A Deep Breakdown Of The NBA MVP Debate
As Nikola Jokic closes in on what is very likely to become a 3-peat, I’ve had a lot on my mind.
Mike O’Connor is the best O’Connor in basketball writing. Previously of The Athletic, you can find Mike on Twitter @MOConnor_NBA. Mike’s writing is brought to you by Body Bio, supplements based on science, focusing on your gut and brain health. Get 20% off E-Lyte, Gut+, and all Body Bio products with promo code FIRECJ at Body Bio’s website.
While everyone else feels frustrated and burnt out with the MVP discourse, I can’t get enough of it. I find the whole thing fascinating, and I think there’s a ton that it reveals about the state of NBA discourse, as well as the media landscape. As Nikola Jokic closes in on what is very likely to become a 3-peat, I’ve had a lot on my mind. Here in this piece, I’ve decided to wade into some incredibly fraught topics and attempt to provide some reasonable perspective.
The end of voter fatigue
Later on, I will discuss certain dynamics that have helped Jokic’s case over the previous two seasons, but one interesting dynamic that has made his case possible this season is the demise of the phenomenon known as voter fatigue. Once an understood, but hardly-considered pattern in these types of awards, numerous media members have made a point this year to mention how silly it would be if Jokic doesn’t win the award on the basis of having already won two in a row.
I’ll get to the validity of voter fatigue in a second, but one thing that is indisputable is that The War on Voter Fatigue is brand-spanking new. It may have been boiling for a while, but it hit a consensus right now, with Nikola Jokic. Comb through the past several years alone and you will find numerous examples of obvious voter fatigue.
After winning back to back MVPs in 2015 and 2016, Steph Curry played 79 games, put up unbelievable numbers, and led his team to 67 wins in 2017, and yet he finished sixth (!) behind Isaiah Thomas and four others in MVP voting that season.
In 2018-19, James Harden was almost inarguably better than he was when he won in 2017-18, but finished second in MVP voting to Giannis Antetokounmpo. After winning his second straight MVP in 2020, Antetokounmpo was essentially the exact same player in 2021 as he was in the two previous years, yet was a distant fourth in MVP voting behind Steph Curry and Joel Embiid (and Jokic, obviously), and multiple media members made it a point to mention that his lack of success in the playoffs in the previous two seasons made his case a bit weaker in 2021 – the (lack of) parallels there to Jokic are obvious.
Regardless, the point is that the history of the NBA MVP is very clear: you win an MVP, it gets harder for you to win the following year. You win two MVPs in a row, and it’s almost freaking impossible to win three. Hence, why Jordan, Kareem, LeBron, etc. never won three straight.
The award has never been a pure meritocracy in terms of going to the empirically best player. Voter fatigue, as well as the narrative surrounding you and your team, have always had enormous impacts on the vote, until now.
The message sent to Curry and Giannis after back-to-back MVPs was clear – “OK, that’s enough for you, for now. It’s time to highlight someone else.” While both players remained MVP-caliber in the following seasons by any reasonable person’s standards, they were effectively taken out of the running so that we could tell a new story that season, highlight a different player’s ascension, and see a different person make the acceptance speech.
Those elements seem to have been cast off as silly, irrational considerations, but I would argue that that’s not a fair critique. Voter fatigue is not a mortal sin. It’s what allows us to shine the spotlight on players who ascended and/or defined the storyline of that NBA season as opposed to just re-stamping the best player. Without voter fatigue, there would probably only be somewhere around 15 players to ever win an MVP, and in my very humble opinion, that’s a boring way to tell the story of the NBA.
My point is essentially that I’d like it to be the way that it’s always been – that the old way is superior for telling the story of that NBA season/decade/era. The rush to do away with it is, in my opinion, the result of a combination of Jokic favoritism as well as the rush to make everything (including the purposefully vague Most Valuable Player award) purely empirical; no media member wants to be yelled at on Twitter for succumbing to a silly bias such as voter fatigue. NBA media has become, perhaps more so than any other sport, a battle for intellectual superiority, and holding onto voter fatigue is seen as anti-intellectual.
As great as Jokic was last season, he was simply irrelevant to the story of that NBA season. There was never a single moment where anyone considered the Nuggets to be a title contender. It wasn’t like he had a fascinating storyline behind him, or a bunch of iconic moments of late-game heroics, both of which helped to bolster Russell Westbrook’s case in 2017 (flawed as it was). Jokic was just mowing down the competition away from the spotlight, carrying an underwhelming roster to the sixth seed in the West and a first round exit. The season just didn’t feel like it was about him – and with him having just won, it would have felt appropriate to highlight someone else for that particular season.
And the irony is, this season is about Jokic! If we’re abolishing voter fatigue, his case this season is pretty unassailable. He’s likely been the best player in the league, and is on the best team in his conference. He has dominated the storylines all season long. Embiid is coming on strong of late, but if the season ended today, Jokic’s case is better in a vacuum – but all I ask is that we acknowledge that the demise of voter fatigue is brand new, and that it will have affected every single great player in the history of the NBA up until Nikola Jokic.
Not race, but something else
In case you’ve been living under a rock, one of the key discussion points throughout the past month or so of the MVP discourse has been whether or not race has played a role in Jokic receiving some degree of favoritism over the past few years. While I’m not going to dismiss the possibility that race plays a small role, I do think that there’s a far more obvious dynamic that is contributing to media favoritism of Jokic.
If you work in NBA media, it’s likely that you love basketball to the point that you would identify as a basketball nerd. And in Nikola Jokic, there is absolutely no player in the history of basketball who better exemplifies the ethos of the basketball nerd. Many media members will openly state that he is their favorite player. He is proof that you can become one of the best basketball players on earth purely with IQ, finesse, and skill – he is a walking one-man version of The Beautiful Game-era Spurs. More so than they love his skin color, they love the way he plays. Jokic’s passing makes the basketball nerds blush far more so than Giannis’ dunks or Embiid’s footwork. And for the basketball nerd, his success is their success; it confirms their basketball world view.
This dynamic cannot be understated – if there’s anything that the success of social media algorithms has shown us, it’s that people absolutely fucking love being right, and having their worldview confirmed to them over and over again. With each passing dominant Jokic performance, he further confirms that their basketball ethos is the right one. That’s why they love him, and advocate for him so strongly.
To put it simply: Jokic is most media members’ favorite player, and so, you move goalposts and make certain arguments that you wouldn’t make for other players, because that’s what you do for your favorite player. It’s an incredibly human thing.
We all hold certain arbitrary and difficult-to-untangle biases in sports debates; pretending otherwise is stupid. For example, here are some of mine, which are pertinent to the MVP race:
I absolutely despise all-in-one advanced stats (more on that in a bit), which have obviously helped to bolster Jokic’s MVP case over the past few years. I have always brushed aside players who are doughy and slow, compared to those who are more traditionally athletic – I was too low on Luka Doncic coming out of the draft, I always found the Joe Ingles hype annoying, and clumsy ass Brook Lopez is my least favorite player in the NBA. And, most importantly, I watch Joel Embiid play basketball every night, and it’s hard for me to fathom that there’s another center who is significantly better than him; saying Jokic is better than Embiid goes against my basketball world view, and that’s where my anger comes from on this subject.
The problem isn’t the existence of biases – on either side. The problem is that the biases have overloaded in one direction; if there were a more even balance of biases – if the MVP voting demographic were more evenly split between basketball nerds and scoring/bag fanatics (often these are former players) – we’d see a more closely contested vote to represent what should have been two closely contested races, and you wouldn’t see the rewriting of unwritten MVP voting rules that have been around for decades, like voter fatigue.
The comfort of quantification
The final dynamic I’d like to discuss here is something I’ll refer to as the comfort of quantification, which ties into the desire to make everything empirical that I mentioned earlier, and explains the appeal of using the transparently absurd advanced stats that get thrown around so comfortably in the MVP discourse. For those who are unaware, Nikola Jokic generally grades out unbelievably well in advanced stats such as RAPTOR, BPM, VORP, etc. – some of these stats evaluate him as being better than prime LeBron or prime Jordan.
While most voters know to take them with a grain of salt, you will hear the majority of them at least make reference to them when outlining their case for Jokic. And ultimately, the appeal of all-in-one advanced stats is that they offer the comfort of being able to provide some sort of calculated, tangible proof for your position; without them, we would be forced to rely more upon our own internal evaluations of each player – it provides comfort at least knowing that, hey, this metric which was created with the purpose of telling us which players are better than others confirms my position.
In what many consider to be a tightly contested race, these metrics might be enough to push certain voters into the Jokic camp. Many will disagree, but I ultimately believe that most of these stats are transparently absurd, both in their formulation and in the results that they produce, and the fact that they play any role at all in the MVP discussion – deciding players’ legacies – is a travesty, and if we were to thoroughly dig into them, most would agree.
Let’s start on the surface, before digging into how they’re made. If I were to put out my own personal, subjective rankings of the top 50 NBA players, and they mirrored the rankings of FiveThirtyEight’s RAPTOR, I would be laughed at and called the single dumbest pundit on NBA Twitter. In the top 50 alone – which should be far and away the easiest group to rank – there are many comically out of place players.
Alex Caruso is 12th. Delon Wright is 18th. Josh Okogie and Derrick White are tied for 20th. Isaiah Joe is 32nd. Alec Burks is 41st. Austin Reaves and John Konchar are tied for 48th. I’m sure many people will call this cherry picking, but if 15-20% of the rankings that your metric spits out are obvious bullshit (spoiler: it’s more than 15-20%), then dare I say, your metric is not very good and should not receive widespread credence from NBA media members.
The first litmus test for these stats should always be common sense. If they directly conflict with common sense, it’s fine to disregard them.
Where these metrics are particularly bad is on the defensive end. RAPTOR, for example, has Jokic rated as the 3rd best defensive player in the NBA. That’s right – this guy is the 3rd best defensive basketball player alive.
Last season, Nate Silver of FiveThirtyEight put out a Twitter thread delving into how, exactly, the metric arrived at the conclusion that Jokic was an elite defender (last year, he ranked as RAPTOR’s second best defender in the NBA). The gist of the thread was this: he’s an excellent defensive rebounder, he contests a high volume of shots around the rim, he gets a fair number of steals, and the Nuggets are better with him on the court vs. off.
Still, overall, RAPTOR sees a lot of complementary evidence. Jokic gets a LOT of *valuable* rebounds, he defends a LOT of shots (at decent efficiency), gets decent numbers of steals; the Nuggets are pretty good defensively with him on the floor and very bad with him off it.
— Nate Silver (@NateSilver538) April 9, 2022
If you are going to sell me on the idea that Jokic is the second best defender in the NBA or anything close to it, you are going to have to present me with some novel information – something that I hadn’t been picking up on. The idea that this can be the basis of a case for a player being a DPOY-level defender is insane.
RAPTOR very obviously overvalues rebounding (which, by the way, one could argue is a separate phase of the game from defense – you can be a great defender and bad rebounder, and vice versa). It also isn’t an ironclad argument to say that a center defending a high volume of shot attempts is a good thing; there are obviously a considerable number of shots that guards may be willing to attempt at the rim versus Jokic, that they wouldn’t against, say Embiid or Rudy Gobert.
Ultimately, I think it has more to do with scheme than anything else, but this entire discussion underscores the point I’m trying to make: all of these metrics are built based upon how some person thinks several more basic stats should be weighted, and those considerations are highly debatable.
Box plus/minus (and VORP, which is based on BPM) provides another such example. BPM places different valuations of stats based on what position you play. For centers, an assist is worth nearly twice as much as it is to a guard. Defensive rebounding is also far more valued for centers, and missed shot attempts are penalized more for centers than it is for guards. In other words: BPM massively overvalues everything Jokic is good at. It inflates the value of his assists and rebounds, and rewards him for being hyper-efficient and not attempting a super high volume of shots.
The idea that centers are getting twice the credit for assists as guards do is backwards; I would argue that, generally, assists from centers are less valuable, because a considerable volume of them comes from dribble hand-offs, where the center simply dumps the ball to a guard, who then creates the shot. Regardless, the point, once again, is that all of these evaluations are debatable, and all we are really observing is someone’s evaluation.
Jokic also ranks first (FIRST!) in the league in defensive box plus/minus, because it’s based on box score stats, which means it heavily weights, you guessed it, rebounding and steals. Jokic also ranks seventh in the league in individual defensive rating, and ninth in defensive win shares.
Individual defensive rating, contrary to what some people assume, is not simply your team’s defensive rating when you are on the court. It is an estimate of how many points that particular player allowed per 100 possessions while on the court. Here is an excerpt from Basketball-Reference’s website on how it is calculated:
“The core of the Defensive Rating calculation is the concept of the individual Defensive Stop. Stops take into account the instances of a player ending an opposing possession that are tracked in the boxscore (blocks, steals, and defensive rebounds).”
Noticing a pattern here? All of these stats are heavily based upon steals and defensive rebounding. Defensive win shares is based heavily upon individual defensive rating. You can find the formula for its calculation here.
A common rebuttal I get from people who use these kinds of stats is something to the following effect: “if they’re all saying the same thing, that has to mean something.”
Yes, it does mean something – that all of these stats are heavily based upon one another and clearly overvalue the same shit. They are essentially Collateralized Debt Obligations. That’s how you get a player who is very, very obviously a barely above average defender being considered one of the best defensive players in the entire league, because he gets a high amount of steals and is an excellent rebounder.
Before I go any further, I just want to stress the degree to which these metrics impact perception and voting. In an interview with NBA TV last year, Stan Van Gundy – who has 30 years of NBA experience and all the credibility in the world – said the following:
“If you look statistically, he leads [Giannis and Embiid] across the board, all the analytics categories, he’s number one, including defensive box score plus/minus and defensive win shares, he’s ahead of both of those guys. So, throw that defensive argument that people have been trying to use out the window.”
Throw it out the window? Because of THOSE two stats? The fact that someone as qualified as Stan can be throwing those metrics around is proof that these metrics have a much bigger impact than they should, and that they have somehow seeped into the discourse without a thorough enough evaluation of their merits.
It’s important to note that there are advanced metrics out there that do not have Jokic as an all-world defender; metrics like EPM and RPM have Jokic as a merely very good defender. Neither of those three metrics provide specific information on how they are calculated, but as I initially discussed with RAPTOR, all it takes is a simple combing through the rankings to find that these metrics absolutely do not pass the smell test. I won’t bore you with my particular gripes, but feel free to comb through them for yourselves.
When it first came out years ago, PIPM at least provided data on how it is calculated; I found it to be interesting, and in theory, the attempt to “luck-adjust” team and opponent shooting output is a step in the right direction. The problem, however, is that in the absence of tracking data, they are forced into adjusting everything to league average, rather than the expected outcome based on in-game probabilities.
For example, the Sixers were excellent at defending the 3-point line for years under Brett Brown for schematic reasons. They provided very little help in the gap, played deep drop coverage with their bigs, and tried to force everything into a mid-range jumper or an attempt at the rim against Embiid. Are you going to punish that team for a schematic decision, which by the way, generally led to good outcomes?
Along that line, they don’t luck adjust mid-range or at-rim shooting. While that’s reasonable in theory, it highlights the fact that an entire team could theoretically be excessively rewarded for the schematic decision to ignore defending the 3-point line at the expense of being better at defending the rim. As I mentioned with the Sixers, they could have certainly played a different defensive scheme, sold out to defend the rim, and let teams bomb away from 3-point range – and with any “luck-adjusted” stat, they would get all of the credit for defending the rim, with none of the punishment for letting teams shoot open 3s, because the metric would “luck adjust” them back to league-average output from 3.
I’m aware of studies that were done 5-10 years ago that indicated strongly that opponent 3-point shooting is strongly tied to random variance. While it clearly still is to a strong degree, what I’m not sure about is whether the league and teams’ defensive strategies have evolved in a way that makes it less reasonable to simply adjust everyone to league average. Over the past few years, the Raptors, Heat, and Bucks have all made schematic choices to allow a high volume of quality 3-point looks, and it’s possible that luck adjusting those teams could allow those players to be overrated.
This, of course, is before even making mention of the fact that PIPM, EPM, etc. are forced into making the same types of box score valuations as BPM does, but we simply don’t know what they are. It’s likely that they are better, given that the outputs seem more in line with common sense, but we simply don’t know.
All of this is to say the following: whenever you use these stats, all you are observing is someone else’s valuation of a combination of various stats. It does not come from some unknown source of magical statistical ingredients that we have no access to. They take the stats that we have public access to, weigh them based on their own opinions (often in ways that any reasonable person could poke obvious holes in), and spit them back out. For all the work that goes into these stats, they likely don’t give us a considerably better idea of who the best players in the NBA are than Points Per freaking Game does.
Last season, I heard a ton of MVP voters voters saying some iteration of the following:
“This was an unbelievably close race! Incredibly close! You could make a great case for any of these three guys! But ultimately I voted for Jokic.”
And yet, 77 out of 100 first place votes went to Jokic. If it was truly that close – and it should have been – there would have been a more even split, and the hive mind wouldn’t have converged on the same “close race, but it’s Jokic” conclusion. It’s hard to say definitively what exactly is putting its thumb on the scale – it is probably some degree of simply liking the guy and what his style represents, and the nakedly preposterous advanced stats likely have an impact, as well. Regardless, if Jokic wins 3 straight (or more!), I can’t help but think that we will look back on this era of MVP voting and say that we got it just a smidge wrong. In an era with this many great players, I can’t help but feel like Nikola Jokic just doesn’t feel like a 3-peat MVP winner. The causes aren’t anything nefarious, and I find the whole thing fascinating, but in truth, my desire to poke holes in it comes from the fact that it just goes against my basketball world view.