7 lineups with your main star over a measly 50 minutes in a season?A very low request.
Not met by Coach D.
SGA only had 5 lineups over 50 minutes and only 1 over 102.
Fwiw, Tatum, 7 over 50, 4 over 100. Even that could be improved upon.
Knicks, 15 lineups over 50 minutes, 8 over 100.
Could you have 7 guys each with 7 lineups over 50 minutes? Absolutely, with focus. Should you? Absolutely imo. What is the excuse for not doing so? Prefer to be rando, sloppy, undisciplined?
Does gentile encouragement of concentration work? How many places, how much? Not enough progress.
Straight talk, coaches and organization are way too casual and inefficient with lineups. Especially in regular season.
I think it would be good to list 7 best lineups of top 7 players and prioritize their concentrated use to get them over 50 minutes, over 100 minutes. A way to start the improvement.
7 biggest minute lineups for Shai without Giddey is a start to the start. Fit Hartenstein and Caruso in them and / or after them.
Decide on your first 7 before or after lineup analysis? Probably work it both ways, constantly.
SGA's 7 biggest lineups without Giddey are combinations of 4 out of 7 players. Not that specialized or profound: 3-5 lineups with the main guys and 2 with K Williams and Jaylin Williams. Anybody could throw that out. Add 3-6 more candidates for minutes and many of these lineups will get even less than 30 minutes and will stay deeply in the "of unknown quality" range. So the playoffs will stay blind guessing after the starting unit.
If you think that individuals and system are enough information and lineup don't matter enough to take seriously, we disagree. 5 on floor create plays / offensive possibilities and the defensive strength or weakness. The differences realized in the various combinations can and often do matter in a game within frequently small margins.
Update: only teams out of 16 in playoffs without 7 regular season lineups over 50 minutes were Thunder and Mavs (who were impacted by major trades). Sixers barely met the criteria.
Do the Thunder know? care? about being an outlier? The answer appears no on at least the latter.
Even though 50 - 100 minutes is not statistically significant, it is the best they have and a minimal goal, to guess about future effectiveness or even just try to become aware of tendencies, good and not.
Thunder lineup analysis and other commentary
Re: Thunder lineup analysis and other commentary
In playoff overall, 3 most used lineups for Thunder were +37. But then the next 4 lineups gave 32 pts back to opponents.
In Mavs series, 3 most used went to -4. Next 4 remained bad at -16.
Other 52 lineups were apparently +20 to pull up to the net 0 but that was probably working from behind and not enough.
7 most used were close to 50% of total time. I'd have preferred more. But I also would have preferred far more concentrated testing in regular season to identify the best 7 and the best within the 7. Only 3 of the 7 were positive against Mavs. Not enough. Not enough proper testing or right guessing.
The performance against Pelicans may have too comforting overall and with starters. Starters did not work against Mavs and Coach D waited too long to change... and then ultimately guessed wrong with Joe. The starters lost badly in first and second games. Despite winning 1st game big, that could have been enough to consider change. But change not made in game 3 (change comes too late for many losing teams) which was again negative and accounted for 3/4ths of the loss. Change in game 4 but too much damage already done. New starters helped very modestly in game 4 win but lost in game 5 and 6, with the starters negative in game 6 being the decisive blow. Picking not "right" alternative and late led to end of story.
The desire would be for better top lineup choices in future playoffs and quicker and better change when necessary. Time will tell.
If Daigneault's orientation and talent is in micro-lineup choices, the organization may want to intervene in the main lineups or more than they have. Still have to prove to me if you can win deep in playoffs with micro-lineup gymnastics alone.
Fwiw, the original starters were bad against Mavs in regular season and chosen alternative with Joe was slightly worse per minute. Though obviously tiny sample, that was the information available to guess from. Maybe you don't change in game 1, but game 4 was too late. The chosen alternative with Joe was good against Pelicans but only 6 minutes test in a series that had room for more disciplined testing of select alternatives but instead, as usual, saw 45 lineups used with more than 75% being pure dinks (less than 1 minute per game), pretty useless for evaluation.
In Mavs series, 3 most used went to -4. Next 4 remained bad at -16.
Other 52 lineups were apparently +20 to pull up to the net 0 but that was probably working from behind and not enough.
7 most used were close to 50% of total time. I'd have preferred more. But I also would have preferred far more concentrated testing in regular season to identify the best 7 and the best within the 7. Only 3 of the 7 were positive against Mavs. Not enough. Not enough proper testing or right guessing.
The performance against Pelicans may have too comforting overall and with starters. Starters did not work against Mavs and Coach D waited too long to change... and then ultimately guessed wrong with Joe. The starters lost badly in first and second games. Despite winning 1st game big, that could have been enough to consider change. But change not made in game 3 (change comes too late for many losing teams) which was again negative and accounted for 3/4ths of the loss. Change in game 4 but too much damage already done. New starters helped very modestly in game 4 win but lost in game 5 and 6, with the starters negative in game 6 being the decisive blow. Picking not "right" alternative and late led to end of story.
The desire would be for better top lineup choices in future playoffs and quicker and better change when necessary. Time will tell.
If Daigneault's orientation and talent is in micro-lineup choices, the organization may want to intervene in the main lineups or more than they have. Still have to prove to me if you can win deep in playoffs with micro-lineup gymnastics alone.
Fwiw, the original starters were bad against Mavs in regular season and chosen alternative with Joe was slightly worse per minute. Though obviously tiny sample, that was the information available to guess from. Maybe you don't change in game 1, but game 4 was too late. The chosen alternative with Joe was good against Pelicans but only 6 minutes test in a series that had room for more disciplined testing of select alternatives but instead, as usual, saw 45 lineups used with more than 75% being pure dinks (less than 1 minute per game), pretty useless for evaluation.
Re: Thunder lineup analysis and other commentary
There are so many things to check. I checked many different time episodes of Thunder lineups and sub-lineups in the Mavs series... but didn't check the first 2 games by themselves initially. But did now.
And the alternative to the starters did great in those first 2 games... in 7 minutes of use. It probably played a role in the move. But later in game is different than starting.
And this decision shouldn't spring from those 7 minutes.
It was 2nd most tested in regular season and did well but that test was still small and could / should have been much bigger. The against the Mavs data mattered and contradicted the general trend.
Yes, this is hard.
There were other lineup candidates for alternative starters. Starters with Wallace or Jaylin Williams or Joe with Wiggins all did well against Mavs in playoffs in limited use. More use might have made the slight difference needed in 1 or more games.
Unfortunately these weren't tested notably against all in regular season or especially agaist Mavs.
Pivotal to test main alternatives, especially against playoff rivals, as much as possible. Too much Daigneault playing with dink lineups that were not main alternatives.
Bottomline is that Coach D did not make some key calls right and will have to do better.
Maybe I am being tough in retrospect. But I looked and tried to lay out the facts, fwiw. A tough look provides things to think about and maybe learn from. Not looking or being too easy is worth less.
And the alternative to the starters did great in those first 2 games... in 7 minutes of use. It probably played a role in the move. But later in game is different than starting.
And this decision shouldn't spring from those 7 minutes.
It was 2nd most tested in regular season and did well but that test was still small and could / should have been much bigger. The against the Mavs data mattered and contradicted the general trend.
Yes, this is hard.
There were other lineup candidates for alternative starters. Starters with Wallace or Jaylin Williams or Joe with Wiggins all did well against Mavs in playoffs in limited use. More use might have made the slight difference needed in 1 or more games.
Unfortunately these weren't tested notably against all in regular season or especially agaist Mavs.
Pivotal to test main alternatives, especially against playoff rivals, as much as possible. Too much Daigneault playing with dink lineups that were not main alternatives.
Bottomline is that Coach D did not make some key calls right and will have to do better.
Maybe I am being tough in retrospect. But I looked and tried to lay out the facts, fwiw. A tough look provides things to think about and maybe learn from. Not looking or being too easy is worth less.
Re: Thunder lineup analysis and other commentary
Aaron Wiggins, large advantage in team results with same4 over JDub, Wallace or Dort.
Celtics massively better with Horford instead of J Brown. Same with D White over Brown.
Same with 5 more and modestly with Porzingis.
Wemby only 1 pt than D Barlow with same 4 but Spurs let the latter go.
Path to link:
https://x.com/JerryEngelmann/status/1837371653602562475
Celtics massively better with Horford instead of J Brown. Same with D White over Brown.
Same with 5 more and modestly with Porzingis.
Wemby only 1 pt than D Barlow with same 4 but Spurs let the latter go.
Path to link:
https://x.com/JerryEngelmann/status/1837371653602562475
Re: Thunder lineup analysis and other commentary
Looking at lineup data can be bouncing around and not fully reading what is being said or not drawing all the lessons from it.
This is only partly digested but here goes on lineups with and without the 5 starters for lineups over 50 minutes.
(Other lineups out there not analyzed in this push.)
Starters were fine, mainly on offense.
Starters without Giddey were even better on offense.
Starters without JDub or Holmgren were fine. Without JDub were surprisingly better on offense, worse on defense.
SGA alone with bench or SGA / Dort with bench were better on offense and defense than all 5 starters. 5 starters not sacrosanct.
No SGA and without one of Giddey or Dort were just neutral. No SGA or JDub was very good. Just Dort JDub and Chet with bench was alright. Alright is either "enough" for planning or shoot higher.
SGA - Holmgren and SGA no Chet were both very good overall. Chet no SGA was meh and probabky should be minimized. SGA - Holmgen only and bench were meh but add Dort and the offense got better.
All bench sucked on offense.
All bench except JDub in was good, surprising because of defense and probably heavily from the bench guys
.
Could build lineups off this data and avoid lineups by checking all 17 (or so) of these findings every time.
Every Giddey and non-Giddey lineup should be analyzed to factor level and deeper to understand what needs to be replaced or enhanced, especially assists and rebounds.
Several versions of 4 previous starters and a new starter should be tested including Wiggins.
This is only partly digested but here goes on lineups with and without the 5 starters for lineups over 50 minutes.
(Other lineups out there not analyzed in this push.)
Starters were fine, mainly on offense.
Starters without Giddey were even better on offense.
Starters without JDub or Holmgren were fine. Without JDub were surprisingly better on offense, worse on defense.
SGA alone with bench or SGA / Dort with bench were better on offense and defense than all 5 starters. 5 starters not sacrosanct.
No SGA and without one of Giddey or Dort were just neutral. No SGA or JDub was very good. Just Dort JDub and Chet with bench was alright. Alright is either "enough" for planning or shoot higher.
SGA - Holmgren and SGA no Chet were both very good overall. Chet no SGA was meh and probabky should be minimized. SGA - Holmgen only and bench were meh but add Dort and the offense got better.
All bench sucked on offense.
All bench except JDub in was good, surprising because of defense and probably heavily from the bench guys
.
Could build lineups off this data and avoid lineups by checking all 17 (or so) of these findings every time.
Every Giddey and non-Giddey lineup should be analyzed to factor level and deeper to understand what needs to be replaced or enhanced, especially assists and rebounds.
Several versions of 4 previous starters and a new starter should be tested including Wiggins.
Re: Thunder lineup analysis and other commentary
I observed what with / without lineups over 50 minutes did. Most were 2 - 4 - many times bigger. I focused on direction and relative magnitude not exact performances. Sure, not all these observations are stable. So test with more discipline and with more priority. But this is what we have from last season. More likely that direction is correct than not though the confidence would vary on each. Maybe add previous season. Certainly add new season.
Ideally the team would review all the tape by combination to understand and evaluate. But this is a start for more work. Teams should have done this much and more but when they use 625 lineups it is somewhat hard to believe they are doing things right and fully after that.
Ideally the team would review all the tape by combination to understand and evaluate. But this is a start for more work. Teams should have done this much and more but when they use 625 lineups it is somewhat hard to believe they are doing things right and fully after that.
Re: Thunder lineup analysis and other commentary
Jalen Williams and Chet Holmgren were #1 and #2 in Darko improvement. Upward is typical for sophomores. But not near as much progress toward end of season. Chet fell back in last month.
How much more upside is there? New data soon.
How much more upside is there? New data soon.
Re: Thunder lineup analysis and other commentary
15 of 20 most used lineups were way better than the team average. Having them is good. Playing them more or way more would likely be better. But Coach D didn't.
His relative lack of lineup concentration is probably the biggest issue on the team.
His relative lack of lineup concentration is probably the biggest issue on the team.
Re: Thunder lineup analysis and other commentary
Popular among Thunder fans to like or love Wallace and Wiggins.
Fwiw, Wallace is on all 3 of the least strong big minute quads and Wiggins on 2.
Know this, think about this, change things... or not, not, not?
7 of 10 most used Wallace lineups were better than the average with him, 6 massively. Only one was barely used 1 minute per game for season. Too busy playing with 290 Wallace lineups. 254 for 9 minutes or less. Only 6 over 40 minutes and none a meaningful test. I simply can't do anything other than call this profoundly inefficient and unacceptable. (I edited out another term but I consider it to be more than just these things.)
How Wallace is used is a major issue in upcoming season. Backup PG or not, which combinations of other wings. Great with SGA, mildly negative without (but used without half the time). SGA without Wallace only marginally less strong than with. No SGA / No Wallace modestly better than No SGA but with Wallace.
Hey Mark, maybe squint at the lineup data more and harder, much more and much harder. If you claim to already, lots could be better wirh Wallace and others.
You do many things well apparently, unfathomable to me that others things are done "wrong" or not done "right". But that situation exists many other places and seems unlikely to change. So...note and move and probably repeat until maybe not.
Fwiw, Wallace is on all 3 of the least strong big minute quads and Wiggins on 2.
Know this, think about this, change things... or not, not, not?
7 of 10 most used Wallace lineups were better than the average with him, 6 massively. Only one was barely used 1 minute per game for season. Too busy playing with 290 Wallace lineups. 254 for 9 minutes or less. Only 6 over 40 minutes and none a meaningful test. I simply can't do anything other than call this profoundly inefficient and unacceptable. (I edited out another term but I consider it to be more than just these things.)
How Wallace is used is a major issue in upcoming season. Backup PG or not, which combinations of other wings. Great with SGA, mildly negative without (but used without half the time). SGA without Wallace only marginally less strong than with. No SGA / No Wallace modestly better than No SGA but with Wallace.
Hey Mark, maybe squint at the lineup data more and harder, much more and much harder. If you claim to already, lots could be better wirh Wallace and others.
You do many things well apparently, unfathomable to me that others things are done "wrong" or not done "right". But that situation exists many other places and seems unlikely to change. So...note and move and probably repeat until maybe not.
Re: Thunder lineup analysis and other commentary
Data suggests Wallace with Joe and maybe less with Holmgren.
Need to see with Caruso and Hartenstein.
With Hartenstein might be more intriguing.
Need to see with Caruso and Hartenstein.
With Hartenstein might be more intriguing.
Re: Thunder lineup analysis and other commentary
Either the team, design, coach and support are enough to actually get to conference finals or it doesn't and maybe isn't. Enough to actually win title or doesn't, maybe isn't.
Show about to begin.
If it not enough, it might be for insufficient somewhere on frontline. Last season there was a problem with opponent center scoring. By far highest opponent positional efg% and PER and highest scoring by a slight amount. PF was 4th strongest position matchup, SF 3rd.
Hartenstein is there now and much could change. But it still has to. Lineups will matter. Systems will matter.
Holmgren won his matchup with offense but the defensive matchup was close to the overall team center matchup and slightly worse in several places.
Jalen Williams won PF with offense but defensive was just ok. Way way better at SF (by 82 games estimates).
SF matchup was a win, mainly on defense but a moderate sized one. Absolutely all of team net scoring margin came from the guards. Front line won at overall measure at SF & PF but not scoring. Because of usage, especially FTAs. How much help Hartenstein helps with that remains to be seen. Either the fairly massive rebound deficits at PF & C decline significantly or they don't. Playing more SG at SF and no Giddey will likely hurt rebounding.
Team is designed for offense and defense but is it particularly well designed to matchup against top 4 or 6 teams? We'll find out more later.
Last regular season, Holmgren was good against Mavs and Nuggets, kinda weak on offense against TWolves, Knicks and Celtics and a disaster on D against Celtics.
Jalen Williams was a disaster against Celtics on both sides of court, about neutral against Mavs, Nuggets and TWolves and explosive on offense against Knicks.
Performance against top 5 or 10 will be more important than against rest of league. Ideally it improves and improves on frontline.
A significant in-season trade seems pretty unlikely but I wouldn't ignore the possibility, depending on the early data and opportunities. I'd be quite interested in a consolidation trade if it upgraded the frontline and improved the top 7 over the top 10.
Have to think more about options, best options but I did think a little about Jimmy Butler.
Show about to begin.
If it not enough, it might be for insufficient somewhere on frontline. Last season there was a problem with opponent center scoring. By far highest opponent positional efg% and PER and highest scoring by a slight amount. PF was 4th strongest position matchup, SF 3rd.
Hartenstein is there now and much could change. But it still has to. Lineups will matter. Systems will matter.
Holmgren won his matchup with offense but the defensive matchup was close to the overall team center matchup and slightly worse in several places.
Jalen Williams won PF with offense but defensive was just ok. Way way better at SF (by 82 games estimates).
SF matchup was a win, mainly on defense but a moderate sized one. Absolutely all of team net scoring margin came from the guards. Front line won at overall measure at SF & PF but not scoring. Because of usage, especially FTAs. How much help Hartenstein helps with that remains to be seen. Either the fairly massive rebound deficits at PF & C decline significantly or they don't. Playing more SG at SF and no Giddey will likely hurt rebounding.
Team is designed for offense and defense but is it particularly well designed to matchup against top 4 or 6 teams? We'll find out more later.
Last regular season, Holmgren was good against Mavs and Nuggets, kinda weak on offense against TWolves, Knicks and Celtics and a disaster on D against Celtics.
Jalen Williams was a disaster against Celtics on both sides of court, about neutral against Mavs, Nuggets and TWolves and explosive on offense against Knicks.
Performance against top 5 or 10 will be more important than against rest of league. Ideally it improves and improves on frontline.
A significant in-season trade seems pretty unlikely but I wouldn't ignore the possibility, depending on the early data and opportunities. I'd be quite interested in a consolidation trade if it upgraded the frontline and improved the top 7 over the top 10.
Have to think more about options, best options but I did think a little about Jimmy Butler.
Re: Thunder lineup analysis and other commentary
The basic message of the Thunder pre-season is that they are strong on factors where they were strong and not strong where they were not strong. Maybe some improvement within factors but that is hard to really tell from pre-season. But basic similarity is a reasonable take from the data and look.
Dead last on FTA. Bottom ten on offensive rebounds and fouls committed. 24th on defensive rebounding rate.
Dead last on FTA. Bottom ten on offensive rebounds and fouls committed. 24th on defensive rebounding rate.
Re: Thunder lineup analysis and other commentary
Fwiw curiosities in Thunder preseason individual data:
I Joe with the worst efg% and ts%(among main guys). Not wearing the 25% usage well at all. Wallace and Dieng also low. Hartenstein and Caruso below league averages.
Hartenstein with highest personal pace, along with but above rest of core. First tier bench were slower (in their blended experience), second tier much slower.
Holmgren 10th overall on offensive rebounding rate, more than 55% less than last regular season. Def. rebounding rate down 10%. New guys won't be as additive if he is providing less.
Leons, outstanding OR%. Reese better on defensive glass.
Dort, Hartenstein and Holmgren with heightened TO rates. Hartenstein and Dort producing assists but not Holmgren. Dieng and Caruso also producing assists.
I Joe with the worst efg% and ts%(among main guys). Not wearing the 25% usage well at all. Wallace and Dieng also low. Hartenstein and Caruso below league averages.
Hartenstein with highest personal pace, along with but above rest of core. First tier bench were slower (in their blended experience), second tier much slower.
Holmgren 10th overall on offensive rebounding rate, more than 55% less than last regular season. Def. rebounding rate down 10%. New guys won't be as additive if he is providing less.
Leons, outstanding OR%. Reese better on defensive glass.
Dort, Hartenstein and Holmgren with heightened TO rates. Hartenstein and Dort producing assists but not Holmgren. Dieng and Caruso also producing assists.
Re: Thunder lineup analysis and other commentary
Hartenstein hurt, roster distribution revealed. 13 6-7 and under, 3 6-9 (at the moment, pending final cuts), 3 more than that (2 healthy). Dieng and / or Jones will likely need to step up.
Just 4 over 220 pounds. Rebounding likely to suck some more.
Just 4 over 220 pounds. Rebounding likely to suck some more.
Re: Thunder lineup analysis and other commentary
Starting 0 starters against 4 or 5, Hawks with the 10 to 1 lead on offensive rebounds.
David Roddy bullies Jalen Williams and others for 7 first half rebounds, 4 offensive.
Hawks take halftime lead despite poor shooting and zero ftas.
David Roddy bullies Jalen Williams and others for 7 first half rebounds, 4 offensive.
Hawks take halftime lead despite poor shooting and zero ftas.