Evaluating Wizards coach Scott Brooks, Part 2: Defense

In part two of the attempt to use objective data in the evaluation of Wizards coach Scott Brooks and his counterparts around the league, let’s dive into defense. Part 1, which looked at playing time, is here.

The Wizards currently rank 28th in defensive rating (points allowed per 100 possessions), which is something of an improvement over last season’s 30th place finish. The big picture Four Factor metrics (listed in order of importance below) provide a picture of why (Wizards rank in each category):

• Shooting — 27th
• Turnovers — 9th
• Rebounding — 5th
• Free Throws — 30th

To sum up, the Wizards don’t make opponents miss, they foul too much, but they do an okay job forcing turnovers (about half a turnover per 100 possessions better than average), and they do a good job on the defensive glass.

The team’s poor performance lowering opponent shooting efficiency becomes a puzzle because the Wizards have the NBA’s best defensive shot profile. This is not a drill. It’s not a typo. I compiled the data, frowned at the result, recompiled the data and got the same thing. So, I went back and downloaded the data again...and got the same thing. Finally, I hand typed the data...and got the same thing.

There’s a shot categorization I keep track of, which I’ll call POOP, short for Poop Shots. (Albert keeps telling me this is a family site and I can’t use the actual name at the top of the column on my spreadsheet. Both of you who’ve listened to every episode of my sporadic podcast know the real name.)

POOP is two-point shots from outside three feet. The reason is simple: the shots are poop. Here’s an illustration in points per shot:

• At-rim — 1.31
• 3-10 feet — 0.84
• 10-16 feet — 0.86
• 16 feet to 3pt line — 0.84
• Threes — 1.10

Most shooting fouls happen in that at-rim area. Shots in the 3-10 foot range are often failed drives that result in an attempt contested by a big or a crummy runner. Shooting percentages are generally crappy on mid- to long-range twos. But you know this. Us stat goobers have been pummeling fans with this stuff for years.

The point is that the Wizards are actually doing some things right on defense. So far this season, just 13.1% of their opponent field goal attempts have been at-rim — far and away the best in the NBA. Number two? The Miami Heat at 18.3%. League average is 25.9%. Somehow, some way, the Wizards are the Secretariat of denying at-rim field goal attempts.

Meanwhile, they’ve allowed the highest percentage of attempts from 3-10 feet, the second most attempts from 10-16 feet, and the most attempts on twos from 16-plus feet. The Wizards allow the 10th fewest three-point attempts (as a percentage of field goal attempts). Somehow, some way, the 49.9% of opponent field goal attempts against Washington have been POOP.

So far, they’ve allowed the second fewest dunks per game, and they’re 11th in combined dunks and layups per game.

With all the kvetching and whining and moaning about Brooks and his inability to craft a coherent defensive scheme, his team — which is largely devoid of quality defenders — has somehow managed to take away both at-rim attempts AND threes. They also do an okay job forcing turnovers and grabbing defensive rebounds.

So what’s the problem? They can’t make anyone miss. Here are their defensive FG% from each shot distance:

• At-rim — 74.6% Rank: 29th
• 3-10 feet — 54.7% Rank: 30th
• 10-16 feet — 48.7% Rank: 27th
• 16 feet to 3pt line — 43.8% Rank: 20th
• Threes — 37.7% Rank: 22nd

A few possible explanations come to mind for the team’s poor defensive results despite being best in the league at limiting high-efficiency shots.

First, it could just be bad luck. If they continue to play as they have, opponents will regress to the mean. The season is still young and sample sizes are small. Over the course of a full season — assuming they’re able to play games again — opponent shooting percentages are likely to move towards league averages.

Second, the Wizards are unusually ineffective at forcing misses. It could be there’s something subtle they’re doing wrong that allows opponents to shoot significantly better than average from everywhere on the floor.

Third, it could be that their excessive fouling is pushing opponent shooting percentages higher by erasing what might have been misses and then allowing more open looks as Wizards defenders dial back the aggression to avoid fouls.

Fourth, it could be that the defensive shot profile itself is just luck — the Wizards aren’t forcing anything, it’s just opposing offenses taking whatever shots they want.

Getting back to evaluating Brooks (and assuming the team’s defensive shot profile is something other than luck), he and his staff have actually done a decent job of installing a defensive structure and inculcating habits that give the team a chance to be good defensively. They’re inducing opponents to take suboptimal shots (and contesting them at an above-average rate), they’re forcing turnovers, and they’re getting defensive rebounds. They foul too much and opponents are shooting a high percentage, but the outline is there for a good defensive team.

That structure extends to the play types captured in the NBA tracking data. The Wizards have allowed the 9th fewest transition opportunities so far this season (as a percentage of possessions). They haven’t been effective (just 21st in transition defense efficiency), but they’ve generally done a good job of getting back and forcing opponents to run a different kind of play.

They’ve been basically average defending pick-and-roll sets, but they rank 22nd or worse in defensive efficiency in most of the other categories tracked: isolations, postups, spotups, handoffs, off screens, and miscellaneous. They’re second best in defending the rare putback attempts opponents have gotten against them so far this season.

Understand, I’m not saying the Wizards are somehow secretly good on defense. They’re not — they have the NBA’s third-worst defense this season. But, this exercise is an attempt to use data to tease out the coach’s performance from that of the players. The data seem to suggest Brooks and his staff have done a decent job of putting players in position to succeed on defense. Ultimately, it falls to the players to make plays.