The NBA season is a marathon, and there are so many ways to consume the sport now it’s overwhelming. But we did — we’re here. I’m glad I get to ramble about basketball, and that some people actually care. There’s so much to dissect, and there’s a constant updated stream of information: game results, trades, new players, guys evolving, etc. This is fun for analysis, and the best part is that instead of pouring over dreary reports or numbers I can watch some basketball to analyze what I’m working on.
There’s the appeal, and I hope I’m continuing the good fight in the advocacy of knowledge and statistics. The anti-intellectual movement all around the globe has been strong; it’s a push-back from decades of experts in economics, policy, and others having a say in society. But if we can show that an objective, number-based system works in sports, hopefully that’ll convince people it’s okay for that analysis in other modes of their life too.
However, question everything, look for the source, and scrutinize the analysts. The truth is out there, somewhere, and if you don’t consider every tool available, you may wander blind. And now let’s look back at the last one-and-a-half weeks in basketball….
How we use clutch stats
There are no official rules about how to vote for Most Valuable Player, so people use their own interpretations, but there are generally a few philosophies that are agreed upon. There’s one topic, however, that’s a little controversial, and I feel like I should address it because it’s not often explained well. Some people use clutch stats as additional fodder in these arguments, stating that a particular player has more value because he helped his team win close games. But that’s flawed.
Clutch stats are not stable year-to-year. A player whose team excels with him in the clutch one year may look horrible in the same clutch situations the next year — there’s a very small correlation. This isn’t about hijacking the notion of value with that of prediction. It’s not about what’s most predictive. For instance, 3-point percentages in general are noisy, and they’re not too stable year-to-year. But the correlation isn’t zero, and it’s reflective of a player’s ability still.
The problem with relying on clutch stats for yearly awards is that the stats are so unstable it suggests they don’t have any connection to a player’s own attributes. If a player has consistently shown he can play significantly better during high-leverage moments, and his teams keep winning close games, then yes, you have some solid reasoning to use those stats. Otherwise you’re letting too much randomness control your decision-making, and you’re valuing something that doesn’t signify anything about the player himself.
Moe money for Harkless
Player incentives are a good idea in theory; you’d want to pay your guys based on how they produce for your team. But the incentives are sometimes based on simple stats, and there are some arbitrary thresholds. Sure, you’d prefer if a player shot above 35 percent on the season, but would it really kill you if he shot 34 percent? That’s the position Moe Harkless was in for the Blazers. If he dipped below 35 percent, he’d lose $500,000 .. and he was at 35.1 percent heading into the final game. With one attempt and a miss, he’d dip below 35%. What did he do? He did not attempt a 3-pointer and, in fact, did not take a shot outside the paint. He loitered inside for most of the game. And who could blame him? Portland’s playoff seeding was set anyway.
Cleveland’s defensive ranking
I hate to keep harping on this, but there’s too much focus on on Cleveland’s defensive ranking, and it’s a misinterpretation of what matters. There have been numerous pieces on how the Cavaliers’ defensive ranking means it’s virtually impossible for them to win the championship because only one other team, the Lakers in 2001, has won with a below-average defense. But there are only a small sample of title winners, and there’s an obvious reason why those type of title winners aren’t more common: it’s tougher to outscore your opponents when you’re deficient at one end of the court. It means you need an incredible offense to shift the odds in your favor — or just luck.
I have a histogram of every championship winner’s net rating since the merger below. You can see that most winners have a rating between 4 and 10 points. Cleveland right now has a net rating of 3.3, despite their excellent offense — that’s why there are so few champions with mediocre defenses. For the Cavaliers to have the mean net rating of a title winner, they’d need an offensive rating that was 8.6 points per 100 possessions better than the average team. Only the Dallas Mavericks in 2004 and the Phoenix Suns in 2005 have ever been that good.
I know this seems pedantic, but if we all write treatises on team’s chances in the playoffs, we should get this right. Cleveland could struggle in the playoffs, and it’s not directly because of their defensive ranking — it’d be because their defense limits their ability to outscore opponents. Focus on their point differential (and how much LeBron James can increase his effort level like last year), not their rankings.
Russell Westbrook is mean
I hope we all don’t forget the shot below: He eliminated Denver from playoff contention with a game-winner from 35 feet. That’s preposterous. And what a heartbreaking turn of events for the Nuggets, who found a new future with Nikola Jokic in the middle. Somehow they finished fourth in the NBA in offensive efficiency — who saw that coming? They’re just behind the Cavaliers, who have LeBron James and Kyrie Irving surrounded by shooters. I feel like this is the underrated crazy stat of the season.
HBox record broken
Well, the final HBox stats are now available, and it’s official: Russell Westbrook has the greatest HBox score ever and he broke the MVP Index. The index is supposed to span 0 to 100; it’s calibrated on the best season ever in the data sample. Since Westbrook smashed that record, I’m going to have to redo it — thanks Russell. James Harden, however, wasn’t far behind, and he virtually tied the best HBox score ever from LeBron James in 2009. And with that, let’s duck right into the awards section of this article, where I explain my reasoning for various items and tackle some of the tougher debates.
Awards
Most Valuable Player
There are four major candidates here: Russell Westbrook, James Harden, LeBron James, and Kawhi Leonard. There are some dark-horse candidates too, like Stephen Curry, Draymond Green, Giannis Antetokounmpo, and Jimmy Butler. Then there a couple guys who could have gotten consideration if it weren’t for injuries: Chris Paul and Kevin Durant. This is a tough race, and this season has the argument for deepest, best MVP pool ever.
I’ve tackled this MVP race many times before, so let me cut to the chase: Westbrook and Harden have been posting statistics that shouldn’t be possible, and it’s the jaw-dropping breadth that astounds me. Both guys are at the bleeding edges of shot creation, creating offenses on their own, while Westbrook is rebounding like a big man and Harden has been wondrously efficient. Both teams depend on their players to extreme ends too — the Thunder depend on Westbrook more, but Harden’s team is vastly superior. So choosing between them is about your philosophy. For me, I’m more interested in Harden’s efficiency and his role on a great offense. And I want to address this: why are we discussing this, “What if Westbrook was on a better offense?” like it’s a hypothetical. Do people not remember him with the contender Thunder-team for years with Durant? He wasn’t even more efficient, even with fewer shots and Durant around.
If you strictly go by the stats, it appears Westbrook has the lead. But, as I’ve analyzed before, Westbrook’s stats are unprecedented, so it’s tough for some metrics to value his season. If you ignore BPM, which Westbrook has broken, Harden and him are virtually tied. It then becomes a matter of how you see his rebounding, and there’s a deep topic there about how he can jump-start fastbreaks, how they rebound better with him, how they clear out and let him take every board possible, etc. Plus, I think his inattention on defense has been overlooked — how does he contest so few shots? He’s still the rock for Oklahoma City though. There’s no easy answer here.
Table: top players by a metric soup
| Player | RPM | BPM | Dredge* | Hbox | Mean | Mean*MP/ 3000 |
| Russell Westbrook | 6.0 | 15.5 | 2.4 | 10.8 | 8.7 | 8.1 |
| James Harden | 4.8 | 10.1 | 4.5 | 10.0 | 7.3 | 7.2 |
| LeBron James | 7.3 | 7.3 | 4.5 | 6.3 | 6.4 | 5.9 |
| Kawhi Leonard | 6.1 | 7.9 | 6.4 | 5.4 | 6.5 | 5.3 |
| Giannis Antetokounmpo | 4.2 | 7.6 | 5.1 | 5.1 | 5.5 | 5.2 |
| Stephen Curry | 7.1 | 7.4 | 3.7 | 4.6 | 5.7 | 5.0 |
| Jimmy Butler | 7.1 | 6.9 | 3.4 | 3.3 | 5.2 | 4.8 |
| Rudy Gobert | 6.7 | 5.9 | 4.2 | 3.4 | 5.0 | 4.6 |
| Kevin Durant | 5.6 | 8.0 | 6.0 | 6.9 | 6.6 | 4.6 |
| Draymond Green | 6.4 | 5.4 | 4.3 | 4.8 | 5.2 | 4.3 |
*Dredge last updated March
There’s a decent argument for LeBron James too. While he’d played fewer minutes than Russell and Harden, I think his defense can close the gap on offense — he’s not offering some world-busting offensive creation like those two, but, of course, he’s still the gold standard for a reason. However, while his defensive skill-level is certainly high, I don’t think his effort has been there all season — he’s one of many reasons they’ve been mediocre on defense. I’ll certainly entertain the notion, and any vote for him isn’t silly. I just can’t find a compelling reason myself.
As for Kawhi Leonard, here’s one compelling reason I found: if you adjust for his bizarre opponent 3-point percentage stats, one could raise his RPM by, say, two points. That’s enough to put him in the lead overall in RPM. And he’s consistently rated well on defense by plus-minus stats for years; most box-score metrics underrate him there. However, he does play fewer minutes, and he’s not the same kind of offensive force the others are. Plus, there’s probably some Spurs-ian magic going on because almost every player sees better numbers with the team. I’m not sure to what extend that’s true with him — he’s undeniably great — but it’s something to keep in mind.
Ultimately, I believe James Harden is the best candidate because of his enormous load for the Rockets, and his efficiency is so valuable — and consistent — for a great team. But the other guys have great arguments too, and I feel like I’m splitting hairs. And if it weren’t for Westbrook’s triple-double shenanigans, I think more people would realize that Harden’s season was simply one of the best ever. The Rockets had a point differential not far from +6.0, which is something you see from a championship winner — and look at that roster.
The last spot’s going to Stephen Curry, and I don’t think that’s controversial. He was the unanimous MVP last season, and he didn’t fall off a cliff. He’s still a court-warping, master shooter. That’s enough to push him over other guys with flawed cases, from the still flowering Giannis to guys who missed a long stretch of time.
1: James Harden
2: Russell Westbrook
3: Kawhi Leonard
4: LeBron James
5: Stephen Curry
All-NBA Teams
The first team is easy. Center is the only spot where I needed to think. But Gobert is Defensive Player of the Year worthy, and he’s transformed into a dangerous finisher the opposing defense has to track. That’s valuable, and no other center can come close.
First team
G: Russell Westbrook
G: James Harden
F: Kawhi Leonard
F: LeBron James
C: Rudy Gobert
By cheating and pushing Butler to shooting guard, which is totally reasonable on the basketball court, the second team is solid too. Draymond has been amazing on defense, and his offense isn’t too shabby either — he has to be included somehow. Davis has been the league’s second-best center, and yes, he’s played most of his minutes there. His defense is pretty great thanks to his condor arms, and offensively there are few centers who can compare. By the way, Giannis is the real deal, and his ability to fill up the stat-sheet in so many ways is jaw-dropping. If I’ve learned anything about basketball and what makes a player valuable, it’s versatility.
Second team
G: Stephen Curry
G: Jimmy Butler
F: Giannis Antetokounmpo
F: Draymond Green
C: Anthony Davis
Chris Paul and Kevin Durant played at an MVP-level while on the court, so in limited minutes they should at least be on the third team. For the last center spot, there weren’t a lot of candidates, and it basically boiled down to DeAndre Jordan versus Marc Gasol. I love Gasol, and his passing is quite valuable. But Jordan is an active force on defense, and his ability to finish inside draws a lot of attention. He deserves it. And it means something that year-after-year Marc Gasol does not rate too well by impact stats.
As for the last guard spot, this might be a surprise, but anyone who knows me understands my adoration of Kyle Lowry. Yes, he got injured, but I believe his production level is so such that it eclipses that of someone like Isaiah Thomas, whose defense is a liability, and John Wall, who is not as valuable offensively because of his efficiency. Finally, I’m giving Paul George the last by the slimmest of margins over Gordon Hayward. I believe most metrics underrated George’s defense; that was the tie-breaker.
Third team
G: Chris Paul
G: Kyle Lowry
F: Kevin Durant
F: Paul George
C: DeAndre Jordan
Sixth Man of the Year
First of all, Iguodala should be an easy vote for everyone, and it’s a shame he hasn’t yet won it and still may never for his time as the key reserve for one of the greatest stretches in NBA team history. I’m afraid people are still too smitten with points per game for this award. After Iggy, I have to go with one of my favorites, James “Bloodsport” Johnson, who, I am delighted to say, has finally become a beloved role player. I’ve always thought he could be a valuable contributor if given the opportunity. He’s always been an athletic force with a decent number of assists and blocks, but he has a high assist rate for a player of his type and he’s added a (somewhat) reliable outside shot. Miami would have never chased the playoffs without him.
For the last spot, I’ll go with Lou Williams, who seems like a curious decision over Eric Gordon, since they’re both on the same team now and Gordon has the bigger role on the bench. But Lou has had superior production overall — he’s a better playmaker, and even in Houston takes more shots per possession. Defensively, both are a mess, but Gordon’s small edge there isn’t really enough to offset their respective offenses. And Lou’s slump in Houston was mostly due to a low 3-point percentage, which is likely to recover.
1: Andre Iguodala
2: James Johnson
3: Lou Williams
Most Improved Player
This one is pretty easy at the top: it’s Giannis Antetokounmpo, who has transformed into a legitimate star, an Sll-NBA player, and someone who could win an MVP some day. After him, the field is wide-open. I’d go with Myles Turner for the runner-up spot because, even if you adjusted for his age, his improvement was massive. Russell Westbrook actually had the biggest improvement in his BPM; however, I believe BPM is overstating his value. He’s still a decent candidate for top-3 though. But for the last spot, I’ll go with someone who for years was a liability on the court — his teams usually got creamed when he played — but has found a role as positive contributor for the Hawks, shooting well and playing better defense: Tim Hardaway.
1: Giannis Antetokounmpo
2: Myles Turner
3: Tim Hardaway Jr.
Table: most improved by BPM (source: b-ref)
| Player | 2017 BPM | 2016 BPM | 2015 BPM | One-Year BPM Jump | Two-Year Weighted |
| Giannis Antetokounmpo | 7.6 | 2.4 | 0.5 | 5.2 | 6.1 |
| Russell Westbrook | 15.5 | 10.0 | 11.0 | 5.5 | 5.0 |
| Myles Turner | 2.2 | -2.3 | NA | 4.5 | 4.5 |
| Terry Rozier | -2.5 | -6.6 | NA | 4.1 | 4.1 |
| Lucas Nogueira | 5.5 | 2.6 | -10.3 | 2.9 | 4.1 |
| Dewayne Dedmon | 1.7 | -1.3 | -3.3 | 3.0 | 4.1 |
| Marreese Speights | 0.7 | -4.0 | -2.6 | 4.7 | 3.9 |
| Julius Randle | 0.0 | -3.6 | -16.8 | 3.6 | 3.7 |
| Spencer Dinwiddie | -0.9 | -5.1 | -4.2 | 4.2 | 3.5 |
| DeMarcus Cousins | 7.1 | 2.6 | 4.7 | 4.5 | 3.5 |
| Kelly Oubre | -2.3 | -5.8 | NA | 3.5 | 3.5 |
| Mike Conley | 5.8 | 1.8 | 2.7 | 4.0 | 3.5 |
| Nick Young | -0.7 | -4.8 | -3.5 | 4.1 | 3.5 |
| Larry Nance Jr. | 2.0 | -1.4 | NA | 3.4 | 3.4 |
| Luc Mbah a Moute | 0.8 | -1.6 | -3.2 | 2.4 | 3.4 |
Rookie of the Year
I’ve talked about this before, so I’ll spare some words for this. Joel Embiid, despite his low minute total, has imparted more value to his team. There is no minutes cap or limit for this award — I’m strictly looking at value. This is pretty undeniable too. If you factor in minutes played and production, Embiid usually comes out on top. As for the rest of the list, Malclom Brogdon was arguably the only competent, full-time rookie who’d you want starting. Dario Saric was only good for half a season, and his playmaking was offset by his poor shooting. For the final spot, I debated between Dario Saric, Willy Hernangomez, and others. Ultimately, I went with Willy because he was genuinely useful on offense and consistently, while his defense wasn’t as bad as suggested. Sure, he only played 1300 minutes, but this was a weak field.
1: Joel Embiid
2: Malcolm Brogdon
3: Willy Hernangomez
All Rookie Teams
Picking the firest team was basically just finding who were maybe the only five competent rookies who played all season. The second team was an exercise in frustration. Besides McGruder, who played well for Miami, it was mostly just “list players who seemed interesting and capable of producing something.” Many rookies were just on a training program, like Brandon Ingram and Marquese Chriss, and should not play for any competitive team.
First team
Malcolm Brogdon
Jamal Murray
Dario Saric
Willy Hernangomez
Joel Embiid
2nd team
Yogi Ferrell
Caris LeVert
Jaylen Brown
Rodney McGruder
Skal Labissiere
Defensive Player of the Year
For this award, it’s a race between Rudy Gobert and Draymond Green, the two indispensable pieces of two of the best defenses in the league. I don’t see any reason to discuss anyone else either. I believe this argument comes down to one central fact: Rudy Gobert is defined by being the current kind of rim protection, but Draymond Green nearly equals him there. And Green can do things on defense that Rudy simply cannot. He’s one of the most versatile defenders in league history, and he led the Warriors to the second-best defense in the NBA despite the team’s loss of Andrew Bogut and Kevin Durant’s injury. They’re pretty close statistically — Green’s ahead by defensive BPM and Rudy by defensive RPM — but Draymond’s the guy whose value is tougher to detect with countable stats. So I’m fairly confident in choosing Green, but both guys deserve to win.
The last spot was the toughest to parse though. There are many players who were definite contributors on defense but they either didn’t play enough minutes or weren’t “anchor-worthy,” meaning you wouldn’t think of them as a one-player defense who defines the team. Thus, I’ll go with the only guy who can be reasonably labeled that: Kawhi Leonard.
I believe two things here, namely that he’s not the best defender in the league anymore, probably due to his larger role on offense, and that his plus-minus stats are misleading. I was actually the first to discuss the flaws in his defensive plus-minus a few weeks ago, and if you do some adjustments for opponent 3-point percentage you can maybe add 2 points or so to his defensive RPM. That’s nothing world-bending, but I’ll take the guy who still executes flawlessly and can be an absolute terror when needed on the perimeter. Everyone else — Anthony Davis needs to master rim protection and boxing out, Dwight Howard has lost a lot of speed, Paul Millsap lost some time to injury and some of his defensive stats have dipped, it’s still easy to fake DeAndre Jordan, et al — has flaws.
1: Draymond Green
2: Rudy Gobert
3: Kawhi Leonard
All-Defense Teams
I shouldn’t have to explain my first guard, Chris Paul — he is tiny but exemplifies how basketball can be about skill and persistence. He has the defensive plus-minus numbers of a good defensive center, and it’s hard to find a reasonable counter to those stats. And no, not even my favorite opponent 3-point percentage looks fluky with him. I’m cheating a bit with the second guard spot, Andre Roberson, who’s often listed as a small forward. But forward is deeper and I’d love to reward this guy, whose work on that end of the court has been underrated this season as Russell Westbrook stole all the headlines. The Thunder have been excellent on defense this season, and with Westbrook focusing on other duties and with players like Enes Kanter in the rotation, it’s largely up to him and Steven Adams to hold down the fort. After that, the rest of the ballot is easy.
First team
Chris Paul
Andre Roberson
Kawhi Leonard
Draymond Green
Rudy Gobert
For the second-team, I gotta include Patrick Beverley, who rates so well in a few stats that I track that he looks like a borderline All-Star because of his defense. He led the league in loose balls recovered per minute, for instance. Then there’s Danny Green, one of the league’s best perimeter defenders on the best defense in the league. I can’t imagine there’ll be many complains about him, so let’s focus on one surprise: Robert Covington plays for the 76ers, sure, but he’s a versatile defender now and he’s one of the preeminent ball-thieves. If his team ever gets to use a healthy Joel Embiid, this duo could form one heck of a great defensive team.
For center, I thought Anthony Davis was a lock for the second defensive team. He’s still a shot-blocking force with Gumby arms, and he’s more refined now than he was when he was younger. The Pelicans were surprisingly great on that end of the court too this season. It’s not like they had any other heavy hitters on the team either, besides maybe Jrue Holiday, who’s a point guard and doesn’t have much of an impact. The last spot for forward was tough, and I realized I needed to have Davis listed as a center to make it work. Paul Millsap and Dwight Howard both had compelling cases, and while Howard’s rebounding numbers dwarfed Millsap’s, everywhere else either Paul had a significant advantage or was close behind. He even defended more shots at the rim than Howard per possession. The value of defensive rebounding is usually overrated, and some of Millsap’s skills are usually hard to rate properly — he’s the last pick for the team.
Second team
Patrick Beverley
Danny Green
Robert Covington
Paul Millsap
Anthony Davis