The Sixers core in lineups

Home for all your discussion of basketball statistical analysis.
Post Reply
Crow
Posts: 6134
Joined: Thu Apr 14, 2011 11:10 pm

The Sixers core in lineups

Post by Crow » Sun Dec 30, 2018 6:12 pm

The current starting lineup of Simmons Redick Butler Chandler and Embid does fabulous at plus 15.9 pts per 100 possessions.
(All future performance data is per 100 poss.)

All combination of 4 of 5 are strong. The two not including Simmons or Butler are somewhat stronger than not including Redick or Embid.

BER, BES are mildly positive. ERS, CER, BCS, BRS, CRS, CES, BCE are incredible. Are there other trios of these 5? If so, they are being uses way less or not at all.

4 core minus Redick or minus Simmons are both horrendous (-44 and -40) is very short minutes. Probably not that bad but they aren't trying it.

4 core without Butler or Embid tried a little more and fabulous (plus 31 for both).

B E S without Redick is horrendous in modest minutes (-37). BES without Chandler is ok (plus 3) in near 100 minutes. BER without Simmons is great (plus 20) in modest minutes. BRS without Embid is mildly positive. There are more trios without somebody. I may backfill later.

For pairs BE, BS are moderately positive. ER, ES, RS, CR, CS, CE, BC, BR are great.
For pairs without, BS without E is moderately positive. BE without is great. ES between moderate and great. There are more of these to do.

I haven't dealt with non-core role players.

Any team that does do this much isn't imo studying lineups enough. Need all this and more, at raw plus minus level, RAPM for lineups, offense, defense, factor level and other discrete stats. I did 36 rows of lineup data. Probably should be 50-100-maybe 150-200 rows. For 1 year and past where applicable.

Are analytic office always and constantly doing this, in some cases or rare to none? I dunno. Are coaches fully aware of this much detail and using it? I dunno... but I doubt it most cases, maybe all.

Have you ever seen all these levels listed in public by anyone? I haven't. I've done close to this sometimes in past but I took it further to make a point and to make the additional point that it needs to go way further. For that more time would be needed. Teams should do that. Probably ahead of some of these they are busy with instead. It would be and is a fundamental mistake to not work lineup data very, very, very hard and thoroughly. And yet if is probably the norm I think. If you KNOW different, say so, please.

Once you have this data, then the task shifts to studying it for tens and hundreds of hours. Then develop policy suggestions. Then study them some more compared to alternatives. Make decisions. Explain the data or at least decisions. See what gets done. Try to get more of the recommendations followed. Assess what was and wasn't done and what happened. Then do it all again as soon as reasonable and practical. How many teams have a full-time lineup management analyst? Might need more than doing nothing else. How many teams have this? Is it 1, 5 or more? I dunno. Maybe some major analytic writer will as around? That would be helpful... if the teams would admit to such coverage or lack of such coverage. They probably won't answer. They probably won't be asked- at least beyond my doing so here and now, fwiw.

Crow
Posts: 6134
Joined: Thu Apr 14, 2011 11:10 pm

Re: The Sixers core in lineups

Post by Crow » Thu Jan 03, 2019 3:40 pm

Analysis: Sixers core generally works together. Issues are revealed by looking at sublineup data and the rarely seen lineup and sublineup data without a player. I know that technique is available at pbpstats.com and nbawowy.con before that and probably used by enthusiasts but I don't recall any major media using it. In sixersscience's write up of how the Sixers do analytics it wasn't specifically mentioned.

More issues are likely to be found on other teams with this method.

Post Reply