APBRmetrics

Celtics lineup results by tier:

5 most used, 812 minutes at avg of +8.1 pts / 100p.

Next 10 most used, 448 min. at +17.7pts.

Other 300, 1260 min at +6.3. These super dinks have been used more than 50% more than the 15 most used and for barely more than half as positive the average result (+6.3 to the average for the top 15 of +11.5).

Limits to playing time of top guys but could overall results be enhanced? I'd say likely. Got to find / go for the optimal mix of starting unit, lineups with 3 or 4 starters and other stuff if necessary or useful.

Should some of the top 5 play less? Probably. Some of the next 10 play more? Yes. Different mix of the rest? Yes. Eliminate many of the superdink 300? Yes.

How dink are the superdinks? Roughly 95% of all lineups have been used an average of 4.2 minutes each for the season.

Everything fine, let Mazzulla do it his way? Or modify to some degree? I am in the latter camp.

Mazzula used 87 lineups in last playoffs. That means 70+% of regular season lineups were not used. "Should" imo be less, probably quite a bit less. 5 most used lineups provide 112% of the net margin. Operating at +16.6 pts / 100p to the average of -2 for the other 82.

Is this lineup analysis meaningful? It is to me.

Does this actual lineup management look optimal to you? Good enough? Don't care?

In 5 minutes I took the 5 most used lineups, made 3 minor adjustments and found a lineup rotation for full 48 minutes and within every player's actual minutes.

Don't have to use just 5 but it can be done and the plan should probably be far from the actual. In the actual 87 lineups played 917 minutes across 19 games, the average lineup was being used barely more than equivalent of half a minute per game. Plenty of room for more selectivity, based on data or judgment or the pre-meditated combination of the two.

Is it not a quibble over a small difference. This is a huge difference with a huge difference of opinion with likely large impacts.

It might have a better chance of having impact if it was coming from the Director of Analytics, the Assistant General Manager or the General Manager. Or even one of the 5-10 most prominent public analysts. But fwiw, I'll say it.

Jaylen Brown, 4 straight years of decline by BPM and O-BPM from fairly modest peak for a "star" (+2.5). More mixed yr to yr on D-BPM but still ultimately down.

Celtics sell at desired price. Supposedly with 3 more years of current owner direction.

Reports are that proposed deal violates league rules on size of principal owner share.

Will variance be allowed or deal modified before or after submission?


The two part sale is another issue. Allow it again after Minnesota mess or change the rules? Two part sale does provide moderate burdensharing for anticipated luxury taxes in next 3 years, lower the real net price a bit. Higher valuation on final piece for presumed additional franchise value growth? Could new owner back out of that? Could existing owner want to stay, presumably after liquidating minority owners?

Questions. Fwiw.

Celtics opponent 3pt shooting:

https://x.com/bballstrategy/status/1907428832715628559

Most used Celtics lineup with Porzingis and starters are slightly negative. Far better team results in Porzingis bench pairs than starters. Playoff plans? Results?

Celtics dead last on FT rate in regular season but 2nd highest in playoffs (before tonight).

Mazzulla has used 43 lineups on 3 games. None in all 3 games, 4 twice, 39 just once. 5 for over 5 minutes and 22 for 2 minutes or less.

5 most used lineups for Celtics were +9 tonight. Rest were -10.

Celtics contention era paused or over until significantly revised?

Will be interesting if Simons lasts at all or for how long. Did not hear any welcoming praise yet but maybe after the deal is official?

On Jaylen Brown and 2 second maxes or supermaxes:

https://x.com/bballstrategy/status/1944483015897596128

Tatum - Porzingis was least effective pair among Celtics 20 most used in regular season and were each's worst pair in main rotation. Each was more than 10 pts / 100p better alone than together.

Was this ever mentioned in mainstream media?

With and without data shows that together was worse on offense and defense than with just one. Pair was far worse on defense together than with just one with either White or Holiday.

Together was only mildly less good than apart in 2023-24. What changed? What was done less well? More investigation would be required. It appears that the problem was pre-All-Star break and was much improved after, but still less good than apart.

They are apart now.

Celtics reportedly shopping Simons and Niang. Not sure what they could get. Expiring. Probably prefer similar but who knows what might be available. Eventually buyouts could come into play for tax savings.

Lot of talk of a gap year but getting back to top contention is not a sure thing for following year. Contention in east is easier task but a step lower than top contention.

Will be interesting to see next moves and results.

New personnel guy named to front office. Not sure of scope of duties but wonder how that affects future drafts. Wonder what leadership thinks of last 2, 5, 7 years of drafting.

I would not have picked Gonzalez (and not close at all) with Kalkbrenner and Broome still on board.

Penda trade package has some appeal but still could have had Kalkbrenner, Broome, K Jones or A Martin. Time will eventually tell which have better performances, fit, value. To my mind, Celtics went for more likely bench or deep bench guys than the forgone who I think had at least more likely potential for bigger bench minutes & impact and maybe starting someday.

The Celtics under Stevens with very poor relative draft performance by craftednba study.

Boucher was a good pickup.

The Niang trade helpful.

Not quite done with the salary shed but close.

Then onto adjustments for return to contention. Probably need something new to win another title.

Do they stick with Brown? Probably. I wouldn't. Have written on that several times, including fairly recently on Twitter.