NBA Week in Review 24-25: the Twilight Stretch

Goodbye, dear 2015/16 NBA regular season — another one has evaporated into the aether. I will miss your Warriors box scores and your Westbrook highlights and your enticing rookie center class and the sadness of your 76ers experimental basketball team. Let’s take a look back at the past week-and-a-half before the playoffs envelop the NBA landscape.

The New Jersey Swamp Dragons

If you haven’t seen Zach Lowe’s article from last week on a name change proposal the Nets had years ago, I suggest checking it out — it’s ludicrous. I can understand the inclination to change a boring name into something snazzy during the time of names like the Hornets and Raptors, but this one is so strange I kinda wish it had happened. It’s at least better than the Thunder. But … what IS a swamp dragon? This is not related to any New Jersey mythology I’m familiar with. Google wasn’t entirely helpful either, although I did get this definition of a swamp dragon from the Discworld series that’s also a succinct description of the franchise: “Swamp Dragons (Draco vulgaris) are small, fly badly and tend to explode, due to the generation of various flammable gases in their internal plumbing.” Your 2016 Nets!

Hustle Stats

I know this might seem like a minor addition, but the stats the NBA will publish for the playoffs — contesting 2-and 3-point shots, deflections, charges, recovering a loose ball, and screen assists — are valuable when used in concert with other defensive stats. First of all, charges have been tracked, more or less, for ten years, and they’re significantly valuable. Strangely, I’ve found in my research that offensive fouls drawn that aren’t charges are more valuable. We’ve also had limited access to contested shots, but I fear the NBA will not provide these stats in a useful way. Contesting a shot isn’t a binary option because being one foot away is not the same as two feet is not hte same as three feet, etc. Plus, contested three-pointers, unless they’re only counting heavily contested ones, are not useful components for defensive analysis, from what I’ve found.

However, the screen assist is something I’ve wanted publicly for years. The screen is one of the most important NBA actions that is not visible in any normal stats. Many plus/minus monsters, like Nick Collison, are excellent screeners, so this would explain some of their value. Deflections correlate strongly with steals, but more information is better and there are likely a few guys who have wildly different ratios of steals to deflections. Finally, the recovery of a loose ball is not only a way to track a possession change, which is valuable, but it’s probably also a good “hustle” proxy. I’ve found a few surprising stats hidden in the play-by-play logs that do the same. Combined together, and we could have a surprisingly powerful way to measure, track, and predict defense. This is all assuming the NBA adopts this for the regular season, so please write to your local representatives to push for this action.

The Quintessential End

I’m not sure how Kobe Bryant’s career ended with the Kobe-ist game possible. Despite all the aches and pains and the futility of the season and the millions watching him and his legacy already in place, Kobe scored an incredible 60 points on 50 field goal attempts with 12 free throw attempts — no one’s had that many field goal attempts since the mid-60’s. He already has the single season shot usage record, and he probably has the single game one now too with a bonkers 62% — that means over 60% of their possessions were used by Kobe. If you think that scoring at the league average TS% with that many attempts is impressive, I’ll point you to this old graph I made showing how high scoring games are typically highly efficient and that Kobe’s last game was outside the low efficiency boundary line — basically, it was one of the least efficient high scoring games ever[1. Kobe had a TS% of 54 and 60 points, and on that chart you can see how far away his game was from everyone else’s.]. However, this isn’t an attack on Kobe. But the way his career ended, with a game designed to hype his fandom and add fuel to his criticism to such an absurd degree, is too crafted — I can’t believe that happened. So long, Kobe, and thanks for all the entertainment.

73

I never thought this record would be broken in my lifetime. I thought the league was too competitive and the salary cap and luxury tax were too restrictive. I know everyone has an angle here and more articles will be written about them, so I’ll be brief. If I ever got to construct a super-team, it would resemble the Warriors. I’d value versatility, passing, shooting, and having no holes on defense. So to see this team emerge, with some of my favorite players like Andre Iguodala, is enthralling. They’re beating the dumb notion of, “Jump-shooting teams can’t win titles/big,” into smithereens.

What’s phenomenal is that no one saw this coming — Stephen Curry wasn’t a number one pick and his early injury problems were worrying; Draymond Green was a second round pick; Andrew Bogut was nearly forgotten due to injuries; and the franchise had been mired in incompetence for decades. There’s something about the way history unfolds where, finally, at the end you step back and realize all that’s unraveled will be remembered decades from now by most hardcore fans because of how important it is, and many of the details will be lost in time but some overwhelming thread of the story will remain. Some of that will be shaped by what happens in the future, like whether or not they win a title and how they do in following years, but for now we’re living in that moment. And it’s awe-inspiring.

My Imaginary Votes

As a bonus, I’ve included how I would have voted for the NBA’s various awards if the commissioner would simply respond to my crayon-scrawled letters.

MVP

Curry is obvious. I have Leonard second because I’m not penalizing him for the San Antonio low minutes program. And yes, Draymond Green is in the top five. Long live the revolution.

1) Stephen Curry

2) Kawhi Lenoard

3) Russell Westbrook

4) LeBron James

5) Draymond Green

Defensive Player of the Year

I think a lot of people are picking Green over Leonard because Leonard won last year, but that’s not a valid reason. And that’s partly why I was miffed Draymond didn’t win last season, as Kawhi missed many games but was so good and still young that I thought it was Draymond’s best opening to win the award.

1) Kawhi Leonard

2) Draymond Green

3) Rudy Gobert

Most Improved Player

No one ever knows what to do here. I’ll just note that Kawhi Lenoard’s shooting improvement alone is special.

1) CJ McCollum

2) Stephen Curry

3) Kawhi Leonard.

Sixth Man of the Year

I am certain the Spurs should win this as a collective, and you’ll see why a few inches below. Andre was the most valuable individual in my eyes, however.

1) The Spurs bench

2) Andre Iguodala

3) Ed Davis

Rookie of the Year

There’s nothing too daring here, like taking Jokic first because of his impressive plus/minus stats. I would caution that Jokic’s defensive plus/minus stats seems way too high and I would bet on him regressing in that area next season.

1) Karl-Anthony Towns

2) Nikola Jokic

3) Kristaps Porzingis

All-NBA

I listed Green as a center because I wasn’t in love with the other center candidates. I’m not “cheating” there; I just wanted to select the best possible players. I don’t think any of my picks are crazy, save for Kevin Love, who I have defended vigorously throughout this week in review series.

1st: Russell Westbrook, Stephen Curry, Kawhi Leonard, LeBron James, Draymond Green.

2nd: Chris Paul, Kyle Lowry, Kevin Durant, Paul Millsap, Al Horford.

3rd: Damian Lillard, James Harden, Paul George, Kevin Love, DeAndre Jordan.

All-defense

I cheated by putting Paul George as a shooting guard, but realistically he can play that position so I’m fine with it and the guards were tricky to sort and manage. I also wanted to make room to shine some light on Jae Crowder, who deserves some acclaim.

1st: Ricky Rubio, Paul George, Kawhi Leonard, Draymond Green, Rudy Gobert.

2nd: Chris Paul, Danny Green, Jae Crowder, Paul Millsap, Tim Duncan.

All-rookie

The second team was a nightmare to figure out because there were a number of half-decent players who all had arguments.

1st: D’Angelo Russell, Justice Winslow, Nikola Jokic, Karl-Anthony Towns, Kristaps Porzingis.

2nd: Trey Lyles, Larry Nance Jr., Frank Kaminsky, Willie Cauley-Stein, Boban Marjanovic.

Bench Greatness

Someone on Twitter brought up an interesting question: the Spurs have an absurd number of players with a BPM of 0 or above with 13. But which team has the most ever? Or is that the record? For the uninitiated, BPM is a box-score metric built from a plus/minus model, and it performs quite well with new data the metric hasn’t seen before. Since 0 is set as league average, it’s remarkable that any team would have 12 or more players with positive values, as it essentially means that they have full roster without any below average guys[2. I have to mention this, however: BPM has a team adjustment so that BPM adds up to a team’s rating. This does create a slight bias in most situations for bench players on great teams. This all depends on how the “raw” BPM adds up before the team adjustment, but bench guys on elite teams typically get a boost here.].

Going back to 1974 when BPM is available, the record is actually 13, but it’s shared by another team: the 2012 San Antonio Spurs. This was, however, a year where they made in-season moves involving Richard Jefferson, Stephen Jackson, and Boris Diaw. Seen in the table below, the Blazers from the other lockout season, 1999, were second with 12 such players. That team was known for its depth — if it weren’t for their epic collapse the next season, they’d could have been used in examples of superstar-less championship squads along with the Detroit Pistons of 2004. Surprisingly, among the other teams listed below is a Knicks team from 2011. That was the year they traded for Carmelo Anthony, so they, like the Spurs in 2012, had a larger than average number of players who found some time on the court. And, naturally, their trading partner shows up too — it was a large trade with a number of decent players. You can also use a filter of 500 minutes played, which ignores bit players who only net a few scattered minutes. The Spurs, again, hold that “record” with 11 such players in 2014, a title year, as well as this season.

Table: teams with most BPM>0 players

Seen graphically, better teams usually have more of these above average players, of course, but it’s interesting to see how fuzzy and wide the curve is. The greatest teams ever have a range of options: one team has only three above average players[3. Surprisingly, this was the 1991 Bulls led by Michael Jordan, Scottie Pippen, and Horace Grant. I imagine most people would guess, say, one of the iterations of the Miami Heat with LeBron James or maybe some recent Oklahoma City teams. But that Bulls team had limited role players like Craig Hodges and Will Perdue while Bill Cartwright was their starting center — somehow he had only 15 blocks on the season and didn’t contribute much of anything else.] while others have ten or more[4. At the other end of the spectrum, the 2015 Boston Celtics, appropriately, had the most above average guys, ten, and still somehow had a negative team rating. That doesn’t even seem possible, but mid-season trades gave them a huge full season roster and a number of highly used guys like Avery Bradley and Jeff Green had significantly negative BPM ratings.]. And I have to mention the two teams there at the bottom as they’re the best candidates for the worst teams ever and they each had zero players with a BPM greater than 0: the 1993 Dallas Mavericks and the 2012 Charlotte Bobcats.

Of course, that’s not the ideal way to measure bench strength for a couple obvious reasons and a few others, although it is interesting. Instead I’ll use the VORP stat, which is simply a product of a player’s BPM and his minutes with -2 as the baseline — if your BPM is below -2, then you create “negative value.” To create bench-only team VORP ratings, I simply multiple a player’s VORP by his games not started divided by his total games. That way I correctly handle guys who are partial starters, and I’m not distorting the numbers with players who are on the bench just a couple games a season completely or over-counting them.

Unfortunately, data for games started only goes back to 1982, but I don’t think anyone will disagree with the leader: it’s the Spurs again, and this time they have five of the top six spots. The aforementioned 1999 Blazers are fourth — they were bringing Rasheed Wallace, Walt Williams, and others off the bench[5. I prorated every team to 82 games so teams from lockout seasons weren’t unfairly penalized for their total VORP.]. The most shocking entry on the leaderboard at first glance is the 2013 Clippers — but that’s because of Doc Rivers the GM has created a few awful benches since then. Back in 2013, they had Eric Bledsoe terrorizing second units along with Jamal Crawford, Matt Barnes, and a few others.

There appears to be a modern bias, however, and it’s probably related to a commonly cited fact/complaint: top guys are playing fewer minutes, which leaves more time for bench players to compile value (i.e. VORP.) Thus, I included a rate version of the bench stat, which is just the BPM of the entire bench using the same games started method. Using that, and the 2016 Spurs hold the top spot by a good margin. Indeed, they have a good argument for best bench ever. Surprisingly, the Blazers in 1999 actually have the second highest rating, and a few older teams move up higher.  I’m not entirely sure which team most people would pick for best bench in NBA history — there are somehow few articles I could find online about this — but I suspect in the three-point era the Bad Boys Pistons are up there, and indeed they rate highly by this method.

Table: best benches

Conversely, just for fun, we can take a quick look at the worst benches ever. By VORP, the worst bench belongs to the 1999 Chicago Bulls, whose cupboard was bare after the exodus of their three-peat core. Usually teams with the worst benches are just plain terrible overall, but there’s one team that’s not just plain terrible: the 2013 Blazers, who had the worst bench BPM ever. To Portland fans, this is vindication because their core was so strong yet the team was far from the playoffs. Once they got rid of JJ Hickson and revamped the bench, they suddenly became one of the strongest teams in the league — it’s no wonder.

The Golden State Warriors dominated NBA headlines, so this flew under the radar: the Spurs probably just had the greatest bench of all-time. I’m not the first one to recognize how great it’s been, and you don’t need stats to see it either — they have some mighty names on the pine. They have one of the greatest bench players ever in Manu Ginobili; Patty Mills has been one of the best bench spark plugs in the league; Boris Diaw is Boris Diaw; David West is a former all-star and is still quite effective; Andre Miller is one of the greatest non-all-stars ever; Kyle Anderson and Jonathon Simmons have been good Spurs-in-training; Kevin Martin is a wicked shooter off the bench; and Boban Marjanovic is downright mythic. In an era with a salary cap and with a franchise without a huge bankroll, San Antonio’s ability to build benches that could beat half the league should be properly recognized. This is historic, and it’s one of the reasons I’ll miss this season.