Nylon Calculus Week 20 in Review: Raptors, Boban, and NBA questions

I know much of this season is being marred by some ridiculous tanking, and the college basketball world is in chaos due to some scandals, but there’s still a lot of quality basketball to watch out there. We shouldn’t be spoiled. I grew up as a basketball fan during that awkward post-Jordan, pre-2005 rule changes era. We don’t have to suffer through glacially paced teams walking the ball up the court just to throw it into the low post so some large, low-skilled player can awkwardly sling the ball against the glass and miss the rim entirely. We don’t have a Finals series end with games in the low 70’s. This is a lot better. And with that, let’s take a look back at the last week in basketball.

Are the Toronto Raptors legitimate?

The league has been dominated lately by the Warriors and the Rockets, the two Western Conference super-powers who have been steamrolling over teams. They both have adjusted point differentials over plus-8 points — before this season, we’ve only seen ten of those teams since 1998. Yet there’s actually a third team over plus-8 points this season, and no one is seriously talking about them as a contender or as another elite team. Toronto has the stats of a top-tier team, and no one is treating them that way — let’s figure out why.

The first thing I’d note is that Toronto has a top-flight bench. That’s been the trend in recent seasons, and it’s especially true this year. It’s been their cache of young guys, like Fred VanVleet, who’s had a suspiciously great RPM because the Raptors have been sensational when he’s played, or Jakob Poeltl, who’s been killing the stat sheet and scoring with great efficiency. They haven’t even had time for Bebe Nogueira, and he’s had scalding rate stats, blocking 3.5 shots per 36 minutes and grabbing 2.1 steals for instance. With that kind of bench production, and especially deep bench production, you can run up the score in blowouts and garbage time.

Even so, the Raptors are still strong at the top; it’s not just the bench. I imagine people are pointing the credit at DeMar DeRozan, because the book on him this season is his improvement, but he’s actually improved only in marginal ways compared to his previous seasons. Pointedly, he’s shooting more 3-pointers, but the entire team is too, and that’s the most noticeable difference offensively from last year, in part caused by an emphasis on utilizing the shot more often. But the biggest improvement isn’t there.

The Raptors have been a lot better on the defensive end. As you can see in the bar chart below, their improvement has almost entirely come from better shot defense (i.e. they’re allowing a much lower effective field-goal percentage.) They’re forcing slightly more turnovers, but rebounding less often. And they’re actually allowing more free throws. When you break it down further, that shot defense improvement is almost entirely coming from a decrease in opponent 3-point attempts. That’s a good sign because that’s a more stable stat. So yes, Toronto has improved this season not just from an awesome bench, but from realizing the power of the 3-point shot and learning to control attempts from both sides of the court. It remains to be seen how this translates to the playoffs — this is where people are most skeptical of them — but they’ve balanced their finely-tuned, deep offense with some truly great defense, and that’s traditionally been the make of a contender.

Defensive matchup data

The NBA’s stats site recently released defender matchup data, which shows who’s been covering whom and what the results were. This is different than the shot tracking defense stats because it’s about which defender is nearest the player in question for the majority of the possession, not just the shot attempt. Actually, this is something I did when SportVU tracking data was new back in the 2015 playoffs when I was looking at who was defending key players like Stephen Curry. I used partial possessions, meaning you got some credit for only “defending” him for a quarter of the entire possession.

You can see some analysis using this data here. You can see the versatility of certain players there (i.e. how they defend players of varying heights.) Players generally defend others around their same height, but there are notable exceptions and some guys have wider ranges than others. Draymond Green is the quintessential example there. We’ll hopefully see a lot more useful analysis like this in the future. With this kind of data, it’s best in the aggregate; let’s hope people won’t misuse it by declaring player X the best at defending James Harden because of ten possessions. (Okay, that will definitely happen a lot.)

Possession counts

In baseball, pitch counts are an integral part of the game. We’re expected to see pitchers, especially young ones, limited by some predetermined pitch count number. Once a player hits a certain load, his chance of injury exponentially increases, the thinking goes. When the body is tired and strained, there are more and more ways issues can occur.

We see this same type of thinking in basketball; it’s just not as well defined. We’re seeing more players skip one of the back-to-back nights, and it’s a rare to see someone average 36 minutes per game, much less 40. Perhaps we’ll see something more regimented in the NBA soon and more accurate than minutes — a possession count limit for players or something that uses tracking, like distance traveled or even a system that identifies when a player is fatigued, based on how they respond and move on the court. I imagine the best teams are already working on that, and it won’t be long before it’s all part of basketball’s lexicon.

National Boban day

In case you missed it, on Tuesday night Boban Marjanovic devoured planets, and the NBA-sphere on Twitter was alight with joy and discussion about how good he could be in a bigger role. He was a crucial player in the Clippers’ win that night on national television against the Nuggets. He only played 15 minutes, but he scored 18 points on excellent efficiency with six boards, including five on offense, somehow. Per minute, this guy is an international terror, and it’s been that way for every season he’s been in the league. People can’t push him away from the basket, and with his 7-foot-3 frame he’ll throw in easy shots or grab the board.

Why isn’t he playing more? There are legitimate concerns about his defense, particularly in pick-and-rolls, as explained by Ben Falk. It’s easy to exploit him, and ideally you have to pick-and-choose the right matchups for him. His per minute stats, by the way, should scale for the most part. Generally, players do a little bit better when they get to play longer stretches on the floor — think about having to play for three minute spurts every other game. It’d be hard to get a rhythm. That’s true until a player hits some fatigue limit, and with Boban that’d probably be under 30 minutes a game but it’s definitely more than his nine minute average for his career. His defensive liabilities would offset a lot of his value, but not everything — his plus-minus stats have generally been good and his production is such that his defensive liabilities would have to be very large.  At the very least, we wouldn’t know until it’s tested, and he deserves at least a chance.

Ricky Rubio

At 27 years-old, Ricky Rubio should be in the midst of his prime and at least near his peak value, but his time in Utah has been disappointing. He’s had the worst season of his career outside of the injury-marred 2015 when he played 22 games. His assist and offensive impact numbers are down, and his defense isn’t what it used to be when he was secretly one of the best defensive point guards in the game. But he’s shot well enough over the past couple months to recoup some of that loss, and he’s actually at a career high in usage. Perhaps the work with the shooting coach has helped or he’s mirroring Jason Kidd and age his turned him into a better shooter. Whatever the case, the Jazz, and the world, need Ricky Rubio at his best, if for nothing else then his unique plays.

Tyus Jones

If you look at the RPM leaderboards for point guards, you’ll notice a few headlining stars and underrated players, but there’s one guy who really sticks out: Tyus Jones is ranked fourth, ahead of Russell Westbrook. He’s the second-highest ranked on defense, and he’s well above average on offense too. A lot of that is a mirage — Kevin Pelton correctly pointed out opponents shoot unsustainably lower rates from the 3-point line when he’s on the court, and there are suspicious patterns with rebounds too — but statistical draft models loved him in college and he’s made great progress this year. Just how good is this backup?

In many ways, Tyus Jones is what I’d call the prototypical backup point guard: he has a low usage rate with a modest amount of assists. He’s fairly small, and he’s a solid, feisty defender. Scouts were worried about his lack of athleticism and size translating to the NBA, but he’s been fine so far — his high steal rate helps a lot. He has fabulous anticipation, as high dribbles and sub-standard passes are not safe anywhere near him and he’s arguably the premier artist of stopping transition attempts with a steal. He’s a good help defender and bigger players don’t steamroll him. However, he’s far from perfect, and you can see him get lost after a screen here and gamble on Ricky Rubio on this play, ceding a layup.

There were real concerns about his shooting proficiency when he was in college even though 37 percent from the 3-point line isn’t that bad — and percentages in those small one-year samples shouldn’t be trusted either, but he hit a scalding 89 percent of his free throws. So far in the NBA, that’s been a mixed bag. He’s an okay shooter from deep so it’s not a liability, but he may not have true NBA range. His passing is usually low risk; he’s a drive-and-kicker who’ll serve up a lot of jump shot assists. Occasionally he’ll have a nifty pass, but they’re usually pretty basic. Overall, while his role has shrunk with all the scorers and ball=handlers they have now,  and his plus-minus numbers have been overrating him, he’s a decent backup point guard who plays tougher defense than you’d think and he was a steal at 29th in the 2015 draft.

For a thorough answer, I’d need an entire, long article with days of research. The entire concept of “clutch” is tricky to deal with, and there’s a lot of data to sift through. But I do have some shorter answers ready to answer this question decently. First of all, there’s the Wyatt Earp effect, where if you have a large number of observations, chances are a few ones will be at the extremes. In other words, by natural random variation some players will have great looking clutch stats and some will have poor looking stats, even if it’s just random.

However, I think there are a handful of players who do perform better in clutch situations. Kyrie Irving has often had pretty crazy statistics in the clutch, and Damian Lillard, his doppelganger, has as well. Carmelo Anthony actually performed really well for years too (and now he’s the forgotten fiddle on a team with Russell Westbrook.) My hypothesis is that the players who most consistently play well in the clutch have games tailor-made for those situation. At the end of close games, the pace slows down, isolations become more common, and stars are fed the ball. That plays into guys like Kyrie Irving.

Of course, there’s a lot of nuance here and room for some wrinkles.  If you go by individual stats, then it’s true guys like Carmelo Anthony look better. But what truly matters is how the team does when you play, and that’s where LeBron James shines, as he’s been part of some of the greatest clutch teams in recorded history.

I’m not sure which players play poorly in the clutch, relatively speaking, but this is a topic that deserves more attention, and it’s one I’ve actually wanted to address for a while.

This is indeed true, and it’s actually the first time since LeBron James’ rookie season where the team has been outscored when he’s on the court — and the first time his team has ever been better with him off it. You can see the trend in the numbers in the graph below. What’s happening now is certainly strange. LeBron’s teams have historically collapsed without him.

Tellingly, Cleveland’s offense has been appreciably better with LeBron on the court this season. It’s the defense that’s been an issue. Per Basketball-Reference, when LeBron’s been on the court, the team has defended worse than the Phoenix Suns, ranked 30th in the league, and with him off they’ve had the same rating as the Utah Jazz, the fourth-ranked defense. Opponents have been rebounding slightly worse with LeBron, and turning the ball over slightly better — that’s all a wash. It just comes down to shooting percentages.

Since shooting percentages are the most volatile stats, you can take this to be mostly noise and hence random. He played a lot with Isaiah Thomas, sure, and there were some awkward lineups with Derrick Rose, but shooting percentages shouldn’t dive like that, and it’s not like you can chalk it up to LeBron’s decline on defense — that’d be like going from having the impact of Hakeem Olajuwon to David Lee.

Next: Nylon Calculus -- Who is the best LeBron-stopper in the NBA?

Naturally, this is a case of opponent 3-point percentage difference: opponents shoot 39% percent with him, and 32 percent  without him. That’s virtually all noise; defenses can’t control it. As this is noise,  I expect the numbers to regress, and we will most likely see LeBron’s plus-minus stats rise back into positive territory by the end of the season. Those raw plus-minus numbers can be fun to dissect, but they’re unadjusted and need a lot of time to stabilize. LeBron hasn’t gotten worse; it’s the stats that are bad.