2018-19 lineup analysis

Home for all your discussion of basketball statistical analysis.
Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Wed Apr 17, 2019 2:28 pm

The worst performing playoff lineups used over 10 minutes so far belong to Pistons, Nets, Magic, Thunder and Spurs. They range in performance from very horrible to insanely horrible... in short minutes.

The biggest minute lineups are pushing 18 minutes per game. Helping prove that there is a rationale for lineup concentration that coaches generally don't bother with in regular season. Raptors have the best one so far amongst the biggest minutes, then Sixers and Thunder bench with George.

2/3rds of the lineups used 10 plus minutes are non-negative. (Only 1 is exactly neutral. I included it this time to get the simpler to say full 2/3rds.)

Dink lineup performance is worse right now at 28% of them positive compared to 30% in regular season.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Wed Apr 17, 2019 3:17 pm

Blazers, Sixers and Nets are the first teams to use 40 plus different playoff lineups. The average is about 30 but not everyone has played 2 games yet.

Nuggets and Thunder are only teams to have 3 lineup selections used over 5 minutes that are negative so far.

About 60% of the lineups over 5 minutes that have a 3pta advantage were positive. About 85% of these lineups with a FTA advantage are positive. Slightly more than 85% with a 3pt fg% advantage won their time.



For playoffs in general (all lineups), shooting percentages are down from field. FTAs and FT% are up modestly. Scoring is down by an average of 5.5pts. Assists down 2. Pace down by a bit less than 1 possession / gm. DR% up 1 percentage point.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Sun Apr 28, 2019 6:05 am

Thunder playoff lineup analysis (with all games):

52 lineups used, only 23% cumulatively positive.

Two most used lineups played half the minutes. That is good concentration. Unfortunately these lineups performed badly, accounting for about 2/3rds of the net point deficit. Add in a third lineup with Schroder in with4 starters and those 3 lineups account for basically the entire point deficit.

Outside one George with all bench unit none of the other lineups really did that much. It was the only lineup that did better than +3 points. With this one exception the rest of the Thunder lineup management can be characterized as bad on top and mostly bad on the bottom.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Mon Apr 29, 2019 1:18 am

Spurs playoff lineup analysis:

3 of 5 most used very bad to horrible results.

28% cumulatively positive is consistent with losing.

68 lineups used seems pretty high, excessive imo.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Sun May 05, 2019 4:28 pm

Brad Stevens used his starting unit 10 minutes per game so far against Bucks. 12 minutes in game 3.

Not enough. Not anywhere near enough. As one of the few successful choices it should have been way more. It could be 30 plus given the desperate need but it should be at least 20 plus. Instead he used 23 lineups in games 3 and only 5 were positive.

Getting Smart back could help, if he is ready.

Dinking a lot with 3 or fewer starters is very unlikely to cut it. They probably should max out on the right 5 and just a couple of variations with 4. Do all or nearly all bench a few minutes if you have to. But max out the right 5.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Tue May 07, 2019 12:04 am

In 2019 playoffs, several first round exit teams used "just" 50 lineups. That is still a lot. Probably way too many. Only 7 playoff teams managed to get positive results from 30 plus % of their lineups cumulatively. The Bucks were highest at 37% positive. The Magic at the bottom with just 20% positive. 2 teams in second round are already over 100 lineups used. Three more are certain to get there soon. 2 more are likely to get there. Only the Raptors might not if they lose the series and sooner rather than later.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Wed May 08, 2019 10:28 pm

Stevens has used 61 lineups against Rockets.

Only 2 used every game. The starters and then the 3rd most used lineup, which happens to be doing terribly. Only 6 more used in 3 games. So 53 lineups used in 1-2 games. Apparently not essential to use every time.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Fri May 10, 2019 10:37 pm

Every team still left in the playoffs have a lineup used 100 plus minutes better than the Blazers' (whose most used lineup is the most used of anybody although it is -6 per 100 possessions)... except the Bucks. The Bucks most used lineup has only been used 52 minutes. I don't know why. Do you?

Raptors and Sixers have the best performing big minute lineups. Even against each other, though the Raptors' most used lineup is twice as strong. One or both must be stomping rotation lineups. Sixers getting positive results from just 23% of lineus. Raptors just 28%.

Pistons went out quick. Didn't use any lineup much over 4 minutes per game. 8 of the 10 most used were awful to horrendous performers. One neutral, one good. Total smash on the top. The dink lineup win % was recent but what happened at the top settled it. Would be pretty hard to worse.

Nuggets so far have about as strong and consistent a set of lineup and sub-lineup results as you could hope for.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Mon May 13, 2019 7:40 pm

https://twitter.com/dr_ilardi/status/11 ... 8448761856

Sixers without Embiid lineup management.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Tue Jun 18, 2019 1:15 am

I've noted elsewhere that 4 of the 5 most used Jazz lineups were horrible in the 2019 playoffs. I just checked and in 2018 playoffs 6 of 7 of Synder's lineups were negative with half being awful. In 2017 it was 6 of 9 negative and almost all awful. The guy has a really bad record for most used playoff lineups.

And you heard it from who else?

Does Lindsey know, care? if not, why not? If it happens again, what then? He said they have a very good team but not a great one. Do they have a very good or great coach? Analytics department? GM?

eminence
Posts: 137
Joined: Sun Sep 10, 2017 8:20 pm

Re: 2018-19 lineup analysis

Post by eminence » Tue Jun 18, 2019 3:58 am

Lineup usage may or may not be an issue for the Jazz, but it's hard to say how much in light of the talent disparity between them and the Warriors/Rockets. Also unfortunately had their fair share of injury issues in the playoffs.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Tue Jun 18, 2019 2:19 pm

You are right that these adjustments would be appropriate. They could be estimated with more analysis. But it is unlikely, given how bad the data is, to get them to ok much less very good or better. It is also 3 years in a row. Regardless of the factors, there is the goal and how far they are from it, with the options being the status quo or change.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Tue Jul 02, 2019 9:34 pm

75% of the 20 most used Raptors playoff lineups were positive. The most used was great. Good job Nick Nurse and Raptors. In the finals it was 65% positive. The most used lineup sucked but 8 of the next 10 were great.

The Warriors were 65% positive for most used 20 but the most used was very modestly positive. In the finals it was 45% positive and the most used was moderately negative.


For full playoffs Celtics' most used 20 were 45% positive. Bucks 55%, Sixers 45%. Rockets 70%. Blazers 65%.. Nuggets 45%

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Sun Jul 14, 2019 10:18 pm

Worst by vast margins playoff lineup used over 50 minutes was the Magic's D. Augustin | E. Fournier | A. Gordon | J. Isaac | N. Vučević. It was mildly positive in regular season but still only 3rd best performer amongst the 10 most used.

They really went with it in game 1 for very large minutes and it sucked. They reduced minutes in game 2 but still made it the mainstay and it sucked even worse. Did they change in game 3? No and it sucked again. Change in game 4? No. It had a good game. But facing elimination with 3 bad games and 1 good, what did they do? They played it fairly big minutes in game 5... and it sucked again.

I dunno if anyone in organization argued for going away from this lineup (coaching, players, management, analytics), but they didn't.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2018-19 lineup analysis

Post by Crow » Sun Jul 14, 2019 10:27 pm

4 of 5 most used Blazers playoff lineups were negative. Two badly, one horrendous. The most used playoff lineup was already performing badly going into the series with the Warriors.

What did they do? They went with it full throttle in game 1... and it sucked. They cut the minutes in game 2 but it still sucked. They substituted Leonard for Kanter in game 3 and got good minutes but not enough. In game 4 it was ok but they were overpowered elsewhere.

They didn't have a chance to test the starters with Kanter in regular season but should they have stayed with it or guessed different in game 1? Or the second half of game 1 or game 2?

Have to adjust, adjust late, you usually lose.

Post Reply