Talent Borrows, Genius Steals: What the Thunder Can Learn from the Rockets

I’ll bet you’ve never heard this one before, but Russell Westbrook, James Harden, and Kevin Durant used to be teammates. Crazy, right?

Kidding. Of course you know that. If you’re even a casual NBA fan, you’ve heard about twelve thousand discussions of the trade that sent Harden to Houston back in 2012. I’m sure Sam Presti would love for all of us to stop talking about it, but it’s a worthwhile reminder how much a past decision can still hang over a franchise 6 years down the line. You can’t change the past though, so while it’s fun to daydream about how many titles the Westbrook-Harden-Durant trio could’ve won, it’s not worth crying over. It is, however, worth pointing out that 2 of the 3 MVP’s who the Thunder once had reached the conference finals this year, and the one who didn’t is the one who’s still on Oklahoma City- and worth wondering why that is.

The easy answer is that Westbrook is the “worst” MVP of the 3. That’s too simplistic, and a bit silly anyways- the worst MVP is still an MVP, and these guys are all top 10 players in the league (that’s why losing two of them is such a painful memory for OKC). It is probably true that if you were going to draft a player to build a franchise around, with the goal of winning a title next season, most NBA GM’s would pick Harden and Durant ahead of Westbrook (I’d guess the order would be something like- LeBron James, Durant, Anthony Davis, Steph Curry, Harden, Westbrook- with Kawhi Leonard slotting in between Curry and Harden if healthy). But I think a team that picked Westbrook in this hypothetical, win-next-season scenario would still be confident in their chances with Russ at the helm- especially since they’d have the chance to build the ideal team around Russ. Even if Russ is the “worst MVP”, he’s still the kind of player you can build a championship team around. The question for the Thunder is how they can get closer to being that ideal team than they were this season.

The Warriors, of course, are hardly a fair comparison for the Thunder or any other franchise; they drafted the two greatest 3 point shooters of all time (undervalued at the time) in Steph Curry and Klay Thompson, found a defensive player of the year and jack of all trades with a second round pick in Draymond Green, and then added Durant only because of a massive one-time jump in the salary cap. Trying to replicate the way the Warriors assembled talent is borderline impossible (although in a strange way, similar to OKC acquiring Durant, Westbrook, Harden and Ibaka in rapid succession). The Thunder shouldn’t try to replicate the Warriors completely- when the Warriors go to their death lineup, they have 5 good to great ballhandlers on the floor, 5 guys who are dangerous off the ball, and 3 of the best shooters in the entire league. They are quite literally lightning in a bottle.

The Rockets, however, started in a situation not too different from OKC post-Durant; at the end of the 2016 playoffs, following a first round loss, Dwight Howard departed, leaving the Rockets with just 1 star. Two seasons they later found themselves one win and one Chris Paul hamstring injury away from reaching the finals.They got there by hiring a coach in Mike D’Antoni who fully unleashed Harden, and building the perfect supporting cast around him.

Following the 2016 defeat, the Rockets brought in 3 point bombers Eric Gordon and Ryan Anderson to join existing 3 and D players Patrick Beverley and Trevor Ariza. D’Antoni declared that Harden, a nominal shooting guard, was his point guard, seeking to get the ball in his hands as much as possible. Surrounded by those elite 3 point shooters and a young but intriguing lob catcher and rim protector in Clint Capela, Harden exploded, leading the league in assists and finished second in MVP voting, behind Mr. Triple Double himself. The Rockets then bounced OKC from the playoffs before being eliminated by the San Antonio Spurs in the second round, following a complete meltdown by Harden in game 6 that the team chalked up to exhaustion. Determined to lighten Harden’s load and add more superstar talent in the offseason, Houston brought in Chris Paul, a superstar who can defend, playmake and score at a high level, someone able to run the offense for stretches to give Harden rest and space the floor and be a second playmaker when playing alongside him.

Jesse Johnson-USA TODAY Sports

That model- surround your superstar guard with a rim rolling center, elite shooters, and a second superstar who can handle some of the playmaking burden- isn’t too far off from what the Thunder have tried to do. Steven Adams can fill the Capela role (and roll) on offense, and might be an even better rim protector. And in Paul George, OKC has the secondary superstar. While PG isn’t nearly the playmaker Paul is, he can run the offense for stretches, and he’s an excellent shooter who knows how to move without the ball. OKC has the two most important ingredients- a second superstar and an elite center- in building a houstouneqsue system around Russ. The problem is OKC has failed to make the most of their 3 stars.

As Oscar Wilde may or may not have said “Talent Borrows, Genius Steals”. The Rockets playbook has proven successful around a similar core, and it’s time for OKC to try to copy as much as they can of that formula, adapting things slightly to fit their own personnel. The good news: that won’t be as much of a sea change as it might seem. OKC took a lot of (deserved) criticism for playing uncreative offensive and running a lot of isolation (ISO) possessions, but the Rockets in fact ran far, far more of them: Houston ran 1280 ISO possessions in the regular season, first by a mile, while OKC ran 959- still the second most in the league, but way less than Houston. The reason that Houston had the best offense in the league while OKC was pedestrian is the type of ISOs each team runs.

The traditional ISO has come to mean selfish and stupid in the NBA- selfish because no other teammate touches the ball, and stupid because they’re ineffective- holding the ball and going right at the player charged with defending you plays into the defenses hands. The defense does not have to make a single decision or chase anyone around the floor- they simply defend the person who they were assigned to at the start of the possession.

Houston ISOs are different. Houston ISOs involve finding the weakest defender on the floor, forcing them onto James Harden, and then letting the beard cook. Watch how Houston patiently screens and re-screens to get Golden States worst defender- Steph Curry- onto James Harden. That drains the shot clock, but Houston believes they only need a few seconds of a bad defender on Harden to get either a great look at the rim:

Or a stepback 3 pointer:

Contrast that to a typical OKC isolation possession: Russell Westbrook brings the ball down the floor, either drives at or backs down his defender, and launches a tough jump shot or is forced to pass out without OKC gaining any advantage. OKC settled for this look against the Jazz again:

And again:

And again:

Sometimes those shots go in, because Westbrook is an extremely talented player who hits difficult shots. But occasionally hitting difficult shots isn’t enough to sustain an offense at the highest levels of basketball. A Russ isolation averaged 0.82 points per possession. If you ran that play 100 times (the average NBA game has about 100 possessions per team) you would score 82 points- a terrible mark. A Harden ISO, in contrast, averages 1.22 points per possession- good for 122 points if run all game, which would be a historic level of offense. Getting a weaker defender onto Harden to start these possessions helps, but Houston also hunts for quality shots even in ISOs- note those Harden plays end in a 3 pointer and a shot at the rim, respectively, while Russ ISOs repeatedly end in a midrange jumpshot.

The midrange jumpshot is the least efficient play in basketball, for pretty obvious reasons- the further you get from the basket, the harder it becomes to score, so midrangers go in less often than shots close to the basket, and 3 points shots go in less frequently than mid-rangers. Because 3 point shots get you an extra point, however, you can hit far less of them and still be efficient- shooting 33% on 3’s get you the same number of points as shooting 50% on 2’s. The Rockets know this, and have tried to eliminate mid-range shots from their offense as much as possible, to the point that their style of offense- all 3’s and layups- has been nicknamed “Moreyball” after General Manager Daryl Morey.

OKC need not eliminate the midranger entirely (by contrast, Durant’s mid-range game has risen substantially the past 2 years in GS at the highest levels by % of total field goals in half a decade), but the Thunder should try to move in that direction. Russ is talented enough to score 1-on-1 at times, and the coaching staff should give him a mandate: “Alright Russ, if you want to get some ISOs in, go ahead, but you have to get to the rim on them.” Russ averaged 8.3 shots in the paint last year, shooting 59% on them, compared to 6.6 Mid-range shots, on which he shot an abysmal 39%. Turn 2 more possessions per game into shots at the rim instead of midrangers, and OKC will score nearly a point per game more on average.

That matters- find another point by replacing Melo ISOs with more pick and rolls, turn a few more 2’s per game into 3’s,and suddenly you’re winning the games you lost by 3 or 4 points last year. Doing the little things right adds up. That’s easier said than done of course- Russ likes to play a certain way. But getting Russ to take less mid-rangers in favor of going to the rim more will be an easier ask than getting him to give the ball up more and become Tony Parker.

It’s a similar story in the pick and roll- Houston pick and rolls end with shots at the rim by the ballhandler, lobs to the center, or kick-outs for 3 pointers. OKC can do all of those things- Russ can get to the rim as easily as Harden, Adams can catch lobs and finish (not as well as Capela, but still well enough), and the Thunder have some players on the perimeter who can knock down 3’s. But they don’t have as many as as the Rockets do, and that’s the problem. The Rockets constantly surround Harden with three above average 3 point shooters and a lob catching center. OKC’s starting 5, in contrast, features exactly one above average 3 point shooter, Paul George, alongside a slightly below average one (Carmelo) and a downright terrible one (Andre Roberson). That matters even on possessions where those wing players don’t touch the ball. Compare these two ISO possessions:

The Harden Iso does not feature a favorable switch- Klay Thompson, one of the best defensive guards in the league, is on him. Harden fights past him anyways, and because there are three good 3 point shooters on the floor, no one helps from the wing. Clint Capela’s presence as a lob threat deters Draymond Green, the theoretical rim protector, from getting in Harden’s way.

Now compare that to Westbrook’s drive. As soon as he gets a step on Ricky Rubio, Rudy Gobert is already standing in the paint, humming I shall not be moved:

Gobert is comfortable sliding that far over because his man, Jerami Grant, shot 29% on 3-pointers this season. Seeing Grant wide open, Westbrook makes the dish, and sure enough, Grant bricks it. Quite honestly, while passing to Grant was the “right” play, Westbrook trying to go 1-on-2 against Rubio and Gobert might have resulted in a higher percentage shot than Grant’s attempt from the corner.

The disparity in wing shooting between these rosters speaks to differences in organizational philosophy. For years, Sam Presti has tended to go for long, athletic guys who will defend well, with shooting as an afterthought. Think Andre Roberson, Josh Huestis, Jerami Grant (via trade), even Serge Ibaka and Russ himself. The Rockets, led by Daryl Morey, high priest of the 3 pointer, have stockpiled wings who can fire away 3’s. Two of those guys, PJ Tucker and Luc Mbah a Moute, came to Houston this year in the hopes of winning a ring- maybe it’s not fair to criticize Presti for not being able to grab those ring chasers (although, when the PG trade first came through, people sure talked like this was a title team). On the other hand, Presti’s preference for defense-first, athletic guys has left him with a roster short on shooting, and that dates back to the Durant era. Westbrook is good enough to still get plenty of points, but the lack of floor spacing has now badly hurt the Thunder in two straight playoff defeats.

When OKC does surround Russ with elite 3 point shooting, the offense starts to reach Houston levels. Per Cleaning the Glass, the lineup of Russ, Alex Abrines, Paul George and Steven Adams posted a 116.4 offensive rating (116.4 points per 100 possessions) across 422 possessions in the regular season- higher than Houston’s league best rating. Russ with a fully spaced floor is damn near unstoppable. Replace Carmelo with Jerami Grant to bring in more defense at the cost of some of that spacing, and OKC still managed a 118 offensive rating, but with a 86.1 defensive rating, across an admittedly smaller number of possessions. The Thunder not turning to those lineups more after Roberson’s injury remaining baffling, as does Donovan not going to them until they were down 3 games to 1 against Utah.

If George is back, the Thunder should insert Abrines into the starting lineup alongside him, and replace Melo with either Roberson or Grant. Russ with two elite 3 point shooters- one of whom can also create his own shot in a pinch- alongside a great rim runner in Adams and a jack of all trades defender in Robes or Grant is close to an ideal starting 5 in the modern NBA. Even if George is gone, OKC should be able to sign another good 3 point shooter with the Mid-Level Exception and build a poor man’s version of this lineup, as I detailed here.

In the wake of OKC’s defeat, many, including me, pined for more ball movement from the Thunder. But that’s a big ask for one season, and maybe counterproductive- Russ is best with the ball in his hands, even if he could stand to make strides off it. What OKC really needs is a shift in mindset- sure let’s keep running a lot of pick and roll, but make sure we’re getting to the rim or finding a 3 off it. Okay, we can run some ISO to get Russ and PG cooking- but let’s generate a switch first, and let’s make sure the floor is spaced properly while they do it. Incremental changes like that can make a huge difference even if OKC never becomes the Spurs with their ball movement.

Talent is the hardest thing to get in the NBA, and OKC has the talent in Russ, Adams, and hopefully PG should he return. On paper, the core of Russ, PG, and Adams measures up pretty favorably to Houston’s core. The difference is in the mindset of the two teams and the supporting cast around those cores. Houston’s principles have led to them decisively outperforming OKC two years in a row despite the similar star talent, and gotten them one win a way from the promised land of the NBA finals. OKC doesn’t need to go all in Moreyball. But heading a little further in that direction is a worthwhile risk, and one the franchise should take as it faces the last few seasons of Westbrook’s prime.