Do people not realize the Celtics have been a ~13 SRS team for almost half a season now?

ardee · Post #1 » by **ardee** » Sat Apr 9, 2022 9:58 am

We're talking GOAT team level for 35 games now. They've been absolutely mauling teams, elite teams. If Timelord comes back I can't see them not making the Finals.

Have we actually seen anything like this before... ever?

(PS there isn't a way of calculating SRS for a stretch of games that I know of. Their point differential, however, for this stretch, has been 13.7. Their point differential for the season is 7, and their SRS is 6.7. That means their SOS for the season is -0.3. I don't know of a way to calculating for a stretch of games that I know of, but even if you assume their schedule for this stretch was TWICE as weak as the rest of the season, that still comes out to a 13 SRS).

70sFan · Post #2 » by **70sFan** » Sat Apr 9, 2022 10:01 am

35 games is way too small sample to call it GOAT-level. 1972 Lakers almost literally didn't lose games in such a span for example.

That said, we have to closely look at them in postseason. They are the main threats for the Bucks on the East.

Owly · Post #3 » by **Owly** » Sat Apr 9, 2022 11:20 am

'71 Bucks posted a 13.87671233 points diff through 73 games (more than double the Celtics sample). And did so with a relatively tough SoS for a great team (SoS skews low for great teams because they don't have to play themselves, incidentally Celtics SoS will be lower and take a greater dent out of their points margin for SRS if we assume they are a circa 14 dif team though in a larger league the impact will be less).

ardee · Post #4 » by **ardee** » Sat Apr 9, 2022 12:21 pm

Owly wrote:'71 Bucks posted a 13.87671233 points diff through 73 games (more than double the Celtics sample). And did so with a relatively tough SoS for a great team (SoS skews low for great teams because they don't have to play themselves, incidentally Celtics SoS will be lower and take a greater dent out of their points margin for SRS if we assume they are a circa 14 dif team though in a larger league the impact will be less).

What do you think the Celtics' real level is? 9? 10?

Owly · Post #5 » by **Owly** » Sat Apr 9, 2022 12:52 pm

ardee wrote:
Owly wrote:'71 Bucks posted a 13.87671233 points diff through 73 games (more than double the Celtics sample). And did so with a relatively tough SoS for a great team (SoS skews low for great teams because they don't have to play themselves, incidentally Celtics SoS will be lower and take a greater dent out of their points margin for SRS if we assume they are a circa 14 dif team though in a larger league the impact will be less).

What do you think the Celtics' real level is? 9? 10?

Don't know. Haven't seen enough of them. Don't know if there's a reason to throw out early season sample.

Normally I'd default to season SRS as baseline (and cautiously adjust from there) but
1) RS seemingly means less these days (also Covid)
2) that is a very uneven split, so I'd want to check to see if the different samples are noise and we should split the difference or they represent distinct e.g. (personnel/health different? rotation different? lineups different? coaching/execution different? etc) teams (or versions of a team). Are statistical differences things prone to luck-based fluctuations or not (and if so does the evidence suggest it was more good luck in the second period or bad luck in the first)? etc

But short version, I wouldn't claim to know.