2017-18 quick observations

[Crow]

Wiggins taking way more 3s, defensive rebounding a bit more, getting a bit less assists. Overall WS/48 and BPM have him at well below average impact as usual. RPM coming soon, supposedly. Slightly negative raw plus minus on the court.

Wolves only 19th on SRS.

But Glen Taylor can use Wiggins as a shiny object to shine up franchise sale price, if / when he gets around to it.

[Crow]

Tatum as the runaway favorite for rookie of year. John Collins Anouby, Markkanen and Jordan Bell should get runner up votes over higher draft picks.

Hat tip to Celtics for Tatum and most moves. WS/48 likes their deep bench far more than BPM does. RPM? Probably more like BPM.

[Crow]

Van Gundy said he was going to use Boban more? Not yet.

There seem to be a lot of teams with a mix of good factor marks and bad ones. 76ers, 4 top 10 factors and 4 bottom 5s.

Magic looking mostly very good. Except mainly on Ross & Biyombo.

Oubre improving and big minute bumps aid that impression. Is he up to average? Still below? Above? Box score metrics don't resolve. I doubt eye test does either but is there any practitioner of that or non-systematic roll-up impressionism that wants to answer this question? (Probably easier to change you mind that way. People have to remember what you said since there is no written summary numeric estimate. And you can always say "things change", since there are no requirement of consistent standards for your ad-hoc roll-up. Brill or lame?) RPM is another vote.

[Crow]

BRef and NBA.com still appear to vary a bit on Giannis' plus minus but the gap has narrowed a lot. BRef gave him a plus 25 in his first game previously, now it is only plus 6 or 7. Bottom line it is still very close to only neutral.

[Crow]

Rubio's ts% is currently exactly where it was last season and near league average. TO% very high as always. Overall boxscore metrics are down some.

[Mike G]

BRef and NBA.com still appear to vary a bit on Giannis' plus minus but the gap has narrowed a lot. BRef gave him a plus 25 in his first game previously, now it is only plus 6 or 7. Bottom line it is still very close to only neutral.

That's a serious correction!
B-Ref still says Bucks are +1.6/100 with Giannis on court, and his On-minus-Off is +19.4
That means the team without him has been -17.8/100 -- when he sits, they are hopeless.

No team has been that bad for a season -- such a team would only win a handful of games -- but the Bucks are heading toward mid-30s.
Turning the worst team ever into a competitive team is not really "neutral", is it?

Tatum as the runaway favorite for rookie of year. John Collins Anouby, Markkanen and Jordan Bell should get runner up votes over higher draft picks.

Isn't Ben Simmons a rookie?
Simmons' On-Off = 4.8, vs zero for Tatum: Celts are just as good without him.
Tatum with better WS, perhaps due to better team; Simmons has better everything-else.
It does look like it will be between these two.

[Crow]

I hadn't checked the accuracy of BRef off court estimate for Giannis, yet. I get a little less. I am not arguing he doesn't have big impact vs being absent just that it is so far insufficient impact to make them a serious force. If everything is built for and about you then it might not be so surprising that things in that ethos don't work when you sit. Are the Bucks without Giannis the worst team in history? I don't think so even though they play like it or near it. That they play like a borderline playoff team with him is the main and somewhat surprising story to me. It has been said LeBron can make any team win 50 plus and except for his first 2 seasons that has been true. When will Giannis win 45? Will it be this season? If it isn't, is that a disappointment?

Simmons probably does qualify as a rookie. I forgot. So you have a race.

[Mike G]

B-R.com doesn't seem to agree with itself. But they do offer game logs with plus-minus.

Code: Select all

Bucks      w Giannis   no Giannis
G   mov    Min   +/-    Min-  +/-
1     8    36.9    7    11.1    1
2   -19    38.5  -17     9.5   -2
3     3    39.3    3     8.7    0
4     9    39.3    7     8.7    2
5    -7    41.6  -10     6.4    3
6    11    38.3    8     9.7    3
7   -19    27.8  -15    20.2   -4
8    -5    32.0   14    16.0  -19
9    -9    38.0    9    10.0  -18
10   -5    40.2   -7     7.8    2

tot -33   371.9   -1   108.1  -32
per48           -0.1        -14.2

The difference with or without him is 14.1 pts/48 min.
Bucks go 96 poss/48, so it's about +14.7 per 100 poss.

[Crow]

Yes I get about -14 from other sources too.

As I said, still a discrepancy and I doubt it is just with one player so I am going to try to avoid using their on/off data or will cross-check it.

[Crow]

J Butler barely positive on raw plus minus. Will be interesting to see the RPM estimate.

[Mike G]

https://www.basketball-reference.com/te ... 18/on-off/
... says the team is +1.9 when he's on court and -9.6 when he's off. That's +11.5 in raw on-off.
When he plays, Wolves offense is 5.0 better and defense is 6.4 better.

[Crow]

The Jazz have an awful offense with no factor strength and 3 really bad ones. They have a top level overall defense but it is composed of 3 average factors and an elite mark on opponent turnovers, a factor of relatively modest impact.

[Crow]

On court plus minus and on / off plus minus are 2 different things. Both of interest. I am highlighting on court. Unless you get strong positive on the court you don't end up achieving much as a team.

[Mike G]

Since team total +/- is always known (as MOV), it's almost impossible to know a player's on-court +/- without acknowledging the off-court +/-.
That is, it takes no extra effort to consider team context as part of basic raw +/-.

The raw part is not knowing substitution patterns. If Butler's backup is especially weak, it exaggerates his impact. Actually, his impact is specific to this team, with this particular lineup management. He might or might not make the Warriors 12 points better, or the Kings.

If players could go 48 minutes every game with no loss of effectiveness, then a team's best 5 could go as long as they want, and their on-court +/- would be much higher. Because they have to rest, one or more at a time, the best players share time with weaker ones, and they have lower +/- in those minutes.

Last year, the Warriors were +6.7/100 with Shaun Livingston on the court. What does that tell us, if we don't know their overall MOV?
They were 7.5 better when he sat. That tells us much more.

[Crow]

Dion Waiters this season: below average efg% & ts%. Replacement level BPM. Worse than replacement level WS/48.