Here at end of March 2021, the Celtics are below .500, and they're just not a very good team.
Overall the Celtics are 12th in offense (111.1 ORTG) and 22nd in defense (110.6 DRTG).
That's a frankly astounding drop from last season. Not only are they significantly worse offensively compared to the 5th best offense in the league last season (110.4 ORTG), but their defense has gone from elite to just plain bad. Last season the Celtics had the 4th best defense in the league (104.5 DRTG).
Could Gordon Hayward have made this big a difference? Losing him certainly hurt, but I don't think it really explains it. Hayward basically didn't play in the playoffs last year, and yet the Celts had a very similar ORTG (108.5, 7th out of 16) and DRTG (103.5, 2nd out of 16) during that run.
So what gives?
I decided to look at a couple of metrics to see how things have shifted. Neither of these metrics is perfect, but it at least gives some sense of what's going on.
RPMThis stat attempts to adjust for circumstances while providing an idea of the impact a player makes while he's on the floor, positive or negative.

(bolded names no longer on the team)
Things that jump out:
- More players in the red than in the green. Not a shock. But that was true last year as well. This is a top heavy roster. Last year that didn't matter so much. Maybe it just hurts them a lot more this year because Tatum, Kemba, and Smart have been worse.
- Daniel Theis has been much less helpful this year than last, but has still been a consistently useful player for the team. Tough loss. By comparison, Tristan Thompson and Rob Williams both look bad by this metric. Although it's worth noting TT has been much worse for the Celts than he was on the Cavs, despite the fact that the Cavs were a deep lottery team.
- As you'll see below, Rob Williams is the player on the team with the biggest gap between production and positive impact per RPM. I think that should tell you something about why he's had difficulty getting so many minutes despite looking exciting and causing fans to clamor for him. Rob makes a lot of highlight plays but he also is still really raw. Daniel Theis is sort of the opposite. Polished and useful but limited.
- Fournier and Kornet seem like they ought to be helpful.
- Brown and Tatum are clearly the best players on the team -- no surprise. Brown has taken a leap, as you'd expect, while Tatum is not quite as good as last year. That seems right as well.
- Kemba actually is a little bit better in this metric than last year, which is surprising, but shows you his importance to the way the roster is constructed. Marcus is a positive, but worse this year than last; again not a surprise.
- One surprise here is Carsen Edwards solidly in the positive. I guess that suggests (1) small sample and (2) shooting and quickness are things much needed in the rotation.
- Grant Williams is bad. I think he looks better than he is. He seems to do the right things on the floor but the team seems to be consistently worse with him out there. I'm growing more and more negative about Grant's chances of being a contributor.
- Semi is bad as well, although to some extent I wonder if that's because he's backing up the best player on the team (though RPM is meant to control for that).
- The backup point guard situation seems bad. Backup guards often look bad in this metric, but the Celts really don't have a guard who is giving them a lot of positive value by this metric.
- Luke Kornet as consistently looked good by this metric over his career, and he's already seemed like a useful player for the Celtics. Hopefully that continues. Maybe he can stick around in a reserve role for minimum money.
Win SharesThis metric is supposed to measure productiveness as a contribution to winning. Players on winning teams garner more win shares, but it's not as straightforward as that. Good players on bad teams can get a lot of win shares, and the inverse is true as well. This metric tends to favor big men and penalize guards due to rebounds and field goal percentage, but it's still useful.

Similar picture here.
- Theis was a big positive last year, less so this year. But still a tough loss.
- Walker looks much worse this year than last by this metric, as you would expect. He's been far less productive overall.
- In general there's a drop-off across the board.
- The hope should be that losing Theis and replacing him with Fournier is a net positive because Rob Williams plays more minutes and thereby replaces Theis's production and then some.
- As I look at both of these lists I really think that the Grant - Semi - Nesmith - Langford backup wing/forward situation is the soft, undercooked part of the rotation that is really undermining things. You wouldn't think that would matter given that the best players on the team play at those spots and get a lot of minutes. But this roster and the team's system are constructed to emphasize wing play. They just need more productive, reliable wings, period. Again, you gotta hope Fournier helps here and that Langford comes back and shows something. Grant, Semi, Nesmith -- I'm not expecting anything positive from these guys at this point, to be frank.
- I think it's interesting that the WS based on previous season production suggest about 41 wins and that's where the team seems headed (82 game pace). With that said, I took a look at this for the past few seasons, and there has been a consistent trend before this year of the Celtics ending up with a good number of wins over and above the previous season win shares of the players on the roster.
- For example, heading into the 2020 season, the cumulative WS of the players from the prior season was 37, but the team ended up winning 48 games, which was a 55 win pace in a shortened season. Similarly, in 2018 there were 39.1 returning WS on the roster, and the team ended up winning 55 games. Point being that fans who expected the Celtics to win 50-55 games despite not seeming to have that many wins on the roster were not wishcasting or looking at things with green goggles. They were basing their expectations on previous performance.
Ultimately, I still can't explain the huge drop-off, especially in defensive production. How do you go from top 5 to bottom 10 on defense despite the fact that Enes Kanter
isn't in your rotation anymore?
Injuries to the best players, and a general drop in production across the board, help to explain it. But it's hard to escape the thought that perhaps the team is just less focused, less motivated, less "into it" this year.
My theory, which is totally not based on any evidence and is mostly just projection:
Due to the weirdness of this season and everything going on in the world of the past year, it is harder for everyone to be as focused and serious about their work. We've all had our priorities jostled and challenged. It's a lot easier to be distracted and to feel like maybe what seemed really important a year or two ago isn't as vital now.
I suspect that dynamic has affected some players more than others. I suspect for some players, it is harder than ever to see basketball as "serious business" instead of a game that people play and watch for entertainment.
Maybe the Celtics have been hit by that dynamic harder than other teams. Maybe the youth of the roster is a factor there.
The only evidence I really have is that the Celts have dropped off on defense much more than on offense. Defense seems like more of a focus / effort area than offense, where you can get by on skill and talent.
But really, I don't know.
