option. And there’s no reason to force it when it’s not ideal.
Second, the reason I think it’s fine for Robinson to switch late in the shot clock — if his primary assignment sets a ball screen — is that doing so will increase the likelihood that Robinson ends up contesting the shot on that possession.
Third, while I’ve sketched the basic outlines of the scheme, there’s still lots of room for variation here. For example, there’s a multitude of ways that you could play on-ball screens with Mitch. It would be my preference to play these screens aggressively against most teams. Drop coverage can be especially effective (see Milwaukee and Utah’s top-rated defenses) because of the way it’s designed to force teams into taking mid-range shots (the on-ball defender trails around the screen which makes it difficult for most shooters to pull a three, and the big drops deep into the paint to dissuade shots at the rim. That leaves room for the ball handler to use a snake dribble to pop off a free-throw line jumper).* But drop coverage is often the forced choice for teams who have good rim protectors who also struggle to defend in space (again, see Milwaukee and Utah). The Knicks aren’t in that position. They have a young rim protector with (perhaps) the greatest recovery speed in the NBA. In my view, they ought to utilize this to employ a more aggressive defense. While I wouldn’t go so far as to trap every ball screen (like the folks in Miami like to do), I would let Mitch play at the level of the screen and hedge against typical matchups.
* Interestingly, Milwaukee allowed the most 3-pointers in the NBA last season, and Utah allowed the fewest. Although, I believe Milwaukee’s high allowance had to do more with other aspects of their defense besides their tendency to play drop coverage — e.g., they were often happy to sag way off anyone they deemed to be a below average shooter (amusingly, the entire Pelicans team last season), and allow them to shoot to their hearts’ content.
(1) Maximizes the number of actions Mitch can affect.
(2) Minimizes the number of times Randle has to play drop, hedge, ice coverages (at least, with his primary assignment as the screener/roll/pop man).
(3) Reduces the amount of off-ball help Randle has to offer.
(4) Switching a lot probably helps out weak defenders like Knox as well.
(1) This defense is complicated, and will require a lot of communication. Players will have to be aware of who is defending screens everywhere on the court, because what sort of help they need to offer will vary accordingly. This may actually work against some of Randle’s weaknesses. As you saw in my last piece, he tends to space out a bit on D
(2) I’m not familiar with any precedents for this particular hybrid scheme. There likely are some, but I’m not familiar with them. A lack of precedent can be good. As we saw last year, a number of teams had success mixing in zone defenses that teams weren’t prepared for. On the other hand, it makes it difficult to fully appreciate the weaknesses of the defense when you haven’t seen 29 coaches game plan to attack and exploit it. And if the Knicks use a hybrid defense regularly, it won’t be novel for long.
Mitchell Robinson and Julius Randle are two extremely talented players who the Knicks have invested in for the next several years. Given that, it’s worthwhile to take the time to think carefully about how to maximize their fit on the court. Because of their particular strengths and weaknesses, that might require going outside the box-and-one (defense joke!!!!) a little bit. And perhaps a hybrid switch-man defense is just the way to do it.
(Shout out to @TalkingNYKnicks on Twitter. Questions he asked me motivated me to explore the central question behind this piece in detail.)