This Week's Chart Dump
The Arena Effect and a Look at Turnovers
This week was a bit busy and I’ve had several ideas about multiple texts but didn’t manage to go in-depth yet, so I decided to use this week’s chart dump post as a sneak peek.
The Arena Effect
I came across this topic on Reddit a couple of days ago and decided to take a quick look at this specific issue. This is a well-known issue, as Seth Partnow wrote about it about 3 years ago on The Athletic.
The Reddit post specifically focused on the Golden State Warriors, and they are the biggest outlier out of all teams, as they and their opponents shoot a lot fewer Restricted Area field goals in the Chase Center than in other arenas.
I’ve done a quick table that only looks at the team’s OWN shots when they play at home or on the road, and the difference for both the Warriors and the Jazz is quite significant in comparison to other teams.
There are several other teams bunched up on the opposite side of the spectrum, and honestly, there is so much to write about this, but I wanted to give a sneak peek into the data for now.
The issue with this phenomenon is that it raises some questions about the deterrence metric I wrote about two weeks ago, or rather any sort of analysis done with either play-by-play data or shot data - since we can see that human error can be quite large.
The issue wouldn’t be only with deterrent metrics actually, but any sort of models and data that use the shot data in their calculation. This kind of discrepancy could cause very inaccurate numbers and skewness (mostly) towards the Warriors/Jazz Defense.
Turnover Breakdown
So yeah, I’ve bad-mouthed the play-by-play data in the previous section, and now I’m going to use it to analyze something else. Forgive me if I’m being a bit hypocritical but the reality is that it’s currently the only data we can work with.
This week I’ve been listening about top players in terms of ball security on Thinking Basketball’s podcast:
Ben has mentioned a lot of interesting metrics and topics in this podcast, and even though he was doing a solo podcast it was an interesting listen because he kept throwing new numbers - maybe that’s why I liked it.
One of the simplest takeaways when observing turnovers is that there is a big difference when a player throws a bad pass or a defender picks his pocket (live ball turnovers) versus an offensive foul or stepping out of bounds (dead ball turnovers).
Giannis is consistently among the worst players in the league when it comes to both turnovers per game and turnover totals, and he is the worst player in the league in dead ball turnovers alone. He is consistently among the worst in traveling calls and offensive fouls.
But this simple distinction between live and dead ball turnovers can make a huge difference in the quality of the shots that follow directly after the turnover, and that’s why we have to be careful when comparing raw turnover numbers.
To prove that, here is another chart that shows the relationship between the share of live ball turnovers and PPP by opponents after a target player has turned the ball over.
Even though Giannis is 2nd in total turnovers, and 6th in turnovers per game, the opponents score only 1.03 PPP after Giannis turns the ball over. The chart also has a moderately strong correlation (0.68 Spearman’s coefficient), which also makes sense, because as the live ball TOV share increases, it is much more likely that opponents have better opportunities, especially since players with high share are point guards operating on top of the key.
The Thieves
On the opposite side of the spectrum, we have the best stealers in the game, and we can divide the steals (AKA live ball turnovers) into two categories:
ball-handling steals - in other words an on-ball steal
bad-pass steals - interceptions
Let’s see what the break down of the steals looks like.
The eye-popping thing on this chart is De’Aaron Fox and Shai Gilgeous-Alexander who are ahead by 0.2 on-ball steals per game ahead of Giannis. Shai is also the overall leader in both total and average steals this season.
But what motivated me for this chart is that whenever I watched Cavs’ games recently Donovan Mitchell always seemed to be everywhere on the court. He is constantly reading the game and silently repositioning himself on the defensive side to jump and pick off the pass. Steals are generally very connected and correlated to the general volume and minutes that a player plays, but this season Donovan has almost more Bad Pass Steals per game (1.4545 - quite fitting repeating decimal number) than he’s had all types of steals per game in the recent seasons.
That’s all for this week, before I wrap this up, here’s a reminder that I’ve touched on the topic of the all-star game this week as well.
I didn’t give any major solutions, and you’ve probably already read multiple times about all the issues surrounding it, but if you still want to read about it, go check it out.