Bucks and RPM

[Crow]

At least one Bucks analyst was publicly critical of RPM or at minimum cautious or unimpressed and dismissive of it to my ears. I think another indicated he didnt use it much to me privately but I can't be 100% sure of that memory.

So what has RPM shown for the current Bucks?

All 5 starters positive on RPM and strongly so. And before acquisition of the 2 non-draftees? Bledsoe and Lopez were both positive in season before and after. Brogdon positive as a rookie, then slipped then returned to positive. Initial RPM showed enough to be a keeper. Middleton progressed from negative to positive in good order. A keeper to be retained. Giannis was a mild negative in year 1 to better and better and now near best. RPM confirms his greatness.

And bench guys? Wilson, negative as a rookie but not terrible. Got better. Conaughton, similar. Ilyasova, positive year before and after signing, though less so now. Hence I guess the signing the RPM positive Mirotic. For bench guys, helping you is great, not hurting you is good and not hurting you much is probably acceptable.

Snell, negative before signing / retaining, maybe shouldn't have been retained but tolerable. Henson, 10th best this season, ok to trade. Dellevediva, bad in Milwaukee, mistake signing, good to trade. Maker, bad RPM before, ok to good to let go.

Newest guys? DD, Trey Duval not looking good but not terrible. Give it time and decide after more time.

Overall RPM has not mislead or failed to capture value levels of these Bucks by much. In my opinion.

Get to good outcomes any way you wish. Different ways to do it.

[Mike G]

Only about 140 guys are positive, which is to say "above average", in RPM. They total about 45% of NBA minutes this year.
That's not quite 5 players per team.
Another 45% of NBA minutes are provided by players who are below-avg in RPM but have positive "Wins" added, according to ESPN. These guys have RPM > -3.125 but <0.
Then 10% of all minutes are with players having negative Wins due to even lower RPM.

Bucks' minutes have been 68% in the A group (RPM>0), 32% in the B group, and not a one from the bottom 10%.

Players with <0 RPM but >3.0 rpmWins are Evan Fournier, Kyle Kuzma, Klay Thompson, Jae Crowder, Bojan Bogdanovich, and Nic Batum.
Huerter, Jeff Green, Ibaka, Jaren Jackson, Barnes, Randle, Collins, Hayward, and Wiggins all have 2.5 to 2.9 WINs and negative RPM.

[Mike G]

Since I was already doing a sort of survey of RPM at the break, I'll share here the correlations between player Min/G and RPM (via ESPN). I sorted by team and by total minutes; then found correlation between MPG and RPM for top 8 players, 9, 10, 11, and 12 (in total min.) for each team.
Numbers shown are the avg of those 5.

Code: Select all

west  corr    east  corr
NOP   .82      Mil   .89
Uta   .78      Tor   .83
GSW   .72      Cha   .82
OKC   .71      Phl   .81
Mem   .60      Orl   .78
Por   .59      Ind   .67
Phx   .57      Det   .67
Sac   .55      Chi   .34
Min   .52      Bos   .33
Hou   .47      Mia   .31
Den   .45      Brk   .17
LAC   .38      NYK   .02
LAL   .33      Was  -.02
SAS   .19      Atl  -.10
Dal   .07      Cle  -.47

avg   .52      avg   .40

East has several of the best and all of the worst. Atl and Cle are both led by rookie PG with team-low RPM.
In general, better teams are toward the top. It could indicate better coaching or the fact it's easier to successfully coach when you have good players.
Anomalies are in the West : the Nuggs and Spurs either disregard RPM or just have quirky situations.
Bucks have nearly "perfect" corr. thru 8 or 10 or 12. Remove Middleton (31 mpg @ 1.88) and it jumps to .93-95. They have no other SF?

[Crow]

How did these correlations end up?

GSW, Tor and Mil were near the top at mid-season.

Did SAS end up near the bottom as they were at mid-season? Brooklyn, "analytic" Brooklyn?", had a garbage correlation between minutes and RPM too.

What were the correlations of RPM to playoff minutes? Who got higher or lower?

[Mike G]

Good questions. Is there an easy way to pluck RPM data? Copying 25 tables isn't really fun.

[Crow]

Generally have to cut & paste to get all of RPM, except when jerry puts full file on twitter or herre.

But there is this alternative http://canzhiye.com/rapm/ that is one page and might be sufficient for this purpose.

[liminal_space]

^ That looks like just RAPM.