Other than the Warriors' dominance, the biggest story over the past couple of seasons in the NBA has been the rise of 3-pointers and offense thanks to "Moreyball" and the Rockets.
Offensive efficiency has certainly dipped for periods of the league, sure, but it's no surprise that the surge of offensive optimization due to analytics around the 3-pointer has led to the league's best collective season ever.
Given that data plus the narrative around the league, you would think there is a strong correlation between teams that limit 3-point attempts and Defensive Rating.
Is Allowing 3-Pointers … a Good Thing?
And yet oddly, the team that has allowed the highest percentage of 3s this season, the Milwaukee Bucks, also sits first in Defensive Rating. The team that allows the second-most 3s, the Boston Celtics, ranks third. The next two — Memphis and Indiana — are both in the top 10.
In fact, there's almost no correlation between 3-point rate allowed and Defensive Rating:
Why is that the case, especially given that 3-pointers overall are more important than ever in the NBA? Is the answer that the worst teams allow a lot higher percentage of open 3s?
That could partly be true, although the league-leading Bucks have allowed the second-highest percentage of "wide open"3-pointers this season. The Pacers and Celtics are in the top four of allowing the most "open" 3s.
Defensive success for these elite teams like the Bucks and Celtics doesn't seem to be about limiting 3s or even open 3s, so what is it?
The answer is simple: It's about who is shooting those 3s.
The Old Is New Again
The Bucks allow the most 3s in the league, but they also allow the fewest shots — by a freaking mile — at the rim. It's a conscious decision on their part and how they built their league-leading defense.
And part of that strategy is allowing opposing big men to bomb away from outside. Take a look at some of their highest 3-point attempt games allowed to opposing players this season:
- Marc Gasol: 4.0 three-point attempts per game, took 12 vs. Bucks
- Al Horford: 3.1 three-point attempts per game, took 11 vs. Bucks
- Luke Kornet: 4.1 three-point attempts per game, took 11 vs. Bucks
- Serge Ibaka: 2.5 three-point attempts per game, took 11 vs. Bucks
- Blake Griffin: 6.8 three-point attempts per game, took 11 vs. Bucks
In one of the Bucks-Warriors games, Jonas Jerebko, who averages 2.5 attempts per game, took a team-high nine 3-point attempts. He made four of them, which is fine, but the point stands: If Jerebko leads the team in 3-point attempts over Stephen Curry, Klay Thompson and Kevin Durant, you like your chances of success.
Let's dive into that Griffin game mentioned above. Milwaukee is bucking the trend made popular last season of switching all pick-and-rolls and instead dropping the big to contain the ball-handler:
This one is not in a pick-and-roll, but Giannis Antetokounmpo, who is guarding Blake, essentially has a foot in the lane:
They played Gasol the same way. I mean, look how far Brook Lopez drops here, leaving Gasol with a wide-open 3 from the top of the key:
Importantly, the Bucks don't necessarily scramble to help, which then leaves the defense incredibly vulnerable to ball movement — that's a huge weakness of traditional pick-and-roll defense and why switching can be so enticing. They're largely fine with staying home on shooters and letting the big men pop.
It's the same story vs. nearly every team. Blazers forward Al-Farouq Aminu, who averages 3.6 three-point attempts per game, took nine in a game against the Bucks this season — he missed them all. Even if he hit them, the math is in the Bucks' favor, even though the overall 3-point math says allowing 3s is bad.
Aminu is a 35.8% three-point shooter this year. That's about average overall, but the Bucks were clearly fine letting him shoot instead of guys like C.J. McCollum, Damian Lillard and Seth Curry.
Switching vs. Traditional
This gets to something Thunder head coach Billy Donovan talked about to our own Matt Moore recently. Here's his quote on switching:
"If you’re switching, are you taking a really good defender off a really good offensive player and putting maybe not as good a defender on [them]? I think switching can be overrated. I think there’s time you have to have physicality to get over and through screens.”
"There are teams that try to manipulate to create switches, like Houston with Harden. They’ll bring guys up to try and create that, but sometimes I think you have to fight it and say, ‘We’re not going to switch.'"
"If we have Paul George or Andre Roberson on him, we want to keep them on James Harden; we’re going to work to do that. If we have someone like Russell Westbrook and we want to keep him on his matchup, we’ll do that. But I think it’s based on personnel, and there’s some times when you’re in a compromised situation, a bind, when you have no choice but to switch and take away what the inevitable is going to be.”
It's a good point, but there's another value of not switching: On a switch, the big man is immediately covered, which means it's more likely the lead guard — the better shooter and likely better offensive player — will keep the ball. The Bucks are doing two main things in their scheme: 1) protecting the paint and 2) forcing the ball to less impactful offensive players.
The data backs this up: The Bucks actually don't allow a ton of 3s, relative to the league, to opposing point guards or shooting guards. But they do lead the league in 3-pointers allowed to opposing power forwards and centers.
3-Point Attempts Allowed to Positions
In general, teams this season that have allowed more 3-pointers to point guards have done worse; teams that have allowed more to big men have done better. The Bucks are just the extreme example of this.
And that split is something switching de-emphasizes.
Sure, the switching strategy has its benefits — no, we didn't just dream about the Rockets flummoxing the Warriors at times in last year's playoffs. Of note, it can limit penetration and removes some responsibility from help defenders.
It remains to be seen, however, how much of that success due to switching was systemic or merely due to novelty. Further, it's possible it requires such perfect personnel that just isn't available on most rosters.
But perhaps the biggest flaw of switching is that it inherently supposes the wrong question. Last year when the Rockets were switching with success, the popular question became, "Is this the optimal way to defend Stephen Curry?"
But that question has no right answers. It just has various levels of wrong ones.
The more important question, and one that the Bucks and Celtics are trying to answer this season is this: How do we get the ball out of Stephen Curry's hands and into the hands of an inferior offensive player?
Of course, the Bucks and Celtics have the benefit of outlier defenders like Antetokounmpo who can close out and help with brilliance. Playing any defensive set, traditional or not, is easier with those guys. But just 10 months ago it was mused whether switching was the utopic form of defense, and perhaps that's not the case. Perhaps it's relative.
It will be interesting to monitor the performances of the Bucks and Celtics defenses in the playoffs; it's not like their defensive schemes are new. It'll also be interesting to see whether teams take a similar approach to, say, the Warriors, and leave Draymond Green wide open all series.
But at least so far this regular season, those teams have found immense success in the comforting, familiar scheme of traditional pick-and-roll defense.
