With the second week of the NBA logged into the books, we’re starting to see the real form of the season, and some of the odd blips are starting to solidify, from the MVP performances of Giannis Antetokounmpo to the Rookie of the Year campaign for Ben Simmons.
There’s a lot of the season we still have not covered, and it looks like it’s going to be a lot more than “yet another Warriors title”. And with that, let’s take a look back at the last week in basketball with a more objective, stats-focused eye.
Cleveland: Everything is on fire
Is it time to panic? For the Cavaliers, doubts are starting to pile up, and fans are beginning to legitimately worry that LeBron James will leave them soon for greener pastures. They’ve lost games against the Pelicans, Nets, and Knicks. Their team, outside of LeBron, is looking downright weak. I have a good news/bad news dichotomy for an answer here. First of all, I did not like Cleveland going into the season because they projected to give a lot of minutes to negative players and many who are poor fits — Derrick Rose, Dwyane Wade and Jeff Green. I’d also say projection systems using RPM overrate Cleveland for the regular season because LeBron has been so much better in the post-season, and RPM is one of the few metrics that includes that data.
On a positive note, remember that they swapped All-Star point guards, and Isaiah Thomas has yet to play. They were going to be worse during the first half of the season anyway — it’s all about how healthy he’ll be when he returns. Additionally, I don’t think Jae Crowder will continue to play this poorly, and one of these days J.R. Smith’s shooting percentages will return to their historic means. This is not a league average offense, and they should improve. I would just temper expectations on how good they can be, unless Thomas returns as good as ever and they figure out their rotation problems.
Rudy Gay and the zagging Spurs
After a curious offseason that I criticized, the Spurs, like always, are looking fine. They had a hiccup against the Magic, but otherwise it’s still the same precision offense with a no-mistake defense that should net them 50-plus wins. Many people have been waiting for the other shoe to drop because they’ve been making the types of moves we see from lesser organizations, tying up Pau Gasol and his statue-esque defense for more years, letting younger valuable defenders leave on good deals, and bringing in Rudy “the decision every team has tried and later regretted” Gay. But their core players and coaching has been so excellent it’s hard to see any kinks anyway.
The Spurs have been zagging while everyone else is zigging. They snatched up mi-drange shooters, like LaMarcus Aldridge, as everyone else has been preparing for a 3-point only future. They loaded up on size as everyone is downsizing. And they’ve gone after older rookies as we all obsess about the age curves of kids who can’t even buy alcohol. I’ve talked about this before, but it’s all the more apparent now. Their voodoo won’t last forever, but for now it’s still working — they’re playing winning basketball without their MVP candidate, Kawhi Leonard, and they’ll once again be one of the best teams in the league.
Your regularly scheduled 3-point defense reminder
I’ve talked about this issue a lot, but teams do not have much control over shots behind the 3-point line, mostly because those are open shots. Thus, it’s important early in the season to identify outliers because history suggests regression to the mean will be at play. Right now the Boston Celtics and the Orlando Magic are the two leaders in 3-point “defense,” meaning their defensive statistics look better than they should. This makes sense the Celtics are doing better without Gordon Hayward than expected (by many analysts outside of Boston, at least), and the Magic’s strong start to the season is, well, a bit shocking. On the other end of the spectrum you have the Indiana Pacers and Phoenix Suns; for the latter I think it’s the case of, yes they are bad but nobody’s quite this bad. There is some control over these types of shots, but not to the extent where the leaders allow 28 percent and the bottom teams allow over 40 percent shooting.
Los Angeles: No Paul, No problem
There are dual storylines going along with the Clippers and their optimistic start to the season. Firstly, I can imagine a critic of Chris Paul using these results to validate a belief they already have about his value. They’ve blasted out of the gate with the best point differential in the league through Sunday, and they do not look like they miss their old point guard at all, even with the unfortunate injury to Milos Teodosic. However, the Clippers were also one of the hottest “over” picks by people doing analytics-heavy predictions coming into the season. I explained it in my preview, but essentially the players they gained over the summer are surprisingly good and there’s a rebound effect with a lot of ball-dominant players like him (i.e. Blake Griffin’s value is higher without him because he gets to use the ball more.)
However, the Clippers will cool off for numerous reasons. They had the best defense in the league on Sunday — and that’s already crumbling. No level-headed analyst in the league would state they’re the best defense. Blake Griffin is also playing the best basketball of his career, but he often misses long stretches of games and if that happens this season they will look quite shallow. Then here’s one telling sign: they have one of the larger 3-point differentials in the league — a much higher percentage than their opponents — and that stat tends to close unless you’re the Warriors. Overall, I’d say the start to the season is a solid evidence there’s still a good club without Paul, but they will not the best team in the league — which should surprise no one except the Rivers family.
The problem in New Orleans is … ?
We’re going to hear a lot about the DeMarcus Cousins-Anthony Davis pairing in New Orleans as the team fights to stay over 0.500. They’ll have tough stretches this season, and the fight at the bottom of the playoff bracket will be pretty fierce. As the two most prominent members of the Pelicans, those two big men will shoulder all the blame. But the problem isn’t with them; it’s with everyone else, save for Jrue Holiday (some nights.) It’s a thin team with very little perimeter firepower, and the franchise has little of the way of assets. It’s downright scary to imagine this roster without those two — so please let’s not blame them because they’ve been playing fantastically.
The effects of continuity and Denver
Good news, Denver Nuggets fans: your defense has performed remarkably better this season. Opponents are no longer draining all their shots against you. The bad news: your offense has fallen greatly. Denver’s turnovers have ballooned, and their shooting percentages are down across the board. The magic of Nikola Jokic starting has worn off. His stats are down, sure, but mainly turnovers are the issue and overall the team performing worse has sunk his overall stats. Wilson Chandler and Jamal Murray’s shooting percentages are catastrophic. It’s a team-wide regression downwards.
The pass-happy Nuggets look a lot more stagnant and those circus passes Jokic was throwing last season are going into the opponent’s hands now. I think people were excited to see Denver with some continuity after discovering some of their best lineups, but an ugly reality may have been exposed: continuity is not necessarily a good thing. Getting used to your teammates and developing chemistry can be useful, but opponents can get used to playing you as well, like how the league adjusted to Milwaukee’s aggressive defense or the Paul George-Roy HIbbert-era Indiana Pacers who eventually disintegrated even though they had the league’s “best” continuity.
As Zach Lowe suggested, perhaps defenses know how to counter Jokic’s passing now. You can see where he’s piling up turnovers and missed shows below. They don’t have a lot of shooting and talent at the guards, so it’s not too difficult to stymie them. Of course, the season is young, and their shooting percentages will probably rebound. But there’s a possibility they can’t recapture the magic of the second half of last season, whether due to regression or defenses catching onto their style or other factors. They had a better game Sunday, though it was only against the lowly Nets; but it’s positive to note their turnovers were down and Jokic had a good game. We’ll see how they respond in the next few weeks, as the real team will emerge.
The star-making of Kristaps Porzingis
One of the joys of the NBA is that you never know where a star will emerge. You’ll watch a favorite player and hope he develops more to the point where an All-Star berth would be truly earned and MVP consideration wouldn’t be bizarre. That’s happening now with the Unicorn, despite the uneven play of his teammates. His usage rate has gone through the roof, which is one of the best precursors to true stardom. I know at some point my Porzingis talk is an overload, but this guy is over 7-feet tall and he truly moves like a guard sometimes. Look at the video below — he drives with the nimbleness you’d expect from a wing, and with his height and shooting ability defending him is a joke. I have no idea if he’ll keep up this scoring binge through most of the season, but watching how he plays I don’t see why you would say this is out of his range. Welcome to the future.
Follies and Fultz
Yeah, so this is happening again: it’s the same old song and dance. A highly-valued 76ers prospect has an injury that was initially reported as minor and has become a bigger issue. Meanwhile, the communication and the PR have been deplorable. I’m really not sure what’s going on here — Markelle Fultz’s shot has completely changed and his shoulder is being reevaluated. His agent and his organization have changed their stories about what’s happening, and they’re not congruent with quotes from his coach and others either. Someone is lying, but the extent to which is unclear.
Philadelphia has a long history of providing inaccurate injury information to the public, even to its own players. Joel Embiid had no idea he was going to be out until the end of the 2017 season. He was told he was going to miss 3-4 weeks, while the public was told it was day-to-day. I’m quite concerned about the team intentionally lying to everyone about what’s happening with the injuries, sure, but the alternative is that they have no idea what’s going on either — and that’s scary. What’s happening with Fultz is quite similar; the last report is that he’s going to be out indefinitely despite the earlier reports indicating otherwise.
There are several scenarios for Fultz. I’ve heard an argument that this has all been fabricated to cover up what an “awful” player he is, but that makes no sense for several reasons. Firstly, we have do have reports of shoulder soreness and either a cortisone shot or a draining. His shot has definitely changed too, And his free throw shooting has completely tanked, which is odd because that’s the most stable stat in all of basketball. He’s not getting “exposed” against higher competition; the free throw line doesn’t move. The explanation is probably a number of factors from mental to physical to miscommunication from the 76ers. I just hope we get to see the fully realized product soon. As Philadelphia fans know, sometimes the situation calls for patience.
Small sample size warning
At this point in the season, I would call it unwise to make grand proclamations for teams. Think about it like this: what would we think of some team’s stats during a two-week stretch in January? Most people wouldn’t even know the team’s stats, besides basic wins and losses. But since these are the only two weeks we have of basketball, we’re obsessing over point differentials and shooting percentages and everything else. All things trend toward the mean, but we mistakenly react to these numbers as the “true” mean.
Let’s bring this back to a tangible example. The simplest and most common team rating system is Basketball-Reference’s SRS, which is just point differential adjusted for opponents. Due to the ubiquity of Basketball-Reference, a significant number of people are already taking these numbers and running with them. These are usually delivered with caveats, of course, but we really don’t know off-hand what these numbers actually mean and how stable they are.
For context, I grabbed the season stats roughly two weeks into the season using the internet wayback machine for the available seasons — just a handful. Then I compared them to the full season results. This was only for 2012, 2014, 2015, and 2016, but it’s enough to get a decent picture of what happens to SRS as the season progresses.
The correlation coefficient between SRS at two weeks and at the full season was 0.67, where the range is from -1 (perfect inverse relationship) to 0 (no relationship) to 1 (perfect relationship.) That suggests that SRS even at two weeks in is surprisingly accurate, but there’s still a lot of noise. The graph below illustrates this better. Basically, if you start off with a good SRS, it’s more likely than you end the season with a good one as well, but there’s still a wide range. Being great now is no proof you’ll even be above average — just ask the 2012 Philadelphia 76ers. They started the lockout season with a point differential of +12 and then hit a differential of +2.7 the rest of the way.
Obviously, because the first two weeks are contained within the full season’s SRS, there’s going to be some correlation anyway, but a correlation is 0.67 is strong — these teams haven’t even played ten games. This agrees with other research too. Kevin Pelton found that you only need about 20 games before the in-season results are more powerful than preseason expectations. Namely, if you’re reading this wondering when we should start caring about SRS and you want a specific number, that’s not a bad one to choose. Just keep in mind that other factors should be (intelligently) considered.
For another illustration of this, see the graph below, which shows the running average point differential of every team last season from the first game of the regular season to the last. I know it’s hard to see a few of those individual lines, and too many colors are similar[4.]. But it’s the pattern of the lines you should pay attention to: you can see a lot of chaos at the bottom there, which slowly transitions into something smoother by game 20. Mechanically, you could actually categorize this like fluid flow, where it goes from turbulent flow to the transitional phase to the laminar phase (laminar just means lines of movement are roughly parallel with each other with few crossovers.)
I was a little shocked by the stability of SRS at this point in the season. The stat stabilizes and corrects itself quickly. This is despite all the external changes that can occur too, like injuries and trades; the correlation is still that strong. I would still say be cautious with the stats, and keep in mind preseason expectations or other ways of identifying team strength. The 95 percent confidence interval after two weeks for SRS — this just means we’d expect 95% of the observations to stay in this range — is still quite sizable at +/- 6.8.
Next: Nylon -- In-game consistency in the young NBA season
This is a topic that could require further study and a wider data source. I could use decades of seasons for this by calculating my own SRS using game-logs, but I wanted to show the basic point here, and I learned something too. You can do a lot more with the first two weeks than I imagined, but don’t go overboard — we’re still in a transitional phase for most team stats, and even by the end of the season we’re still blind in many areas.