State of the Celtics: Is Avery Bradley an elite defender?

It’s been a thing for years, but I've never really known what to make of Avery Bradley’s defensive numbers. It hasn’t been since the 2013-14 season that a Celtics team has allowed fewer points with Bradley on the court. So when I started researching for this piece my theme was simple: Avery Bradley is actually a bad defender because of ____ (unknown reason I was about to find). And honestly, one look at his stats would have affirmed that it wouldn’t be hard to find. Bradley is 92nd amongst shooting guards in DRPM, (that’s Evan Turner territory). When he’s on the court, the Celtics’ defensive rating jumps 21.6 points (113.5 to 91.9), and opposing teams shoot, rebound, and assist better when he’s on the court. In the last five minutes of a game with the teams’ scores within 5 or fewer points, Bradley’s net rating is a -31.3, with his defensive rating skyrocketing to 31.3. In the last three minutes of a game when the scores are within 5 points or more, Bradley’s net rating gets event worse (-44.7), and his defensive rating jumps up to 132.1. Individually, players shoot 4% better on shots that he defends (including 8.4% in shots less than 10 feet) and shoot 3.9% better from shots greater than 15 feet away.

Those aren’t the numbers you’d usually correlate with a player coming off a first team all-defense award, but it’s become something of a running mystery. However, after spending an entire weekend combing through every game from this season, I’ve come to realize that Bradley’s numbers and performance are a more complex issue than I initially suspected.

At this stage, we’ve all grown accustomed to Bradley making plays like this.

What Bradley lacks in strength he makes up for in cat-like reflexes that highlight his lateral quickness and next-level anticipation. To understand why Bradley’s defensive numbers are atrocious, it’s important to understand that there’s no trump card in this situation. Bradley’s struggles are a nexus of factors that derive from team schemes, lineups, matchups, and bad tendencies.

At 6 foot 2, Bradley is the size of smaller point guard, but because he plays alongside the 5 foot 9 Isaiah Thomas in the starting lineup, he’s constantly asked to defend bigger and stronger wings. According to Basketball-Reference, out of Bradley’s top 10 five-man lineups, only two of them don’t involve either Isaiah Thomas or a three-guard lineup. These different small-ball lineups can lead to some misleading numbers. For example, Bradley allows 8.4% higher shooting percentage to opponents when he defends guys within 10 feet. In his case, the fact that the players he defends can have six inches or more on him mean that those shots are virtually unguardable because they can shoot right over him.

Here, once Batum is able to gain post up position on Bradley, there’s really not much he can do. At 6 foot 7 with a 7-foot wingspan, Batum will be able to take this shot effortlessly regardless of Bradley’s contest.

Another subset of lineups has been the issue of rebounding. Because the Celtics don’t have a traditional big, teams that do have had their way with the Celtics to start the year. One thing Celtics fans seem to have forgotten is that, even with Jared Sullinger, the Celtics were 26th in defensive rebounding last year and 21st overall in rebound rate. However, Sullinger was very good at boxing out bigger guys and going to get boards himself, so Boston didn’t necessarily have to send in wings to crash the glass. This year’s front court traded that luxury for versatility on the perimeter, but to make up for its lack of a true rebounder, bigs are only asked to box out while wings—specifically Avery Bradley—are asked to dive in and scoop the board. This forces Bradley to play the ultimate balancing act of being in position to rebound while also staying glued to his man, which is usually the best backcourt player on the floor. For the most part, Bradley has actually handled it pretty well, but things can go a little astray if a team is able to get an offensive rebounding.

Here, Bradley was guarding Fournier before he dived in to get a rebound. But when ORL got the rebound, Bradley got stick on Biyombo, leaving a big to have to run out to defend Augustine. Again, this isn’t a conclusion that Bradley’s numbers are bad because he goes for rebounds, but it is an example of how his added responsibility can hinder how his presence defensively is interpreted. Being asked to guard bigger guards isn’t his fault, and neither is being asked to rebound, but both can have a negative effect on his defensive impact. However, even that doesn’t tell the whole story.

*Stats via NBA.com

These plays make up about 87.1% of the shots that Bradley defends. As you can see he’s either average or below average in every shot category he defends. Again, we find ourselves in a situation where there’s no clear answer. Sometimes you just get matched up with a unstoppable player:

And other times you force a guy to do something, and he beats you anyway:

But Bradley also has some issues on his own. I have a running theory that ever since Bradley had his shoulder surgery back in 2012, he has tried to avoid as much contact on the pick and roll as he can. This can lead to him taking weird angles at instances such as this one, when he basically puts Olynyk in a one-on-one situation with James Harden then tries to recover.

And on this play, he tries to avoid a collision with Zeller, which makes him just a millisecond late in defending the Batum jumper.

To paint a clearer picture, here’s a quick possession of another known defensive stopper, Patrick Beverly. Beverly, who’s similar to Bradley in size and defensive style, is defending Curry when a pick comes his way. Instead of trying to get around it, Beverly goes straight through the pick, which closes up any opportunity for Curry to get a clean look.

These are subtle things that don’t catch your eye initially, but in a league where guys can get up shots as fast as you can blink, the difference between circling around picks and going through them can be crucial.

So what do we make of this? The numbers say Bradley is a below-average defender, while the reputation suggests he’s one of the best league. The eye test says he’s great and flawed at the same time. In the same way I found a bunch of clips of bad Bradley defensive possessions, someone could find just as many good ones. However, I think the combination of his added responsibilities of being a rebounder and scorer combined with his shoulder injury, tough matchups, and some bad tendencies draw an interesting distinction that’s important to make. Bradley is no longer playing defense at an elite level and at times even an average level, but he’s still an elite defender. When it comes down to it, Bradley still has the ability to do stuff like this when the game matters:

In the last full playoff series that Avery Bradley played (the 2015 series against the Cavs), the Celtics’ defense was better when he was on the court, and the Cavs’ efg% dropped from 56.8% to 46.9% when he was in the game. The data show similar results if you go back all the way back to his second-most-recent full playoffs back in 2012, where the defense was 12.9 points better when he was on the floor ( 90.2 to 103.1).

Over the course of an 82-game season, Avery Bradley may no longer play at an elite level defensively, but when the playoffs come and the responsibilities become more narrow, the defense is consistently better with him on the court. Should an elite defender be able to play at high level regardless of what’s put on his plate? Should all elite defenders be graded at the same level? Or, even a more fundamental question, what makes an elite defender? The answer to all those questions requires a subjective test that has no conclusive answer. But how many players are you trusting more than him to defend Steph Curry with the game on the line?