Hey! Like the article? ← Do us a solid and give it a share. Everyone wins, especially you!
It’s still summer and the only thing going on for hoops junkies right now is the exhibition circuit leading up to the FIBA World Cup (never have I been this excited to watch Puerto Ricans play basketball). The vast majority of American/Canadian basketball media have spent a countless amount of virtual ink discussing who will make Team USA’s final 12, like it’s the most important debate since Magic or Bird.
Like any good back-and-forth about which basketball player one should choose over another, there are a wide variety of dissenting opinions – some reasonable, others insane. Personally, I’m a big fan of following the cold hard numerical facts in these situations, because at least that bias is based on, ya know, quantifiable evidence.
There are plenty of people that turn their noses up at the practice of using analytics to settle such debates, usually because it goes against the narrative they’re irreversibly all-in on for some emotional reason. Such people can now turn their attention elsewhere, as they want nothing to do with what I’m about to break down, fun and pointless as it might be.
Still with me? Good on you.
Why Use Analytics to Settle Debates?
In the past, I’ve explored using an all-encompassing metric to choose the East and West All-Star teams. I find the heavy bias and favoritism in the fan and coach voting silly at times and I hate the way it skews how we view a player’s career. Using analytics instead, to me, is a reasonable solution to the problem.
The most recent example of how egregious voting can be that comes to mind is Kyle Lowry’s season last year, in which I believe he easily deserved All-Star and All-NBA consideration, but missed out on both due to his shaky reputation (only reasonable opinion I’ve heard on the subject). Because of this, a guy who was a top-ten statistical standout last year according to the vast majority of metrics (and the emotional leader on a scrappy, no-one-believed-in-us playoff team if you absolutely need the narrative fuel) will always have his career year overlooked. That’s why I put more faith in carefully crafted and well thought out numbers to decide things of this nature than the whims of voters who may or may not deserve their say in the matter.
Letting nERDs Pick Team USA
All that digressing was simply to say that I was curious to see what Team USA would look like if we stripped all the emotional biases, discussions of role, and where these guys went to school, and based the decision solely on something measurable.
Since I’m a company man, I've decided to use numberFire’s own nERD metric to make the selections. A player’s nERD score is an all-encompassing number - developed from per-possession efficiency stats - that denotes how many wins above or below .500 a team would be with said player as one of its starters. Reasonably, a team filled with high performers in nERD would be better than one with lower ones (case in point, 12 of 14 Spurs had nERDs above zero last season, while only 2 of 17 Sixers did).
Making up the team with just the top 12 nERD scores wouldn’t quite work position-wise, so I decided to follow the All-Star voting process a bit. I picked the top player at each position for a starting lineup, picked a backup for each of them, and then finally chose two “wild cards” to fill out the squad. For the positions, I used the generally accepted ones where obvious and defaulted to basketball-reference.com’s position estimates where it wasn’t (for example, Gordon Hayward plays more small forward than shooting guard). Finally, I melded power forward and center together, since there is only one true power forward left in camp (Kenneth Faried).
Also worth noting, since Derrick Rose only played 10 games last season and his -1.9 nERD is not very representative of who he is as a player, I used his last “healthy” season of 2011-12 - when he played 39 games - for the exercise. Considering reports from camp have ranged from “Wow, he’s so healthy and explosive” to “Oh no, he’s sore,” I figured that choosing the campaign where he was kind of healthy but not really seemed right.
Here’s everyone who has been part of the player pool, ranked by nERD. This is based on the 21 players that have been in camp, ignoring cuts, injuries, and withdrawals.
First of all, a starting frontcourt of Anthony Davis and Andre Drummond is very unrealistic, considering Drummond is only 21 and probably won’t make this team just yet, but darn if that isn’t a scary/drool-worthy combination. Otherwise, it’s kind of hard to find too much fault in this as a balanced squad. Plumlee surprisingly makes it, so maybe Coach K is onto something with him, although I personally prefer Faried there (missed out by only 0.1). It’s interesting to note that using this method could be a way of justifying the initial cuts of John Wall, Paul Millsap, and Bradley Beal.
Current Team Selected by nERD
PG: Stephen Curry
SG: James Harden
SF: DeMar DeRozan
PF: Anthony Davis
C: Andre Drummond
In this iteration, Faried and Chandler Parsons step in to replace the injured Paul George and the recently withdrawn Kevin Durant, while everything else remains the same. That means that the cuts from the 16 players left in the pool at this point would be Kyrie Irving, Klay Thompson, Rudy Gay, and Gordon Hayward. Again, I have no issue with the choices made by the numbers and my arbitrary selection system here (natch). Irving and Damian Lillard have somewhat redundant skill sets, as do Klay and Kyle Korver. Rudy Gay was a late addition to even out practice squads after Durant pulled out and Hayward is rumored to be firmly on the chopping block. For those reasons, every one of those guys being omitted from the final roster would make sense.
Way to go, math.
This method is pretty limited, of course, because it doesn't account for team needs, chemistry, attitudes, age, affiliations, or - most importantly - the thoughts and desires of Team USA's coaching staff. Of course, that last one is something fundamentally missing from 99% of the arguments on the matter. None of us knows exactly what Coaches Krzyzewski, Thibodeau, etc. are going to do. It certainly doesn't make debating it any less fun, whether we're right or wrong.
With that in mind, now that you know what our algorithms say, what's your opinion?