APBRmetrics

By Darko projection,

Kevin Durant is 3rd lowest on "DPM improvement".

He is 10 years off his Darko peak. He broke below his long-term trend this season.

What is a reasonable trade price? Extend? How rich? What usage level?

Other very big declines include:

Embid, D Fox, T Murphy, Jrue Holiday, D Murray, Issac, Booker, Beal, George, Markkanen, Randle, Ingram, Simmons, Porzingis, Lamelo Ball...


In next tier: LeBron, Adebayo, Andrew Wiggins, Claxton, Kuzma, VanVleet, Irving, D White, Middleton, Cason Wallace, Brunson, H Jones, Ayton...

Among not so published improvements:

Furphy (on defense), B Boston, Ighodoro, Mamukelashvili, Okogie, Olynyk, R Holmes, Aldama, Gueye, Podziemski, Dunn, Aaron Wiggins, Huff, M Plumlee, Okongwu, J Collins...


Better known cases: Jerome, Zubac, SGA, Cunningham, Grimes, Braun, Pritchard, N Reid...

By Darko, "internationals" (direct from European play) make up 7 of top 40 and 5 of bottom 34. Not much different from overall percentage of players.

You hear about the top, not the bottom. The top does include 4 of top 10,

Full break-out into 4 or 5 tiers would be interesting.

Looking at top 100, I count 14 internationals. Less than their overall percentage. 7 on bottom 100.

Is there any playoff insight here?

Of the 28 declines in 1st post, few are in playoffs. 2 Celtics are the most notable names.

Of the 23 improvements in 2nd post, Zubac is the most notable to me.

These lists were not targeted on the playoffs, though it offers a few messages. They are retrospective and meaningful for off-season and next season decisionmaking in minor, medium and major cases.

The strong / weak RAPM thread may be more playoffs relevant.

Of the 10 biggest DPM improvements in entire league, 7 were PF or C.

S Mamukelashvili and B Carlson in 2nd ten.

Zero full-time PGs in top 10. 2 in second ten and a couple combo guards bring the representation back up.

Jabari Walker is another improver to watch next season.

Maybe Scheierman.

Considering Darko information is better than not imo. Using 7 parts as I have been recently better than 1 or 4. At least 11 parts available, most with trend graphs. I might go to 9 or 10 part analysis.

A few examples:

Barnes
https://x.com/bballstrategy/status/1942280580026233294

Houstan
https://x.com/bballstrategy/status/1942017674537812095

Rollins
https://x.com/bballstrategy/status/1942024860743725233


A few others out there report Darko but only a few and never to my knowledge deeper than the overall projection. So much more easily available and imo necessary to review.

By "Darko improvement", 13 of the 16 biggest improvements over this rookie season were for guys 6-8 and taller. Almost half were non-negative on defensive Darko and that was probably a bigger part of the improvement. None were estimated positive on offensive Darko.

The other 3 were traditional to very small sized PGs.

It is possible that the starting Darko assumptions for talls and not talls are not discrete or optimal. It is possibly this way in part due to use of draft order for starting Darko and GM drafting order
of not talls and talls.

Do I overrated bigs in draft? Possibly. It is a common opinion. But the answer to the question will be based on facts, not opinions. And this Darko data is supportive of picking talls over smalls.

All 3 of the smalls in top 16 did it in part by being slightly positive on defensive Darko.

May be better to lean toward bigs than away from them and lean away from guards who aren't the best.

20 of the 25 worst rookies on Darko improvement were perimeters. That is only modestly worse than proportional but it is worse.

Every single one of the bottom 25 rookies on improvement were non-positive on defense.

Something to consider. That teams are probably not in general considering adequately.

If I am rating or picking a not tall, I would favor better defenders within that group.

Talls were twice as common in top 16 rookie improvers compared to proportional and half as likely to be in bottom 25.

By "Darko improvement", the top 10 were 7 Centers, 2 PGs and 1 wing (broadly defined). Top 20 goes to 10 centers, 6 wings, 4 PGs.

Center by far the most listed, PGs pull up to even representation in top 20. Broad wing group only half of an even share.

Nearly half of the best improvements by rookies but it does happen with older players.