2023 all-stars

[Mike G]

Top 70 players according to some stats.
https://www.basketball-reference.com/le ... anced.html
Separated by Conference. Currently named all-star selections = *

Code: Select all

xyzW   Western  Tm   Pos   min  perW   WS   bpmW      PER   WS/48   BPM
10.2  Jokić   * DEN   C   1512   9.1  10.2  11.4      32.2   .324  13.3
9.6   Dončić  * DAL  PG   1714   9.6   8.3  10.9      30.5   .233  10.3
8.1   Shai    * OKC  PG   1673   8.1   7.8   8.6      27.1   .223   7.3
7.2   Sabonis * SAC   C   1658   6.5   7.7   7.5      23.0   .224   5.8
6.8   Lillard * POR  PG   1404   6.5   6.7   7.1      26.3   .229   7.1

6.6 Markkanen * UTA  SF   1645   6.4   6.8   6.5      23.0   .197   4.5
6.3   LeBron  * LAL  PF   1529   6.7   4.7   7.4      25.1   .149   6.7
5.9   Curry   * GSW  PG   1287   5.5   5.3   6.9      24.7   .198   8.0
5.8   Morant  * MEM  PG   1436   5.9   4.8   6.6      23.8   .162   6.1
5.2   A Davis   LAL   C    967   5.4   5.1   5.1      30.3   .253   7.7

xyzW   Western  Tm   Pos   min  perW   WS   bpmW      PER   WS/48   BPM
4.8   Fox     + SAC  PG   1584   5.3   4.6   4.6      20.6   .138   1.9
4.7   Edwards + MIN  SG   1977   5.4   3.3   5.4      17.9   .080   1.4
4.6   Bridges   PHO  SF   1930   4.0   4.7   4.9      14.9   .118   1.0
4.5   Grant     POR  PF   1738   4.6   4.5   4.5      17.4   .123   1.1
4.5   Gordon    DEN  PF   1389   4.5   4.8   4.2      20.1   .165   2.2

4.5  Dinwiddie  DAL  PG   1776   4.3   4.6   4.4      16.5   .125   0.9
4.2   Şengün    HOU   C   1294   4.6   4.0   4.0      21.4   .147   2.4
4.1   Gobert    MIN   C   1366   4.0   5.1   3.3      18.8   .179   0.6
4.0   Zion    * NOP  PF    956   4.2   3.6   4.3      25.3   .182   5.7
4.0   Looney    GSW   C   1218   3.2   5.3   3.6      17.2   .208   2.0

xyzW   Western  Tm   Pos   min  perW   WS   bpmW      PER   WS/48   BPM
3.9   Russell   MIN  PG   1691   4.0   3.6   4.2      16.1   .103   0.9
3.9   George  * LAC  SG   1296   4.2   3.2   4.4      19.9   .120   3.1
3.8   Leonard   LAC  SF    924   3.7   3.6   4.1      23.4   .186   5.7
3.7   Murray    DEN  PG   1428   4.0   3.5   3.6      18.1   .118   1.0
3.7   Wood      DAL   C   1219   4.2   3.5   3.5      20.8   .136   1.8

3.7   Kessler   UTA   C   1018   3.6   4.3   3.3      21.3   .201   2.6
3.6   Ayton     PHO   C   1311   4.2   3.6   3.1      19.8   .131   0.6
3.6   McCollum  NOP  SG   1658   3.8   2.5   4.4      15.9   .072   1.2
3.5   Murphy    NOP  SF   1462   2.7   4.1   3.8      14.0   .134   1.1
3.5   Barnes    SAC  PF   1639   3.2   4.2   3.2      14.3   .122  -0.4

xyzW   Western  Tm   Pos   min  perW   WS   bpmW      PER   WS/48   BPM
3.5   Jackson * MEM   C    945   3.4   3.7   3.4      21.8   .188   3.6
3.5   Booker    PHO  SG   1002   3.6   3.2   3.7      21.7   .151   3.9

Davis has played more minutes than Jackson, along with being about twice as good.

The East picks depart from the numbers a bit earlier, and they dig just as deep.

Code: Select all

xyzW   Eastern  Tm   Pos   min  perW   WS   bpmW      PER   WS/48   BPM
7.8   Tatum   * BOS  PF   1827   7.8   7.3   8.3      24.6   .193   5.9
7.1   Embiid  * PHI   C   1323   7.5   6.7   7.3      30.7   .242   8.3
6.6   Giannis * MIL  PF   1374   7.0   5.5   7.2      28.2   .191   7.7
6.6   Randle  * NYK  PF   1900   6.5   6.0   7.2      20.8   .152   4.0
6.5   Durant  * BRK  PF   1403   6.6   6.0   6.9      26.5   .204   6.9

6.4  Mitchell * CLE  SG   1563   6.0   5.8   7.4      22.6   .177   6.3
6.3   Butler    MIA  SF   1287   5.7   6.5   6.6      25.3   .244   7.4
5.9  H'burton * IND  PG   1372   5.4   5.4   7.0      23.2   .190   7.2
5.8   DeRozan * CHI  SF   1740   6.2   5.8   5.3      21.6   .159   2.2
5.6   Brunson   NYK  PG   1697   5.5   5.5   5.7      20.1   .155   3.0

xyzW   Eastern  Tm   Pos   min  perW   WS   bpmW      PER   WS/48   BPM
5.5   Vučević   CHI   C   1733   5.3   5.3   5.8      19.3   .147   2.9
5.3   Adebayo * MIA   C   1687   5.8   5.3   4.9      21.0   .150   1.9
5.3   Garland   CLE  PG   1598   5.1   5.4   5.6      19.8   .161   3.3
5.2   Claxton   BRK   C   1409   5.1   5.9   4.7      21.6   .202   2.9
5.2   Plumlee   CHO   C   1503   5.1   5.8   4.8      20.7   .184   2.6

5.2   Allen     CLE   C   1560   4.7   6.2   4.7      19.1   .191   2.1
5.1   Harden    PHI  PG   1249   4.7   5.2   5.5      22.2   .200   5.6
5.1   Siakam  + TOR   C   1622   5.5   4.5   5.4      20.6   .132   2.9
5.1   Irving  * BRK  PG   1478   5.4   4.6   5.3      21.9   .148   3.6
5.0   Young     ATL  PG   1634   5.9   3.9   5.3      21.6   .115   2.7

xyzW   Eastern  Tm   Pos   min  perW   WS   bpmW      PER   WS/48   BPM
4.7   Mobley    CLE  PF   1779   4.5   5.0   4.5      16.9   .136   1.0
4.7  Porziņģis  WAS   C   1369   4.9   4.5   4.7      21.6   .157   3.1
4.5   VanVleet  TOR  PG   1627   4.3   4.2   5.2      17.3   .123   2.6
4.4   Brown   * BOS  SF   1650   5.2   3.6   4.5      19.6   .105   1.4
4.4   Murray    ATL  SG   1716   4.8   3.4   5.1      18.1   .094   2.0

4.2   Turner    IND   C   1319   4.3   4.2   4.2      20.0   .154   2.5
4.1   Lopez     MIL   C   1541   3.8   4.4   4.2      16.6   .138   1.4
4.0   Barnes    TOR  PF   1747   4.3   3.4   4.3      16.5   .093   0.8
4.0   LaVine    CHI  SG   1683   4.5   3.3   4.1      17.5   .094   0.8
3.9   F Wagner  ORL  SF   1692   4.3   3.6   3.8      16.9   .102   0.2

xyzW   Eastern  Tm   Pos   min  perW   WS   bpmW      PER   WS/48   BPM
3.9   Holiday * MIL  PG   1346   3.9   3.6   4.1      18.5   .127   2.3
3.8   Harris    PHI  PF   1556   3.5   4.0   4.0      15.8   .123   1.0
3.8  Bogdanović DET  PF   1548   4.3   3.1   3.9      17.9   .095   1.0
3.7   Hield     IND  SF   1704   3.5   2.9   4.8      14.9   .081   1.6
3.6   Capela    ATL   C    972   3.7   4.2   3.0      22.5   .209   2.3

3.6   Portis    MIL  PF   1258   3.7   3.9   3.3      18.6   .148   1.2
3.6   Brogdon   BOS  PG   1096   3.2   3.7   3.7      18.8   .164   3.0
3.5   Trent Jr. TOR  SG   1534   3.6   3.5   3.6      16.0   .109   0.5
3.5   Robinson  NYK   C   1022   3.0   4.3   3.3      18.8   .203   2.6

Butler is 6th in the league in BPM, 8th in PER, and 2nd in WS/48.
He's played as many games as Embiid -- 6 G more than Jrue -- and is 3rd in WS in the East.

EDIT Feb.10: Siakam, Fox, and Edwards added by commish (+) to replace injured players.

[Crow]

J Jackson 15th on EPM, Haliburton 16, George 17, Zion 18.

Butler 9, Harden 21.

Fox 42. Brunson 95. Paul 49. LaVine 53.

Holiday 29. Randle 33. DeRozan 39. J Brown 41. Adebayo 47.

Could have been worse. At least some undeserving old heads got moved out- for effectiveness and / or limited games.

Wonder if it was pure coaches or manipulated a bit by league.

Adebayo 35 on DARKO, 22 on RAPM, 67 on luck adjusted RAPM.

DeRozan 75 on DARKO. 88 on RAPM, 58 on luck adjusted RAPM. DeRozan should not have made it and this better be his last time.

Fox, 70th on DARKO, 210 on RAPM, 155 on luck adjusted RAPM. Fox was not unfairly snubbed imo and probably should never make it.

Brunson, 87 on DARKO, 308 on RAPM, 108 on luck-adjusted. Not a snub.

J Brown, 17 on DARKO, 175 on RAPM, 196 on luck adjusted. Probably shouldn't have made it. But good team reward.

Holiday, 10 on DARKO, 394 on RAPM, 380 on luck adjusted. Maybe it is his last time.

The GAME is about ratings and money. The coaches probably are influenced by that.

Randle 156 on DARKO, 110 on RAPM, 114 on luck adjusted. Mostly New York / Thibs gift.

J Jackson, 40 on DARKO, 23 on RAPM, 4 on luck adjusted. Plenty of reasons to go with him. Enough games to not make it a deal breaker. Off injury and not a slacker, game avoider... like Kawhi and others.

[Mike G]

Brunson, ... 308 on RAPM, 108 on luck-adjusted.

That's quite a difference that "luck" makes. Are we sure this means anything?

[Crow]

No, I am not sure.

I have mixed feelings about luck adjustments. I understand the concept and potential improvement on average but it will make mistakes too with individuals. Luck adjusted only for 3 point shooting is limited adjustment, maybe useful but also selective and limited on application of the concept.

Regression and regularization already act on presence or asserted presence of "luck".

[Crow]

14 All-Stars (I think) on top 6 teams in each division. Average of almost 1.2. 6 on next 8 teams. Average of 0.75. 4 on bottom 10 teams. Average of 0.4.

[Crow]

Pretty good chance Edwards replaces Curry. Davis might end up replacing Williamson.

[Crow]

Is Irving still eligible to play for East? Will he?

[DSMok1]

No, I am not sure.

I have mixed feelings about luck adjustments. I understand the concept and potential improvement on average but it will make mistakes too with individuals. Luck adjusted only for 3 point shooting is limited adjustment, maybe useful but also selective and limited on application of the concept.

Regression and regularization already act on presence or asserted presence of "luck".

Within a single season, the influence of shooting luck by the opponent is a very large portion of the variance observed. Without this luck adjustment, I would consider single-year RAPM utterly useless.

With the luck adjustment, it is still very noisy but perhaps more useful.

Ryan Davis's approach to luck adjustment is to replace each opponent's points scored from 3 and from the FT line with their career average from that location on the court. I think that's a fair way to do the adjustment.

Is still consider single-year true RAPM very noisy just due to the unavoidable covariance issues. With a prior of some sort (ie. EPM) it can be even better.

[Crow]

"Within a single season, the influence of shooting luck by the opponent is a very large portion of the variance observed. Without this luck adjustment, I would consider single-year RAPM utterly useless."

Is there any opponent shooting luck adjustment in BPM? To my understanding there is not. Is single year BPM "utterly useless" on same basis? At least on defense?

On other metrics:
Doesn't appear to be in DARKO DPM. In LEBRON. Unclear for RAPTOR, so without explicit confirmation I assume not. DRIP appears to not do it from explanation, though talk of maybe doing in future. Obviously not in the simplest boxscore metrics.

[Crow]

I don't know how they will handle Irving now being in West but theoretically he could replace Curry in West and then be replaced in East by Butler, Harden or Claxton.

[DSMok1]

"Within a single season, the influence of shooting luck by the opponent is a very large portion of the variance observed. Without this luck adjustment, I would consider single-year RAPM utterly useless."

Is there any opponent shooting luck adjustment in BPM? To my understanding there is not. Is single year BPM "utterly useless" on same basis? At least on defense?

On other metrics:
Doesn't appear to be in DARKO DPM. In LEBRON. Unclear for RAPTOR, so without explicit confirmation I assume not. DRIP appears to not do it from explanation, though talk of maybe doing in future. Obviously not in the simplest boxscore metrics.

Random fluctuations of opponent shooting have a very large impact on RAPM because of how it is constructed, but have comparitively little impact on box score metrics that don't look at individual lineup performance, just team as a whole.

It's similar to why looking at lineup performance without adjusting for shooting luck is almost meaningless. Far too small of a sample size to balance out random fluctuations.

Now, even at a team level there can be impacts from shooting luck--for instance, the discussion around why Celtics opponents seemed to shoot worse than "expected" from 3 for several years in a row.

[Crow]

BPM has all the random fluctuations of opponent shooting, with a player on the court and not. Maybe it reduces "luck" impact, maybe not for every team. But rating a player based on shot defense for on and off is a distortion of its own.
(Rating a player based on shot defense for on VERSUS off may be a distortion for RAPM.)

Nobody luck adjusts own shooting and nobody justifies doing it just for defense.

Nobody luck adjust steals but there is certainly random fluctuations / luck.
Fouls have random fluctuations / luck. Everything has random fluctuations / luck. Luck adjustment is questionable. Selective luck adjustment is very questionable.


It is better to look at and consider selective luck adjusted and non luck adjusted than just one. Not easier, but probably better.

[Mike G]

Do we assume that 'at the team level', opponent 3FG% is not a matter of luck so much as defensive strategy and personnel?

The Spurs are giving up an unconscionable .401 3fg% -- fully .041 above the league avg. Month after month at .380 to .420.
If an individual player on the floor sees opponents shoot .380 from the arc, is he considered to have been 'lucky'? Aren't players as well as teams responsible for opp FG% ?

Note, this may not be the best example. Pop may be stealthily winning the Wembayana sweepstakes here .

[Crow]

Durant out, another "snub" in.

[Crow]

17 of 19 currently at +4 BPM or higher made All-Star game.

A Edwards joins J Brown with lowest BPM estimates in group at +1.2. Yes, Fox is higher at +1.7.