Nylon Calculus Weeks 1 and 2 in Review: Luka Doncic and win projections

After a late start to the season, I have finally joined the NBA season. And it’s just in time — the Warriors have gone bonkers, playing at some level of basketball not seen in this universe where Stephen Curry and Klay Thompson have tuned their precision down to the atomic level. Anthony Davis has been honing his gargantuan potential. The Western Conference itself has some upstart young teams high in the standings, like the Denver Nuggets, while stalwarts like the Rockets and Thunder are staring at the leaderboard while lying prostrate on the floor with ugly records.

The Eastern Conference will finally have a new victor, as the Cavaliers are being outscored by ungodly amounts. Then there’s the paradigm shift of the league with scores being taken from early 1990’s Nuggets games. This is a strange season so far, and there’s a lot to analyze. And I have not yet gotten to the LeBron James experience in Los Angeles, so let’s take a look back at the last week-and-a-half in basketball.

LeBron James in purple and gold

I should preface this section with my bias: as a lifelong Portland fan, the Los Angeles Lakers have always been the enemy, the “evil empire.” I cannot effectively remove this layer of perception. I’m looking through rose-tinted fan glasses, and I see the rivals gaining a superstar through little merit of their own. As a fan of the NBA in general, I do get concerned with power concentration and the general level of competitiveness. My issue then is that it appears the Lakers grabbed a generational superstar not because they had the best crop of promising youngsters or a home-grown core capable of winning a title but simply because of the prestige still remaining from their past and the related potential of cap space. A Los Angeles fan would whip up an argument based on league interest, the talents of Lonzo Ball, and other factors. I know I’m in a different basketball schema, but at least I’m sharing this weakness.

Regardless, I think that any objective NBA fan would realize this season for the Lakers is not one where they can be considered title contenders. It’s a waiting season. The players and fans can dream of being underdogs and cite all the upsets in NBA history, but realistically the primary goal should be developing the young core for 2020 and beyond.

Of course, the Lakers need to be attractive enough to grab more premier free agents by being competitive too. Remember, Paul George opted to stay in Oklahoma City, and Kawhi Leonard fled to Canada. Even though James has entered their basketball lives, the team can’t assume the next superstar will be just as easy.

The Western Conference is weak enough that the playoffs won’t be tough to crack, and there is some idea of synergy putting LeBron James, and all his passing talents, with a group of long-armed, athletic youngsters. But I’m afraid this will be a waste of his prime — and by the time his career is over, we’ll all realize what a special commodity that was.

Cluster roles

In case you missed it, here’s a very nice breakdown of offensive roles and how to define them in the NBA. Basically, Todd Whitehead used play type stats from the NBA with a clustering analysis package to organize players. Scoring shouldn’t be summarized into a single number: too much information is lost. For instance, one stat I’ve tried to emphasize is the proportion of assisted, or unassisted, shots. After all, if an assist has value, then consequently an assisted shot must have an associated inverse value. Being an efficient scorer isn’t the same if you’re a highly dependent scorer, and the article is a great way to show who is and in what ways.

The Eastern Conference power vacuum

The Cavaliers weren’t the Warriors of the East snatching up an unfair number of top seeds, but LeBron James’ bewildering finals appearance streak has dominated the East to the point where it warped Finals discussion talk. Now we can talk about the Raptors in the Finals without our tinfoil hats.

I’m not a fan of tiers — there are often very real differences between teams in the same tier and the lines between are often arbitrary where you can lose the valuable, granular information thinking only in tier terms — but I’ll use that here to frame my discussion. I think there are two mainstays at the top with the Raptors and Celtics, who have shown they can fly past 50 wins with ease, and then after that most projection systems have a large drop-off. The Celtics hit 55 wins despite key injuries, and while Kawhi Leonard is an injury concern himself the Raptors are deep enough and have already shown they can play well without DeMar DeRozan.

After those two teams, there’s a large middle class, but two teams have a lot more potential than the others. The 76ers, obviously, have Joel Embiid, whose injury history is well-known. With their youth and the potential of Ben Simmons and their role players, there is a looming shadow of a contender here. The Bucks are the other team on my radar, and so far they’re the only undefeated team left. I’m higher on them than the Pacers, a team whose projected record was similar, because they have a young superstar with a new, developing coaching system, while Indiana is a good candidate for a regression back down to earth after surprising everyone last season. Even though LeBron James left for the west, the east may finally be catching up — these teams could be excellent.

Everyone is scoring like it’s 1969

As everyone has noticed, scoring in the NBA is on some nearly unprecedented run for the modern league. This has been talked about before, so I’ll be brief. First of all, as a survivor of the 1999 season who suffered through some traumatic Knicks-Heat games where the final scores limped out of range of the Einstein-Bose condensate phase, I welcome these high-scoring binges. In fact, after being scolded by older generations about the failures of NBA offenses in the early 2000’s, this should be a welcome reprieve, and everyone complaining about these triple-digit games clearing 130 should look back at those dark days.

Lastly, while everyone is pointing to the rule changes or the ever-growing love of the 3-pointer, I think there’s one general cause here. If you think back to some of the basics of how field-goals are influenced, many factors are well-understood. So yes, it does matter who’s shooting them, how far away the basket is, and how close the defender is. There was also a sharp trend that followed the shot clock. As a possession wore on, the possession got less efficient. The shot clock itself isn’t defending the ball, of course; but there is some truth of the seven-seconds-or-less adage of shooting quickly before modern, smart NBA defenses equipped with long-armed terrors can set up. Plus, when you run down more of the shot clock, it’s just inviting more danger of turning the ball over. Maybe it’s counter-intuitive, but shooting early can be more efficient.

Are we sure about Boban Marjanovic?

This is Boban Marjanovic’s fourth year in the NBA, and he’s already 30-years-old. At this point, I feel like most people have come to a consensus on who he is and what he can provide for a team. He’s a paint-eating center who will get exposed on defense if you leave him out too long, like potato salad. But just like potato salad, I want to question his value. He has an unprecedented combination of usage, efficiency, and rebounding. In fact, you have to relax the filters significantly before you find players with similar career numbers. Even if you’re giving up points on the other end, at some point you’re impressive enough on offense you can balance the scale and provide some value.

Have we been underrating him? He’s an unorthodox player in the present-day NBA, and coaches can be risk-averse. It’s better to play it safe and use the commonly agreed upon “good” players rather than gamble on someone like Boban for long stretches of a game. I will admit that I don’t exactly believe he’s a secretly valuable player and there’s a lot more to the game than efficiency and rebounds. But I’d at least like to see him unleashed onto the NBA world before he gets any older because we will never truly know otherwise. Plus, he can dunk without jumping and while fading away.

https://twitter.com/NBA/status/1052781523576020993

Luka Doncic: The early returns

As a gift for Luka Doncic’s prodigious talents, he was drafted onto the Dallas Mavericks, and while they (again) have had issues with out-scoring opponents, they’re at least bereft of talent to the point where he’s been given full run and a long leash. That’s resulted in some decent basic stats which I’m sure will be used for some awards candidacy, like averaging about 19 points per game with a few assists and rebounds. But obviously, we can do better than that. How is the Slovenian Savage?

So far, Doncic hasn’t been electric as NBA player, but things do look rosier once you account for his age. He’s shooting often and he does appear to be capable of hitting long-range shots in the league. He’s rebounding, at least on defense, mightily for someone his size, and that’s a fairly stable stat. The darker side of the coin: he’s not getting to the rim often and consequently not getting a lot of free throws either, and unless your jump shots are pure fire it’s very hard to be efficient. Once you factor in his turnovers, and you can see why people are lukewarm on him already — but again, he’s still only 19-years-old.

Turning to the recordings, Doncic does not have the benefit of flinging open 3-pointers. He’s already taking some tough shots, and sometimes they go in and sometimes they do not — but at least the shot is there and he does appear confident and capable of making them. As a quick proxy for 3-point difficulty, he’s indeed one of the league leaders in pull-up 3-pointers. He’ll improve there. I’m more concerned with his ability to finish near the rim. The NBA does have higher caliber athletes and rim protectors. Note that 2-point percentage does usually go down significantly when players transfer from the Euroleague, and the free-throw rate follows. By the physical evidence, however, I don’t think he’s failing there. He does have the ability to drive to the rim even against good teams, and he can work off-the-ball and finish inside. But having the actual ability to convert those shots at a good clip is another matter entirely you can’t always glean from the video — we’ll need larger samples from games for that.

As for the other skills, we’ve already had highlight plays, like the pass in the video below where he completely fakes out a defender. The passing wizardry is expected, and his assist numbers are most definitely respectable for a rookie wing player on a bad team. So far, the Mavericks are playing better with him on the court, but those numbers early on are extremely noisy. Defensively, he is not a stopper you’d use on an all-star scorer, but he appears to have good awareness and instincts so he may be better than expected, like, for example, poking the ball out to a teammate after a missed ball. I went through a lot of video on defense for Doncic, and I didn’t see many egregious errors. He appears attentive for a rookie, and overall there’s a lot of promise. The turnovers are the biggest negative for the present, but, again, that’s pretty common for a young player. Let’s not forget how tough it is for anyone to come into the league and succeed right away. He’s on the right path, so let’s be patient.

Predicting the now: 2019 wins predictions

Every year, a group of fans try to predict the records of every NBA team before the season starts with quantitatively, increasingly advanced approaches. Overall, the results are competitive with the best prediction sources out there, and the top analysts usually do significantly better than major sites. It’s a well-known niche in some circles, and you can learn a lot from these numerical wizards.

You can see the ongoing contest here on the old forum. A few individuals have thrown in their own numbers compiled by whatever means necessary, and those are tabulated alongside a few well-known sources from gambling lines to ESPN. I decided to grab all those numbers and coalesce them into one graphic, including those major sources. Below you can see the set of 30 histograms for each team. The histogram was chosen to show the spread and the patterns of the various teams and their predicted win totals.

Since I fixed all the axes to the same limits, the spread and shape of the histograms are instructive. There’s a team like New Orleans with a narrow range of variation in their predicted win totals. In fact, 13 predictions were within two wins, which is downright bizarre and makes for a sharp peak. What’s curious is that, while there haven’t been many changes with the team, they will have a full season of Nikola Mirotic and a whole new point guard in Elfrid Payton. Yet the prognosticators still all agreed. The Grizzlies had a narrow range too even with some legitimate concerns about the durability of Mike Conley, who missed a large chunk of last season.

At the other end of the spectrum you had Minnesota, and any NBA fan paying attention understands why. The Jimmy Butler situation has been dicey. There’s no easy way to input his minutes for this season, and if Butler goes then the rest of the team surely changes too. The rumors surround the entire team too and not just Butler. For a more subtle but interesting spread, you have the Warriors. Given their talent, it’s easy to compute 70-plus wins, but due to a number of factors, many of which can be quantified, that can slide down into the upper-50’s. The big question, again, is minutes for their stars. Even if your metrics and adjustments are abnormally precise, if you don’t know how many minutes everyone plays, you won’t get the prediction correct.

In some ways, it’s remarkable how close some projections were for teams. For example, with the 76ers one could assume they’d have at least a modest amount of variation, and it’s easy to understand why. Joel Embiid’s health and his future, in general, have a wide reach. He could play heavy minutes and get invited to an all-NBA team, or he could play 20 games and receive another debilitating injury. Compounded is the fact that they’re still a young team in general and there’s a lot of variance in predicting, for example, Markelle Fultz.

Plus, I still think there’s a legitimate question in how good Joel Embiid is and what he means for the team. I know there’s a lot of data spouted out about the team’s numbers with and without him, but those are raw, unadjusted stats without the context for the starters or who he’s playing against. He’s valuable defensively, yes, but how refined is he and how close is that to a Defensive Player of the Year caliber player? And how valuable is his offense, given that he’s a ball dominant big man without a gift for passing? The analytics sage of ESPN Kevin Pelton projected them for 47.5 wins, while one person had them nearly at 60 wins. Yet nearly every other prediction was in the low 50’s. The low roster turnover explains most of that, but that range still seems too narrow to me.

As Nylon Calculus has developed, I hope there’s been at least one crucial lesson: never trust a single number. A win projection all averaged and even rounded can look tidy and convincing. But many modern projection systems use simulations of thousands of possibilities for a season, and there are so many assumptions, factored in or not, that any number given for the output is highly dependent on a few key choices. Paradoxically, to understand this world, we have to think of it in stochastic terms: the future is a varied mess of possibilities, and it’s our job to make sense of the maelstrom.