After some lightly contested discussion on Liberty Ballers Slack about various young players who may or may not have gotten a terrible whistle in the playoffs, I decided to jump into the numbers to try and get a sense of if this is true or not on the large scale. All data is combined from the 2015-16, 2016-17, and 2017-18 seasons.
Total Personal Fouls per Game
First things first, let’s take a look at total fouls per game (includes both teams) over the course of the season. You’ll note three rather distinct ranges.
I’m guessing that first quarter of the season is when the new points of emphasis are over-enforced, combined with general rust. The last quarter I believe is when tank commanders really start to exert their influence on the game. For that middle section, it appears as if the average is about 22 fouls per game, with the season long average at almost exactly 25.
Next, we look at just the playoffs. One caveat to note, is that unlike the regular season, the game numbers to not perfectly match up. Certain teams (Warriors) might only play eight games in the first two rounds, whereas others might play 14. No adjustment is made for that.
Maybe first rounds series have more “hack-a” or just are situations where the aggressive team is giving maximal effort at all times with no let up? I’m not sure — compelling though. Although fouls per game is interesting(ish), really fouls per time played is what we want.
Personal Fouls per 48 Minutes of Play
Just based on these numbers, rookies average about 0.5 more fouls per 48 than vets do over the course of the season. However, I think that a binary rookie/vet situation might not be exactly what we want. Here’s a plot of years in the league against average fouls/48 for players who accumulated at least 48 minutes of playing time in one season.
Pretty noisy - certainly no clean and obvious trend besides it appears most high minute players have a foul rate below 5/48. Let’s now look at the playoffs using the same methodology.
Hmm. Something certainly appears interesting in those first 4-5 games that more or less 100% correspond to a first round series. The fouls/48 going up compared to the regular season is expected as rotations shorten.
Again, still kind of a mess. Let’s look at one more type of plot before drilling down a bit more.
Percent of Personal Fouls and Minutes
I think total fouls and fouls/48 are potentially not the most useful tools in this specific instance. I am more of the opinion that looking at the percentage of all fouls compared to the percentage of all minutes played might be a better option. This allows for relative comparisons, taking playing time into account.
Unsurprisingly, as the season goes on, rookies get more playing time (% total minutes) whether because they earned it or because the tank needs additional crew members. You will also note that rookies have a % foul/% min ratio greater than 1.0, whereas the vets are the opposite. Not that surprising, rookies tend to foul more often based on their lower skill level regardless of ultimate ceiling (see Jaren Jackson Jr. and/or Wendell Carter Jr.). Figure 8 presents the same methodology, but with playoff data. Due the relative rarity of rookies playing 20+ games in a playoffs season, I’d likely disregard the last data point or two there.
Based on Figure 7 and Figure 8, I don’t see much of a smoking gun of any sort. Further evidence for a lack of evidence is presented in Table 1.
The changes are both minuscule and in the opposite direction of what you would expect if rookies were getting routinely hosed. However, and this is a huge however, we can only account for fouls called. If Landry Shamet is covering Kyrie Irving in the playoffs and is whistled for a foul at the rim, we can “see” that in this data. If Kyrie challenges Landry at the rim on the following possession and there is hard contact but no foul called, this data does not show that Landry got the short end of the ref stick. I even ran the %F/%M ratio again for players with less than four seasons of experience compared to those more than four years and got essentially exactly the same numbers. Now obviously there are other considerations to be made - maybe certain position groups get called for more fouls, etc. Even so, only based on the data I have and the results I presented, it seems as if on a large scale, rookies do not magically get called for more fouls on a time basis than they do in the regular season. If I missed something obvious, let me know.