APBRmetrics

https://x.com/bballstrategy/status/1842684140875141147

https://x.com/bballstrategy/status/1843323386745008183

Rough initial lineup plan (not ordered in use):

Cur Podz Kum Gre TJD 15 min
C Podz Wig K G 6
C Mel Pay W G 5
C Mel W G TJD 4
C Mel Mood G TJD 2
C Pay Podz G TJD 2
C Pay Mood G TJD 2
Mel Podz Mood K Loon 6
Mel Podz Mood K TJD 6

Minutes (at full use, may be less) :


Curry 36
Melton 23
Podziemski 33
Payton 11
Moody 16
Kuminga 33
Wiggins 15
Green 36
TJD 31
Looney 6

This is just a first draft and probably could be better. But it is a workable start.

More players could be added in a second draft.

Kerr has used 75 lineups.

One over 20 minutes. 4 total over 10 minutes. 12 for 5+, 63 dinks.

Little or nothing of what I suggested has gotten trial.

Will look at sub-lineup data later.

Most used 5 man moderately negative then 8 positives.

All 20 of 20 most used pairs are positive, 70% over +10.

I am liking my projection of 52 wins for them, highest in contest.

4 of 5 most used lineups are great but 4 of next 5 are bad.

1 lineup used not quite 2.5 minutes per game and then only 2 more over 1 minute.

4 of 10 most used player pairs are negative.

331 lineups used tells me that don't really know what they are doing and may not care enough.

An average of 10 brand new lineups every game.

Curry - Green is about +7. Last title year it was +13 and the others were +12 to +18. The core is not at same level. Half.

They almost certainly won't trade Green in next month but the question will be should they have.

Green was near +3 BPM to almost double that in title seasons. +0.5 right now.

Is there a different, better formula to win again with Curry? Have to think about it to guess.

Who is +12 or better with Curry? Hield, Payton and Podziemski.

How much have they been used together? 11 minutes at +64.

Would you try it more than 11 minutes? I would. With Green, Kuminga, TJD, Looney, Anderson or somebody else. Like 100+ or 200+ minutes with each pronto. And combos without Payton or especially Podziemski.

Schroder - Curry? By far worst pair in Curry's 10 most used.

No Curry lineup used more than 2.5 minutes per game. Only 2 much more than 1 minute per game. 157 tried? Cuz why? Doesn't really know or doesn't care? Incredible.

Is there anyone in Warriors organization who has said this? If not, why not?

Curry Hield Kuminga Green Jackson-Davis has been tried zero minutes. I'd certainly try it for several hundred minutes and evaluate. I'd expect it to work. Quad without TJD has been tried and worked great.

https://x.com/bballstrategy/status/1883563869442134211

https://x.com/bballstrategy/status/1883574587163828502

As I expected, Kerr plans for Butler to anchor the non-Curry minutes, beyond the starting unit. Will look for lineup data right away and after 10+ games. Who else appears to each group will be interesting to see / check.

2 games with Butler:

4 most uses lineups positive, 7 of 10 positive.

5 lineups over 2 minutes per game involving 9 players. None yet over 6.5.

Most used: S. Curry - J. Butler - D. Green - B. Hield - M. Moody

Best of bigger minutes:
J. Butler - D. Green - B. Hield - B. Podziemski - Q. Post. Tiny sample but no Curry.

10 of 10 most used pairs positive, 8 strongly. 17 of 20 positive.


As good of team results as could be reasonably hoped for.

Lineup management successful but 33 lineups in 2 games.

58 lineups in last 5 games.

5 of 7 most used positive. Probably should squeeze out a lot of the dinks. But Kerr won't. 595 lineups for season. No lineup over 1.5 minutes per game for season. Clearly committed to dinkathon.

New "death lineup", apparently to close.

Crow wrote: ↑Mon Feb 24, 2025 2:50 pm No lineup over 1.5 minutes per game for season. Clearly committed to dinkathon.

If you have 6 players and use them equally in all 6 possible combinations, you'd have each lineup on the floor for 8 minutes.

If your rotation is just 7 players, there are 21 combinations of 5; giving all 7 equal minutes, each comb. would avg 2.3 minutes. At 11-player rotations, equal distribution of minutes yields an average of 6 seconds (.10 min) per lineup per game; less than 10 minutes over 82 games.
When some lineups run much more than this, other lineups run even less. Some not at all.