This is going to be a quick article about a different way I wanted to visualize shot data. I have already done this using shot charts, as I wrote about in this article. Here, I am going to talk about the new way I went about presenting the same data, and some of its advantages and limitations.
Above are the two different shot graphs that are produced with my new project. In the last one I did, we could get a really good picture of where each player was shooting from because of the chart, but this graph provides us with more data as to how good a player is a scoring at each distance or zone of the court. Jimmy Butler is the player I chose to outline this, because I think he is interesting to look at as someone who is overrated as an individual scorer but underrated as a playmaker and help defender. These charts do a good job of illustrating the scoring ability of Butler from an efficiency standpoint. He is a good scorer in the paint and around floater range, but he isn't as good as advertised in the midrange. He is also a below average three-point shooter, which makes him below average as a scorer in terms of efficiency. He is also even below the average of the location eFG%, meaning he misses more shots that the league average would from each distance. The advantage of graphs like the ones above is that we can do this kind of analysis of where a player seems to score efficiently from, and where they do not.
There is some noise in these numbers when referring to a player like Butler. High usage guys who create a lot of their own shots tend to be shooting tougher shots. So, a higher proportion of Butler's midrange shots are going to be off the dribble or contested, which does not factor in here. This highlights the limitation of the only thing that counts is make or miss. Another limitation is that it is more difficult to visualize where the shots are coming from using these graphs. The bars don't do a good enough job of showing us volume from each area.
What I would recommend going forward is to look at both this shot graph and a shot chart that was described in the last project. Looking at a new player would look something like this:
These charts show the shot distribution and accuracy of Carmelo Anthony. As we can tell from just looking at his accuracy numbers, he is above average in operating in the mid-range. He shoots a lot of shots in the mid-range, as we can tell from our shot chart, but the issue is that he doesn't make those shots at the level of a league average three or look at the rim would earn, meaning the looks from mid-range overall are inefficient. He is also below average in terms of finishing at the rim, which isn't good because like most players Melo shoots a lot from close range. He also shoots a good amount of shots from 3, which he is slightly below average at. A limitation to this is that Carmelo tends to take tougher shots than the average player. However, since he isn't a star player like a Jimmy Butler, why is he shooting those tougher looks? I was surprised to see his eFG% is lower than his location eFG%, but how good he is in the mid-range isn't enough to make up for below average at the rim and from three, the two most efficient spots. After this assessment using both the shot chart and the shot graph, we can determine on a good team that already has shot creation, Carmelo Anthony would not help them offensively.
This wouldn't be a fun project without some fun shot graphs, so here are some of my favorites:
This is why Mathews is one of my favorite players in the league:
The big I picked to illustrate how weird their charts look:
Duncan Robinson is insane at what he does:
Wiggins is always fun to look at for any project:
A good example of why sample size matters:
Underrated shooter from (very) long distance:
Last but not least, the insanity of Damian Lillard's shooting ability:
The code for this project can be found here.
Comments