Warriors vs Rockets Playoffs Series Preview: A deep dive on an old rivalry with brand new faces

This Golden State Warriors season can, for all intents and purposes, be divided into two distinct time periods: Before Butler (BB) and the Butler Era (BE).

The Warriors during the BB period (spanning from October 23, 2024 to February 7, 2025 — the day before Jimmy Butler’s Warriors debut against the Chicago Bulls) was, suffice to say, wracked with absolute mediocrity. After a 12-3 pump fake to start the season, the Warriors fell off precipitously, with an injury to key offseason acquisition De’Anthony Melton not helping matters in several regards. They had fallen a game below .500 (25-26), with the only aspect of their play keeping them afloat being their ninth-ranked defense (113.2 points allowed per 100 possessions without garbage time included). Their offense was a toothless bite (112.2 points per 100 possessions — 21st), easily stifled by opponents focusing all of their attention toward the Warriors’ sole source of consistent shot creation in Steph Curry.

That problem was even more pronounced whenever Curry sat on the bench for rest or was sidelined. Amid renewed calls for the Warriors to make a move at the trade deadline, it was looking like they were headed toward another “wasted” season with their franchise icon, undisputed best Warrior of all time, arguably the best point guard of all time, and — at the very least — a top 10 player of all time.

A mid-December trade for German guard Dennis Schröder only managed a whimper. Schröder was an imperfect fit, a pick-and-roll guard who was acquired by a team with a widespread reputation for infrequent usage of the pick-and-roll, nor one that had the personnel to run a spread pick-and-roll-heavy offense that would’ve allowed Schröder to thrive. Suffice to say, Schröder was unable to fulfill the role of the much touted “second banana” that Curry needed in order for his burden to be relieved, if only for a tad bit.

Then came the trade for Jimmy Butler — in the process, shipping Schröder and a 2031 second-round pick to the Detroit Pistons; and Andrew Wiggins, Kyle Anderson, and a 2025 first-round pick (that will now convey due to the Warriors making the playoffs) to the Miami Heat. Butler made his debut against the Bulls on February 8, a rousing success that ended in a Warriors win to mark the start of the BE.

The Warriors ended their regular season on a 61-win pace, a 23-8 record, and managed to climb as high as fifth in the Western Conference — all the way from as low as 12th — before settling for seventh. A victorious play-in battle over the Memphis Grizzlies has now earned them a spot back in the playoffs — and a date with an old rival with new faces.

The Warriors and the Houston Rockets faced each other five times during the regular season, an unusual number of matchups made possible by the NBA Cup. The Warriors were denied a trip to Las Vegas by the Rockets in their quarterfinal matchup — and it happened to be a historic one for the Rockets, who had previously lost to the Warriors 15 consecutive times dating back to February 20, 2020. It was a monkey off of their backs, and while the Warriors were able to get their lick back in a February 13 matchup (one of two Rockets games involving Butler), their recent April 6 matchup was perhaps a daunting preview of the kind of trouble these Rockets could present the Warriors, in the form of Curry being limited to 3 points on 1-of-10 shooting, all while being defended by a revolving door of wings in Amen Thompson and Dillon Brooks and a stocky, physical guard in Fred VanVleet, an old NBA Finals foe of Curry’s.

Head coach Ime Udoka — another former Finals foe — has his young wards playing a brand of defense that is unforgiving and unrelenting. Making use of physicality and length, the Rockets finished the season as the fourth-best defense in the NBA with garbage time eliminated (110.8 points allowed per 100 possessions), while equally being stingy in the half court (94.9 points allowed per 100 half-court possessions — fourth).

But perhaps the key statistic with the Rockets — and one that has helped alleviate their problems trying to score (116.9 points per 100 possessions — 15th; 93.2 points per 100 half-court possessions — 22nd) — has been their offensive rebounding. They rebound 35.3% of their missed shots (96th percentile), the best offensive rebounding rate in the league. Their 18.1 second-chance points per game is a full 1.5 points better than second place (16.6), which happens to be the same difference between second and eighth on the list. Their 27.2 putback points per 100 missed field goals/free throws also overwhelmingly tops the league. What they lack in offensive creativity, imagination, and shotmaking, they make up for in winning the possession battle and giving themselves plenty of chances to score in a single possession.

But the BE Warriors do have something going for them, despite the April 6 matchup that may have seemingly tipped the scales toward the Rockets. In the final 31 games of the season with Butler in tow, the Warriors have been:

• Seventh in offense (119.4 points per 100 possessions)
• Thirteenth in half-court offense (100.3 points per 100 half-court possessions)
• First in defense (109.9 points allowed per 100 possessions)
• Fifth in half-court defense (95.0 points allowed per 100 half-court possessions)
• Third in point differential (plus-9.5)

Not to mention — having a stable five-man lineup that can be considered their undisputed best: Curry, Brandin Podziemski, Moses Moody, Butler, and Draymond Green. In 459 recorded non-garbage-time possessions, this grouping has:

• Scored 120.3 points per 100 possessions
• Allowed 103.1 points per 100 possessions
• Has outscored opponents by nearly 18 points per 100 possessions

This may paint a picture of rainbows and butterflies, but it’s still not a perfect one. Questions remain concerning the rotation, namely how Jonathan Kuminga and Buddy Hield fit into the picture, and how the playoff inexperience of key rotation pieces (Podziemski and Quentin Post, in particular) may play a significant part. There’s also the question of how the Warriors can scheme Curry toward a position to succeed despite the copious number of defenders and “junk” defenses Udoka may throw at him.

Worthy to note: Butler took only seven shots and scored only 13 points in their last loss to the Rockets. Butler’s penchant for upping his level of play and flipping the switch when the stakes are high have been on full display as of late; they’ll need more than seven shots and 13 points from him in every game of this series.

How will the Rockets defend the Warriors?

As aforementioned, Udoka should and will dedicate his most trusted defensive assets toward stifling, slowing down, and stopping Curry in his tracks. The widespread belief when it comes to superstar offensive players is that they can never really be totally “stopped” — rather, they can be slowed down, made inefficient, and forced to bleed for their points, figuratively speaking (although, in Curry’s case, that has been quite literal, as evidenced by the several scratch marks he has suffered on his arms). It’s quite rare for a total or near-total lockdown to happen — but that’s exactly what the Rockets were able to accomplish on April 6.

Take note of how Udoka had Thompson, Brooks, and VanVleet guard Curry below:

This shouldn’t be surprising to anyone who has followed Curry’s career. It’s no secret that teams have been provided some leeway in terms of how physical they can be against him, especially off the ball, where defenders have resorted to all kinds of tricks — top-locks, hugs, holds, etc. — to annoy him at the very least and knock him off his rhythm at most. At this point, Curry’s quite used to that kind of treatment and sort of lets it happen, as if he’s resigned to the fact that he won’t get any help from officiating. The Warriors shouldn’t expect that to happen in this series and instead find ways to free Curry from his off-ball shackles — while also making the most out of the opportunities he gets whenever he touches the ball.

(That is, if the Rockets and Thompson allow him to touch the ball in the first place.)

For example: in the last game against the Rockets, the Warriors attempted to involve Alperen Şengün in the action against Curry by having Şengün’s man (Gary Payton II) set the screen. Unless the Rockets have complete trust in Curry’s defender to navigate screens, they will most certainly not choose to play deep drop coverage against Curry whenever Şengün is involved. The coverage may range from screen-level meetups (alternatively termed as “soft” hedges), or more aggressive forms of pick-and-roll defense such a hard hedge/show, or even an outright blitz/trap.

The latter, more aggressive form of coverages is what the Rockets had Şengün play against Curry. Şengün hounds Curry all the way back to the half-court mark in the possession below, but Curry was able to get the ball to Podziemski, who then dished to Payton for a 3-on-2 numbers advantage:

Payton goes up to set a screen for Curry to involve Şengün in the possession below. As expected, Şengün opts to jump out and put two to the ball against Curry, once again freeing Payton on the roll. While Şengün isn’t particularly known for his rim protection, the fact that he’s been drawn up to defend the direct ball-screen action means nothing else is left on the backline to rotate and capably deter shots up close:

But what if the Rockets have Şengün guard someone who isn’t as proficient at setting screens, nor as effective as a connective passer? That was Udoka’s line of thought when he had Şengün guard Moody to start the game and at the start of the third quarter. It’s clear Udoka was daring the Warriors to make Moody a more involved party — dare him to score, see if he can make his shots, compel him to attack closeouts, and force him to create an event off the ball by cutting and coming off of screens.

Curiously, the latter was what Steve Kerr had in mind to start the second half:

Moody’s shot above is representative of a rather “duh” statement, but one that will be important for the Warriors to succeed: the non-Curry/Butler contingent will have to make their shots, or otherwise create impactful events.

Regardless of Şengün’s involvement in Curry pick-and-rolls, the Rockets aren’t afraid to send extra bodies on Curry’s drives, especially if they deem the other four players on the floor unworthy of attention and, therefore, unable to take advantage of Curry drawing an additional defender toward him.

There’s the distinct possibility that Udoka won’t be aggressive with his pick-and-roll coverages early on, as evidenced by how he elected to play a high form of drop coverage against Curry in the Warriors’ Finals series against the Boston Celtics in 2022, before relenting by resorting to other forms of coverages such as switching Al Horford onto Curry and putting two to the ball. Udoka may have Şengün do the same to start the series and adjust accordingly. Other than that, Udoka may be content with switching actions that do not involve Şengün or Steven Adams (i.e., switching 1-to-4).

While the prospect of trying to involve Şengün and catching him with his pants down (figuratively) while simultaneously freeing Curry from the lengthy vice grip of someone like Thompson may sound daunting, it’s certainly not impossible and more doable than some people might think. The key lies in Butler — more specifically, his ability to drive and touch the paint, draw attention to himself, and allow Curry and Draymond to work off of his off-the-dribble game.

Note above, in the “blind” handoff action between Draymond and Curry, how Şengün was initially drawn in by Butler’s drive, only for Butler to kick out to Draymond, who then quickly hits Curry on the handoff while also setting a solid screen on Thompson to wipe him away from the equation, with Şengün in virtual deep drop position and therefore unable to close Curry’s space on time.

Draymond’s screen is key in the possession above — as will screens by Warriors bigs throughout this series. A huge part of why Curry was unable to shake loose of his defenders during the April 6 matchup was the ineffective screens being set for him, which is both quite unusual as the Warriors are typically a good screening team and a testament to the Rockets’ corps of capable screen navigators.

Good screening can make a world of difference between the successful Draymond and Curry action above, and instances where Curry is forced to rush his shots and take them off balance because his defender was unaffected by the screen, such as the possession below:

Good, fundamental screening allows Curry to not only get good looks on his threes — it allows him to create advantages off the dribble, where teammates can feast and score against a tilted defense. Again, getting Şengün involved as much as possible will be key, especially when Udoka will have him “defend” whoever the Rockets perceive as the least threatening offensive player on the floor. Whenever Payton’s on the floor, he often gets that designation.

As such, Payton setting screens for Curry on and off the ball — such as the “away” screen below — maximizes not only Curry’s ability to create an advantage, it weaponizes Payton on offense, especially with his ability to get to the most valuable spots on the floor (that goes for other screeners who aren’t as offensively potent otherwise, such as Draymond and Kevon Looney).

Kerr and his staff may have to get a bit creative in other aspects, such as the fact that Udoka will most likely be content with switching off the ball. Switch-beating concepts such as quick slips toward the rim will be handy to have in their back pocket, especially when Curry gets the brunt of the defensive focus.

Other than Şengün, the other pressure point the Warriors will most certainly ask questions of is Jalen Green, who is the Rockets’ least-capable perimeter defender in the starting lineup and overall rotation — and, by default, will be an obvious switch-hunt target. The Warriors may have reps of Curry trying to bring Jalen up to switch onto him through guard-guard screens — but something to look out for is how often they’ll have Butler, known for his prolific switch-hunting, target Jalen in early offense, courtesy of “Pistol” screens on the side.

Jalen’s warts on defense aren’t as pronounced on the ball. Rather, he has had trouble when guarding off of it. Granted, the Rockets aren’t going to station him on Curry off the ball, but in the event that Jalen has to guard a cutter or shooter, the copious amount of information he has to process quickly may overwhelm him:

It will behoove the Warriors to test the Rockets’ ability to stay disciplined with their coverages and rules against Curry. They were able to stay largely focused on April 6; however, it’s often difficult to maintain a consistent level of focus over the course of a prolonged seven-game series. Even the smallest of cracks can present within the sturdiest of walls — it’s the Warriors’ job to hammer at those cracks until they cause the wall to shatter completely.

But in order for that to happen, reads have to be recognized almost instantaneously. Not only do the correct decisions have to be made — they must be made with speed and decisiveness.

When Curry and Butler cross paths with each other in the instance below, note how Butler’s quick slip results in two defenders attaching themselves onto Curry, with Butler being left alone as a result:

That should be a clear pass by Podziemski to the cutting Butler, who has no one obstructing his path toward the rim. But Podziemski makes the read a beat late — all the time in the world for an elite defender in Thompson to recover and deflect the pass, especially when the pass is a floating lob instead of a straight-line bullet:

But when the reads are made quickly and decisively, the windows are there for the Warriors to take. The Rockets may have top-line defenders and the scheme to bother and annoy — but like every team that has faced Curry, they’ll find it difficult to be completely immune to the confusion that Curry often sows.

While the Rockets are a defense that doesn’t force a ton of turnovers (14.4% opponent TOV rate — 16th), the Warriors did manage to shoot themselves in the foot during their April 6 meeting: 20 turnovers that were turned into 19 points by the Rockets. Suffice to say, the Warriors cannot fall into that rabbit hole.

One last note about the Rockets’ defense when it comes to a specific pairing, which is quite unorthodox in the modern NBA: the two-big duo of Adams and Şengün, which the Rockets threw out for a total of 334 possessions during the regular season, and a pairing that has kept opponents to 95.7 points per 100 possessions. In the nine minutes and fifteen seconds of the Adams-Şengün pairing on April 6, the Rockets outscored the Warriors by five points, including a defensive rating of 63.5.

What was quite intriguing about this duo against the Rockets: several possessions of zone, especially against a lineup that had no Curry and several non-spacing personnel.

While Payton was able to get a three in during the possession above, the Rockets are willing to live with that shot, if the payoff is getting the ball out of the hands of Butler. The possible inclusion of Post in these lineups is intriguing, in the sense that he is a natural zone-busting weapon uniquely occupying a seven-foot frame. With Post (in place of Payton), Hield and Podziemski surrounding them, Butler and Green will have a bit of room to work with against the double-big zone.

(Additionally, lineup data supports the assertion above. The Podziemski-Hield-Butler-Green-Post lineup has been outscoring opponents by a whopping 34.5 points per 100 possessions, albeit on a small sample size of 85 possessions. Replacing Hield with Moody nets a bigger sample of 248 possessions and is outscoring opponents by a smaller-yet-still-significant margin: 16 points per 100 possessions).

How will the Warriors defend the Rockets?

Another huge reason why the Adams-Şengün duo has worked for the Rockets: with these two on the floor, the Rockets rebounded an astounding 50.3% of their own misses. In other words, they categorically rebound half of the shots they miss with their two bigs on the floor (which, in 334 possessions, has put up an astronomical offensive rating of 125.7).

That, plus the Rockets’ overall offensive-rebounding excellence across the board (boosted by an unholy combination of athleticism, positioning, and timing), provide them with one heck of a ballast against what was otherwise a half-court offense that ranged from mediocre to impotent — despite technically finishing the season with a higher overall offensive rating (116.3) than the Warriors (114.9). Their aim against the Warriors, as it has been against every other opponent, is to win the possession battle, get more shots up than their opponent (their 93.4 field-goal attempts per game led the regular season), and win games behind the philosophy of quantity over quality.

That isn’t to say that they don’t have quality offensive players. Şengün is a post-up big who will get plenty of touches in the blocks. The Rockets score a decent 1.008 points per possession on Şengün post-ups that include passes to teammates, per Synergy tracking. But here’s the catch for the Rockets: against the Warriors this season, those Şengün post-ups have netted the Rockets only 0.60 points per possession. Suffice to say, the Warriors have largely slowed down the Turkish big man’s post-up endeavors, especially whenever Draymond takes the challenge in single coverage.

Which is why the Rockets have done all they can to get Şengün away from Green as much as possible, with methods such as “snug” inverted screens (snug because they are set near the rim, often below the free-throw line) to force switches onto smaller defenders.

And — especially in the double-big lineups, where the Warriors have had Butler guard Şengün — having Şengün come over to screen and force a switch, which the Warriors were all too willing to acquiesce:

(A brief aside: snug ball screens for Thompson — a noted non-shooting threat — have also been dialed up. The Warriors will most certainly switch these if the action entails wing-wing defender involvement or Draymond-wing involvement. Not switching can net devastating results, especially with Thompson’s rising profile as an on-ball scorer).

The Warriors also haven’t been hesitant to send doubles against Şengün to get the ball out of his hands. Those doubles have often come from the top, which signals their general confidence in rotating behind those doubles and also their relative indifference when it comes to the Rockets’ shooting competence. It’s generally a better alternative to doubling from the weak side (otherwise known as doubling from the bottom), since sending an extra body from the top would allow everyone to see the floor and rotate accordingly in unison.

Jalen profiles as the Rockets’ other off-the-dribble scorer and will most certainly command specific coverages. Just like Şengün, the Warriors will test Jalen’s ability to make decisions out of a pickle, with knowledge that the Rockets aren’t a particularly gifted passing team (dead last in the regular season in terms of assist rate — 54.9% — although that has had somewhat of a positive effect on their turnovers, of which they rarely cough up, as evidenced by their 13.8% TOV rate, eighth in the league).

But just like how the Warriors should and will involve Jalen in action, the Rockets will have Jalen attempt to get Curry on a mismatch, due to Curry often being the default weakest defensive link. The Rockets will coax him into defending and spending energy on that end of the floor — which is less energy for him to use on the offensive end.

As can be seen in the possession above, the Warriors were willing to give Jalen the switch onto Curry. Something to look out for is if they’re willing to do the same during the upcoming series — or have Curry hedge/show to counter the mismatch hunting.

VanVleet is the Rockets’ highest-usage pick-and-roll operator, but the Rockets are scoring a middling 0.958 points per possession on pick-and-rolls with VanVleet as the ball handler. The Warriors could opt to play conservatively, keep themselves out of rotation, and have their screener defenders play some form of drop coverage. That is a viable coverage given how VanVleet has been subpar as a pull-up shooter from beyond the arc: 31.5% on four pull-up-three attempts per game during the regular season; and inside the arc: 41.8% on two pull-up-two attempts per game.

But if the last matchup was a predictor of how they’ll cover ball screens for VanVleet, they won’t hesitate to mix in aggressive hedges/shows or traps, not unlike those they unholstered for Jalen.

Who will round out the rotation?

While this section generally aims to see who Kerr will trust beyond the starting five, the true aim of this section is to answer the question regarding the rather large elephant in the room: Will Jonathan Kuminga see minutes in this series?

In four regular season matchups against the Rockets, Kuminga averaged the following:

• 21.3 points
• 6.8 rebounds
• 1.8 assists
• 54.5% on twos, 42.1% on threes, 59.1% on free throws
• 58.5% True Shooting

Kerr infamously left Kuminga out of the Warriors’ last regular-season game against the Los Angeles Clippers and their play-in game against the Grizzlies. While guesting for a local radio station, Kerr stated his reasons for such a decision.

“I have to read what we need. I have to read who’s playing well together. And to be very frank, the Jimmy/(Kuminga) combination has not been great and that’s the tricky part here. And I’m playing Jimmy 40 minutes a night because Jimmy is, he’s one of the very best players in the league. And there’s a reason we’re here and we’re in the seventh seed… And the reason is Jimmy and so my job as a coach is to keep the team on a path that I think gives us the best chance to win. And for now, that’s why I’ve been leaning into the lineups I’ve had.”

Based on the lineup/on-off data, Kerr is correct in his assertion that Butler-Kuminga lineups just haven’t given the Warriors the juice they’ve needed. Cleaning The Glass, which eliminates garbage time from their data, has the Warriors being outscored by seven points per 100 possessions in 242 possessions of the Butler-Kuminga duo. PBP Stats presents slightly kinder numbers, but still doesn’t paint a good picture, with the Warriors being outscored by nearly five points per 100 possessions in 123 non-low-leverage (i.e., non-garbage time) minutes of the Butler-Kuminga duo.

With that in mind, should Kerr listen to the on-off data? Or should he pay heed to the fact that Kuminga has had success against the Rockets this season?

In my opinion, Kerr should find Kuminga a couple of minutes against the Rockets, if only to see if Kuminga does have something to offer the Warriors. Kuminga is someone they don’t have an abundance of; he profiles similarly to the athleticism and length the Rockets have on their roster.

However, Kerr may want to separate Kuminga from Butler and keep Kuminga attached to Curry, who has been the only Warrior able to coexist with Kuminga and unlock some of the skill sets that make him valuable — specifically, as a screener.

Not to mention — Kuminga’s rebounding will be important and will be a deciding factor behind whether Kerr will keep him on the floor or pull the plug (assuming he chooses to plug Kuminga in at all). To Kuminga’s credit, he tallied seven rebounds in the April 6 matchup.

Kuminga’s minutes during this series will depend on whether he can consistently string together “positive” reps like the ones above and minimize “negative” reps like the ones below:

There’s been plenty of resentment concerning Hield and his tenure as a Warrior — specifically, about his shot selection, apparent lack of situational awareness, and his nature as a streaky shooter. However, Hield happens to be the only one other than Curry and Butler who can draw attention and attract multiple defenders on the ball (not to the same extent as Curry and Butler but is easily a clear third place in that department). More often than not, he is the only Warrior other than Curry who can draw attention from multiple defenders out on the perimeter, both on and off the ball.

Suffice to say, a Hield that makes his shots will make a world of difference against this Rockets team:

Payton and Looney (as a counter for the Rockets’ rebounding) will see minutes, while Post will be important as a stretch five who can put pressure against Şengün and Adams, especially against their zone alignments. I also wouldn’t count Gui Santos out as a rotation piece, although as the series progresses, I wouldn’t be surprised if Kerr shortens his rotation and Santos becomes the first casualty of that decision.

Shoo-ins:

• Gary Payton II
• Buddy Hield
• Quinten Post
• Kevon Looney

Conditional/situational:

• Jonathan Kuminga
• Gui Santos

X-factor: Jimmy Butler

There’s no need to think long and hard about who the x-factor will be. Butler, as previously mentioned, did not get enough shots up and seemed to “float” at times during the April 6 matchup. He upped his level of play over the last two games against the Clippers and Grizzlies. With Curry getting the brunt of defensive attention, Playoff Jimmy will need to shine and make himself more of a presence, because Curry can’t exist in this series on an island.

The lowest hanging fruit: pairing Butler with Curry in as many two-man actions — or three-man actions with Draymond being the third man — as possible.

As previously stated, Butler’s drives will be key, not only in terms of rim scoring but also overall rim pressure and coaxing the Rockets’ defense into mistakes. Possessions like the one below — in which Draymond sets the screen for Butler with an empty side to force a late Şengün switch — can get Butler to the rim and allow Curry to attack from the second side:

Prediction

The intrigue this series has generated is already off the charts, if only for the fact that, in several ways, it is a fresh perspective on what is effectively an old rivalry. Of the two teams that comprise the rivalry, only the Warriors, in the form of Curry, Green, and Looney, retain memories of their previous battles with the Rockets, who were a much different team with an entirely different identity. These new-age Rockets, on the other hand, are a mishmash of old rivals from other teams (Brooks, a former Grizzly; VanVleet, a former Raptor; and Adams, a former Thunder) and a crew of young players heading into their first playoff series of their NBA careers.

When seasoned veterans square up with first timers, it can go one of two ways: the veterans flash their experience and lord it over the greenhorns and their collective inexperience; or the young’uns prove to be stronger, faster, and have more juice in their legs — not to mention, have the audacity to openly defy their elders and send the senior citizens off into the sunset.

Perhaps serving as a preview of how muddy, physical, and rocky this series could turn out to be, the Warriors — in their five regular-season games against the Rockets this season — were only able to score 104.2 points per 100 possessions. In turn, they kept the Rockets offense to 103.0 points per 100 possessions. Such a miniscule point differential could be representative of how this series will be decided: on the slimmest of margins.

While I don’t count out the possibility that these Rockets live up to their seeding and turn out to be the better team, the Warriors have the best and second-best players in this series, with perhaps the only thing stopping them from being at their best is age — and the fact that the Rockets possess the physicality to take advantage of that age over the course of a seven-game series.

I do think this series can go seven, and it can tilt either way. But I’m leaning toward experience and playoff pedigree slightly edging out youth and audacity.

Warriors in 7