The Warriors-Celtics Finals rematch was a battle of defensive schemes and coverages

The Celtics were inflexible and stayed to an old principle that didn’t work. The Warriors returned to what made them NBA champions.

The marquee 2022 NBA Finals rematch between the Golden State Warriors and the Boston Celtics was much anticipated — if only for what transpired half a year ago between two of the league’s top defenses.

In one corner were the Warriors, a team that finished just behind the Celtics in terms of points allowed per 100 possessions. Their defensive rating of 107.6 was 4.7 points better than league average last season, with a half-court defensive rating of 91.2 that was third in the league and was 4.4 points better than league average, per Cleaning the Glass (which eliminates end-of-quarter heaves and garbage time).

The Warriors’ brand of defensive versatility allowed them to play multiple schemes and pick-and-roll coverages. If the situation warranted it, they would play drop to minimize help and keep ball-screen possessions a two-on-two endeavor. They certainly had the personnel and lineup configurations to switch almost every screen, on or off the ball. They had the wherewithal to rotate behind aggressive coverages such as screen-level meetups or traps. To throw off opponents, they would even employ funky zone configurations and “junk” defensive alignments such as a box-and-one or a triangle-and-two.

To summarize, they had every ounce of what made a team a championship-caliber defense.

In the other corner, the Celtics were the best defensive unit in the league last season, leading the league in terms of points allowed per 100 possessions. They limited opponents to the tune of a 106.9 defensive rating, 5.4 points better than league average. Their 90.4 half-court defensive rating also led the league and was 5.2 points better than league average.

They had the distinction of being the switchiest team in the league, buoyed by a roster that had the capacity to switch almost every on or off-ball screening action. Per Second Spectrum, the Celtics switched the most off-ball actions last season (15.2 per 100 possessions). They were second behind the Miami Heat in terms of the rate of ball screens they switched (28.6 per 100 possessions).

In theory, they were built to stifle a Warriors offense that was predicated on a multitude of off-ball movement and screening actions. We’ve seen the formula before from teams like the Cleveland Cavaliers and the Houston Rockets — switch everything, keep things flattened, and force the Warriors to rely on late-clock shot making.

Of course, the Warriors short-circuited such plans on their way to titles. With the Celtics, the method with which they used to bypass the Celtics’ switching tendencies was simple: run continuous ball screens for Steph Curry and punish a defense that wasn’t keen on being put in rotation.

This season, the Warriors defense (112.4 overall defensive rating, 13th; 96.7 half-court defense, 18th) has experienced a precipitous fall-off. A bunch of factors have contributed to this: post-championship malaise, fatigue, and personnel downgrades, to name a few. The trademark defensive versatility that powered them to a title has been sporadic — sometimes outright absent.

The Celtics, on the other hand, are on track to become the most efficient offense in NBA history (121.0 offensive rating, 7.9 points better than league average). Their excellence in the half court (107.8 offensive rating, 11.6 points better than league average) has been the main driver of their early season success.

Much like the Warriors, however, their defense has largely struggled to return to its elite status. They’re seventh in overall (non-garbage-time) defensive rating (111.5) — good, but far from transcendent. They’re just below the Warriors in half-court defense (96.9 defensive rating, 19th). The absence of their main defensive anchor in Robert Williams III has certainly played a huge part, while the recent loss of Al Horford — a coverage-versatile big — hasn’t helped them in any regard.

This first regular season matchup was intriguing in the sense of how much carryover there was going to be from the Finals, especially in terms of defensive coverages. How will the Warriors — without their best wing defender in Andrew Wiggins and having lost key defensive personnel over the offseason — cover Jayson Tatum and Jaylen Brown? How will they account for the Celtics’ three-happy tendencies, one that has been the main driver of their league-best offense?

On the Celtics’ end, will they do anything different against Curry ball-screen possessions? Will they continue to switch actions endlessly, as was their brand last season? How will they account for the absence of two of their most important defensive personnel?

The answer to the question of ball-screen coverage came faster than expected — and to a certain degree, it was quite shocking in the sense that they elected to do nothing different:

Off the split action in which Jordan Poole curls around the Kevon Looney flare screen, Poole makes use of Looney’s screen anew. Blake Griffin — a somewhat scheme-limited center — is in deep drop. Jaylen Brown is no slouch when it comes to screen navigation, but Looney catches just enough of him to free Poole for a pull-up bomb.

Drop coverage does have its uses: as aforementioned, it prevents a defense from having to be put in rotation, spending less energy defending swing-swing ball movement. It also results in a favorable shot profile if done correctly — fewer shots at the rim (the Celtics are second in opponent rim rate), more mid-rangers forced (they allow the highest mid-range rate) and fewer three-point looks allowed (fourth in opponent three-point rate).

It was somewhat crucial for the Warriors to establish some sort of mid-range game to beat the Celtics’ drop coverage. Klay Thompson — who finished with 34 points on 26 shots (63.2 TS%) — took it upon himself to be the Warriors’ resident mid-range merchant.

Although it’s not ideal nor sustainable for Thompson to take mid-range pull-ups off of several dribbles, he was able to make it work against the Celtics’ drop — and accordingly punished it:

These are the kinds of shots the Celtics will happily take on the chin, but just like with every coverage, it comes with risks. If teams are able to establish a mid-range rhythm, the gamble turns into a risky endeavor. Drop may not put your team in rotation, but there are still exploitable gaps.

Thompson manages to touch the paint against a drop that has more or less contained him — but Brown falls ever so slightly behind, which puts all the pressure on Griffin to navigate the middle ground. Brown tries to peel off toward Looney but is too late to stop the pocket pass and too small to affect Looney on the finish.

Even Poole had success in the mid-range against drop — which may have come from reading the scouting report on the Celtics’ defensive shot profile, which states that they allow the second-highest shots from floater range (interestingly, just behind the Warriors):

Even while having extensive footage on what happens to defenses when they play drop coverage on the greatest shooter of all time — some of which most likely include their own experiences during the NBA Finals — the Celtics were still adamant about putting all the pressure on the on-ball defender and electing to drop their big.

While the Celtics may possess arguably the best corps of screen-navigating point-of-attack defenders in the league, it still doesn’t justify this kind of possession:

While the Celtics had flashes of shutting down the Warriors’ off-ball movement and screening through their switching, the Warriors had more success than expected at getting their motion offense going. Possessions rarely died on a whimper, with the ball continuously popping, personnel moving around like a perpetual motion machine, and hubs relentlessly hunting for cutters and shooters curling around screens and handoffs.

On the other hand, the Celtics offense had massive trouble creating advantages. Some of their troubles came through unfavorable variance — open shots they’d normally make didn’t fall in — but it can’t account for how crisp the Warriors were in terms of rotations, while also taking into account conditional switching — i.e., switching when it was acceptable, and fighting over screens whenever switching wasn’t the answer.

The Celtics are fond of using off-ball wide pindowns for Tatum like the one above to force defenses to make tough choices: do you switch a smaller defender onto him, or do you fight over the screen in an effort to stay home?

The Warriors chose the latter option — and were brilliant in its execution. Curry “tags” Tatum on some sort off-ball hedge in order to give Thompson ample time to recover, after which Curry himself recovers back toward Marcus Smart. With no advantage created, Tatum tries to isolate against Thompson, but is forced into a turnover.

The Warriors were able to run the necessary shooters off the line, gave the Celtics a bit of their own switching medicine, navigated and fought over screens with verve, were sharp in their rotations and “X-outs”, and let the personnel who they were fine with shooting the ball chuck from outside.

Tatum struggled mightily. He managed only 18 points on an inefficient 6-of-21 from the field. Again, some of the shots he took were shots he’d normally make — but others were misses against stifling defense, especially whenever he tried to force things inside.

Remember when I mentioned at the top of this article that the Celtics were second in the league (28.1%) in terms of opponent rim rate? The team above them: the Warriors, who allow only 24.8% of their opponents’ shots to be taken within four feet of the rim — 10 percentage points better than league average, per Cleaning the Glass.

In this battle of defensive schemes and coverages, the Warriors notched an important victory. Regaining their identity of defensive versatility — against the team with the best record in the NBA, no less — is an important stepping-stone at this juncture of the regular season.

They were able to limit the best offense in the NBA to an offensive rating of 104.9 and a half-court offensive rating of 93.3 — no doubt the best defensive performance of the season.

This is the kind of defense that won these Warriors their fourth NBA title in eight years. Consistency will be key as well as proper usage of personnel and correct mixing and matching of lineup combinations. More importantly, this win will go a long way toward team morale.

If anything, this win proved that the Warriors still have the juice to contend for another championship. But to do so, they will have to continuously cross their t’s and dot their i’s, as they did against the Celtics.