As a warning before starting off, this one is going to be longish and a medium amount of math. Check out my previous article on the topic for a little review on what the basics are. Also, this article will primarily present the information, with a follow-up article working on the analysis. All data is through the Utah game.
Teamwide Baseline for Shooting
I think that a reasonable way (your opinion might be different) to measure what the expected baseline is for team shooting is to just sum up all the makes and all the attempts across each game. There will clearly be some variation and nuance that is missed with this approach, but I want to go with a broad strokes method first.
Figure 1 is a just a plot of makes and misses, and Figure 2 is a density map of all attempted shots. Both show the spatial pattern of the shots, but in slightly different ways.
If you’re especially a glutton for punishment, flip through Table 1 - all the players and their made and missed shots are listed.
2018-19 Shot Makes and Misses
|Amir Johnson||2PT Field Goal||Made Shot||32|
|Amir Johnson||2PT Field Goal||Missed Shot||17|
|Amir Johnson||3PT Field Goal||Made Shot||4|
|Amir Johnson||3PT Field Goal||Missed Shot||3|
|Ben Simmons||2PT Field Goal||Made Shot||101|
|Ben Simmons||2PT Field Goal||Missed Shot||82|
|Dario Saric||2PT Field Goal||Made Shot||28|
|Dario Saric||2PT Field Goal||Missed Shot||37|
|Dario Saric||3PT Field Goal||Made Shot||21|
|Dario Saric||3PT Field Goal||Missed Shot||47|
|Furkan Korkmaz||2PT Field Goal||Made Shot||8|
|Furkan Korkmaz||2PT Field Goal||Missed Shot||5|
|Furkan Korkmaz||3PT Field Goal||Made Shot||9|
|Furkan Korkmaz||3PT Field Goal||Missed Shot||11|
|Jimmy Butler||2PT Field Goal||Made Shot||18|
|Jimmy Butler||2PT Field Goal||Missed Shot||7|
|Jimmy Butler||3PT Field Goal||Made Shot||2|
|Jimmy Butler||3PT Field Goal||Missed Shot||2|
|JJ Redick||2PT Field Goal||Made Shot||51|
|JJ Redick||2PT Field Goal||Missed Shot||50|
|JJ Redick||3PT Field Goal||Made Shot||48|
|JJ Redick||3PT Field Goal||Missed Shot||86|
|Joel Embiid||2PT Field Goal||Made Shot||137|
|Joel Embiid||2PT Field Goal||Missed Shot||124|
|Joel Embiid||3PT Field Goal||Made Shot||24|
|Joel Embiid||3PT Field Goal||Missed Shot||48|
|Jonah Bolden||2PT Field Goal||Made Shot||6|
|Jonah Bolden||2PT Field Goal||Missed Shot||1|
|Jonah Bolden||3PT Field Goal||Made Shot||1|
|Jonah Bolden||3PT Field Goal||Missed Shot||6|
|Landry Shamet||2PT Field Goal||Made Shot||17|
|Landry Shamet||2PT Field Goal||Missed Shot||12|
|Landry Shamet||3PT Field Goal||Made Shot||29|
|Landry Shamet||3PT Field Goal||Missed Shot||54|
|Markelle Fultz||2PT Field Goal||Made Shot||69|
|Markelle Fultz||2PT Field Goal||Missed Shot||79|
|Markelle Fultz||3PT Field Goal||Made Shot||4|
|Markelle Fultz||3PT Field Goal||Missed Shot||9|
|Mike Muscala||2PT Field Goal||Made Shot||11|
|Mike Muscala||2PT Field Goal||Missed Shot||10|
|Mike Muscala||3PT Field Goal||Made Shot||19|
|Mike Muscala||3PT Field Goal||Missed Shot||31|
|Robert Covington||2PT Field Goal||Made Shot||22|
|Robert Covington||2PT Field Goal||Missed Shot||22|
|Robert Covington||3PT Field Goal||Made Shot||31|
|Robert Covington||3PT Field Goal||Missed Shot||48|
|T.J. McConnell||2PT Field Goal||Made Shot||22|
|T.J. McConnell||2PT Field Goal||Missed Shot||15|
|T.J. McConnell||3PT Field Goal||Missed Shot||4|
|Wilson Chandler||2PT Field Goal||Made Shot||8|
|Wilson Chandler||2PT Field Goal||Missed Shot||5|
|Wilson Chandler||3PT Field Goal||Made Shot||5|
|Wilson Chandler||3PT Field Goal||Missed Shot||6|
Now that we generally see where the shots are taken from, lets go a little deeper and see if we can get a smooth estimate of shooting percentage by location. We divide the density of shot makes by shot attempts to get this. Again, this is not a perfect or exact count, and there are certainly parts of the floor that are over or underestimated in terms of shooting percentage. Further, this is not predictive, just descriptive. I still think it has utility in terms general trends.
Unsurprisingly, the Sixers appear to shoot pretty well at the rim (thanks Ben and Joel) with lower percentages in the midrange R and L of the key. Of note, it seems as if we do not shoot appreciably better from the L corner than we do from above the break. Also, the L baseline short midrange seems to be where Joel takes (and makes) a fair amount of his short jumpers, certainly more than the R baseline short midrange.
Incorporation of Lineup Data - Area FG%
The first comparison will be Joel on vs. Joel off. He’s been playing at an MVP level, and this offense only charting will severely undersell how good he is. Figure 4 is the same area FG% estimation, but only based on shots taken with Joel on the court. It looks very similar to Figure 3, since he is currently averaging an insane number of minutes, and therefore most the shots taken by the team are taken with him on the court (1072 to 464).
Figure 5 looks rather different, partially due to less data and therefore a little less resolution. However, the more interesting part is Figure 6, where the Joel - Off FG% are subtracted from the Joel - On FG%. That allows us to get a spatial estimation of FG% differences based on Joel on and off the court. Keep in mind that number of attempts does not factor into this part.
Couple things stand out based entirely on observation and conjecture (I will numerically verify next article after turkeying)...
- Sixers finish better at the rim with Joel in the game
- Feel like those three point areas are DHO with JJ and Joel
- The nail-ish area is where Joel likes to hit that little jumper
- The L corner three is often directly adjacent to a Joel post up when he’s on the court and likely to not be an open look based on help distance
I don’t necessarily believe that any differences in FG% are inherently good or bad - I think that you expect some change in play style when a 31.5 usage rate 34.9 minute per game monster leaves the lineup. However, the ORtg difference of +7.2 to -12.8 is indicative of the non-Embiid lineups getting vaporized.
Now, lets replicate Figures 4, 5, and 6 with Ben instead of Joel.
Observational takeaways based purely on the eye test...
- Sixers shoot vastly better from corner threes when Ben is on the court
- Ben’s presence does not seem to negatively impact certain areas of FG% as much as Joel’s presence does. Again, I think this is because Joel is usually the one taking the shots, whereas Ben’s usage rate is much lower.
- Figure 6 and Figure 9 appear nearly opposite in large sections of the court — I’ll look into this based on data where Ben is on and Joel is off and vice versa.
Things to Come
This was mostly an informational gathering article that I thought was worth publishing. Next up, I want to look at the specific shots and shooters that comprise the differences in the Ben and Joel on/off splits. If anyone has suggestions on avenues of investigation, let me know in the comments.
For the colorblind readers - are these more legible than the previous plots?