APBRmetrics

By this playoff RAPM, Harden and Paul were also slightly negative combined but separately.

BOS, HOU... one may got out by 2nd round. Neither likely to win conference finals. Warriors with 4 of top 8. Cavs 3.

Crow wrote:Amir had strong positive raw plus minus on 7 of his 9 most used regular season lineups including all of the top 4,(and the starter group). He was good in more than just a few. ..

After Isaiah went down, Amir played 9, zero, and 6 minutes. As a very low Usg% guy, the lineup suffers without some high Usg players. The pressure is on everyone to create some shots, and Amir became dead weight.

A good coach in fact does not use inefficient lineups for many minutes. If you can see all the Amir-containing lineups down to, say, 20 minutes, what % of them were above the Celtics' average?

Amir played 1282 minutes alongside Isaiah in RS. Celts were +8.0/100 in this time.
But his total raw on-court was +8.7 over 1608 min. Coach hardly used him when Thomas wasn't on court; yet the team was better?

the lineup suffers without some high Usg players...

howso? any statistical proof for this statement?...

if a team trotted out a lineup of say:

C - tyson chandler
PF - jamychal green
SF - luc mbah a moute
SG - andre roberson
PG - shaun livingston

what would actually happen? how would this lineup "suffer" - statistically i mean...

8 of Amir's 11 regular season lineups over 20 minutes had a clearly better net rating than Celtics average. Your cite that he was even better without Thomas than with further reduces the case against him regular season. In playoffs only 1 lineup over 20 minute use. Looks like little and scattershot use then. Scattershot use is usually bad use... by Steven's history and league average history with dink lineups compared to larger ones.

In terms of playoffs estimated RAPM leadership by the top guys, San Antonio would be a laternal move for James. Everything else but the unthinkable a downgrade unless a top 8 player goes with him or 2-3 other very good playoff players end up there, with at least one coming from a move. For Kawhi, he might find better but is unlikely to try.

By several available metrics last season, nobody should be particularly interested in Deron Williams and they don"t seem to be. Playoff RAPM however rated him well on both sides of the court. Maybe just small minute fluke. But I'd watch the tape and see if there is a case for him.

Playoff performance vs regular season, Amir had about a 75% drop in assist rate, a third increase in fouls given and a 1/3rd cut in defensive rebounding. All these things may involve him playing poorly AND possibly affected by others. Need to look at his touch rate and some rebound in details.


Touches down about 25%. Defensive rebound chances about cut in half. Need to watch tape to evaluate causes.

This was his worst playoffs ever. He had been fine the year before. Something happened. Him, teammates, coaching of him. Does he play as poorly next playoffs? Bounces back halfway, all the way? Time might reveal.

Mike G wrote:Amir made 12 million last year, 2nd on the team. In RS he played 20 mpg, #6 or 7 in the rotation. In PO he was 10th/11th at 10 mpg.
Maybe they were tired of paying huge $$ for a guy the coach couldn't find minutes for?

Even if a player compliments a few lineups, if he's an albatross in other lineups, he can be a drag on the team as a whole.

Glancing at some lineup data, I don't get the impression of the limitations you suggest, but whatever. What concerns me more is the "huge $$ for a guy the coach couldn't find minutes for" bit, implying that his contract represented bad value for the team.

Last season Amir Johnson played 1608 minutes, 8.1% of available team minutes, and RPM shows him to have been the 20th ranked player in the NBA at 3.8 (Jeremias says here that straight RAPM has him the 24th ranked, what would suggest an approximately identical contribution) and his salary was $12 million.

Now, I would have thought it in some sense obvious that getting the 20th most productive player, on a per possession basis, for $12 million, even playing only about 60% of a starter's minutes would represent good value, but I guess it's not.

So, here's a pretty straightforward thought experiment/calculation that makes the case. Let's take the top 7 players from last year's Celtics team, in terms of minutes played, and substitute these with 9 "Amir Johnsons" (what it would taker to equalize the number of minutes).

These 73.3% of team minutes then would be played at AJ's RPM of 3.8 (and cost $108 million) and for the remaining 26.7% let's make a conservative assumption and say that the RPM is at rookie/"replacement" level of -3.

The overall team strength you get from this is +9.9, basically a team that in any normal year would be expected to win the NBA finals (and one that would give the GSW a serious contest).

Of course, such a team would probably see a luxury tax be paid. The bottom 7 minute-playing Celtics cost them $25.6 million, so adding this to the hypothetical team of 9 Amir Johnson's, the total salary would be$133.6 (what is less than GSW is on the hook for this year, but, again, a conservative assumption was made about player strength for the "bottom 7", hence an unconservative assumption about cost).

But anyway, I hope the basic point is plain: according to "naive" RPM rules (i.e. the weighted average of RPM implying team strength) Amir Johnson (and his contract) in fact represented value.

But to link this to the larger point of the thread, in this off-season Danny Ainge dumped his three most productive players in RPM terms (both in terms of on court productivity, and productivity per dollar spent) and in exchange (in terms of the primary targets: Hayward and Irving) got players that were strictly inferior, on both metrics, for virtually all pairwise comparisons.

Something changed.

14 yr GM itch? Tired of listening to analytics?

If you have a solar/battery/gas hybrid car that gets 95 mpg when the sun is shining, tops out at 25 mph when it's overcast, and only will go 20 miles after sunset -- is it a good car to drive across the country?
There's something inherently more useful about a car that is capable of going 900 miles in a day, compared to one that goes 300 on a good day. Someone should look around for Stevens' reasons why Amir should only get 20 minutes a night, or 10.

All re: regular season.

Post all-star decline? Not by simple measures. Poor against east top8 in regular season? Slight decline, normal. Better off bench? Nope. Better as C? Emphatically on hoopsstats, not really on 82games. Go with 82games because by stint.Worse in high scoring games? Nope. Trouble shooting in clutch in regular season? Hell no, great. On/off teams stats almost all better with him on, except team attempted / made freethrows. Clogging? Middle of pack team win % on court.

Playoffs: Biggest minute lineups with Bradley. His worst regular season pairing. Great 2nd most used r.s. lineup J. Crowder | A. Horford | A. Johnson | M. Smart | I. Thomas not used or hardly at all in playoffs. A. Bradley | J. Crowder | A. Horford | A. Johnson | M. Smart 4th most used in regular a season, barely tried in playoffs.

Asked a BOS media guy for reaction.

Kevin O'Connor dinged Amir for defensive inconsistency.

Probably true and probably true for most. For effort or effort implied by results.

I'd note it may be true but KOC is a Boston fan and has rationalized some things in the past iirc.

Mike G wrote:If you have a solar/battery/gas hybrid car that gets 95 mpg when the sun is shining, tops out at 25 mph when it's overcast, and only will go 20 miles after sunset -- is it a good car to drive across the country?
There's something inherently more useful about a car that is capable of going 900 miles in a day, compared to one that goes 300 on a good day.

This automobile analogy fails inspection on multiple counts.

https://www.youtube.com/watch?v=9CdVTCDdEwI

I don't think the Celtics expect this trade to make them immediately better.

Also, unless I missed them, no comments on the swap of Bradley for Morris? The trade was a net positive in RPM.