Nylon Calculus Week 9 in Review: PER 2.0 updates

INDIANAPOLIS, IN - NOVEMBER 24: Victor Oladipo /

As much of the western world readies itself for the holidays and end-of-year-festivities, the NBA is going deep into its season. We’re seeing teams hit 25 wins, and the All-Star game is right around the corner. Before you know it the Warriors will be collecting the title and we’ll be ready to reward the Kings with Luka Doncic (or perhaps it’ll be the Rockets and Deandre Ayton.) But before that happens. let’s take a look back at the last week in the NBA.

Victor Oladipo’s surprising rise

After the now infamous Paul George-Victor Oladipo/Domatas Sabonis trade, I was among the most ardent critics. I saw Oladipo as an expensive scoring point guard who wasn’t particularly effective scoring or creating, and had never capitalized on his defensive potential. Sabonis was a mid-first round pick who didn’t have a good rookie season, and George as an All-Star in his prime (who could leave after a year, yes) — it seemed clear-cut to me. But both guys in Indiana have played exceptional basketball, and Oklahoma City has floundered. Did we miss something? Were the Thunder holding him back?

One of the popular theories now about Oladipo is that Westbrook was holding back his development — and you can understand how that theory would arise due to Westbrook’s historic usage patterns. But Westbrook has only been a part of a small portion of Oladipo’s career, and much of the projection of his 2018 season stemmed from all those other minutes. In fact, we can just look at how he did last season when Westbrook wasn’t on the court, which was a large chunk of time: 562 minutes. According to the irreplaceable site NBAWOWY, Oladipo’s usage rate did indeed balloon to something north of 30 percent,, which is what’s happening now, but he was decidedly inefficient. He was near league average in terms of efficiency when his MVP teammate was on the floor, but his usage rate was actually below average — that happens when your backcourt mate takes 41 percent of all possessions, and it’s a natural trade-off. What happened in Indiana wasn’t natural.

Table: Victor Oladipo (via NBAWOWY)

With Westbrook169054.419.4-0.0194.8
Without Westbrook56250.830.20.3095.5

It’s hard to blame Westbrook for stifling Victor, as it doesn’t explain why he played so much worse without the guy on the court too. His Indiana stats are just in another stratosphere, and it’s virtually only because of his shooting percentages. All his other stats are near his career norms — sure he’s blocking more and netting more steals, but his turnovers are up a bit too. Since shooting stats are the most volatile of all stats, you can see why I’m a bit skeptical here. Plus, the failures of the Thunder right now have little to do with Paul George; it’s the other guys, like Westbrook’s decline, Patrick Patterson not being able to hit his shots, and Carmelo Anthony being on the court.

There’s precedent for Oladipo going off the map and then regressing again. A few players, like Goran Dragic, had magical seasons and then fell back to earth, or the others who had fantastic first halves of the season (and slipped in the second half), often leading to a surprise All-Star appearance, like Jrue Holiday in 2013. Oladipo’s also at that transitional age of 25-years-old, where he’s on the edge of being old enough where we should still expect significant improvement. I’m afraid, as Tom Ziller suggested, this is a mirage. We’ll see what happens eventually, but for now I would heed caution — this sunny story could slip back into mediocrity.

Freaky free throw boards

My how things have changed — when Russell Westbrook shocked the league with a triple-double average, it was partly due to his prodigious free throw rebounding rate. It was bizarre to see a point guard lead the league in defensive free throw boards with 98. But this year he’s regressed heavily, as several people are ahead of him. Giannis Antetokounmpo stole his free throw board superpowers, and the lanky forward has been a transition terror all season because of his ability to rebound and push the ball. By the way, Russell has a plausible chance of missing the All-Star game while averaging a triple-double again — what an odd season.

The state of the Timberwolves

It wasn’t long ago when one would be criticized at the mere suggestion that a Tom Thibodeau-led team could have a poor defense. But now we’re going on two seasons in a row where the Wolves are one of the worst defenses in the league at roughly 3.0 points per 100 possessions under the league average. And this is even with the import of two of his favorite defenders, Jimmy Butler and Taj Gibson, and the export of one of their biggest liabilities in Zach LaVine. Now if you bring up Thibodeau, you get responses about how he’s already an outdated thinker and how the current NBA has passed him on by.

Digging into their stats, and they’re failing on the two major components of a disciplined defense: they’re allowing high shooting percentages and collecting a low number of defensive boards. No surprise, but their lack of a rim protector is apparent, as they allow both a high number of shots at the rim and a high percentage. That has killed them. Unfortunately, while they didn’t inherit Thibodeau’s great defense from Chicago, they did inherit his propensity to play his starters huge minutes. Through last week, four of the 20 players who had hit 1000 minutes were on Minnesota, while Jeff Teague isn’t far behind, having missed a few games due to an injury. I can understand playing Jimmy Butler huge minutes, but Taj Gibson? And their bench hasn’t been pathetic: can we get Nemanja Bjelica, who’s been destroying teams from behind the arc with a 72.5 true shooting percentage, more than 17 minutes per game?

On the flip side, the Wolves are winning games because of their offense. Their scoring is pretty balanced between their three cogs — Jimmy Butler, who’s lessened his load this season; Karl-Anthony Towns, one of the best offensive centers in the league; and Andrew Wiggins, who’s quietly had a terrible season but is impossible to discuss rationally with fans — but their merits are indirect. They aren’t shooting well, but the team gets to the line at a super high rate and they keep their turnovers down while rebounding their own misses. It’s actually quite a strange team. The team still has potential though. If they can have a winning record (and a positive point differential) with horrid rim defense and poor shooting from Andrew Wiggins, then there’s a lot of room for improvement. We just can’t assume Thibs will have the magical touch anymore. That era is already in the rear-view mirror.

Bring in the old boss

How perfect is it that a Mike D’Antoni team looks like the best competition for the Golden State Warriors? The Warriors bludgeoned the league by downsizing and taking the seven-seconds-or-less dictum to its extremes. They set a few 3-point records, and played large stretches of the game without a traditional center. Who could upend that dynasty, besides the guy who popularized, or arguably invented, the style? Through last week, the Rockets have won 13 straight games, they tacked on the 14th last night, and hopefully this will all culminate in an epic Conference Finals. What they’ve been doing to teams, like the Bucks below, isn’t fair.

All-Star selections: West and East disparity

Tom Ziller broke down the voting gymnastics in the Western Conference surrounding Russell Westbrook and the possibility he could not be selected. What’s most troubling is not how plausible this is; it’s how loaded the west is, once again. This is even with Kawhi Leonard and Chris Paul missing large sections of the season, or past borderline candidates who deserved a consideration like Rudy Gobert or Mike Conley going down with injuries. (Going by my HBox metric, it’s a pretty even split when you try to take 12 for each conference, but that could change with injuries and it’s still top heavy when you look at the eight to ten best guys.)

The Western/Eastern conference disparity is glaring during these All-Star games, and the league has done nothing to address it, even just for the game itself. They solved a problem that didn’t exist with the co-captains tweak this year. This is one we need to address because we still value All-Star selections highly, despite all its flaws, and it’s not fair to guys with worthy seasons who are excluded because they’re in the wrong conference. We’re going to have a major “snub” in the West this year, at the very least, and one guy will look out of place in the East — and that’s what we can expect every season until this is fixed or the disparity, which has lingered for years, disappears.

Instant classic: The 76ers-Thunder triple-overtime

In case you were gone last weekend, you may have missed what’s arguably the game of the season so far: a triple-overtime game between the 76ers and the Thunder. The game had everything — Russell Westbrook had a triple-double, Ben Simmons was channeling his inner Magic Johnson with hook shots, Joel Embiid played an astounding 49 minutes and survived, Andre Roberson missed a layup, and the Thunder somehow won another close game. Ultimately though, the winner was probably Joel Embiid, who had a huge block on Westbrook, egged on the crowd like a pro wrestler calling for support after a tussle with Carmelo Anthony, and had the best parting shots after the game.

Rejoice: Milos Teodosic is back in our lives

After a nightmare of a season beset by injuries, the Clippers at least have Milos Teodosic back. It’s one thing losing Chris Paul, a franchise guy who’s not only the cornerstone but most of the foundation itself; but they lost his replacement Patrick Beverley for the season too. You don’t want to be without the replacement’s backup too — the Clippers were in a free-fall for a reason.

Milos is a great one-handed passer, tossing off quick assists like this one. You aren’t supposed to jump and then pass, but he pulls it off well here, delaying for a while during his hang-time. He’s naturally a maestro in the pick-and-roll. You can see him dart around a pick here, going to the exact right place to draw the defender enough to create a wide open dunk for Montrezl Harrell. He’s good at threading the needle too. Once he gets more playing time with DeAndre Jordan and they learn how to play off each other, I can imagine those two will cause a lot of damage.

PER 2.0

About a year ago, I developed a tweak of the popular metric PER by introducing new, relevant statistics, like field goals you were assisted on, and modifying the base formula a little so high volume, inefficient scoring wasn’t so valuable. I’m not entirely satisfied with the results — there’s only so much you can do with recordable actions from the box-score and play-by-play. But it was a fun experiment, and the new information isn’t useless; it’s a natural extension of how PER was built, debiting and crediting all the possible actions on the court.

You can view the 2018 full results here as well as PER 2.0 for every season going back to 1997. And, of course, I have a table below for the current leaders, besides I know that’s the most pressing concern for people now. Yes, LeBron James and Giannis Antetokounmpo are neck-and-neck, just as they are with the real PER — both guys are prodigious stat-stuffers who create well for themselves and others. For reference, a season with a PER 2.0 over 40 is roughly a once-in-a-season phenomenon, and All-Stars are usually over 25.

Table: PER 2.0 leaders, 2018 (min. 500 MP)

LeBron James1081CLE31.538.7
Giannis Antetokounmp982MIL31.438.5
Anthony Davis883NOP28.035.2
James Harden962HOU30.934.5
Nikola Jokic640DEN24.832.2
Clint Capela679HOU26.832.1
Stephen Curry750GSW27.731.2
Andre Drummond959DET22.730.6
Kevin Love823CLE24.628.4
Karl-Anthony Towns1026MIN23.027.2
DeMarcus Cousins1064NOP23.227.1
Enes Kanter647NYK23.327.0
Steven Adams814OKC20.926.6
DeAndre Jordan883LAC19.426.2
DeMar DeRozan927TOR23.125.8
Kyrie Irving933BOS25.625.8
Joel Embiid723PHI24.125.3
Jimmy Butler1008MIN20.724.9
Kevin Durant827GSW25.124.9
LaMarcus Aldridge934SAS24.624.8
Kristaps Porzingis764NYK24.124.6
Damian Lillard1039POR23.824.4
Rudy Gobert539UTA19.124.3
Kyle Lowry893TOR19.523.8
Alex Len546PHO19.323.3

Funnily enough, the guys helped the most here are big men like Nikola Jokic. I actually wanted to make PER 2.0 to penalize big men like Clint Capela who were efficient scorers but who didn’t create their own stats — whoops. My tweak in how efficiency is handled probably affected guys like him. Plus, in Capela’s specific case, he has more high value blocks and rebounds than others, like a fewer number of free throw defensive boards. I’m not quite too sure what to make of Alex Len — but hey he does well in PER too. Again, this has to do with efficiency. In statistical plus-minus models, efficiency is usually regressed heavily because it’s so noisy.

For some more random fun, I got chase-down block leaders for 2018 in the table below. (Definition: block within five seconds of a possession change and within five feet of the rim.) As scary as it is, DeMarcus Cousins, all 270 pounds of him, is the player most often running other people down and swatting away their attempts. I’m sure many of those possessions end with the offensive player splayed out on the ground in pain. John Wall’s second, and I wouldn’t be surprised if he were first at the season’s end — he’s like a track athlete in a point guard’s body. The “record” for a single season, by the way, is 33 from Josh Smith in 2006.

Table: chase-down block leaders, 2018

DeMarcus Cousins1064NOP9
John Wall608WAS8
Kevin Durant827GSW6
Will Barton881DEN6
LeBron James1081CLE6
Marquese Chriss604PHO6
Serge Ibaka694TOR6
Danny Green753SAS6
Bismack Biyombo458ORL6
Jarell Martin445MEM6

Here’s another miscellaneous stat: who has the most layups/dunks immediately after a defensive board they grabbed? (Definition: defensive rebound to an unassisted basket within five seconds and within three feet of the rim.) Yeah, it’s really not his season: Russell Westbrook isn’t the leader here for the first time since 2014 (you can see why I named the stat after him.) Giannis is instead, and LeBron James, who has the highest on record with 21 in 2006, is second. But look out for David Nwaba, a second-year shooting guard who’s been rebounding like a frontcourt player, who’s on the leaderboard despite only playing 383 minutes.

Table: Westbrook leaders (DRB to fast break inside shot), 2018

Giannis Antetokounmpo982MIL10
LeBron James1081CLE8
Jonathon Simmons852ORL6
Russell Westbrook1023OKC5
Brandon Ingram928LAL5
David Nwaba383CHI5
Kevin Durant827GSW4
Eric Bledsoe620TOT4
Tyreke Evans835MEM4
Rondae Hollis-Jefferson693BRK4

And finally, let’s revisit the “Russell,” which is just a block recovered by the defense. It’s no surprise that the leaders here are the block leaders too, but it’s still good to keep tabs on which players are good at actually controlling their rejections. Blocks are a lot less useful when the offense gets the ball again and then hits their next shot. So kudos to Andre Drummond, once known for “empty” big man stats, piling up a large ratio of Russells to blocked shots.

Table: Russell block leaders, 2018

Anthony Davis883NOP5033
Myles Turner666IND5333
Marc Gasol1015MEM4230
Kevin Durant827GSW5128
Giannis Antetokounmpo982MIL4128
Clint Capela679HOU4928
Kristaps Porzingis764NYK5028
DeMarcus Cousins1064NOP4526
Karl-Anthony Towns1026MIN4226
John Henson610MIL4225
Joel Embiid723PHI4425
Rudy Gobert539UTA4125
Andre Drummond959DET3624
Brook Lopez617LAL4324

I don’t purport that any of these stats are perfect, or even that they should be heavily relied upon. PER 2.0 is more of an experiment. In theory, it should be more useful than plain ol’ PER; it does have more information. The Russell blocks are fine — it’s a simple enough definition that I can precisely pull them from the play-by-play. But for Westbrook’s and chase-down’s I needed to make some arbitrary decisions on distances and time — I’m not subjectively collecting these.

Next: Nylon Calculus -- Projecting the top 23 and under NBA players

This is at the very least a showcase of what’s still possible with public data and some creativity. You don’t need access to SportVU data or expensive software programs to explore new grounds. The advanced stats revolutions started at home, propelled by mostly hobbyists, and that will remain a crucial force going forward — as long as we all don’t give up.