Foul Trouble or Troubled by Fouls

The playoffs can be an odd time for the statistically-minded. As Billy Beane famously said, “my [stuff] doesn’t work in the playoffs.” This doesn’t mean the underlying factors identified through data-based analysis cease to matter. More, since it’s a small, biased sample, weird things happen. Among those oddities were Games 1 and 2 of the Miami series, where the Heat took two seeming strengths of this Hornets team and turned them upside down. Game 1 featured one of the most surprising offensive rebounding performances of all time[1. Miami recovered 43.8% of rebounds from their own missed field goals — I prefer to separate rebounding into field goal and free throw rebounding as missed free throws result in far higher defensive rebound rates — while Charlotte allowed opponents to recover only 17.9% of missed field goals during the year. Going back to the 1997 playoffs, this is the fourth highest single-game discrepancy between a team’s regular season rate of conceding second chances and their performance in one post-season contest.], while Miami’s shooting was scalding in both contests and their 63.6 effective field goal percentage through two games was the third-highest of any team in the first two games of a playoff series since the 1984 postseason.

Particularly on the shooting front, this fell into the “just one of those things” category, as Miami seemingly could not miss. On the other hand, it also reflected why regular season trends frequently don’t hold in the playoffs. Moving from studying aggregates from a relatively balanced schedule across the whole league to the intricacies of a specific matchup can do that.

In this case, the Hornets to a large degree have been bamboozled by Hassan Whiteside. Whether as a rim-runner in pick-and-roll play, attacking the offensive glass, or protecting his own backboard, he’s been a menace to the Hornets. Reducing his effectiveness will probably be a necessity if the Hornets want to pull the upset. Sadly, they seem ill-equipped to neutralize the center from a personnel standpoint. With Nicolas Batum injured and Marvin Williams’ effectiveness coming and going, smallball is an option which seems to play more to Miami’s strengths, while none of Charlotte’s bigs are likely to be able to take on Whiteside directly. The best chance is Al Jefferson, whose post-up trickery has flummoxed Whiteside, but whose own questionable defense and conditioning following an injury hit year make him likely to give back on one end what he takes on the other.

There is one thing Jefferson, or any other Hornet going to the rim might do, though. Get Whiteside in foul trouble.

Many, notably ESPN’s Amin Elhassan, have observed that Whiteside’s defensive effectiveness often appears to fall off a cliff when he gets in or near foul trouble. This is somewhat understandable — if picking up another foul means a long stretch on the bench, it is almost natural to avoid putting one’s self in position to commit that foul. Of course, shying away from contact is also a good way to play substandard defense, especially on the interior. The question is, does this charge stick to Whiteside, and if so is he any better or worse than other players?

With the enormous caveat that correlation does not prove causation[2. There are any number of chicken-and-egg type arguments why Whiteside getting into foul trouble might be a result of opposition offensive tactics and skill which also lead to poorer defensive performance rather than the foul trouble itself being the cause of the the defensive decline.], Whiteside’s defense does appear to suffer when he starts to accrue fouls. For these purposes, I defined foul trouble as being in a situation where getting one more foul would like send him to the bench[3. How to treat players in foul trouble from a coaching standpoint is a much-discussed but not fully explored area. On the one hand, observation indicates coaches are liable to “foul out their own players.” On the other, some research suggests players perform worse when playing through foul trouble. I’ll split the difference and suggest that general research on the topic should probably take a back seat towards evaluating the individual player — their importance to the team, their likelihood of picking up the “next” foul and their ability to remain effective while avoiding that foul.]. For convenience sake, I broke his time on the court into two samples, one where he had accumulated fewer personal fouls than the number of the current quarter and one where he had picked up at least as many fouls as the quarter number. Initially, I wanted to look at situations where Whiteside was playing with more fouls than the quarter number, but it turns out he payed very little under those conditions, suggesting Erik Spoelstra himself has a hard and fast rule to pull Whiteside.

The results were startling. During the regular season[1. All stats via NBA.com play-by-play logs.]:

• Opponents had a 47.5 effective field goal percentage when Whiteside was on the floor and not in danger of foul trouble. This rose to 51.5 percent when the fouls started to accumulate.
• Relatedly, 29.9 percent of opponents’ shots came from within the restricted area in non-foul trouble situations, compared to 32.6 percent percent when he was in danger. Accuracy rose slightly as well from 55.5 percent to 56.3 percent.
• Whiteside’s block rate on shots taken from the paint outside the restricted area drops from 12.2 percent to 8.2 percent of shots, which is accompanied by a rise in accuracy from 41.9 percent to 45.5 percent on these “short mid-range” shots. Combined with the above, this suggests less efficacy as rim deterrent[5. Repeating the caveat about inferring too much, as the causality could conceivably run either way in terms of fouls causing worse defense, or a defense already compromised whether by good offense or other issues could lead to fouls.] once Whiteside starts to pick up fouls.
• While there is little change in Whiteside’s own rebound rate, the Heat as a whole drop from recovering 76.8 percent of opponent misses to 72.3 percent with Whiteside in the game with and without foul worries respectively.
• All told, this results in the Heat posting a collective defensive rating of 90.2 when Whiteside is not in danger, rising to 102.7 when he is. You read that right, a difference of 12.5 points per 100 possessions.

That’s strong stuff, but it stands to reason that a team’s defense might decline when its paint anchor is forced to ease up, so the Whiteside numbers should probably be compared elsewhere. To do so, I selected 9 other centers. To some degree, this is cherry picking, but I wanted to compare apples-to-apples in terms of role to the extent possible, so I searched for players who played almost entirely at the five and have at minimum non-terrible defensive reputations.

Looking at these other centers[1. Tim Duncan, Andre Drummond, DeAndre Jordan, Dwight Howard, Ian Mahinmi, Rudy Gobert, Pau Gasol and DeMarcus Cousins.], I found similar types of declines to be common, but not really the magnitude of those afflicting Whiteside. Perhaps the closest is DeAndre Jordan, another prodigiously athletic big man whose defensive reputation might be somewhat overblown. The Clippers’ defensive rating jumpers over 10 points from 88.8 to 99.5 when Jordan gets in foul danger. The chart below summarizes some of the findings:

So, short version, and (again) with the caveats that there are all kinds of other things going on between lineups, opponents, questions of causality and so on, the charge appears to stick. Though it seems commonplace for having a center in foul danger to correspond with decreased defensive performance, Whiteside and the Heat stand out at least this year. So while commentators might question whether coaches are too conservative with players in foul trouble, Spoelstra is probably right to do so in this case because he’s simply not getting the same player after Whiteside picks up a foul or two as he is before[1. Quick finishing tangent: these sorts of coaching decisions are one of the factors that play into the general messiness of RAPM and RPM-style metrics. If these sorts of declines are persistent, than Spoelstra is saving Whiteside from playing a higher proportion of his minutes in a less effective state. Or perhaps he’s not being protective enough. Either way, another coach might make this decision differently, thus affecting Whiteside’s overall “score” in one of these metrics with little or no change in his underlying skill. It’s likely not a huge deal, but just a worthwhile reminder why it isn’t just statistical noise that make one-numbers estimates rather than precise measures.].