clock menu more-arrow no yes

Filed under:

Players As Assets, Part 1: Who Are The NBA’s Most Valuable?

New, comments
Orlando Magic v Philadelphia 76ers Photo by Drew Hallowell/Getty Images

There’s something discomfiting about discussing NBA player value. Good roster-building strategy requires thinking in terms of an “asset” that consists of a player’s on-court contribution, his contract and salary cap effects, along with intangibles such as how hard he practices, how well he gets along with others, how much he cares about winning, etc. etc. Yet at the same time these are human beings, with thoughts, emotions, loves, losses, and everything else that comes along with our shared humanity.

I make this rather obvious point by way of apology. It is my intention to discuss the value of NBA assets. But there’s no avoiding that the most convenient way to talk about the single most valuable asset of the Denver Nuggets is to refer to that combination of play, contract, and intangibles as “Nikola Jokic” — this despite the fact that Nikola Jokic is, in actuality, a person rather than an asset. I actually toyed with referring to the person as Jokic and the asset as “Jokic” or *Jokic* or something, but it got too unwieldy. So I’m just calling the asset by the person’s name, and I hope that, with this introduction in hand, no one will misunderstand me. An NBA big shot recently pointed out to me that treating players exclusively as assets is even more problematic in the NBA than in other sports; a player like LeBron James sees himself, quite rightly, as every bit as much a mover and shaker as a coach or GM or owner. If someone like that gets even a whiff that he’s seen as merely a piece to be moved around the chessboard, well, it’s not going to enhance his willingness to stay with that organization or give them 100% effort.

OK, my throat is properly cleared! Now we can ask the question: what player/contract combinations are the most valuable in the NBA? All will be revealed, but first the ground rules:

  1. Value will be measured in dollars. A $10M player in 2018-19 is one who, as best we can tell, delivers $10M of value this season under the current salary cap rules.
  2. We will use the predictions of how good players will be, in dollars, for this season and future seasons, from CARMELO. If you think the CARMELO predictions are worthless, you can stop reading now, as this article will offer you nothing of interest. If you think CARMELO is useful but biased toward/against certain players or player types, you can adjust what you see here to unskew CARMELO. If you think CARMELO is not obviously biased but that it’s predictions are noisy and not always accurate because the world is hard to predict, well, welcome to the club, that is my view as well. We work with the tools available. Personally I think the predictors at 538 do a fantastic job creating CARMELO, and they are moreover kind enough to share their findings with the world. Maybe someone can do it better, but I sure can’t, so I can only offer them kudos!
  3. We will measure a player’s value as follows:
  • for each year they are under contract, take their CARMELO value and subtract the cost of the contract; then sum the years. So if a player is under contract for $17M this year and $18M next year, and his projected value is $23M and $27M, then his asset value is $6M + $9M or $15M for those two years.
  • if a team or player has option rights, assume these will be exercised if and only if doing so is optimal. So if a team has an option for an extra year at $$10M and the player is forecast to be worth $14M, the option adds $4M to the player’s value. I should note that this significantly undervalues the options. In real life, options can be exceptionally valuable because of the uncertainty in the value of the underlying asset. For example, say that a player is expected to be worth $16M for one year, and the team has the option to sign him for $15M, a million less. In our reckoning the option is worth that $1M. But the real truth is closer to: the player is 1/3 to be worth (in anticipated value at the moment the option keep-or-drop decision is made) $6M, 1/3 to be worth $16M, and 1/3 to be worth $26M. So the option is 1/3 to be worth zero (if his value is low the option won’t be exercised), 1/3 to be worth $1M, and 1/3 to be worth $11M. On average the option is worth $4M, not $1M. We ignore this subtlety because taking account of it is complex. Indeed the research that developed the “Black-Scholes-Merton” model for estimating such option values won a Nobel Prize in Economics for its aforementioned inventors. If you’re a young person and figuring out problems like that sounds interesting to you, consider a career in the lucrative and fascinating world of finance!
  • If a team has restricted free agent rights for a player who is predicted to be worth more than the max such players can be offered, we assume the team will keep him and pay the max.
  • If a team has RFA rights for a player who will in expectation be worth less than the max but more than the one-year Qualifying Offer, we assume he will be offered, and accept, the QO. Of course in actuality this rarely happens. But the obvious alternative is to assume the player signs an RFA deal for exactly his fair value, which would make him worth zero to his current team whether they match or not. And that’s clearly wrong; the RFA matching rights certainly have some value. But it’s hard to say how much; we’d need to guess that players are offered, say, 80% of their fair value or something. We went with the QO approach as an approximation of the RFA matching-rights value; if you have a better idea that you think will make a meaningful difference, let us know in comments.
  • We assume that unrestricted free agents will be signed for their fair value, or for the max if their value is above the max. If they are below the max, they are a break-even so it doesn’t matter whether they sign for their current team or not. For above-max players, we assume they are 50% likely to re-sign with their current team.

Finally, big props to my friend Andy, who did the analysis for this and the upcoming additional articles that exploit this data.

OK, ready for the list? Here it is:

  • Ben Simmons: $381M
  • Jayson Tatum: $291M
  • Giannis: $257M
  • Donovan Mitchell: $246M
  • Lonzo Ball: $231M
  • Karl-Anthony Towns: $199M
  • Nikola Jokic: $198M
  • James Harden: $186M
  • Russell Westbrook: $170M
  • OG Anunoby: $124M
  • LeBron James: $119M
  • Anthony Davis: $107M
  • Victor Oladipo: $100M
  • Robert Covington: $99M
  • Dejounte Murray: $91M
  • Kyle Anderson: $89M

The top ranks are dominated by two kinds of players: giant superstars, and very promising players on their rookie deals. That shouldn’t come as a surprise, indeed if it had come out any other way, we’d know the whole approach was misguided. It will probably surprise some that the former group captures so much compared to the latter, but really, it has to be that way. Not that many players are worth a lot more than $35M per year. And if you’re not, and you’re not on a rookie deal, you’re probably fully paid or close to it. If you are worth a lot more than the max, then you will probably finish pretty high, but... suppose you’re worth $50M/year, and have one year left on your current deal, and are 29, and are earning $30M. Well, you’re worth $20M for this year, 50 - 30 = 20. Then you’ll sign a $35M max deal for 4 years, let’s say. You’re past peak so maybe your value for those seasons goes 45-40-35-30. You are then worth the sum of those, which is $170M, minus the sum of the contract, which for simplicity let’s say is 35x4 or $140 million. So that’s +30M, except your team is only 50-50 to keep you, so the value of that to your current team is $15M. Add this to the $20M for this season, and you get $35M of total value, making you around #50 on the list even though you, at $50M of value, are probably a borderline top-10 player. Sensible, right?

Next: if your favorite player or prospect is not on the list, or is too low, or is below another player or prospect you like less, the likely culprits are:

  1. The guy you like better has a shorter contract, or more expensive, or with less team optionality. Jimmy Butler is a tremendous player, but with only this season left on his deal, and with him only perhaps 50-50 to stay with the team that trades for him, well, the net value of acquiring him is very high but not elite. If he were signed for 4 more years at $20M/year, it’d be a different story.
  2. The CARMELO model thinks the higher-ranked player has more value (or is more promising), or the lower-ranked less, than your view; in this case it’s quite possible you’re right and CARMELO is wrong; after all, it’s just a computer model. The guy who sticks out most on this list is Lonzo Ball, whom 538’s forecast method really likes a lot. Maybe the system is wrong about his future, and, if so, then he’s ranked too high.
  3. You are underestimating (or, perhaps, CARMELO is overestimating, but probably the former!) the expected impact of player age/experience on future performance. E.g. CARMELO expects meaningful decline in leBron’s play over the next four years, so even though he is still unbelievably great, and still underpaid at his max salary, he doesn’t make it into the top 10 in overall contract value.
  4. Something has happened in the past few weeks that either strongly affected player value, or that you think did so. These CARMELO forecasts were made before the season began. So if a player is off to a hot or cold start, his value may have changed, and you know about it but these forecasts don’t. The obvious example is the injured D. Murray of the Spurs; clearly his value is way down, which is why I included a 16th player in my Top 15. Note also that CARMELO forecast the rookies based solely on college/international play; if Luka is killing it already, then he may well deserve to be Top 15 right now, and there just wasn’t any way for the model to take that into account.

On the flip side, CARMELO doesn’t suffer from recency bias the way we fans do. After two games, I think many Sixer fans felt Ben Simmons was obviously on his way to becoming a top-5 NBA player a few years from now. Which, as you can see, is what CARMELO expected. Then his back tightened up and he had a 5-for-20 shooting night, and now I’m seeing hints in LB comments that perhaps we should look to trade him! As I write these words, he is destroying the Hawks, so maybe the recency-bias pendulum will swing back to pro-Ben after tonight.

Let’s talk more about Ben. Ben Simmons: get excited! The method says that Simmons is the most valuable asset in basketball, and by a substantial margin. By this method he is worth more than the rest of our team combined. I will do more posts based on this data, including one comparing teams and one ranking all the Sixers, but you can see from this list just how thoroughly he dominates. So, look, you can just decide you don’t like the CARMELO forecasts. But if you take it seriously as an objective measure, then Ben’s development is not only the most important thing to happen in Sixerland in any given week, month, or year. It’s, like, more important than the next 5 or 10 things combined. More important than Markelle and Landry and Dario and Zhaire, all combined. This is the nature of basketball — look at Cleveland, where losing a single player was the difference between a Finals berth and the current worst-in-the-NBA squad. That’s the kind of player the model thinks Ben Simmons is going to be. If it’s right, and if he’s happy in Philly, then the Sixers are going to be relevant for a decade or more. And whether Jonah Bolden steals minutes from Amir, whether Muscala is as good as Bjelica would have been, you name it, all that stuff is trivial by comparison.

Beyond that, I can’t say much. Clearly Ben cannot shoot at this point in his career. The 538 model, based on all the data it has about past players, thinks he’s going to be ridiculously amazing. Does the model reach that conclusion because it “thinks” he’ll eventually shoot? Or because it “figures” he’ll be that good without shooting? I don’t know! Is it possible the model, trained on historical data, doesn’t realize that being a non-shooter is so much more devastating today than it was in the past? Yes, that is possible — the folks at 538 are smart, but they are not omniscient; maybe they have an important error baked in. If so it should show up in their overrating of other non-shooters, so we can see if we think that’s a pattern. Anyway, my point here is just that I can’t answer any of that because it’s not my model. I report, you decide!

The next-most interesting thing for Sixer fans is the high placement of Mr. Lord Robert Covington. It won’t surprise anyone who’s read my work before to see that Rock is the second-most-valuable player/asset on the team. If that seems wrong to you, please read one of my eleventy-seven other articles on Cov. Short version for the uninitiated:

  • Defense is half the game, and Cov is a top-5 defensive wing, arguably the very best
  • Shooting is about half the other half, and Cov is a very effective shooter; more so even than his solid 3P% suggests because he shoots in high volume and when closely guarded, thus creating extra spacing compared to other shooters who are shorter or just less able to ignore a hand in their face.
  • The remaining quarter of the game includes dribbling, passing, rebounding, dunking, screen setting, and other stuff, and Cov is probably below the average starter at this package of skills, but not by all that much, and it’s only a quarter of the game, after all.

Upshot is, he’s terrific, he’s in his prime, he’s underpaid by a lot, and so he’s a top-15 asset in the NBA, and that’s not coming from Homer Randy, that’s coming from a computer model that’s never read a word I’ve written! Everyone who wants to include him in a trade package for anyone other than a few ultra-elites should reconsider. Dario, Chandler and Miami pick for Butler? Count me in! But not if Cov is in the package.

What about the other top value players?

The Class of ‘17

  • Jayson Tatum: $291M
  • Donovan Mitchell: $246M
  • Lonzo Ball: $231M
  • OG Anunoby: $124M

Yes, Markelle is not here, believe me, I noticed! Even if the #1 pick of the year never becomes anything, 2017 was a hell of a draft. And of course there are lots of others from this class who rank high but just not in the top 15.

What else? Well, Jason Tatum is #2 on the entire list, behind only Ben Simmons. Frustrating; we could have had him! My sense is that the number of Philadelphians who wanted us to take Tatum in the draft is far exceeded by the number who will write LB comments over the next few hours saying we should have taken him. But everyone’s an expert in retrospect, and that includes me: damn, I wish we’d taken Tatum! He’s going to have a great career playing for our arch-rivals. Mitchell is also clearly of exceptionally high value; like tatum, he’s already really good, so you get 3 more years of cheap, excellent play, followed by 4 more of likely-underpriced play, and even after that you’re 50-50 to keep him. Lonzo we discussed; he is a similar story to these guys, not as good right now but clearly 538 thinks his future is almost as bright. And then there’s OG; I was shocked when the Spurs didn’t demand him back in the Kawhi trade, but as you can see 538 thinks his contract, all by itself, is a more valuable asset than Kawhi’s, so I guess I understand why Toronto was able to hold firm without needing to bluff. Did I mention that OG went just a couple spots ahead of Anzejs Pasecniks? Couldn’t we have traded up just a few more slots somehow? I know, I know, hindsight!

Still-Improving Stars

  • Karl-Anthony Towns: $199M
  • Nikola Jokic: $198M
  • Victor Oladipo: $100M

This has Jokic and Townes in a virtual tie; personally, I am much more confident in the Joker. I grouped these three together based on where they are in their careers, but do note that the big men are shown with twice the value of Dipo. As much as I admire Victor’s game, I buy that for Jokic; he is perhaps the best offensive center in basketball, and while people say he’s a terrible defender the on-off stats do not support that view.

Megastars In Their Prime

  • Giannis: $257M
  • James Harden: $186M
  • Russell Westbrook: $170M
  • Anthony Davis: $107M

Giannis actually deserves his own category; megastars who are still young and improving. He’s probably not as good as Harden just yet, but he makes a lot less money, and he’ll probably catch up by next season, if not next week! Westbrook seems a little high to me given his massive paycheck; he’s a tremendous player but I would have thought he was pretty fully paid on his current arrangement, rather than $166M underpaid. Harden feels to me just about right. I’m guessing Davis is lower than would be expected because he misses so many games with injury.

The Old Guard

  • LeBron James $119M

Seems a little light to me — I’d rather have him and his contract than OG and his (admittedly far cheaper) deal! But you can see how it’s hard to make it way higher than this; even saying Lebron is worth $70M a year, which is almost the whole salary cap, that makes his value over the next four seasons about (70 of value - 35 of salary) x 4 = $140M; so 538’s $119M estimate is not exactly assuming the King is going to fall off a cliff, production-wise.

Pop’s Kids

  • Dejounte Murray: $91M
  • Kyle Anderson: $89M

A real bummer about Murray’s injury; terrific young defensive player. I still expect him to have a great career, but missing one cheap year plus likely being less than 100% next season cuts into his value a lot. As to Kyle Anderson, I conclude the following:

  1. He was a ridiculous bargain this offseason and the Sixers should have gone after him even though it would make it harder to add a max player. And of course San Antonio’s decision not to keep him is extremely puzzling to me. But that was true of their whole offseason, from adding Beli to trading Kawhi for DeRozan, whom I see as a fine basketball player but close to worthless as an asset due to his high pay.
  2. The high rankings of Anderson along with Murray, Ben, and I suppose Lonzo — plus the model’s long standing love of Marcus Smart — suggest that 538 really believes that young players who are good at everything but shooting are likely to develop into big stars. In particular the model seems to share my view that defense is half the game. I know this is not a widely held opinion, so feel free to take the position that all these defensive players aren’t really worth anything, and it’s just a big coincidence that their teams do well when they are on the floor and poorly when they’re off. I’ll keep making the case for Covington and Simmons and Kyle Anderson!

The final obvious topic for discussion is the players who didn’t make the top group. In most cases the reason is obvious; e.g. the player was injured last year or has only one contract year left or, like Kawhi, has both issues. Sometimes a player is overrated because he’s all-O, no-D, in other cases the player is just paid so much they don’t add much value above their comp. I already mentioned that we shouldn’t put much weight on the forecasts for current rookies like Ayton, Doncic, and Trae Young. Others, like Curry and Dame, are among the very next few on the list; I’ll do a post that goes deeper into the rankings soon; probably Part 3 or Part 4.

In Part 2: The Sixers! Whither Joel Embiid? Does CARMELO think Dario has real value? What about young guns Fultz, Shamet, and Smith? Coming soon! Meanwhile, I eagerly await your thoughts on the Top 15 in comments.