NBA teams as snowflakes

the devil is in the difference

I like NBA shot charts. Hex charts. Scatter charts. Density charts. 3D charts. A&D charts. Whatever. The weirder, the better.

One thing I do not like about shot charts, though, is that there’s not much variation within them. It can be difficult to make meaningful player or team comparisons using shot charts because they all kind of look the same since the NBA has, for the most part, figured out where the most productive parts of the court are.

For instance, when you look at the most common areas teams shoot from, it’s hard to reach many conclusions because the overall distribution of shot attempts for each team looks so similar.

Part of the problem is that some of the most productive parts of the court (i.e., the restricted area) have less surface area than the least productive parts of the court (i.e., the mid-range), so traditional shot charts often paint an incomplete portrait of what’s really going on. The only real takeaway you can have from looking at a chart like the one above is that teams favor shots at the rim and shots from beyond the arc. While that’s true on a macrolevel, charts like these obscure some of the more subtle and interesting aspects of what makes teams unique.

To see something more revealing, we just need to adjust our perspective.

Instead of plotting the overall frequency of shot attempts, we can just plot the differences between how often a team shoots from a given area and how often an average team does. That’s what the chart¹ below shows.

These plots still show the relative frequency of shot attempts from a given area. But the crucial difference is that each chart is relative to the league average as opposed to itself. In other words, denser and darker areas indicate that a team takes a higher share of their shot attempts from the corresponding location relative to how often the league at large does. Meanwhile, blank areas indicate that the team takes either a smaller or equal share of their shot attempts from that area compared to the league average. So if a team’s distribution of shot attempts across the court mirrors the league’s distribution to a T, their chart will be blank.

In this version, there’s lots of variation between teams since we’ve stripped away the things they have all have in common and emphasized the things that make them unique. You can think of these charts as illustrations of what a team likes to do that other teams don’t. In other words, these charts show what makes a team special. For that reason, I’m planting my flag and calling these Snowflake Charts from here on out.

Given the number of players that have switched teams via trades or buyouts, I thought it would be fun to highlight a few points of interest before things get too muddled by all the personnel changes.

For example, on Orlando’s chart, there’s a noticeably dark, dense area at the top of the arc where Nikola Vucevic was making a living off of wide open, pick-and-pop threes. If I remake this chart at the end of the year I expect to see a similar one on Chicago’s chart, provided that the Bulls use Vucevic in a similar fashion. Also of note on Orlando’s chart is the heavy amount of color in the mid-range area, which makes sense considering the Magic have taken a higher share of their attempts from that area than any other team.

Meanwhile, in Indiana, there’s hardly any color in the mid-range area, a stark departure from last year. Under first year head coach Nate Bjorkgren, the Pacers have made it a point of emphasis to excise mid-rangers from their diet. Only the Rockets and Jazz take a smaller amount as a percentage of all their attempts.

Even though we’re looking at things from a 10,000 foot perspective, we can see the fingerprints of individual players on a team’s chart. There’s Zion Williamson’s utter dominance on the left side of the hoop in New Orleans. Luka Doncic dancing with his defender on the block in Dallas. We can even see all those stepback threes James Harden is fond of taking above the break, right of center.

Fans should be able to pick up on the subtleties in their team’s charts better than I can, but I’d like to briefly touch on a couple that stood out to me.

---

Memphis has been in love with the floater for two straight years now. In each of the last two seasons they have taken a higher share of their shot attempts from the short mid-range area than any other team. Why? Because they have a bunch of guys who can take and make that shot. Ja Morant, Tyus Jones, Brandon Clarke, Kyle Anderson, and several other players on their team feast on the floater. Meanwhile, the sparsity of color around the three point line, particularly above the break, illustrates the lack of high volume three-point shooters they have on their roster. Getting Jaren Jackson Jr. back from injury at the end of April should help on this front.

One of the reasons I’m a big fan of these types of shot charts is that you can really see when a team — or more often a high volume player — favors one side of the court more than the other. That’s true in Sacramento where the purple blob running down the left side of the lane traces De’Aaron Fox’s (known lefty) flight path to the basket. Like the Grizzlies, the Kings also subsist on a full diet of push shots and short mid-range floaters thanks to Richaun Holmes and Tyrese Haliburton.

If I had made a chart for Phoenix and didn’t see a dense area at the elbow I probably would have just thrown this whole idea for a post in the trash. That’s because few players have a more identifiable signature shot than Chris Paul. He’s been taking and making the elbow jumper for years. As usual, he’s been ultra-efficient on these shots this season. Paul is shooting 51 percent on all mid-rangers, which would put him in the 95th percentile of all players this season according to Cleaning The Glass.

In Portland, Damian Lillard is stretching the limits of what’s considered a good shot. This year, he’s taking and making more deep threes than ever before. Also notable on Portland’s chart is the Blazer’s relative reliance on shots just inside the three point line. The Ringer’s Kevin O’Connor recently published a video on CJ McCollum’s evolution as a scorer and how he’s taken a smaller, but still meaningful share of his shots from the mid-range this season compared to previous ones. You can see that on this chart where there are some pockets of dense areas a foot or two inside the arc. Also, I like that you can see Carmelo Anthony’s old school baseline jumper game on this chart. It’s those little things that can get lost when you’re aggregating data at the basic shot zone level.

---

As much as I like these charts, they have their limitations.

All they show you is where a team takes shots from more frequently than the rest of the league. They don’t say anything about the shots a team doesn’t take. Nor do they say anything about how a team generates the shots they do take or who takes them.

Also, these charts don’t tell us what a successful team looks like because teams are successful for reasons other than their shot distribution. How shots are generated and who is taking them is just as important — if not more — as where the shots are coming from.

However, I think these Snowflake Charts are a nice improvement over traditional shot charts because they highlight what actually stands out when we watch a game. Instead of painting all teams with same brush, these charts acknowledge and emphasize each team’s quirks. One of the things that’s fun about being a fan of the league as a whole is recognizing the individual differences in styles between teams and ultimately that’s all these charts are trying to call attention to.

1. When I’ve made this type of chart in the past I have binned them into hexagons. The concept is the same: darker areas indicate that a team takes a high share of their shot attempts from that spot relative to the league average. Binning the data like this helps with the overall visual aspect, but we lose some granularity. I think I actually like the hex version better when viewing all 30 teams at once and the non-hex version when viewing just one team at a time. Sound off in the comments with your thoughts.