Fwiw, it would be possible
to find the lineup set with the highest correlation of metric data to minutes, within whatever position assignments and minute constraints are considered "reasonable".
Find it, study and consider actions.
And then reason from best possible lineup set from lineup data to date.
Compare. Combine approaches.
Do something from analysis and a plan instead of the haphazard looking micro-moment coaching chaos approach.
improved players from 2023 to 2024
Re: improved players from 2023 to 2024
For the Thunder, working from BPM to minutes rigidly would involve 9 players and not Dort.
Working from best performing to date
biggest minute lineups til you have a feasible rotation would involve 11 players but only 6 lineups.
Wiggins would play way more and Dort way less.
This approach hypothetically would triple the combined actual minutes of these 6 strong lineups and eliminate all dinks, which way underperformed the 6. In real world, you can't / won't go that far but you could double their minutes and still have time for necessary or strongly desired change-ups and farting around.
The ability to double or triple the minutes of your 6 best tested lineups shows how far from theoretical optimum the current practice is. If you believe in small sample raw lineup performance.
If you don't, you can get most of the way to similar place going by BPM or other metric. You mainly just deal more harshly with Dort's minutes. Given that he is in only 2 of the best tested, best performing and only 1 of top 4, it is not that different and might turn out even better.
Now, if you want to go from current Coach given minutes to an apparently more concentrated, more optimal lineup rotation, you can do that mostly the same way but giving Dort some minutes back from Wiggins and / or Joe and Wallace.
Pick a method. Improve on current chaos any one of 3 ways.
Working from best performing to date
biggest minute lineups til you have a feasible rotation would involve 11 players but only 6 lineups.
Wiggins would play way more and Dort way less.
This approach hypothetically would triple the combined actual minutes of these 6 strong lineups and eliminate all dinks, which way underperformed the 6. In real world, you can't / won't go that far but you could double their minutes and still have time for necessary or strongly desired change-ups and farting around.
The ability to double or triple the minutes of your 6 best tested lineups shows how far from theoretical optimum the current practice is. If you believe in small sample raw lineup performance.
If you don't, you can get most of the way to similar place going by BPM or other metric. You mainly just deal more harshly with Dort's minutes. Given that he is in only 2 of the best tested, best performing and only 1 of top 4, it is not that different and might turn out even better.
Now, if you want to go from current Coach given minutes to an apparently more concentrated, more optimal lineup rotation, you can do that mostly the same way but giving Dort some minutes back from Wiggins and / or Joe and Wallace.
Pick a method. Improve on current chaos any one of 3 ways.
Re: improved players from 2023 to 2024
(back to the previous tangent)
Correlations of player minutes to their various available summary stats, in the 2023 playoff rounds:
It's possible that this sample of 16 teams (and ~80 games) is unusual; or likely it reflects an agreement between what coaches believe and what these stats detect.
I'm not suggesting that PER is better than BPM (for example), but I do feel disturbed when my own eWins are not well corroborated by coaching decisions. So this is comforting.
There is a very small minutes factor in eWins; starter/sub is a bit heavier. See the previous Celtics' breakdown, and the nearly random Win Shares allocations throughout the postseason.
Correlations of player minutes to their various available summary stats, in the 2023 playoff rounds:
Code: Select all
rd PER WS/48 BPM e484
1 .47 .14 .34 .59
2 .60 .42 .53 .68
3+4 .57 .35 .45 .72I'm not suggesting that PER is better than BPM (for example), but I do feel disturbed when my own eWins are not well corroborated by coaching decisions. So this is comforting.
There is a very small minutes factor in eWins; starter/sub is a bit heavier. See the previous Celtics' breakdown, and the nearly random Win Shares allocations throughout the postseason.
Re: improved players from 2023 to 2024
Yes and yes.
Ironically, both finalists I think have great coaches, as far as adjusting their rotations according to performance. But after 3 rounds, they had made up their minds?
Separating and averaging the winners and losers in 1st and later rounds:
Code: Select all
tm rd PER WS/48 BPM e484 avg
W 1 .47 .20 .36 .60 .41
L 1 .47 .07 .31 .58 .36
W 2-4 .62 .45 .51 .69 .57
L 2-4 .56 .32 .48 .71 .52Re: improved players from 2023 to 2024
Season at 87%, and it could go a bunch of different ways:
https://www.basketball-reference.com/
Magics with Banchero, Suggs, and both Wagners.
Knicks also with 4; Bulls and Rockets 3 each. That's 14 of top 31.
Biggest deficits:
James Harden monthly splits:
https://www.basketball-reference.com/pl ... plits/2024
About 35 mpg and 21% Usage every month.
Code: Select all
eWin+ per36: tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
2.45 Shai G-A OKC .627 32.0 5.8 4.8 2.76 .610 30.4 4.8 4.5 2.28
2.37 A Sengun Hou .575 22.2 10.5 4.1 1.98 .588 16.5 11.7 3.8 1.43
2.24 J Collins Uta .605 16.9 11.4 1.0 1.46 Atl .585 14.2 7.9 1.1 .89
2.24 A Edwards Min .569 27.1 6.0 4.2 1.87 .555 22.3 5.9 3.6 1.44
2.23 DiVincenzo NYK .589 19.3 5.0 2.7 1.29 GSW .589 11.8 6.1 3.8 .75
2.13 Jal Williams OKC .616 21.6 4.5 3.7 1.47 .594 15.6 5.2 3.2 .97
1.92 Hartenstein NYK .648 11.0 13.0 2.5 1.40 .560 8.2 11.9 1.8 .84
1.89 J Walker Por .553 11.7 10.9 1.0 1.00 .490 9.5 7.6 1.6 .37
1.88 J Kuminga GSW .586 20.1 6.6 2.0 1.29 .588 15.6 5.9 2.5 .81
1.83 J Smith Jr. Hou .559 14.0 9.8 1.3 1.12 .508 11.9 8.7 1.2 .70
eWin+ per36: tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
1.82 D Robinson Mia .616 16.9 3.3 2.9 .78 .530 12.1 3.8 2.0 .31
1.81 P Banchero Orl .543 21.7 7.9 4.3 1.51 .517 18.4 7.8 3.2 1.16
1.81 F Wagner Orl .565 20.9 6.8 3.3 1.47 .581 19.1 4.8 3.1 1.07
1.78 S Barnes Tor .558 17.7 8.5 4.3 1.38 .518 13.9 7.3 4.2 .98
1.63 K-A Towns Min .618 25.8 9.7 2.6 1.98 .607 21.8 9.0 4.2 1.59
1.60 C Sexton Uta .602 22.5 3.7 4.6 1.32 .604 19.8 3.3 3.4 .92
1.60 J Duren Det .645 15.7 15.0 2.1 1.59 .642 12.5 12.8 1.3 1.13
1.47 J Allen Cle .647 19.5 12.6 2.3 1.90 .658 17.3 11.7 1.6 1.58
1.45 A Caruso Chi .600 12.1 5.0 2.9 .80 .580 8.1 4.6 3.5 .42
1.37 T Maxey Phl .566 22.9 3.5 4.4 1.51 .597 21.4 3.4 3.2 1.25
eWin+ per36: tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
1.29 J Brunson NYK .580 27.6 4.0 5.2 1.86 .587 23.9 3.8 5.4 1.61
1.26 J Green Hou .535 20.2 5.8 2.9 1.15 .529 19.4 4.1 3.1 .90
1.25 M McBride NYK .583 14.5 2.7 2.3 .73 .470 8.3 2.3 2.7 .14
1.24 I Zubac LAC .660 15.9 13.1 1.4 1.57 .646 13.8 13.0 1.1 1.20
1.18 C White Chi .568 17.4 4.8 3.7 .90 .568 13.8 4.4 3.4 .69
1.17 J Suggs Orl .590 16.5 4.9 2.7 .98 .520 12.9 4.8 3.5 .67
1.17 A Nesmith Ind .628 14.4 5.0 1.3 .86 .558 12.5 5.5 1.5 .55
1.17 K Murray Sac .569 14.2 6.2 1.3 .94 .593 13.3 5.9 1.2 .70
1.14 A Dosunmu Chi .592 13.8 3.5 2.8 .59 .562 11.0 3.9 2.9 .33
1.13 J Smith Ind .688 19.4 11.5 1.4 1.78 .557 15.2 11.0 1.5 1.19
1.13 M Wagner Orl .661 22.3 9.6 1.7 1.75 .616 18.5 8.7 2.1 1.33
Magics with Banchero, Suggs, and both Wagners.
Knicks also with 4; Bulls and Rockets 3 each. That's 14 of top 31.
Biggest deficits:
Code: Select all
eWin+ per36: tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
-3.36 D Lillard Mil .584 22.8 4.5 5.2 1.45 Por .630 30.4 4.9 6.0 2.17
-2.63 Jr Holiday Bos .601 14.1 5.8 4.2 .83 Mil .580 20.6 5.4 6.9 1.48
-2.54 J Harden LAC .613 17.2 5.5 6.5 1.15 Phl .595 20.3 6.5 9.0 1.72
-2.42 K Durant Phx .622 26.2 6.8 3.7 1.79 .664 31.1 6.7 4.2 2.29
-2.12 S Curry GSW .609 27.8 5.0 3.8 1.73 .647 30.4 6.4 5.2 2.24
-2.10 Kris Dunn Uta .546 8.4 5.5 5.1 .51 .598 16.8 6.2 6.2 1.41
-1.90 D Garland Cle .560 19.2 2.9 5.2 .86 .578 22.2 2.9 6.8 1.45
-1.88 J Poole Was .519 15.9 3.0 3.1 .59 GSW .564 22.0 3.3 4.3 1.05
-1.82 B Lopez Mil .597 13.4 6.3 1.5 .93 .623 19.0 7.6 1.3 1.37
-1.82 J Butler Mia .608 22.6 5.9 4.0 1.71 .629 25.3 6.9 5.0 2.24
eWin+ per36: tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
-1.66 C Wood LAL .558 11.8 10.2 1.3 .83 Dal .612 22.1 10.5 2.2 1.77
-1.64 LeBron LAL .611 24.3 7.5 5.8 1.79 .575 27.2 8.3 5.5 2.17
-1.53 J Clarkson Uta .514 15.7 4.1 4.1 .50 .551 20.3 4.5 3.9 .96
-1.46 A Gordon Den .590 15.5 7.7 2.8 1.09 .604 19.0 8.3 2.9 1.46
-1.38 T Young Atl .569 22.1 2.7 7.4 1.12 .560 24.0 3.1 8.1 1.49
-1.35 J Jackson Mem .542 22.4 6.4 2.0 1.54 .602 23.1 8.4 1.0 1.89
-1.32 D DeRozan Chi .569 20.4 4.2 3.7 1.35 .580 23.6 4.7 4.1 1.61
-1.31 B Beal Phx .583 18.4 4.7 4.1 .97 Was .585 23.6 4.3 4.7 1.43
-1.11 K Thompson GSW .561 18.4 4.2 1.9 .91 .572 21.6 4.5 2.1 1.20
-1.10 N Powell LAC .626 18.3 3.6 1.0 .88 .600 22.4 4.1 2.1 1.21
-1.10 M Bridges Brk .552 18.6 5.0 2.8 .99 .579 19.5 4.5 2.8 1.22
Code: Select all
mo. FG% 3fg% TS% Pts Reb Ast O-DRt
Nov .458 .410 .639 15.3 4.4 6.2 +10
Dec .466 .433 .661 20.0 5.2 9.6 +16
Jan .434 .396 .616 16.5 5.1 9.6 +12
Feb .423 .398 .632 18.2 5.9 7.3 + 2
Mar .415 .341 .578 15.6 4.8 9.6 - 1About 35 mpg and 21% Usage every month.
Re: improved players from 2023 to 2024
Both Lillard and Holiday way down, as individuals, by this measure. Context less run for, optimized for them. But teams wanted them, should probably have expected that.
Re: improved players from 2023 to 2024
Those guys and others are in the age range that's expected to decline.
Smoothed over by using 5-year avg (rather than single year), the diff column is total gain or decline by an age group, in eWins/484 min. A change of .01 would be like a WS/48 change of .001Players are peaking at age 27, though you may say they 'plateau' from 24 to around 30. Those are pretty small avg changes in that span.
There is selection bias here. Players who didn't play last year aren't included; nor those who didn't return this season -- thus declined more than those still in the sample.
Smoothed over by using 5-year avg (rather than single year), the diff column is total gain or decline by an age group, in eWins/484 min. A change of .01 would be like a WS/48 change of .001
Code: Select all
Age 2024 Min diff players (most minutes)
19 20 12494 .20 Sochan JSmith Duren Branham Daniels
20 21 22644 .18 Banchero JGreen Sengun Ivey Kuminga
21 22 31613 .12 Edwards Barnes FWagner JWilliams Cade
22 23 37986 .06 CWhite Maxey KMurray Vassell Haliburton
23 24 43638 .02 Luka Bey Poole Ayo JJJr Keldon Dort Nic
24 25 45733 .02 Tatum Shai Reaves MPJ JAllen HJones Trae
25 26 35406 .02 Fox Ingram Bam JCollins Lauri JMurray
26 27 41041 .01 Doman Mikal DJMurray Brunson JBrown DLo
27 28 33423 -.02 Jokic Hart GAllen Kuzma AGordon Towns
28 29 29158 -.03 FVV Giannis DWhite TPrince Nurkic JGrant
29 30 17466 -.05 ADavis KCP DFS NPowell Niang SloMo Beal
30 31 25119 -.07 Gobert Tobias HBarnes Bogdan THJr Jonas
31 32 8486 -.13 Kawhi CJ Middleton DPowell Kleber Olynyk
32 33 16851 -.20 Dame Vucevic George Jrue Klay RJax Dray
33 34 7890 -.24 DeRozan Harden Butler Bojan Morris*2
34 35 14326 -.22 Durant Curry Lopez EGordon W'brook Love
35 36 3239 -.15 Conley Ingles McGee
36 37 3626 -.11 Horford JeffGreen Lowry WMatthews
37 38 1551 -.17 Paul PJ Taj
38 39 2182 -.36 LeBron
There is selection bias here. Players who didn't play last year aren't included; nor those who didn't return this season -- thus declined more than those still in the sample.
Re: improved players from 2023 to 2024
Well, Tyrese Maxey at least made the top 20.
His 3fg% was down .061 from last season; 2fg% down .012, eFG down .044. But he shot more often.
Biggest gain was assists, which doubled in 30% more minutes.
His 3fg% was down .061 from last season; 2fg% down .012, eFG down .044. But he shot more often.
Biggest gain was assists, which doubled in 30% more minutes.
Re: improved players from 2023 to 2024
catching up -- Final rankings
Code: Select all
eW+ per36 rates tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
2.40 Jalen Brunson NYK .583 28.7 3.9 5.2 2.02 .587 23.9 3.8 5.4 1.61
2.38 Donte DiV. NYK .593 19.1 4.8 2.5 1.23 GSW .589 11.8 6.1 3.8 .75
2.34 Anthony Edwards Min .566 26.5 5.8 4.1 1.84 .555 22.3 5.9 3.6 1.44
2.33 Jalen Williams OKC .614 21.6 4.5 3.7 1.46 .594 15.6 5.2 3.2 .97
2.21 Shai G-A OKC .622 31.5 5.8 4.7 2.69 .610 30.4 4.8 4.5 2.28
2.18 Jabari Walker Por .537 11.7 11.3 1.1 .98 .490 9.5 7.6 1.6 .37
2.16 Alperen Sengun Hou .575 22.0 10.4 4.0 1.93 .588 16.5 11.7 3.8 1.43
2.12 John Collins Uta .612 17.5 11.2 1.0 1.42 Atl .585 14.2 7.9 1.1 .89
2.08 I.Hartenstein NYK .656 11.7 12.6 2.7 1.36 .560 8.2 11.9 1.8 .84
2.08 Paolo Banchero Orl .535 21.5 8.0 4.2 1.51 .517 18.4 7.8 3.2 1.16
eW+ per36 rates tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
2.02 Jonathan Kuminga GSW .587 20.4 6.6 2.1 1.30 .588 15.6 5.9 2.5 .81
1.96 Franz Wagner Orl .566 21.0 6.6 3.1 1.46 .581 19.1 4.8 3.1 1.07
1.90 Payton Pritchard Bos .594 15.6 5.1 4.2 .99 .534 13.5 4.7 3.0 .50
1.87 Jabari Smith Hou .564 14.3 9.2 1.3 1.06 .508 11.9 8.7 1.2 .70
1.78 Tyrese Maxey Phl .565 23.2 3.5 4.4 1.57 .597 21.4 3.4 3.2 1.25
1.73 Duncan Robinson Mia .608 16.7 3.3 2.8 .74 .530 12.1 3.8 2.0 .31
1.71 Scottie Barnes Tor .558 17.6 8.5 4.1 1.37 .518 13.9 7.3 4.2 .98
1.71 Miles McBride NYK .587 15.0 2.9 2.3 .75 .470 8.3 2.3 2.7 .14
1.66 Collin Sexton Uta .596 22.3 3.7 4.5 1.30 .604 19.8 3.3 3.4 .92
1.50 Jalen Duren Det .642 16.0 14.6 2.0 1.53 .642 12.5 12.8 1.3 1.13
eW+ per36 rates tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
1.49 Jalen Green Hou .534 19.9 5.9 2.9 1.16 .529 19.4 4.1 3.1 .90
1.41 Moritz Wagner Orl .664 23.5 9.8 1.8 1.80 .616 18.5 8.7 2.1 1.33
1.34 Kevin Love Mia .587 18.6 13.5 3.4 1.89 .542 13.7 12.3 2.9 1.20
1.33 Moses Moody GSW .571 15.1 6.1 1.2 .90 .596 11.9 4.5 1.7 .35
1.31 RJ Barrett Tor .570 20.9 6.3 2.6 1.23 NYK .522 18.5 5.5 2.5 .90
1.30 S. Fontecchio Det .596 13.4 5.5 1.5 .61 Uta .491 12.0 3.9 1.5 .24
1.30 Cam Thomas Brk .546 23.1 3.8 2.4 1.25 .556 20.7 3.5 2.4 .95
1.29 Deni Avdija Was .586 15.0 8.2 2.9 .96 .527 10.8 8.8 3.0 .70
1.29 Ivica Zubac LAC .659 16.3 12.7 1.3 1.54 .646 13.8 13.0 1.1 1.20
1.29 Jarrett Allen Cle .651 19.5 12.3 2.3 1.83 .658 17.3 11.7 1.6 1.58
eW+ per36 rates tm Eff% Sco Reb Ast e484 Eff% Sco Reb Ast e484
1.28 Alex Caruso Chi .607 12.1 4.9 3.1 .72 .580 8.1 4.6 3.5 .42
1.27 Jalen Smith Ind .672 19.7 12.0 1.4 1.77 .557 15.2 11.0 1.5 1.19
1.26 Jamal Murray Den .580 23.5 4.8 5.2 1.67 .565 20.5 4.6 5.4 1.35
1.25 Ayo Dosunmu Chi .599 14.3 3.6 2.8 .59 .562 11.0 3.9 2.9 .33
1.24 Jake LaRavia Mem .530 14.9 5.9 1.9 .77 .528 7.8 5.1 1.4 .03
1.23 Amir Coffey LAC .597 10.9 3.7 1.3 .31 .487 8.0 3.0 2.5 -.09
1.20 Keegan Murray Sac .565 14.7 6.1 1.2 .92 .593 13.3 5.9 1.2 .70
1.18 CJ McCollum NOP .588 21.2 4.8 3.7 1.44 .536 19.3 4.7 4.9 1.19
1.17 Aaron Wiggins OKC .657 16.3 5.4 1.8 .98 .601 12.2 5.5 1.7 .53
1.16 Daniel Gafford Dal .717 16.2 10.8 1.5 1.63 Was .722 17.0 9.9 1.5 1.33