APBRmetrics

Mostly on offense.

https://twitter.com/bballstrategy/statu ... 6qndOINbTw

Celtics used 1 lineup last regular season more than 1.5 minutes per game. It was great but used far too little in regular season (5 min / g) and playoffs (7).

2 other most used playoffs lineups were used under 1 min / g in regular season and way way under. One bad, one barely positive. Both ended up sucking. Poor testing, Poor / virtually blind selection, poor results. Used too much for what little they knew / got from them.

Celtics currently using 4 lineups for 1.5 / plus minutes per game. Most used at close to 7. Better but far from optimal. The change suggests to me acknowledgement to some degree of the extreme inappropriateness of last season's lineup distribution.

Marcus Smart had a horrible October, strong November and now a pretty bad December (worst offensive rating month, best defensive rating month but only a little better than league average).

6 of 9 seasons has been negative on BPM. But BPM has issues, especially on defense.

DARKO projections has been pretty steadily up for his career at currently stands at +2. But he did slip below his trendline early this season before recovering. The average rate of projection increase though has never been lower though still positive.

Things to watch over next 6 - 12 months. They do have options but probably aren't changing soon.

EPM and DRIP are in between DARKO (highest) and BPM (lowest). Use an average or weighted average or pick your own point or give up trying to evaluate or settle for estimating him probably somewhere in the range of the 2nd highest quartile.

Celtics most used lineup down to 5 minutes per game from 7 min/gm earlier. Lineups over 1.5 min/gm down to just 3 from 4. Wrong direction, imo. Doesn't really matter in their current situation, except bad habit to be this low concentration, this scattered.

High minutes concentration in regular season for some other teams suggests less change in rotation quality going into playoffs than most other contenders. Should look at that on 538.

Celtics have plenty of room to increase top lineup concentration, if they decide to.

Celtics most used lineup down a little more to effectively 4.5 minutes per game for season. Still just 3 over 1.5 minutes per game.

Smart - Brown very mildly positive in over 1000 minutes. Brogdon - Brown slightly negative in about 650 minutes. Most used quad is flat neutral.

Updates to original comments on twitter as referenced in first post of this thread:

Actual wins above expected for first time in long time.

3rd best win% against Sagarin top 10.

Still very low on 4th quarter scoring (25th place) this season.

Nothing special on points scored directly at rim or further out in paint. Quite weak on ft rate. As they were last season. Average on fast break points. 3pt fg% up to 8th. Will probably need to keep that strength up given the other rankings.

No longer consistent across 8 factors. 5 elite- 8th or better. 3 now below average.

I won't update some of the other player specifics.

8 man rotation tonight in a double overtime game? 3 played well. Celtics drop 3rd in last 4 games.

4th quarter scoring issue, another overtime, another loss.

Starters with G Williams, slightly negative for season. With Brodgon or R Williams, terrible. Both near -16pts / 100 p and both in 5 most used. The design and coaching perspective calls for these but the results don't.

The 5 best performing Celtics lineups close to or above equivalent of 1 minute per game for season all have D White. 1 each for G Williams and R Williams, none for Brogdon.

D White at 28 min / gm.

Smart - Brown is the least strong raw pair in 15 most used and 10pts / 100p worse than last season. Worse net team shooting and rebounding among other things. Almost 1400 minutes of use this season. Modestly better lately and not really "fully fixed".

Celtics with only 1 chance in 4 at #1 seed in East according to BRef. Will that be very important?