Failing Up

In light of the [insert any superlative you want here] trade in NBA history, I’ve been bombarded with “what the hell happened?” messages. So, I figured I’d write a quick primer on what the hell did, in fact, happen. [1]

To start, we need to establish just how good Luka Dončić (25) and Anthony Davis (31) are.

DARKO and EPM are the two most respected publicly available player projection models. The easiest way for someone unfamiliar with one-number metrics to gauge their credibility is to simply look at the top players in each and see if those rankings pass the smell test. That doesn’t mean these metrics are infallible, but it strongly suggests they’re on the right track.

So, we’ll go with this: Luka Dončić is approximately a +6 player, and Anthony Davis is approximately a +4 player. That seems fair. In plain English, Doncic is a perennial MVP candidate, and Davis is a perennial All-NBA player.

Since Dončić and Davis are in very different phases of their careers, and DARKO projections show career trajectory, let’s take a look at an aging curve. This one is from over a decade ago, but the rough shape is all that matters.

As you can see, Luka is a +6 player who might still be improving, while Davis is a +4 player who’s probably just about to fall off a cliff.

Now, let’s take a quick intermission for the oopsies portion of the program to address the elephant in the room: the guy who’s about to get obliterated for probably the worst trade since Trae Young and a first were dealt for Luka Dončić (before Nico Harrison was with the Mavs).

You’ve probably seen this quip about how bad Nico Harrison was at his job at Nike. I mean, it’s on Wikipedia, so it must be true:

“During his work at Nike, he botched a 2013 presentation to Stephen Curry, where according to Harrison he may have called him Seth, and the presentation used was made for Kevin Durant. This, along with not offering Curry a signature shoe, caused the superstar to switch from Nike and sign with Under Armour.[4]”

The name mix-up is one thing, but using KD’s presentation? That seems like an immediate firing offense, and the fact that he somehow ended up running an NBA franchise makes me think he knows where the bodies are buried. I know life operates more on cronyism than meritocracy, but even that seems like a bridge too far.

OK, back to our regularly scheduled “analysis.”

How can we quantify the difference between a +6 and +4 player over the next few years? Well, it’s about the same as the difference between a good starter like De’Aaron Fox and a league-average starter like Coby White. In the NBA, it’s nearly impossible to make up that kind of surplus value in one max salary slot, which is why players of this caliber are essentially never traded. Instead, they’re typically held onto far past their prime. Then overpaid. Then they get a statue outside the arena. [2]

In the interest of full disclosure, I should mention that an unprotected 2029 first-round pick is heading from Los Angeles to Dallas in the deal. That pick could be valuable, given how awful the Lakers’ roster is outside of Luka and LeBron (40!). But realistically, Luka Dončić, cap space, and sunshine should be good for 45 wins a year for the foreseeable future. And while there isn’t much room for the Lakers to improve outside of this trade, Anthony Davis, cap space, and sunshine was more like a 35-win team.

So that’s it. A story of failing upwards—for both Nico Harrison and the Los Angeles Lakers.

[1] Spoiler alert: there was probably some crazy shit that went down behind the scenes, possibly related to sweet tea consumption, but since this blog doesn’t know anyone in the Mavs or Lakers organization, I’ll have to deal with exclusively publicly available information.

[2]

2025 Win Projections

This is the third year I’ve posted my win projections publicly. Last year, my delta was 7.1, while Vegas had a delta of 6.4. The previous year, I had a delta of 5.9 compared to Vegas’s 6.3. So, we’re about even? I will continue betting on the delta of my projections against anything publicly available, including Vegas.

Vegas is much better this year. Last year, I was off by an average of 2.7 wins per team; this year, it’s only 1.6! Even though I "lost" last year, I think Vegas’s improvement is pretty incredible. I know the limits are so low that no one is betting more than a used Toyota on season win totals… but still. Markets aren’t efficient, but they’re improving quickly, and it’s important to have a healthy respect for them [1].

To be honest, there weren’t many surprises this year. One thing I found interesting is that some of my best bets from last year are also some of my best bets this year—but in the opposite direction! The Kings and the Pistons were among my strongest unders last year, and this year they’re some of my strongest overs. Meanwhile, the Raptors were one of my top overs last year, but this year, their lack of depth makes them one of my best unders.

But really, this exercise isn’t about beating the market. It’s about analyzing teams’ various strengths and weaknesses—where their depth falls off, which players might be poised for a breakout year, and who might be nearing a decline. I’ll share some data from my process that highlights overall weaknesses. The chart below shows three fairly simple but illuminating metrics. This is a team’s age, creation, and spacing weighted by projected minutes [2]. I didn’t project these metrics with an aging curve, but they still provide insight into where a team might be deficient [3].

One of the first takeaways is that contenders like Philadelphia and Milwaukee are relying on extremely old supporting casts. That almost never ends well. This’ll be the third year in a row that I’m betting on the geriatric Bucks to go under [4].

The usage calculation also tells a few stories. Teams like the Raptors, Thunder, and Magic have prioritized defensive versatility and transition (to varying degrees of success!), but they may lack half-court creation when the game slows down. And, of course, the 76ers will have plenty of shot creation due to the addition of Paul George.

The spacing metric highlights some questionable team-building decisions by the Raptors and Lakers [5], who have suspect spacing on the wing and non-shooting bigs. Meanwhile, it praises teams like the Celtics and Knicks, who have spacing at the 5, and the Mavericks and Cavs, whose guards can create unassisted 3s.

To be clear, these stats aren’t terribly important without context, and internal improvement will shore up many of these perceived issues. However, teams need a sufficient balance of various aspects of the game, and creation and spacing are two of the most noteworthy.

Without further ado, let us NBA.

Atlantic		Central		Southeast
Celtics	58.8	Cavaliers	51.8	Magic	48
Knicks	52.2	Pacers	49.1	Heat	40.6
76ers	50.5	Bucks	47	Hawks	37.4
Raptors	27.2	Bulls	29.8	Hornets	29.7
Nets	19.8	Pistons	29.4	Wizards	18.7
Northwest		Pacific		Southwest
Thunder	59.6	Kings	48.5	Mavericks	52.8
Twolves	50.6	Suns	44.8	Grizzles	47.5
Nuggets	49.2	Warriors	42.6	Pelicans	45.6
Jazz	25.3	Lakers	39.2	Rockets	43.7
Trailblazers	21.7	Clippers	34.4	Spurs	34.7

[1] Unless the election is already over.

[2] Creation is USG%, spacing is 3%*volume.

[3] And rookies never really provide spacing anyway.

[4]

[5] As you can probably tell, I’m trying to avoid commenting on the core trying to sag for Flagg.

2024 NBA Draft: A Shot in the Dark

As I’ve written before, the NBA draft is a high variance affair. It’s not that analyzing it is a waste of time (though there are likely quite a few inefficiencies), it’s that the end result is destined to be looked at with such strong hindsight bias that entire careers are made (or ruined!) on outlier outcomes.

The strength of my “analysis” is not so much in my ability to scout, but my ability to project how important that type of player is even if he hits various outcomes. For example, with someone that has a huge range of outcomes like Rob Dillingham, I recognize that he is likely to be a fairly bad (and overrated) NBA player. But Dillingham’s upside outcomes [1] are very exciting to me. He has enough shake and foot speed that I can envision high end outcomes being similar to Trae Young, with even some upside to be a better shooter than Trae (who is awesome, but not as wet as you think). That doesn’t mean I’d want to draft him, but thankfully since I’m just here to play armchair quarterback, I care more about completing the exercise for posterity than actually drafting the player for my team.

So I’m writing a little something publicly on the record this year since I spent a few dozen hours looking at the draft. The most notable thing to remember about this draft is that without top end talent third-tier prospects are much closer in quality to fourth-tier prospects than first-tier prospects would be. And there are no first-tier prospects in this draft. And there may not be any second-tier prospects, either. The result of that should be fairly obvious but bears repeating: the top of this draft is going to wind up looking very silly, very often. I think once you get to the back end of the lottery you wind up with a pretty solid class, but I haven’t scouted enough of the second round to really comment on the depth of it. So, congratudolences on binking the first pick, Atlanta!

Instead of doing a full writeup on the entire draft where I paraphrase the same shit that everyone else is on the internet, I’m going to have GPT4o summarize each prospect and give some color on where I might have something interesting to say or disagree with the consensus. Within tiers, my rankings are randomized (so you haven’t lost the ability to count, I’ve just averaged the numbers next to player within that tier).

1. Donovan Clingan (7'2 C, UConn, ESPN: 3) ScoutGPT says: “A big presence in the paint with strong defensive skills. Questions remain about his ability to keep up with the NBA's pace, but his size and shot-blocking are impressive.” To me this is clearly the top prospect in the class. Due to the volume of other quality mid to late lottery prospects, chances are a couple of guys will eventually surpass him as NBA players. I certainly have no idea which ones, so I’ll stick with the two time NCAA champ who was a dominating college defensive player and has enough passing to make decisions in the short roll and enough touch to punish mismatches.

Upside: Rudy Gobert

Downside: Ivica Zubac

Probably: Jarrett Allen

Verdict: Buying

2. Ron Holland (6'7 F, G League Ignite, ESPN: 13) ScoutGPT says: “An energetic and versatile defender with a developing offensive game. Excellent at scoring in transition and making timely cuts​.” I expected not to like Holland based on his poor outside shooting and high turnovers. But after watching him play, I think his passing is pretty good and he's willing. He's got a great motor on defense and he shows enough event plays with good enough tools that I’m interested. And his worst skill, shooting, is one of the highest variance things to project, and his form looks okay to me. He would benefit greatly from getting into a good system where he’s not asked to do much primary creation early on.

Upside: Jaylen Brown

Downside: Josh Jackson

Probably: Tyreke Evans

Verdict: Buying

4. Stephon Castle (6'6 G, UConn, ESPN: 6)

ScoutGPT says: “Big guard with good size and physicality. Versatile defender, confident ball-handler, and patient playmaker. Needs to improve his shooting.”

With the best nickname in the class, Bouncy is a fun prospect as the guard version of Andre Iguodala.

Upside: Jrue Holiday

Downside: Josh Green

Probably: Josh Richardson

Verdict: Holding

4. Rob Dillingham (6'1 G, Kentucky, ESPN: 8)

ScoutGPT says: “A dynamic scorer with elite ball-handling skills and creative shot-making ability. His size is a concern defensively, but his offensive talent is undeniable​​.”

I typically hate guys like this, because they often get overpaid (Herro, Poole, Simons), but his footwork and hand eye coordination are incredible. Most likely outcome is a bench gunner, but his high end outcomes would have him as one of the best dozen or so offensive players in the league. It’s troublesome that he is guaranteed to be one of the 10 worst defenders in the league, but guard defense isn’t nearly as important as other positions. While everyone would love to have a team with as much defensive versatility as the Boston Celtics, the reality is that’s a sacrifice most teams would be willing to make for high level creation.

Upside: Trae Young

Downside: Bones Hyland

Probably: Brandon Jennings

Verdict: Holding

4. Reed Sheppard (6'2 G, Kentucky, ESPN: 4)

ScoutGPT says: “A reliable two-way player with excellent shooting and playmaking abilities. Defensively tenacious despite his size​.”

Sheppard is quite an interesting prospect, but one thing I think the market has wrong is that he’s “defensively tenacious despite his size.” He’s got good IQ, but he is a horrific defender. He's small and he runs smack dab into the middle of every screen. He does have good steal and block numbers, but by my estimation most of that came as a result of gambling.

Upside: CJ McCollum

Downside: Seth Curry

Probably: Jason Terry

Verdict: Holding

6.5 Matas Buzelis (6'9 SF/PF, G League Ignite, ESPN: 5)

ScoutGPT says: “Versatile offensive player with good size and athleticism. Effective scorer with a smooth 3-point stroke, but needs to improve consistency and ball-handling​.”

My doppleganger, I have similar thoughts on Buzelis as just about any of these mystery box forwards: NBA sized player who gives good effort and nothing is broken, roll the dice!

Upside: Chandler Parsons

Downside: Matas Bustzelis

Probably: Kyle Anderson

Verdict: Holding

6.5 Alexandre Sarr (7' C, Perth Wildcats, ESPN: 2)

ScoutGPT says: “A strong defender and shot-blocker with good hands and physical play on both ends. His offense is developing, but his defensive skills are NBA-ready​​.”

Sarr strikes me as a pretty good prospect as a garbage big man, but he seems to have aspirations to be much more than that, and his offense is very far away.

Upside: Bam Adebayo

Downside: Jared Jeffries

Probably: Zach Collins

Verdict: Selling

8.5 Tidjane Salaun (6'9 SF/PF, Cholet, ESPN: 9)

ScoutGPT says: “Athletic and versatile forward. Solid defender with potential to become a 3-and-D player. Needs to develop consistency in his shooting and add strength​​.”

When I watch Salaun play defense I can see the outlines of a guy who knows how to move at the NBA level. The shot isn’t broken and he’s large. We’ve kind of reached the portion of the program where this type of mystery box is a worthwhile pick.

Upside: Rashard Lewis

Downside: Sekou Doumbouya

Probably: Naz Reid

Verdict: Buying

8.5 Nikola Topic (6'6 G, Crvena Zvezda, ESPN: 10)

ScoutGPT says: “An elite passer and pick-and-roll maestro who can score in the paint. Not the quickest guard but plays with a calm, composed style​.”

A fascinating prospect who got to the rim relentlessly against weak competition and is probably the third best passing prospect I’ve ever scouted [2].

Upside: Lamelo Ball

Downside: Kendall Marshall

Probably: Greivis Vásquez

Verdict: Buying

10. Zach Edey (7'4 C, Purdue) ESPN: 16

ScoutGPT says: “A towering center with significant impact in the paint, but questions remain about his defense and speed at the NBA level​.”

Of course Zach Edey is too big and slow footed to play PNR defense in the NBA, of course, of course. BUT MAYBE the best player in college basketball in at least a decade, who also happens to be extremely good at the biggest barrier to entry in the NBA (size!), is the most obvious smash pick possible and the tail upside of having a very tall person with a little skill and toughness is good.

His college stats as a sophomore compare favorably with the 10th pick in the 2008 draft, Brook Lopez. As much as Edey struggles with foot and processing speed, I’m not sure that he’s worse than Lopez in those aspects, and he’s certainly bigger. Lopez has carved out an awesome NBA career as he perfected 2.9-ing [3] and became a serviceable 3 point shooter. As a good free throw shooter with a high release, Edey could, in theory, become a similar version of the same player. And with his physical advantages the sky might need to be the limit because if there was a ceiling he’d probably hit his head on it.

Upside: Brook Lopez

Downside: Boban Marjanović

Probably: Jonas Valančiūnas

Verdict: Buying

Here are a few others that I watched and would be interested in drafting: Devin Carter, Kyle Filipowski, Yves Missi, Isaiah Collier, Jamal Shead.

Finally, I’ll add a little color to a few other notables that are absent from my top 10.

Zaccharie Risacher (6'9 SF, JL Bourg, ESPN: 1) is an interesting prospect in the same way that any mid-first round prospect with good dimensions, good shooting, and very little experience contributing to high level winning basketball would be. He’s a pretty similar prospect to Salaun, I just happened to like Salaun’s size and the way he moves on defense a bit better. The fact he's spent so much time mocked at the first overall pick can only be attributed to Jonathen Givony selling his soul for a baguette. I don’t think he’ll fall out of the top 10 on draft night, but I wouldn’t be surprised if he slipped out of the top 5 despite all the smoke that he’s the favorite for the top pick.

Dalton Knecht (6'5 SG/SF, Tennessee, ESPN: 7) is an interesting prospect in the same way any late-first round prospect with good college production and very little upside due to being extremely old would be. The track record of old players being picked in the lottery is bad. I asked ChatGPT to come up with a list of the last 10 players who were at least 22 years old on draft night and drafted in the lottery:

Year	Player	Age	Pick
2022	Keegan Murray	22	4
2022	Ochai Agbaji	22	14
2021	Chris Duarte	24	13
2021	Davion Mitchell	22	9
2019	Cam Johnson	23	11
2016	Buddy Hield	23	6
2015	Frank Kaminsky	22	9
2014	Doug McDermott	22	11
2018	Mikal Bridges	22	10

It’s not like there aren’t NBA players on that list, and you might even suggest that you’re more likely to find someone who doesn’t fall out of the league if you draft someone old, but only Bridges has outperformed his draft position. And most importantly, there are zero All-Star games on the list, and the base rate of becoming an All-Star from the lottery should be 25%.

So yes, when it comes to politicians and NBA lottery picks I am ageist. From a macro prospective, I’d even argue that in a draft that’s weak at the top you should become even less risk averse. No gamble, no future, right?

[1] In this instance I’m not really worried about top 1% outcomes, which are too hard to wrap my head around, I’m looking at top 15% outcomes. [2] Josh Giddey is the first.

[3] 2.9-ing is staying in the paint as long as possible on defense while clearing your feet before defensive 3 seconds is called. Here’s an article from back in the day explaining how Brook Lopez has perfected that art to become one of the best defensive players in the NBA.