A lot of movement occurs over the course of a professional basketball game. The ten players on the floor have to run back and forth as they switch between trying to score on one basket and trying to prevent the other team from scoring on the other basket. Some players (and teams) cover more ground than others. Stephen Curry, for instance, is known for his relentless movement on offense, while James Harden has picked up a reputation as a rather lethargic perimeter defender.
But that’s just anecdotal evidence. Among the data available to the public on stats.nba.com is the average speed and total distance traveled of NBA players and teams since the 2013-14 season. Let’s use these facts to answer some relevant questions.
I’ll start simple: which players have covered the most ground on offense over the past seven regular seasons?
The dashed line running diagonally through the graph represents the expected offensive distance traveled for a player based on how many minutes they played. I plotted some notable points above or below this line, and none of the results are particularly surprising.
Most of the points above the line are guards, whether it’s high-energy point guards like Damian Lillard and Kemba Walker, or shooting guards who primarily operate off-ball like JJ Redick. I included Stephen Curry’s point because he’s one of the first names that come to mind when you think of players who cover a lot of ground offensively, but he’s actually not as far above the line of best fit as I expected.
Unsurprisingly, big men tend to travel less distance offensively than expected based on their minutes played. They’re not running around the perimeter like guards, so nobody would expect players like Marc Gasol, DeMarcus Cousins, and DeAndre Jordan to accumulate a bunch of miles traveled relative to their smaller, more agile peers.
The two players who clearly don’t fit into either category here are LeBron James and James Harden. They’re clearly not below the line of best fit because they’re big men. It’s easy to explain Harden’s placement on the graph — he operates in isolation more than any player in the history of the game, and these plays frequently feature him dancing behind the arc and stepping back for a contested 3-pointer. He’s not exactly running around the entire court like JJ Redick. LeBron, meanwhile, is not known for his activity off-ball. He works best with a player like Kyrie Irving who can also handle the ball. During those plays, LeBron typically isn’t moving around as much as he normally would. I also imagine these numbers are higher for him during his first stint in Cleveland when he was in his athletic prime.
Despite those two discrepancies, the trend still seems clear. I went ahead and used the position designations on stats.nba.com to see if this apparent relationship between position and average offensive speed still holds up. We also haven’t talked about defense yet, so I’ll throw in average defensive speed as well.
The result of an ANOVA test on this data allows us to reject the null hypothesis and conclude that the variance in average offensive speed between positions is unlikely to have occurred due to chance.
There’s also a positional trend for average defensive speed, although it’s not quite as significant as the offensive side of the ball. Versatile wing players listed as both guard and forward (including LeBron James, Giannis Antetokounmpo, Paul George, Jimmy Butler, Andre Roberson, etc) tend to be the most active on defense. I imagine that their versatility entails more active defensive assignments that force them to move around more than they otherwise would. Andre Roberson, for instance, was1 one of the league’s best perimeter defenders and his average defensive speed was a whopping 4.36 miles per hour, the second-highest in the league among guard-forward hybrids.
Offensive speed also tends to be higher on average than defensive speed. That makes sense, because the defense is tasked with just defending their basket, while the offense must move around more to evade and infiltrate the defense.
Anyway, it’s not enough to simply adjust for position, though. Let’s go back to offensive speed: there are other variables that will obviously have an impact on our numbers. JJ Redick and James Harden are both classified as guards, but they play very differently. Harden is one of the more ball-dominant players in the league, while Redick primarily creates his shots without the ball in his hands. This is part of the reason Redick’s average offensive speed is 5.01 miles per hour while Harden’s is a measly 3.99 mph. So, let’s take a look at the relationship between average offensive speed and the percentage of minutes in which a player does not have the ball in their hands (aka their off-ball percentage).
The vertical dashed line represents the league average off-ball percentage (91.6%), while the horizontal dashed line represents the league average offensive speed (4.55). Using this visualization, I’ve mentally separated the ‘outlier’ players into four categories: high-energy on-ball guards, high-energy off-ball guards, less active ball-dominant players, and less active big men. There are obviously many players in that pile of dots on the right side of the graph who don’t fit into any of these categories, but I won’t focus on them.
The players on the top left often have the ball in their hands and have a very high average offensive speed. There are ten players who play off-ball less than 90% of the time and still maintain an average offensive speed of at least five miles per hour. Three of them play/played for the San Antonio Spurs (Murray, Mills, Parker), seven of them, including Mills, are bench guards who are therefore able to conserve more energy, and the last player is just Steph Curry. He’s special, of course. Ish Smith and TJ McConnell both epitomize the scrappy and active backup point guard — their placement on the graph is not at all surprising. I’m not sure if some part of Spurs’ offensive system explains why three of their point guards are also around that same spot, but it’s possible. And Curry is just … Curry. The most energetic superstar I’ve ever seen.
The bottom left is less dense and consists of the star players who are less active but still ball-dominant. This is mostly just LeBron James and James Harden, which we’ve already discussed. I threw in a label for the late Kobe Bryant, although his data is pretty useless because it only contains the final two seasons of his illustrious career. At that point, he was coming off of an Achilles tendon tear and was nowhere near the same player he once was. I would guess that a prime Kobe would be a faster Harden on this graph. This is also a good place to point out that average offensive speed has little to do with a player’s actual speed. Before his injuries, John Wall was widely considered the fastest player in the entire league. His relatively slow average offensive speed is indicative of his play style, not his top speed.
The top right contains premier shooters who work to create separation off of the ball. JJ Redick, Landry Shamet, Klay Thompson are all in the lump below Davis Bertan, who is hitting 42.4% of his 8.7 three-point attempts per game this season. Buddy Hield also certainly fits the description of an elite perimeter shooter and therefore covers plenty of ground without the ball in his hands.
Finally, the bottom right basically exclusively consists of inactive big men. DeMarcus Cousins, Marc Gasol, and Hassan Whiteside headline the group of players who don’t have the ball in their hands as often as the guards and forwards of the world, and certainly don’t move around as much as them.
Also, slightly off-topic: it’s not at all surprising to see Terrance Ferguson with the highest off-ball percentage in the league over the past seven seasons. You could even argue that his off-ball percentage should be higher. The following play demonstrates his lack of offensive value:
Down by four with under 40 seconds to go, Chris Paul pokes the ball loose from Tatum to initiate a 3-on-2 fastbreak. He kicks it to a wide open Ferguson on the wing who, for some unknown reason, immediately dumps it off to a tightly guarded Dennis Schröder in the corner. By the time Schröder has the ball, the entire fastbreak advantage is lost. The Thunder proceeded to lose by one point. One could describe Ferguson as a worse Andre Roberson, who has the second-highest off-ball percentage in the league over the past seven seasons. The Thunder sure like their lockdown defensive wings who provide little value offensively.
But I digress. There’s plenty of more that can be done with this data, but I think this is a good stopping point. In the future, I might revisit this idea but focus more on the defensive side of things. The ten players who cover the most ground per game on defense are Jimmy Butler, Bradley Beal, Paul George, Klay Thompson, Stephen Curry, Ben Simmons, Andrew Wiggins, Trevor Ariza, Jrue Holiday, and Lonzo Ball. Almost all of those players are pretty damn good defenders.2 Could the data be used to quantify defensive value to some extent? Or is it just indicative of role? Probably the latter, but it might be interesting to look deeper into.