clock menu more-arrow no yes

Filed under:

Various thoughts on the Celtics’ defense

New, comments

Is the “bubble effect” real? Should Daniel Theis and Tristan Thompson play together? Is the roster depth forever doomed by injuries?

Los Angeles Lakers v Boston Celtics Photo by Maddie Meyer/Getty Images

The Celtics’ defense can’t be this bad, can it? Well, maybe it isn’t, but it sure feels pretty bad to watch sometimes. One thing to remember about the Brad Stevens era is that the 2015-16 season probably broke our collective brains when a thoroughly average Celtics roster held the league’s fourth-best defensive rating before getting mopped by the Player of the Month, 60-win Atlanta Hawks in the first round.

Four All-Stars. Four.

Here’s the roster that held that 103.6 defensive rating:

Before you say it, the answer is no; Avery Bradley’s defensive stats have pretty much never been good.

The current Celtics’ defensive rating of 109.8 lands them at 12th in the league after playing the 10th toughest schedule so far. To be honest, I don’t care too much for the actual number, but instead, the long list of factors that add up to it. These are sort of scattered, so I’m breaking them up into small sections, starting with:

The Bubble Effect - maybe it doesn’t exist?

Generally, I’ve been a believer in the “bubble effect” (no fans, fewer distractions, more or less neutral sites) as to why so many players have had 40, 50, and even 60-point explosions in the COVID-era of the NBA. But the more I look at things, the less convinced I am.

The past three seasons haven’t shown any significant change, but there’s a huge jump before that. This could have happened for a lot of reasons, but I’m only interested in what didn’t happen, and there was no pandemic in 2017-18.

I’m sure the lack of crowds has had an impact on the game, but the more I look, the more this look like a natural progression of the league. High usage players have been getting a lot of shine in the past few years, and this season is an extension of what’s already been happening. My interpretation is pretty simple: it turns out you can jam the ball in your best player’s hands every possession and good things will happen.

This doesn’t mean those players will necessarily shoot every time down the floor, but the ball always finding them means the defense always has to react, and that’s where the Celtics seem to get caught. Back door cuts and long swing passes around the perimeter have hurt the Celtics this season for reasons I’ll cover soon.

The Daniel Theis/Tristan Thompson Lineups

Before Kemba Walker’s return to the starting lineup, the TT lineup was about -20 in their time together. This is what happened after one—ONE—game against the Cleveland Cavaliers:

  1. The obvious takeaway is that the Cavaliers’ early high defensive marks are absolutely fraudulent (they gave up far fewer points when their guards were injured).
  2. Kemba’s return seemed to unlock something that made that pairing work. Theis and Thompson are both fantastic screen-setters, which is the perfect tool for Walker’s skill set.
  3. Maybe these lineup stats are useless if they can change so drastically after one game. Alright, maybe not useless, but take them with a grain of salt. Any stat that can flip on it’s head after 20 minutes might not be the indicator of success we think it is, at least not this early in the season.
  4. Either way, that defensive rating is an atrocity. My take is that the Celtics are much more vulnerable on the perimeter with TT on the floor, specifically on shots created by long passes around the perimeter. Theis is a solid perimeter defender, but doesn’t have the speed to catch up with swift passing. Thompson, well, hasn’t impressed me. There’s plenty of time for him to figure it out, but he hasn’t done much on either side of the ball with any consistency. At this point, I’d rather start Robert Williams. Or Grant Williams. Or Semi Ojeleye! Anyone with the mobility to switch and close out quickly will do.
Boston Celtics v Detroit Pistons Photo by Chris Schwegler/NBAE via Getty Images

Injuries

Bostonians don’t accept injuries as excuses to lose games in any sport, and you know what? You’re all wrong. Playing games with fewer good players makes the team worse, alright? I don’t make the rules.

You could say this team is cursed to eternal injury woes, but you could also say that a short offseason adds a lot of unnecessary risk for teams that went on deep playoff runs and none of this is surprising. The 2011-12 lockout season’s schedule was reduced to 66 games, but with more games on average every week. This article has a lot of data terms that are entirely over my head, but it’s pretty clear that a condensed schedule leads to more injuries.

Playing through one stretch without Tatum hurts the defense for obvious reasons, but missing Kemba for extended time hurts as well.

“But Jeremy, Kemba isn’t exactly the strongest defender.”

Sure, but one-on-one defense isn’t the only factor here. For example, Kemba has the lowest turnover percentage of any of Boston’s guards. It’s a much smaller sample size, but it’s still consistent with Kemba’s career. The only consistent rotation players (meaning we’re excluding Carsen Edwards for now) with lower turnover rates are Brown, Tatum, and Ojeleye. Teague and Smart have only been marginally worse, but those turnovers still add up to easy points on the other end.

Factor in Teague’s dreadful offense and you can see why adding Kemba makes such a big difference.

Smart’s injury is the next test. Actually, test is too generous of a word. This is trial by fire, except the fire is the plague and the trial is Brad Stevens figuring out if they’re better off playing guard-less lineups or getting what they can out of Teague and Tremont Waters. It’ll probably be a mixture of both.

Waters’ numbers have been awful in his limited opportunities, but I’m still a believer. His 35.1% assist percentage is easily the best on the team. His 39.0 turnover percentage is perhaps one of the worst statistical feats I’ve ever seen.

I have no short term solutions to the injury bug. The TPE will continue to stalk in the shadows and strike when we’re not looking.

Also, nobody in their right mind is making a trade to cover a 1-2 week injury.

Romeo Langford should be back soon, who deserved more credit as a strong team defender last season. Not much else to say there.

Turnovers

This is the real issue if you ask me. Even when the turnover differential isn’t bad, some of the mistakes they make are really, really careless. For the record, I’m counting “scoring on your own basket twice in a one-point loss to your hated rivals” as a turnover problem rather than a rebounding problem even if it isn’t rational.

Even so, I don’t have to fudge the numbers to talk about how the Celtics have played their way out of multiple wins by not taking care of the ball. They rank 23rd in the league with over 15 turnovers a game, but the numbers don’t really do it justice. They went on a 7-0 turnover run in the fourth quarter against the Lakers in an extremely winnable game. They won the turnover battle 15-14 overall, but it doesn’t matter if one string of mistakes unravels an entire quarter.

For what it’s worth, the turnovers don’t seem to have a strong correlation with wins and losses:

Turnovers aren’t exactly a defensive metric, but it goes without saying that giving the ball away gives the other team more chances to score. Even when the defense is good, the extra possessions kill them because of how many good transition players there are in the league now.

Touching on the return of Romeo again, he also had a low turnover rate (10.8%) last season on very low usage. Now, an increase in usage is what could potentially change that, but Romeo’s strength as a slasher comes from his pretty solid handles with the ball. I don’t envision him as a distributor when he comes back, but he’d at least make some nice plays around the basket instead of getting stripped.


I’m not sure where the light at the end of the tunnel is. This has obviously been a weird season and all I can say is that it’s just going to keep being weird. All I can hope for is that Brad spreads the minutes around on this 16-games-in-27-days gauntlet in February so that everybody can make it through in one piece. The Celtics look too good to be 10-8, but the entire league (outside of the current top four) looks so unstable right now that I can’t worry about win/loss record. Not yet.