The NBA's Biggest Whistleblower

My Referee Database is now available for everyone. Plus, pouring some room temp water on Dyson Daniels' DPOY campaign

Welcome to another edition of The F5.

Once a month I remind everyone that the F5 is a reader-supported publication. There are no ads. No gambling promos. Just a few charts, some Q&As, and my 2 cents. This newsletter only exists because some readers have decided that my 2 cents are worth more than that. If you feel the same, then consider a subscription to The F5 for the cost of one iced coffee a month ($5).

Paid subscribers receive a tutorial every week on how to recreate the graphics and analysis found throughout this newsletter. I’ve published over 60+ tutorials on how to work with NBA data and for $5/month you get access to the full archive of tutorials plus anything I write going forward. It’s a pretty good deal, imo.

---

Which referees are the biggest whistle blowers?

This seems like a simple question on its face, but once you dig into the data you quickly learn there are factors beyond surface level impressions that influence how tight or loose a game gets called.

For instance, a referee that’s officiating a back-and-forth game between the Raptors and the Magic — the league’s two most foul-happy squads — are naturally going to call more fouls. Not because that ref likes the sound of their own whistle, but because the game requires it.

Or perhaps a referee is paired with two lesser experienced officials. That might encourage a more senior official to call more fouls than they typically would. This also works the other way, too. A lesser experienced referee might defer to their more experienced crew chief.

So to find out and once and for all which refs tend to call a tight game I took a more calculated approach. Using Ken Pom’s Fouls Above Average (FAA) for college basketball as a guide, I looked at how many fouls a referee calls relative to an average NBA referee by controlling for factors like:

• the margin of victory (close games have more fouls — particularly at the end)

• the number of possessions in the game (more possessions means more fouls)

• the foul tendencies of the teams that were playing

• the referee’s position (crew chief vs. referee/umpire)

• the other members of the officiating crew

I tried to control for each referee’s height to test ESPN’s Tim McMahon’s theory that shorter refs call more fouls, but I couldn’t find a source for referee combine data.

Anyway, here are the results:¹

You can interpret the FAA value as the number of fouls a referee calls above (or below) what you would expect an average referee to call in the same situation per 100 possessions.

When building a referee metric, a good sniff test is to see where Scott Foster ranks. The more uptight he looks, the better your metric smells. To no one’s surprise, Foster has the highest FAA over the last five years, which makes him the NBA’s biggest whistleblower.

What Fosters +2.31 FAA means in practice is that all else equal we’d expect to see Foster blow his whistle about two more times than a league average referee over 100 possessions. Hypothetically, if you replaced Foster with Intae Hwang (-2.39 FAA), then you’d expect nearly five fewer whistles. (I say “hypothetically” in the most literal sense because Hwang has never served as a crew chief, which is Foster’s typical assignment.)

It’s hard to say how much of this is actionable from a team perspective. There are a million things a coach has to deal with in the lead up to a game. Knowing that the referee assigned to their game is likely to call a couple more fouls than a typical referee shouldn’t radically alter their game plan.

And yet, I can say from first-hand experience that coaches do look at this stuff. They want to know everything there is to know about the officials. “Know thy enemy,” and all that stuff.

I have a memory of when I was with the Knicks and we were getting ready for a game against the Brooklyn Nets. When our coach found out that Zach Zarba was the crew chief they muttered, “Are you kidding me?” The room went quiet for a moment before they continued their thought, “Zarba’s from Brooklyn!”

Whether knowing referees’ personal biographies makes a difference in how a coach prepares for the game is a question that I’m not qualified to answer.

Still, I do think there’s value for teams to use some kind of referee data if they aren’t already. For instance, only a handful of referees are willing to call defensive three seconds while the vast majority are not. Knowing which ones are likely to turn a blind-eye on those calls might give your defense a little boost without costing you much.

Similarly, knowing which refs tend to call more fouls inside the restricted area could help you decide whether to leave your rim protector in the game when he’s already in foul trouble.

At the very least, knowing which refs are notorious for handing out technical fouls to coaches is something I’d want to know if I was on the sideline.

I’ve created a website with all that information and more, which is now free and available for everyone.

Sort refs by their FAA. See how many times their crew’s calls have been challenged and overturned. Toggle between totals, per game, and per 100 possession values. Look at the data over a one-, three-, or five-year window. Click on a given ref to see their percentile rankings.

You can even download the data if you want.

I’ll be updating the database weekly throughout the rest of the regular season.

---

Dyson Daniels leads the league in steals. So why are the Hawks better on defense when he sits?

Dyson Daniels is having an eye-popping statistical season on the defensive side of the ball. He’s averaging three steals a game and leads the league in defensive stops². He’s also 2nd in betting odds to win Defensive Player of the Year behind only Evan Mobley.

And yet, the Atlanta Hawks are performing better on defense when Daniels is not in the game.

The Hawks are holding opponents to 117.5 points per 100 possessions when Daniels plays. That number improves to 112.0 when he sits.

A five-and-a-half point swing isn’t enormous but it’s not what you’d expect from someone who’s in line for a first team All-Defense selection.

If your first instinct is to blame Trae Young for Daniels’ poor on/off numbers then I would encourage you to listen to something other than the Ryen Russillo podcast.

Over on databallar.com we can break down Atlanta’s performance with and without both players to see how much of this is a Young-related issue.

The Hawks have a 116.5 Defensive Rating when Daniels and Young are both on the court. That number jumps up to 120.4 when Daniels is on the court without Young. Meanwhile, the lineups with just Young have a 113.2 Defensive Rating. So at first pass it’s hard to find a way to blame this one on Young.

So the question remains: what’s causing the Hawks to perform worse on defense when Daniels is on the court.

I think much — if not all — stems from the fact that Daniels has quietly been one of the worst high volume rim protectors in the NBA this season.

There are 71 players that have been the closest defender on 150 shots at the rim this season according to Second Spectrum tracking data. Daniels Ranks 65th in opponent field goal percentage at 74 percent. The only players that allowed their opponents to score at the rim more efficiently and on more volume than Daniels are Josh Hart (four inches shorter than Daniels), Guerschon Yabusele Yabusele (plays out of position); and Nikola Vucevic (is a piece of Swiss cheese).

You don’t get to three steals a game without being a bit of a gambler. And when it goes wrong it often leaves Daniels in poor position to attempt a solid contest.

To be fair to Daniels, it’s probably not all his fault. The situation the Hawks are in isn’t doing him any favors. He’s likely being “credited” as the closest defender on more shots than he would be if the Hawks were fully healthy. With an injury to Jalen Johnson, Daniels is sometimes the Hawks second biggest player on the court and the only form of resistance at the rim outside of their center. Lord know Georges Niang isn’t blocking anything.

1. I put all the refs with less than 50 games officiated into a single group for the purposes of the regression. As a result, they all share the same FAA value — a slightly positive +0.3.

2. A catch-all defensive stat that sums up steals, blocks recovered by the defense, and offensive fouls drawn (including charges).