Nylon Calculus Week 23 in Review: Sophomores and awards

There are only a few games left in the season, but a lot is left to be resolved. Golden State has regained a lot of their mojo, but there are still questions about Kevin Durant’s health and how much of their magic from last season they can bring into the playoffs. The Spurs are great, as usual, but how many of their core players play defense — can they beat one of the many elite offenses out there with Pau Gasol in the middle? Likewise, Houston has been breathtaking on offense, but defensively they could breakdown quickly. Then there’s Cleveland, whose team has been taking on water for weeks. But that’s a more complicated story. And with that, let’s take a look back at the last week in basketball.

Cuyahoga on fire: Cleveland is failing

In my season preview of the Cavaliers, I noted my suspicion of projections that had the Cavaliers winning 60 games. LeBron James’ monstrously high RPM was a result of their title, and I did not feel that we’d see the playoff version of LeBron very often in the regular season. As good as LeBron has been, it appears my suspicions were correct because they’ve been a (relatively) weak team besides him. One could blame injuries, but a number of their core players have had a long history with getting injured; this isn’t out of the norm. Their defensive problems aren’t coming out of left field either, because they weren’t great last season and lost two of their best defensive players in Matthew Dellavedova and Timofey Mozgov.

Sometimes with a poor defense you can point to a number of extenuating circumstances and argue that in a different world, they’d be better. But the Cavaliers are terrible at the most controllable actions: they’re second-from-last in creating turnovers, they have the third-worst ratio of opponent free throws to field goals, and they’re one of the worst defensive rebounding teams too. Watching them, you’ll note a shocking number of problems for a title-contender.

Ultimately, I don’t think what we’ve seen from the Cavaliers is an anomaly. They’ve never been a great defensive team in this LeBron era, except for a few short stretches. They were actually 0.7 points per 100 possessions below average in defensive efficiency in 2015, and they were only 1.9 points better than average the next year when they won the title. They’re at 1.6 points below average now. That’s not out of line with what we’ve seen from them in the past or their personnel — LeBron’s their best weapon but he’s not putting in anywhere close to 100 percent effort on that end of the court for the regular season. If you’re mediocre on defense, you have to be spectacular on offense to make up for the deficit; and they haven’t been. If LeBron doesn’t turn the dial again for the playoffs, we shouldn’t assume they’re going to walk into the Finals. And we really shouldn’t assume a team can flip a switch and play better in the playoffs, as Kevin Pelton noted. They have real issues.

The last Rose

A couple days ago, Derrick Rose suffered yet another knee injury, and it’ll complicate his and his own team’s plans this summer during his free agency. But I think he and Rajon Rondo, the other prematurely washed-up ex-all-star, should both consider a better basketball environment for who they are. Both players need the ball in their hands to be effective, yet they’re not effective enough anymore to have the ball for most NBA teams. What can they do? Try China, where they can be paid handsomely by a basketball loving culture and play real minutes more competitively. I imagine they’d prefer the states, but an NBA team would be making a stupid decision wanting to start either guy and they’ll be frustrated if they try to fight for minutes here. It’ll be better for everyone involved if they change hemispheres, as harsh as that sounds.

Russell Westbrook’s historic production

I updated Dredge recently, and while I wasn’t surprised Westbrook led all players in both HBox and MVP Index (which is a factor of minutes and HBox), I was a bit taken aback by his production level. Currently, he has the highest HBox on record, which goes back to 1974, and his MVP Index is a little over 90. It’s a scale calibrated so that the best mark all-time is 100, and with a good performance over his last remaining games there’s a chance he busts that number too. Previously, the best HBox was done by LeBron James in his historic 2009 season with 9.7.

Other metrics are identifying Westbrook’s as historic too. There’s the old standard PER — he’s not “best ever” there, but he’s north of 30, which is rare. Basketball-Reference’s BPM, however, sees his production as unparalleled: the best ever BPM mark was 13.0 from LeBron James in that 2009 season, while Westbrook’s at 15.3 right now. BPM, like HBox, uses interaction variables, and the ones BPM uses are what makes Russell unique. He has the best combination of USG%, TRB%, and AST% ever, and BPM relies on the stats USG%*AST% and AST%*TRB% pretty heavy with its regression. HBox, however, uses several more interaction variables so that an outlier like Westbrook doesn’t completely stand out (I use BLK*AST as well as a couple other combinations with blocks, and that probably helps here because it’s one statistic that Westbrook isn’t dominating). He’s still so dominant in multiple categories that he has the highest HBox ever, but if you’re wondering what the main differences are between the metrics, that’s one important distinction. And not all destinations have one road: if not for Russell, James Harden would have the highest HBox ever, thanks to his monstrous efficiency and high rate of assists on a great team.

The problem with building any metric is that it can only be tested on data it has seen. That may sound uninformative, but if the data’s never seen someone with stats like Russell Westbrook, then how can it evaluate him? They can only use abstractions and projections based on players with different stats, so it doesn’t completely connect. It’s tough to gauge the value of someone who pads his rebounding totals so that he looks like a rebounding power forward on paper, yet can jump-start fast-breaks on his own after the rebounds too. That’s happened with similar players, but never to this extent, and it’s stressing some of the metrics we all use.

For an example of how BPM responds to Westbrook’s incredible stats, I calculated his BPM for 2016 and then found the difference between that and a version where one of his stats from 2017 was inserted into the formula. For instance, if his DRB% in 2016 had been 28.6%, as it is in 2017, then his (raw) BPM would increase by 0.5. That’s actually not a lot, so let’s turn to another two more likely culprits. If you do the same with AST%, his BPM increases by 1.4 points, while for USG% it increases by 1.5. And if you include team TS% — it’s a factor in BPM, meaning your own TS% looks better if your team is worse — then it increases by another point. In fact, AST% is linked to USG% too, because it’s a proportion of teammate field goals you’ve assisted, so if you take more shots your AST% goes up too. Due to those variables, Westbrook’s BPM has exploded, making him look like a god among mortals.

Old Faithful

Last week, the Grizzlies clinched a playoff spot by beating Mavericks, who have been mathematically eliminated. This will somehow be the seventh season in a row that Memphis makes the playoffs, and while Dallas missed just back in 2013, it’ll be their first losing season since 2000. That’s impressive, but of course the model of consistency is the Spurs, who haven’t missed the playoffs or slipped under 0.500 since 1997. That’s 20 years, and somehow they don’t yet have the record for most consecutive playoffs appearances; Portland had 21 in a row from 1983 to 2003, while Syracuse-turned-Philadelphia had 22 in a row from 1950 to 1971.

The Spurs have the most consecutive winning seasons record, but it’s based on how you define it — if you discount the Utah Jazz’s 41-41 season in 1985, then the Spurs now own the record; otherwise it’s the Jazz with 21 seasons. Either way, with how they’re managed and how well Kawhi Leonard is playing, both longevity team records will be theirs soon enough.

Please consider LeBron for MVP

I know the MVP race is crowded, and I know James Harden and Russell Westbrook have more interesting cases — and Kawhi Leonard’s team is superb, once again. But LeBron James absolutely deserves consideration for Most Valuable Player. Yes, he leads the NBA in RPM wins, but even if you hate that metric, isn’t it easy to imagine the Cavaliers absolutely failing without him? This guy is a national treasure, and let’s not treat him like Shaquille O’Neal or Michael Jordan and take his greatness for granted. Another MVP would not be out of the question for his brilliance.

The Rookie of the Year argument for Joel Embiid

There’s a philosophical disagreement nested within the arguments for Rookie of the Year: how much does playing time matter when the quality level is so much different? There are a few rookies out there who could win the award simply for playing competently for a large chunk of the season (namely, Dario Saric and Malcolm Brogdon.) But Joel Embiid is the rookie everyone will remember, and he was by far the best rookie, even if for only 786 minutes. When people argue that point, they’re implying the scale we should use is a nonlinear one, where production at an elite level should be valued much higher per minute than one at an average level or below. That’s what I had in mind when I created the MVP Index. Using that, Joel is the leader by a good margin using HBox: 0.17 where the next closest is Dario Saric at 0.04. Keep in mind that HBox is regressed based on minutes played, so Joel is actually being underrated here.

If you use the old Kevin Pelton trick of summing every rookie’s PPG, RPG, and APG numbers, and discounting Embiid because he doesn’t have enough games to qualify, then it appears Dario Saric will win. Recency bias will help him too, as he’s played much better since the All-Star Break. This is a weird case, but there is totally solid backing for supporting Embiid here.

Portland’s restoration

Not too long ago, the Trail Blazers were on the outside looking in, fighting for the playoffs and trying to leapfrog the Denver Nuggets. They’ve now comfortably taken the No 8 seed after a lackluster first two-thirds of the season. The first explanation people latch onto for this is the trade of Jusuf Nurkic, which even I’ve examined as a catalyst for their great recent play. But Mason Plumlee, whom they lost in the trade, was a good player too, and they’ve had no major additions either. The truth here is tougher to see, but I think there are some reasonable explanations.

First of all, one cause here is a boring one, and it’s regression to the mean. Pre-season, Portland actually had the lowest variation among all the win predictions: all from 40 wins to the mid-40’s. This was true no matter the source either, from ESPN’s summer forecast with 46 to Vegas (represented by oddsmaker) with 44.5. The team made few changes, with the exception of small forward, adding Evan Turner and losing Gerald Henderson. It was surmised, quite reasonably, that they wouldn’t stray far from their previous season’s mark of 44 wins. Yet by the end of February, they were 24-35. But what we knew about the Blazers suggested a team stronger than that, and we should have expected a regression toward the preseason expectation line anyway.

Of course, we do have to consider the changes to the team in the past few weeks. I would have awkwardly argued that Evan Turner’s injury was a blessing in disguise, but he got injured a bit before their resurgence and has come back, although his minutes are lower. Nurkic fever is the most popular explanation, and it has been documented thoroughly. His own recent injury will be a test for that theory, unfortunately. He’ll be reevaluated in two weeks, and in that time we might get a satisfying answer, or at least some evidence, about why Portland has been so good lately. And if Nurkic wants to play in the playoffs for his new team, he’ll have to pray that it’s simply regression to the mean and they’ll keep performing this way, rather than his own influence.

Sophomore of the year

We’re in the midst of heavy award season talk, and while there’s not much more room to discuss a new award, I thought I’d introduce this one as an alternative means to evaluating young players: the Sophomore of the Year Award. We can celebrate how rookies perform, but too often the best prospects have barely played or they’re still developing. This is more of an issue now than in the past because the best talent has usually only had one year of college and, consequently, is quite young. Thus, I thought the sophomore designation would offer a better balance between youth and real peak talent.

To aid my sophomore of the year quest, I’ll be using my metric HBox, which goes back to the 1974 season. It’s not a perfect metric — nothing is — but it’s better than looking at the points per game leaders in a season and I admittedly have not watched those 1970’s game live. You can see the top players each season by HBox’s MVP Index. Things aren’t perfect — hi, Slick Watts — but it nails most of the picks pretty well.

Table: Top Sophomores by HBox

For these imaginary awards, I’ll go with the above table for many of the picks, but I’ll explain the ones I changed by hand. First of all, Slick Watts probably isn’t a bad decision — he was in the lead by a decent margin, and he was a pretty good defender, suggesting the stats maybe didn’t even capture all of his value. The next tough case is 1978, where Lonnie Shelton had a narrow lead over Adrian Dantley. However, I again decided to stick with HBox’s decision, as the more research I did on Shelton, the more I saw him as an “intangibles” type player who was undervalued in his day, while Dantley’s raw stats overstated his value. I’ll still with Cedric Maxwell too for 1979; Jack Sikma was the closest.

Moving onto the 80’s, I’ll stick with Michael Ray Richardson for 1980. While Mo Cheeks and a couple candidates are more famous now, Richardson was a force back then, and he was the only leading sophomore to make the All-Star Game. I’ll finally change something though, and it’s for 1982: Rick “McNasty” Mahorn over Johnny Moore. By some measures, it was his best statistical season. For 1984, I think stats can overrate LaSalle Thompson, so I’ll go with James Worthy, who would have surely won the award back then anyway thanks to his number one pick status and his team being the Lakers. Charles Barkley had an outstanding HBox score in 1986, but I think he’s overrated by those stats and Hakeem Olajuwon, who made the All-Star Game over him, was the better overall player.

For the 90’s, I wish I could change 1992 so that Gary Payton takes the award, but he had two lackluster seasons to start his career before the Glove as we know him developed. I’ll dispense with my usual Allen Iverson hate and flip Antoine Walker for him in 1998; the two were neck-and-neck anyway, and Walker’s versatile-looking stats for a big man cause him to be overrated. Otherwise I’m leaving the decade unchanged. Some of these decisions aren’t tough, like in 1991 when David Robinson had the stats of an MVP, much less a sophomore.

In the next decade, 2000 is a tough one: Paul Pierce, Dirk Nowitzki, and Vince Carter all have great cases, and obviously they’re all gone all to prove that their sophomore seasons weren’t flukes. But I think I should go with Vince, who at the time was definitely considered the best player — he’s the only one picked to an All-NBA team, and to the All-Star team — and had the stats comparable to any sophomore. Pierce and Dirk were probably more loved by HBox due to their higher 3-point rates, and their efficiency too, but Vince drew a lot of defensive attention back then. Other than that, I’m fine with every other decision for the decade. Some of those earlier seasons, like 2001, don’t feature the best player in the draft class, but it’s about who’s the best for that particular season.

For 2013, Ricky Rubio, Chandler Parsons, and Kawhi Leonard were all evenly rated, so I’ll go with Kawhi because his defense isn’t well covered by HBox and the Spurs limit minutes severely. Similarly, I’ll go with Anthony Davis for 2014; his defense had some issues back then, sure, but the same can be said of Damian Lillard to a higher degree. Finally, for 2016, Nikola Mirotic is a class of player who can be overrated by HBox because he has a high 3-point rate for a big man, and I think I’ll go with a subtler player, Rodney Hood, who has so far peaked in his sophomore season.

Table: Sophomore of the Year

The Sophomore of the Year award isn’t an idealized version of Rookie of the Year like I had hoped. Sometimes the best players have a sophomore slump, or they’re just injured like Michael Jordan when he broke his foot. But in this era where the best prospects are rarely upperclassmen, this should at least serve as a reminder that our views of rookies and what they accomplish should be modified. The “Rookie of the Year” tag can follow a player forever, and it makes a guy like Michael-Carter Williams more enticing, but there obviously real problems with the award.

Next: Nylon Calculus -- The most erratic free throw shooters of 2016-17

A number one draft pick might play in the league as a teenager, screwing with our notions of what rookie seasons mean, and while looking at what players do in their second year isn’t ideal either, it’s another way to think about the league. After all, it’s unfair that Nikola Jokic’s team didn’t realize how special he was until now, and that he wasn’t given an opportunity to play big minutes as a rookie. But he’s still a great player — regardless of whether or not he has any awards to back up that claim.