Nylon Calculus Week 17 in Review: Trade deadline fallout and historical metrics

BOSTON, MA - FEBRUARY 11: Lebron James
BOSTON, MA - FEBRUARY 11: Lebron James /

As we look back at the wreckage of the trade deadline, let’s try to learn from the disaster that became the Cavaliers season and how it formed. I am all for taking risks and swinging big. It makes the league more interesting, and you need some risk if you want a chance at a title. So let’s remember that Cleveland’s situation is most likely the negative extreme of what could have happened. That’s just life, and we can’t predict everything. We could see some surprising fallout from the recent dealings, but that doesn’t make our initial diagnoses incorrect. We should definitely try to learn from what happened, but let’s not over-correct either. And with that, let’s take a look back at the last week in basketball.

Cleveland makeover

As panic gripped Cleveland, the trade deadline provided the catalyst they needed to take the risk and rebuild their roster. In a matter of hours, most of their off-season wheeling and dealing was erased, and they brought in an entirely new lineup. Just as a reminder, they dealt Isaiah Thomas, Jae Crowder, Dwyane Wade, Derrick Rose, Iman Shumpert,  and Channing Frye for George Hill, Rodney Hood, Jordan Clarkson and Larry Nance Jr. We’ve already had many pieces produced that covered all the details of the trades and the ramifications, but I want to cover a few of the finer points I found curious.

First of all, I understand why the Cavaliers were forced to make deals like this. They’re afraid of losing LeBron James. They have to remain competitive or he’ll likely walk. That’s a death knell for the organization, so you can imagine how strong their motivations were. Smarter, more patient teams would have waited for the values of Jae Crowder and Isaiah Thomas to regress upwards — one dude was a coveted 3-and-D player and the other was an MVP candidate in many people’s eyes a year ago — but they didn’t have time.

Thomas’ case is particularly vexing. Everyone loved him and what he did last year; he was an NBA darling. And quite swiftly he’s this malignant tumor the Cavaliers had to excise if they ever wanted to play winning basketball again. People who evaluated the trade as fair for Cleveland, or more, now look silly, including myself. Did we not know enough about Isaiah Thomas’ injury? Perhaps, but if it’s more severe than we thought, should the Cavaliers have waited until he was more properly healed, if possible? And isn’t that damning for Boston, who had him play through an injury and sold junk goods for a younger model?

Then there’s Jae Crowder, who was undeniably a valuable piece of Boston’s success before last summer. Even if you thought his 3-point shooting would leave him, he should have still remained a good defender. It was no mirage either: the eye tests liked him, he had a good steal rate, and he was a large, versatile, athletic defender. As others have noted, his mother died before the season started, and with Kevin Love playing center for much of the season his duties were perhaps a bit outside of his range — though I’d note Boston used a lot of unconventional lineups too. This is the guy more people should have been talking about because there was no obvious injury problem. What happened to him, and can he recover?

The success of the trades will mainly stem from them riding the team of their net negative players. Dwyane Wade and Derrick Rose were terrible ideas from the beginning, and to the surprise of no one outside Cleveland, they failed there. They will miss Channing Frye though, as they team actually did play well with him. The core issues remains unresolved though: LeBron James should play power forward going forward, especially because he’s aging, and Kevin Love is not a viable option at center. The hand injury probably killed any possible trade. They definitely need a defensive center for the middle, as Tristan Thompson has gotten worse (physically, he has not been 100 percent.) The trades were a short-term fix for a team that feared losing their legend, and it’s not the kind of overhaul that’ll make them competitive against the Warriors or Rockets. When you are that afraid of your present, you can lose sight of your ultimate goals.

Elfrid Payton and his hair traded for a mere second-round pick

I thought I’d touch on this trade quickly. Phoenix, home of the worst point guard rotation in the league, traded a second-rounder for promising athletic point guard. And yet, I didn’t love this for the Suns. As I feared a while ago when Payton was a fun young prospect with a unique look and defensive chops, his outside shooting has put a low ceiling on his value. It’s one thing if your point guard has an iffy 3-point shot; it’s another if he can barely hit free throws. And I wouldn’t take too much stock in his 37 percent success rate from behind the arc this season. He’s at 30 percent for his career there, and 33 percent on long two-pointers. I hope, like everyone does, that he improves and makes a career for himself, but he’s like a polar bear in a warm world. There’s no place for him anymore.

Okay but how good are the Cavaliers now?

We can spend a lot of time hand-wringing over what the Cavs should or shouldn’t have done, but that’s already in the past. We got one great game out of them on Sunday, but let’s not set this in stone after one game. Who are they now? They have a new starting point guard, and it’s one I personally “endorsed” to address some of their issues. George Hill, at his best, is exactly who they need: he’s a defensive long-armed point guard who can cover multiple positions, is a great spot-up shooter who can, when needed, pick up a lot of ball-handling duties. It’s like he was tailor-made for a contender. The problem, of course, is that we don’t know is that George Hill still exists because of his ho-hum season in Sacramento. He could have a lingering health issue, or he could be in the grips of lottery team malaise. We’ll see.

Rodney Hood is another major rotation piece now. Hopefully he’ll replace J.R. Smith in the starting lineup, as Smith has been playing like a 13-year veteran who doesn’t have much left for the league. Hood is an injection of scoring and a little bit of playmaking. But based on who you ask, he’s either a negative or a promising young scorer. He showed a lot of promise in his sophomore year as a sweet-shooting lefty who could make the right pass. He was a beloved surprise back then. But there are concerns about his defense, and he doesn’t get to the rim a lot, depressing his efficiency. That he’ll still be an upgrade over their other pieces says more about them than how good he is.

The other big addition is Larry Nance Jr. He was a sweet late first-round pick years ago, and he’s the sort of glue guy who doesn’t do one thing really well but checks off a lot of boxes at once including some of the smaller things that don’t show up in traditional stats. He’s consistently played either frontcourt position, and he’s an athletic finisher who just barely doesn’t have 3-point range. Some metrics peg him as well above-average — he has a higher RPM than LaMarcus Aldridge and Blake Griffin and sits right behind Kevin Love and Kristaps Porzingis — which can actually be more impressive for a player of his type on a bad team. He could potentially look a lot better next to LeBron James.

The last piece is Jordan Clarkson, and I don’t think he’ll be a major addition. His biggest negative might be his contract. But what’s remarkable is how similar his stats are to Rodney Hood’s — except Clarkson passes more and gets to the rim more often. He’s another gunner, and Cleveland is hoping his efficiency will be raised on a better team. Overall, the Cavaliers still have a lot of holes, and their defense will definitely not be fixed — as Kevin Pelton noted, this is a lower upside team that has less of a chance to beat the Warriors. But if Hill and Nance can defend well with the other guys hitting 3-pointers from outside, it’s enough to fight through the playoffs to make it to the finals. And if that keeps LeBron around for a little while longer, it’ll all be worth it.

Emmanuel Mudiay

In a three-way trade, the Knicks-Mavericks-Nuggets played musical chairs with perimeter players. This was not a blockbuster trade, but I do want to focus on Emmanuel Mudiay, A long time ago on a continent far away, Mudiay was an intriguing prospect who opted to play in China over the NCAA and had the skill, athleticism, and size that tantalized draft-niks. Instead that shooting deficiency we saw overseas carried over to the NBA, and he’s a shot-happy guard who can’t shoot, can’t pass and has a boatload of turnovers — and his defense doesn’t cancel out those negatives.

The good news is that he’s still quite young. I know people are giving the Knicks around a Bgrade for this trade because they’re taking a flyer on a young guy who has some tools, but here’s my fear: they’ll ignore most of his weaknesses, they’ll hold onto him for a while, and they’ll play him over better players and kill a lot of their lineups. That seems like a very New York Knick-thing to do. Keep a pin on that for the future — we shouldn’t be encouraging the Knicks to go after guys like him.

Another week, another Team LeBron injury

I wanted Kemba Walker to make the team, but not like this, and not on dry land! He’s the injury replacement for Kristaps Porzingis. This is bittersweet. And it’s yet another new guy for LeBron James’ team. This is getting weird. Did someone put a hex on him? Did he dunk on a witch? Or did he make some deal with the devil that all his potential injuries would be transferred to his teammates? I’m getting concerned. Please no more major injuries for a while. (And, oddly, LeBron James appeared to get hurt early on in Sunday’s game, but of course he was okay, that cyborg.)

Kristaps Porzingis, a player too grand for our world

I know I should address the Porzingis injury, even though I don’t have the heart to, but it’s a major storyline. This is devastating news. He’s a special player because of his unique combination of size, skill, and athleticism. I don’t want him to lose any quickness at all; the ACL is a notoriously rough injury. But there is cause for optimism here. The sports medical world has gotten much better, and guys have been coming off major, supposedly career-altering injuries looking as fine as ever. Porzingis is also still quite young, and that’s a hugely important variable here; his body can still do a lot of complete, lasting healing.

The effects of this injury will permeate the Knicks, of course, and much of the league. They could be major sellers this offseason, and they could be yet another team looking for draft picks. Porzingis won’t be back until next year, and thus they’ll playoff odds could be doomed for two years now. What will they do? We all have concerns about the Knicks and their decision-making. But if Porzingis comes back healthy, they’ll have him, Frank Ntilikina, a lot of cap space, and better draft picks. They still have a future. They just need patience. But do we trust the Knicks?

Advanced stats in the pre-stats era

As we’re drawing to the close of this decade, I’m a bit disappointed with how little the basketball community has done with advanced stats for the past. Sure, there are some nifty statistics now and numbers for just about everything, but we don’t have much for the past, aside from one recent invention. Also, no offense to the people who created PER and Win Shares, but we should be moving past those; they were created a while ago and our understanding, and techniques, have improved greatly. But what more could be done?

Let me explain the challenge here. Creating a decent estimate for a player’s value with the full box-score feels impossible already. There’s so much of the game that’s not covered adequately, and the basic stats are blind to entire areas of the game, like a large swath of defense and other vital features, like screens or box-outs. Now imagine you don’t have steals and blocks to evaluate defense, and you only have total rebounds, not offensive and defensive. And on top of all that, you don’t have turnovers, not even on the team level. How exactly are you going to create a metric with that much missing information? The only thing you have for defense is the team’s defensive rating and fouls. That’s it.

That’s true for the time span of the 1952 season to 1973 — before that, you don’t have minutes played, and then rebounds disappear a year earlier. You still have the old standards of Win Shares and PER, however, but that’s a bit misleading. I think many people don’t understand those metrics are missing data like everyone else is, and they’re vastly different metrics because of it. Wilt Chamberlain’s “record” for PER in 1963 should not be compared to the best numbers after 1974; they’re on different scales.

We do have some tricks available though. Let’s start with an easy one. Usage rate is a popular stat, and some will note it’s not possible to calculate before turnovers were tracked. But we can calculate the same stat without turnovers, and in fact in many situations “shot usage” is more relevant anyway. It’s just the proportion of your team’s own shots and free throws you’re taking. Then there are a couple of powerful interaction stats, which I discussed last week. Yes, they’re prone to overfitting, but they improve results and it’s not like there are better alternatives in the prehistoric NBA era. There’s the playmaker measure, which is an assist rate multiplied by shot usage — Basketball-Reference uses AST%, which is the percentage of teammates you assisted, and I use assists per 100 possessions. Finally there’s the “ball-boarder” measure, which is the square-root of assist rate multiplied by the total rebound rate.

Those are some powerful statistics that boost versatile players who create for others and themselves. But what else can we do? We’ve got to get creative. For example, we can consider assist rate by free throws attempted per 100 possessions. This is the drive-and-kick measure, and you can think of it as a metric who create plays for others through penetration or, to a lesser extent, post-ups. Free throws are a rough proxy for drives or at the rim attempts. As you can see in the table below, the player with the top marks are all point guards or lead initiating stars on offense, like LeBron James. And the top players, of course, are Russell Westbrook and James Harden.

Table: 1952 to 2018

PlayerSeasonTeamFTA per 100 possessionsAST per 100 possessionsDrive-and-kick
Russell Westbrook2017OKC14.714.7216.5
James Harden2017HOU14.314.8211.8
Russell Westbrook2015OKC14.212.5178.2
Magic Johnson1991LAL10.017.2171.4
Magic Johnson1990LAL10.815.4166.4
Magic Johnson1987LAL10.315.9163.2
James Harden2018HOU13.212.3162.7
LeBron James2010CLE13.711.5157.8
Dwyane Wade2007MIA14.810.6157.1
Russell Westbrook2016OKC10.315.1155.8
Magic Johnson1989LAL9.416.4153.6
Chris Paul2009NOH9.515.7149.6
Kevin Johnson1990PHO10.214.5148.5
John Stockton1990UTA7.419.4143.8
Kevin Johnson1995PHO10.713.3142.6
Magic Johnson1988LAL9.015.8141.6
Kevin Johnson1991PHO10.613.4141.4
Spud Webb1986ATL10.713.2141.4
LeBron James2009CLE13.510.4140.4
Russell Westbrook2011OKC11.512.2139.8

The first big man who appears, by the way, is DeMarcus Cousins, so take that as you will. As I found with a similar methodology last week, the top values in the league have been increasing over time. They were actually fairly steady for a while, and then the 1980’s and Magic Johnson hit. After that, we saw another explosion after the league changes and the paradigm shift of the seven-seconds-or-less Suns in 2005, and the top marks have been rising ever since. Thus, while I would have concerns about the measure’s use over the time frame of the NBA’s entire history, but within the bounds of 1952 to 1973 that time series issue is not present.

There are other combinations you can test too, of course, like rebound rate multiplied by free throws, or even a player’s foul rate interacting with other variables — it’s not like there are a lot of defensive stats that can be used.One may need to become inventive while still guarding against overfitting. There are lot of issues with the data, not just limited to the small number of stats available. In the pre-1966 data, there are some players who have their totals available, but it’s not split by the team. Thus, it’s trickier to obtain some statistics because a team may be missing a player or two.

Next: The Encyclopedia of Modern Moves

My method will have many drawbacks, but there’s a lot to improve upon. Based on my initial findings, stats like PER and Win Shares do an awful job at predicting wins before 1974. There are still some issues to address, and there’s some data to clean, but it’s a low bar to clear. There’s an opportunity for a better metric for those players in the first years of the league, and with any luck and some elbow week that may be seen in a coming week.