Playing big and small: Visualizing effective height
By Ian Levy
Height has always been a valued commodity in the NBA. If a team was seeking a certain package of skills, the tallest player possessing them was usually the choice. However, for decades, teams were often willing to make sacrifices in the skill department to add an inch or two to their lineups. The success of the Golden State Warriors and their versatile small ball lineups is redefining how teams think about size and the way it affects combinations of offensive and defensive skills when constructing rotations.
Given that pure height is not valued in the way it once was, I thought it would be interesting to look at the actual effective height of each team. The interactive graph below shows the minute-weighted average height for each team at each position. It uses listed heights from player pages at Basketball-Reference and Seth Partnow’s player position estimates here at Nylon Calculus. Obviously, neither is a perfect data set but they are close enough to serve our purposes here.
[Ed. Note: Steve Fenn was nice enough to take my visualization and improve on it. What you see here is his version. Thanks Steve!]
The impact of a few individual players are quite obvious. For example, the minutes Isaiah Thomas — one of the shortest players in the league — plays at point guard for the Boston Celtics give them by far the shortest point guard rotation in the league. Draymond Green has the same effect on Golden State’s power forward rotation. In fact, if you filter the visualization to show just the Warriors we can really see the structure of their small ball lineups.
Golden State’s shooting guard, small forward, and power forward positions are really tightly clustered. There is only about an inch and a half’s difference between their average shooting guard height and average power forward height. This means that across three positions they are using players of similar height. The implication (which we largely know to be true from observation) is that they are of similar athletic ability as well. This is essentially a picture of positional interchangeability.
The graph above creates a visual comparison but for those who like some aggregation and numeric values, I also built the table below. It shows each team’s average minute-weighted height at each position compared to the league average. The last three columns group those position columns in backcourt (point guard and shooting guard), frontcourt (small forward, power forward, center), and the total for all five positions.
While we think of the Warriors as a small team, they’re actually only about a half-inch shorter than the league average. They have the smallest frontcourt in the league, narrowly edging out the Atlanta Hawks, but their backcourt is the third-largest. Stephen Curry and Klay Thompson are both big for their positions, but having Shaun Livingston come in to sop of some of those backcourt point guard minutes really pushes things to the extreme.
The smallest team in the league, in the aggregate, is the Boston Celtics. Thomas is a big part of this, but they’re frontcourt is also slightly undersized and Avery Bradley is on the smaller end for a shooting guard. The Milwaukee Bucks and Minnesota Timberwolves grade out as the two biggest teams in the league. With backcourts and frontcourts that run nearly two to three inches bigger than the league average. Neither team is particularly good this season, but they’re built to some degree on that older ideal — skilled players at each position, all a little bit bigger than their positional peers. Obviously there is still some player development to be done in order to make that a reality.
Hopefully, these visualizations help establish a little more context when it comes to discussing size and how it is used in building rotations. I have a few more fun things planned with this data that should be coming in the next few days. In the meantime, have fun, play around, ask questions, share with your friends, and let us know if you find something interesting.