Rahat: So as you all hopefully are aware, I’ve been writing a ‘what went wrong’ series on the Houston Rockets’ 2016 season, trying to dig into the numbers surrounding their catastrophic fall from grace. You don’t have to look too hard to figure out the source of the problem: the Rockets were abysmal on defense. That was obvious and plain to see. But what I’ve found perplexing is the paradox of the team’s offense. My recollection is of a mash unit playing pick up ball, with four guys watching James Harden dribble the air out of the ball. But by the numbers, the Houston offense was very good – 7th best in the league to be exact. Last year, the team’s offense was 12th in the league. The surprise wasn’t so much that the offense had improved from last season as it was that it ranked so relatively high among its peers this year.
The only publicly accessible numbers I had at my disposal are NBA.com’s crunch time statistics which measure a team’s performance in close and late scenarios. Those numbers reveal the Rockets shot 40.4% overall in crunch time, good for 17th in the league, and 30.8% on 3’s, good for 14th. The team’s overall efficiency in crunch time was 12th – a decline, but still not as bad as what I was expecting.
As I’ve now exhausted the extent of my own abilities for these purposes, I decided to reach out to Richard Li for his expertise. Is there some way to tally Houston’s performance against playoff teams and measure the decline relative to the decline of other teams? Is there something else I’m missing?
Richard: I don’t think you’re missing anything. The premise we’re operating under, I think, is as follows: The Rockets offense looked awful, but is statistically above average–what gives?
One possibility you and I talked about is that the offense this season might have been decent on average, but was maybe very inconsistent. The numbers, though, don’t support this hypothesis. The standard deviation for the Rockets ORtg (points per 100 possessions) from a game-to-game basis is 1.1 more this year than last year. That’s a little bit more variance, but nothing to write home about.
I also took a look at the difference in ORtg between wins and losses, thinking maybe there was a big disparity, again supporting an inconsistency hypothesis. The numbers don’t corroborate. The difference in ORtg between wins and losses was almost identical this year as last year, around 10.
Since you mentioned crunch time, I also dug up performance in close games. Last year, the Rockets were 17-6 in games decided by five points or less. This year they were 17-12. I thought there might be something worth mentioning here, until I saw that the Rockets record in non-close games this year was 24-29. Last year it was 39-20. So you could argue that the Rockets were better in close games this year, relative to their overall performance, compared to last year.
I couldn’t find a dataset that identified ORtg against playoffs teams.
What’s happening, I think, is a case of the eye not seeing what the mind believes. What we saw offensively this season is the same we’ve seen for the past three years. But the expectations were much higher this year and the team clearly failed to meet them. Thus, inconveniences that we simply accepted or brushed off in years past became objects of scorn and criticism this year.
Let’s be honest here, the description of a pick-up ball offense in which four guys watch James Harden dribble until the ball is deflated could be an accurate synopsis of any Rockets season over the past three years. Last year, at the beginning of the playoffs I believe, Bill Simmons had a podcast in which he was so perplexed by the Rockets’ lack of offensive organization that he suggested then-coach McHale must “just roll a ball onto the court during practice.” And this was at the end of a regular season in which the Rockets finished with a high seed and would make it to the WCF.
The fact that the only players who seem to “thrive” in our “system” are 2015 Josh Smith and 2016 Michael Beasley should speak volumes about what’s wrong with our offense. These are players who basically march to the beat of their own little drummers. It’s impossible for them to play synchronized basketball, and that’s why they succeed on the Rockets, because the entire team isn’t synchronized and hasn’t been for quite some time. So players like Smith and Beasley just improv to their hearts’ content and are then encouraged to do it more. No wonder everyone else looks so fed up. They want to act out Hamlet and are forced to be part of bad sketch comedy.
Rahat: So the numbers we have available bear out that despite the eye-test, the Rockets have a consistently good offense. However, you’ve been very assertive this year in your stance that the team would do best to trade James Harden. Naturally, if we had data to assess efficiency against superior competition (and measure the corresponding drop-off), we could better know if the offense truly was “consistent”. But controlling for that unknown, is the offense really a priority? If the team just improved its defense schematically and brought back the same offense and style of play, wouldn’t everything be okay, despite the aesthetic concerns? Or is your stance that this Harden-led offense, while good, is necessarily capped at a ceiling below the upper echelon?
Richard: The data I have doesn’t disaggregate for performance only against playoff teams. But the fact that the overall standard deviation for ORtg is not remarkably different from last year’s indicates that the comparative drop off against playoff teams, if any, from last year to this year, was not big. Otherwise, the difference in overall standard deviation would be considerably larger.
The Rockets have a consistently good offense, but there’s a caveat. From last year to this year, the Rockets average ten fewer passes per game. As a team, they move slightly less per game, despite Harden moving slightly more per game (meaning the other players are moving less). I think these drop offs are direct consequences of being James Harden’s teammate. He himself is great at putting the ball in the bucket, but he de-inspires (I just made up that word) the people around him. I think we can all agree that, no matter how offensively great one player might be, a team’s offensive productivity is greater with everyone contributing instead of just that one player.
Defense is also affected by Harden’s ability to de-inspire his teammates. No need to beat a dead horse here. No one wants to try hard on defense with Harden on his team. Twenty-three seconds of good defensive play vanishes into thin air with one Harden ‘ole! From the data, it definitely appears that defense was affected by the Rockets woes more than offense. If James Harden is “feeling it,” it’s basically taking two steps forward and one step back because the other players become more disengaged watching him ball hog and, more importantly, smoke a cigarette on defense. Better to take three steps forward and no steps back.
I think this is representative of the floor and ceiling you allude to, in terms of having James Harden on your team. He’s so good offensively that you know your offense will be at least decent, and thus making you competitive. But the more Harden does his thing, the more disinterested his teammates become. Sooner or later the marginal benefit of more James Harden is nullified by the marginal loss of everyone giving less and less of a shit. When this happens, nothing can make the team better. And thus we hit the ceiling of a team led by James Harden.
To answer your question, neither offense nor defense are a priority. They are reflective of a deeper ill, which should be the priority. This team’s culture sucks. Harden is a part of it, BUT HE MIGHT NOT BE THE REASON IT SUCKS. Contrary to popular belief, I do not advocate for trading Harden. I do advocate for changing the team’s culture. If Harden is at the root of the team’s culture, only then should the Rockets trade him. If bad coaching or a GM dictum is responsible for the team’s culture, then those aspects of the team need to change.
Rahat: So essentially, the team’s offense is consistently good, better overall than last season’s, but even more Harden-centric, with less passing and movement, except from Harden. (Harden’s usage percentage has also increased slightly from 31.3% to 32.5% this year, though that’s a negligible difference).
Your theory is that these offensive circumstances directly contributed to the decline in defensive production by inducing player apathy. Some might argue that such a hypothesis is speculative and unquantifiable. The rebuttal would be, however, that with largely the same personnel and the same scheme, and with the common sense evidence we have regarding interpersonal dynamics, the theory that guys aren’t trying as hard due to a lack of involvement is likelier than an overall decline in abilities from top to bottom on the roster.
What’s interesting is that had the Rockets had the type of year we were expecting, I’d be assessing the differences between their 7th ranked offense and the very top offenses in the league. Instead, I’m cautiously satisfied with the offense, but intrigued by your theories regarding the effects upon the defense.
Richard: I, of all people, hate when data can’t be used to substantiate theory. Unfortunately, I think this is one of those cases. As human beings, we understand that morale and cohesiveness affect a group’s product. If we had psychological survey data from across the NBA (I would actually love to see this) we could begin to measure concepts such as chemistry and motivation, and correlate those indicators to team performance. Absent that data, all we can do is speculate based upon the data that we have.
We know that the Rockets’ offense became slightly better and that this occurred with less passing and movement. We know the Rockets’ defense fell into an abyss. Our speculation is that the former is related to the latter based upon qualitative evidence and common sense.
If our speculation is not true, and the way the Rockets’ offense operates is not related to the team’s defensive productivity, then I would be begrudgingly satisfied. I say begrudgingly because there are still questions of sustainability if everything depends on one guy and the comparative ease of defending one player over the course of a seven-game playoff series as opposed to an entire team. If our speculation is true, then how the Rockets’ offense operates, regardless of its productivity, is not satisfactory because it affects team defense in such a way that the best possible outcome for the team is limited.