Deep Dives: Production By Position

The most visible and most often criticized decisions made by NBA coaches are the most fundamental: which five players on on the floor? While these choices are basic necessities, they are far from simple. Constructing a lineup is not simply picking the five best[1. “Best” in itself being a nebulous concept.] guys and letting them sort it out. There are harmonies and synergies of skills which need to be addressed, and fundamental roles than need to be filled for a unit to have a chance of success[2. Though we are getting closer, we still don’t quite have the statistical vocabulary to identify these lineup necessities. That said, the basic concepts are clear from moderately careful observation, with the requirements being things like ball-handling, creation and spacing ability on offense; rebounding, rim protection and ball pressure on defense.]. Picking players to fill out a lineup generally means assigning them to one of the broad and overlapping roles which are known more generally as positions.

Positionality remains a complex topic in the NBA with the emergence of scoring point guards and small ball fours, muddied further still by teams like the Warriors intentionally blurring those lines with five-man squads composed of similar interlocking and interchangeable players capable of filling multiple roles at once. So while in a precise sense, a term like “shooting guard” or “power forward” my not mean much, they can serve as useful shorthand for the particular bundle of skills brought to the table by each of the first, second, third, fourth and fifth largest player on the court at any one time.

A key part of winning basketball is putting players in the best roles and spots for their abilities[3. Especially from the standpoint of the players themselves, who get paid based on production.], maximizing their ability to succeed. While this maximization doesn’t always equate with gaudy individual numbers, there does tend to be more than a little overlap between efficient accumulation of statistics and aiding a team towards victory. However, the right position for each player can be an elusive thing. Sometimes even when the ideal role is known, it isn’t really available to for a team due to other needs. Despite the growing body of evidence that Zach LaVine[3. We’ll get back to him in a moment.] is best utilized as a “shooting guard” in a non-primary ball-handling role, if Ricky Rubio is out with injury, LaVine might be the best choice at Minnesota as a lead guard with Andre Miller being on the back nine and Tyus Jones not yet ready for the rigors of the NBA.

Still, being able to identify the spots and the floor and places in an offense where players can best contribute can go a long way towards making a team the oft-sought whole greater than the sum of the parts. Which brings us back to LaVine. In broad strokes, the allure of playing him at the point is understandable. Explosive athleticism, good handle, passable court vision. There are things to work with. Of course, those same things, along with his nice shooting stroke, could also contribute to a very successful off-guard, especially for a team in need of some deep shooting and ball-handling from the wing spots. Comparing LaVine’s production[4. Albeit in a somewhat limited sample at the two entering Tuesday night’s contest.] in each role reveals just how effective he has been when not asked to run an offense himself:

Add in the fact that initiating offensive sets for his teammates is a skill most charitably described as “work-in-progress” for LaVine and there can be little surprise that the Wolves are better served at present with him at the two. In fact, prior to Tuesday night’s loss to Orlando, Minnesota was 35 points per 48 minutes better with LaVine at shooting guard (+28.8) than with him at the point (-6.2). Both numbers are likely to regress towards zero as the sample size increases, but the starkness of the difference does serve to illustrate the point nicely.

Across the league, many players are having great success playing outside of their “standard” positions. In fact, among players with at least 100[1. LaVine at the two does not qualify, though his +6.8 DRE at shooting guard would place him comfortably between Durant and Griffin. Wait, what?] minutes played at a given position, seven of the top ten in terms of our DRE box score plus/minus stat have been accumulated from players “playing up” a position:

Many of these instances are easily identifiable as situations where the team context of a player is radically more favorable with them in a seemingly odd position. Green at center generally means Golden State’s lineup of death in on the floor. Bosh in the pivot has mostly been used with Miami’s blitzkrieging second unit, while Davis at the five means yes Ryan Anderson and no Omer Asik/Alexis Ajinca for the Pellies. But that’s kind of the point, as a player’s production in a position is a function of that lineup context.

This isn’t to suggest coaches should simply “sort by positional DRE” and go from there. Many of the units will have relatively low minute totals for good reasons. Golden State isn’t trying to repeat as champions in December, so asking Green and Harrison Barnes to bang with players 40+ pounds bigger for more than a few minutes a night might be unwise. Similarly, some of these gaudy offensive numbers can come at the expense of defensive efficiency[9. Which DRE as a box score metric doesn’t capture especially well.] in the usual (for non-Dubs) offense-for-defense tradeoff of going smaller. But knowing things like the fact that C.J. McCollum falls off a cliff when deputised into being Damian Lillard’s backup at the point, his True Shooting Percentage going from a robust 59.3 percent to a well-below average 47.0 percent while his turnovers skyrocket from 1.3 per 36 to 4.5 per 36 when moving from off to point guard, can only help Portland’s coaches and management as they assess the specific needs of their team going forward.

And finally, breaking production down by position allows for the construction of benchmarks. The table below shows the league average production from players at each position to this point in the season:

There is certainly much more to players’ contributions than what shows up in the box score, but the above at least provides a per minute baseline for comparison for determining “are we getting enough from the power forward slot?” and similar conundrums.

Individual per minute statlines have been added to the Positional Playing Time page of our stats section, please check them out.