Cause or Correlation: A Look at Pitt’s Rebounding

A year ago I wrote about TPA (Total Points Added) and what, if anything, that was telling us about the Pittsburgh Panthers. At the time (and it continues) advanced analytics largely replaced box score stats in terms of how players were discussed and compared. The article was a reminder that TPA and other advanced metrics were designed to quantify different aspects of the game, or attempt to aggregate multiple aspects of the game into one number.

More recently, and with the accessibility and relative low cost of video capture, we’ve seen the rise of video clips as an additional supplement. I used them quite a bit last year as many of the clips were largely an extension of professional scouting videos (mostly for visiting players) that I was already creating and editing.

Individual player analysis (McGowens’ handle, Toney three-pointers) and even situational analysis (screen-and-roll) represented a bulk of the articles in which I used video clips. While I am still writing similar articles this year, I’ve taken the time to caution more about compiling clips just to support a narrative. Specifically, I want to make sure I avoid presenting something as a cause when it’s really correlation. Let’s all get on the same page real quick.

Causation, or cause, is an observed event or action that appears to have caused a second event or action. A correlation is the relationship between two variables used to describe or predict information. Here’s an example.

The other day I sought to answer the question:

Is there was some video evidence behind the notion that Pitt rebounds better with junior forward Terrell Brown on the floor as opposed to fellow forward and graduate transfer Eric Hamilton?

Now statistically, Hamilton hauls in 14.6% of available rebounds while on the floor as opposed to Brown who only snares in 11.1% of available rebounds, per sports-reference. But the question isn’t about the individual, rather the team. Also, the entire exercise was more about feeling than numbers anyway.

One of the first things that stands out when Hamilton is on the floor is how subpar of an interior defender he is, and he’s particularly ineffective as a post defender. That shouldn’t surprise anyone that’s followed Pitt throughout the season. Opposing teams have had more success finding quality shots around the basket with Hamilton on the floor as opposed to Brown.

I have to take a step back here. What I want to answer is A + B = C, but this is probably A + D = E. An important distinction when trying to marry video clips to a question or narrative. The fact there potentially aren’t as many defensive rebounds available isn’t exactly the same thing as why the Panthers don’t rebound as well, but there’s definitely some correlation here.

Additionally, teams also seem to get more quality shots around the rim as Hamilton leaves his feet very early to contest, biting on the smallest move at times.

To be fair, Brown can be overly aggressive in trying to block shots, but he’s more like a pogo stick, capable of a decent second jump to contest. Hamilton is an all-or-nothing leaper.

Again, I don’t think the availability of rebounds is the cause for why the team doesn’t rebound as well with Hamilton on the floor. So let's keep looking.

As I said, Hamilton is an all-or-nothing leaper and that applies heavily to his rebounding technique. He battles and bodies his man in the post, but when the shot goes up he just gathers for a two-legged leap. This allows his man to do the same, a battle Hamilton doesn’t win very often. He does create a wide base to leap from, but a little more use of the posterior and fanning out his arms to seek contact would probably box out his own man more often.

Brown on the other hand, usually initiates contact but fails to consistently create a wide base.

It’s not that Brown doesn’t find his man, he just uses an arm and shoulder to try and fend off his man instead of going a little wider and using his lower body to anchor his man. One of Brown’s biggest weaknesses has always been his hands - not particularly soft or strong - so he often can’t secure rebounds at the top of his jump, something that Hamilton is above average at.

Wait, I just found part (it’s seldom ONE thing) of the answer to something that I already cited statistics for above. Hamilton is the better individual rebounder, albeit neither is particularly very good. But this isn’t really what I am after.

So here it is, the cause in my opinion (we are talking perception here). Brown tends to get a little more of his body, especially his upper body, on his own man limiting their ability to be the primary rebounder. This allows his teammates to crash the boards more freely as they aren’t contending with Brown’s man as often. So while Brown himself isn’t snaring in boards at the same rate as Hamilton, he does a better job of accounting for his own man. Of course since I wouldn’t consider either even a good rebounder, the above isn’t always the case. Here’s a more direct example.

A head coach or player can say in the post game, “We didn’t do a good job rebounding today.” Naturally, grabbing X clips that show the team getting out rebounded makes some sense, right?

Answering that can be a little murky because if one finds that in every clip a teams 7-footer is grabbing all the rebounds they might label the cause as size differential. However, if in all those same clips no one even bothers to try and box him out, is size really the cause?

That was the goal of this article, aside from attempting to identify the root cause of a feeling, I wanted to demonstrate some of the pitfalls when using video clips to support a narrative AND labeling it as the cause. It’s more likely that you’re going to see a video clip used for visual aid and likely an example of correlation rather than causation.

But hopefully, the next time you watch a game and wonder, “Why are the Panthers so bad against the zone (which will happen later today)?” you’ll be able to pick out causes and correlations and know the difference. Which in turn will move you one step closer to being able to differentiate between explanation and analysis.