LIneup trends around the league

[Crow]

Wizards most used lineup was -34pts / 100p before a trade finally stopped it.

9 of 10 most used negative.

20 of 20 most used pairs negative. Only 1 much better than -10, at -6. 5 were near or worse than -20.

394 lineups used. 7 above the equivalent of 1 minute per game.

About as bad as could be done.

[Crow]

Griz, 4 lineups much over 0.5 minutes per game for season. Top one is short of 3 min / gm.

Total is up to 634. Average lineup use is less than 5 minutes for season.

[Crow]

I tend to cite lineups per minutes per game for season because it is understandable and I think good lineups "should" play most of the time.

But the reality is that few big minute lineups play more than half the games in a season.

Over last 5 years, the league has 4-9 lineups that average the equivalent of 7+ minutes per game for season. But most play half or less games, so they are getting 14+ minutes per game when used.

Very few seem to pay much attention to lineup detail. But I mention anyways for now.

[Crow]

I will review this thread more, as I do when I have time, to check my work and extend it. Maybe comment more later.

Anybody have any new comments on any of the posts, if interested in the topic? Do you have an on-going analysis on league lineup decisions and trends? Most important points to you? Anywhere you turn to for that? Links?

[Crow]

61% of lineups used over 100 minutes across the league are positive. Decent but not really amazing.

There are only 20 lineups averaging over about 3.5 minutes per game. Down almost 1/3rd from March 2024.

Teams average a bit less than 7 lineups tested for 50 minutes or above for season, two weeks later than last season.

The gap between average performance of biggest minute Thunder lineups and the dinks and super dinks has gone up massively from last season. A great set of averages but could be better orcwsy better with less emphasis on dinks.

[Crow]

Crow
Wed Jul 19, 2023 3:29 pm

"Imo, the ideal distribution of minutes to lineups for a team trying to be it's best would be something like:

1 lineup of 800-1000 minutes (your best performer)

2 lineups of 400 minutes (your best alternatives)

4-6 of 200 minutes

maybe 5 more at or near 100 minutes

then about 500-1000 minutes of situational calls.

Last season, only 3 teams had a lineup used 8 minutes per game for season. Only 12 had a lineup even over 400 minutes as a top lineup. Instead of 6-8 lineups at or above 200 minutes as in my model, the actual average was 0.7. Instead of 12-14 lineups over 100 minutes, the actual average was 3.3.

My model might not be exactly right for every team, especially if heavily affected by injuries or change, but the actual average is way short of desirable for every team. There would be a middle ground, giving some priority to development of / experimentation with young or new, but that middle ground of concentration would still be higher to far higher than the current norm."

This is the start of a model that could be built team by team to test the theory of greater concentration and more exacting selection.

Could with time by redone for a handful of teams or more or hypothetically all. Could do at start if season and compare as able but incompletely to actual. Could be redone several times during a season based on new data and new thinking. Reviewing and revising the lineup distribution several times based on actual data to date in a formal, rigorous way is part of my lineup management philosophy. Not a fan of just winging it moment to moment beyond 1-3 lineups.

[Crow]

Wed Jul 19, 2023 5:46 pm

"The simplest way to improve is probably just to take the best coach chosen lineups and increase their utilization greatly...

Close to 60% of the most used 100 lineups have been used less than 200 minutes, Any over +5 or +10 should probably see 50-100% increase in utilization (in perceived favorable circumstances based on split analysis) for awhile and see what happens. There were almost 20 cases in league were a +7.5 lineup over 100 minutes failed to get 200 minutes. Warriors 1, Celtics 1, Heat 3, Bucks 2, Griz 2, Suns 1, Lakers 1, Clippers 2... Nuggets 0"

82% of the most used 100 lineups have been used less than 200 minutes to date, so the end result is likely close to what have been observed in past.

Almost 15 current cases in league were a +7.5 lineup over 100 minutes has so far failed to get 200 minutes.

[Crow]

"95.5% of the champion's net playoff margin has come from their 5 most used lineups."

Worth repeating.

[Crow]

In regular season and playoffs Holiday - Tatum, no Brown has been twice as good as with Brown but the full trio played twice as much in regular season and 3x in playoffs. That appears to be a monstrous inefficiency...

This season Holiday- Tatum, no Brown is still better than the trio but only slightly. The news is that the minutes switched from the full trio playing twice as much in regular season as without Brown to now 2.3 times as much for without Brown.

Maybe the team did similar research, maybe Mazzulla did it on instinct. I don't know if they saw this. But they followed the mesage of this post.

In other news, Brown or Tatum without Holiday and the other is almost twice as good as 2 or 3 together. Will the significant usage of such continue in playoffs? It definitely did not last playoffs despite the regular season results showing the same thing on pretty healthy sample sizes.

[Crow]

Coaches, teams can "care about" 1, 3, 5, 10 or 20 lineups. Testing "adequately" is another thing; but they should decide if they care first, then try to test or plan to test. It can't be more than 20. It would be difficult to care about and test more than 10. And they don't. Elite teams test an average of just 3.6 lineups over 100 minutes over a regular season.

At game level they can / will use about 10-30 lineups. They could be the same or quite different and coaches choose the latter, when imo they should mostly choose the former. The average team uses just 16 lineups for 10+ games. Less than 3 for 20+ games. Do the Coach and team know or care? I don't know if they know, but I conclude that they don't really care enough to change the pattern. 0.4 lineups per team get 40+ games. Just 1 got 60+. (Thunder starters.)

Lineup analysis & management matters a lot whether the effort and care is there or not.

Coaches substitute players. It changes lineups. Sometimes for micro-management of matchups, sometimes for rote player minutes management. Have a 10 lineup game plan? That actually matters or lose yourself and any plan 6-10 minutes in? I don't know. I generally doubt it based on the data. Should they? I'd say so. Doesn't absolutely have to be same lineup plan every game but it should be different for good reason, a better reason than the general reason of trying to get half-decent testing done for evaluation in advance of playoffs.

About 70% of lineups used 20+ games are positive, 40% very positive. Worth finding and using more. Way more.

If a team uses a lineup less or way less than 10 times in a season, they aren't serious about it. Even 20 times is pretty light.

More lineup management philosophy.

[Crow]

Current Blazers starting unit is positive. Less than 100 minute test before tonight. Almost as much use in last 5 games as previous rest of season. Concentration came, but quite late. Will it stay?

Ayton, Grant and Hendrrson have mixed results with this core. Ayton more often positive than the other two.

What's random, what is signal? Far more disciplined lineup management would have some value.

[Crow]

Estimated help or harm of teammates in a particular lineup available at databallr.com, via wowy selection and constructing and clicking on a particular lineup.

[Crow]

Grizzlies reach 777 lineups used before firing Taylor Jenkins. Almost 150 brand new in last 3.5 weeks. As if previous 600 some weren't enough.

3 of 5 most used negative. Biggest at about 2.5 minutes per game for season. 2 more over 1 minute / gm. 1 more over 0.5 minutes. 773 super-dinks.

Will this behavior change? I hope so. Radically. Was it part of the decision-making? I don't know. Not likely to hear much. What counts is whatpositive.


18 of 20 most used pairs were positive. But about half (including 3 of the 4 most used) were considerably stronger than the other half. Raw data is not the last word but there were / are things to investigate further and evaluate and perhaps alter.

777 lineups would be a bit more than trying EVERY combination in a 4×4×4×4×3 matrix by position. That does not show strong judgments. It does not produce familiarity in specific lineups or much even in families of similar lineups. It did make coaching decisions prominent. Maybe too much so.

To state the absurd: on average lineups were used 4.5 minutes for season (or take it to another level- less than 4 seconds per game).


But being an active coach might attract some teams wanting action. If openings are made, he might be a candidate for Raptors, Wizards, Magic, Pelicans and maybe Suns.

[Crow]

First game with new Griz coach:

24 lineups used only 1 above 4 minutes. Modestly less concentrated than the game before that. 10 never before used as usual. At least for 1 game it is the same style lineup distribution trends.

Edey back with starters and he and they did very well.

Game lost in the lineup management weeds.

[Crow]

7 more never before used lineups in new Griz coach's second game.

Only 1/3rd of the 18 selections were positive.