22-23 MVP Tracker

[Mike G]

Jokic, Giannis A. and Tatum in top 4 on DARKO projection (with 2nd place Embid). Doncic 17th.

Do you have a link to this?

[Crow]

https://apanalytics.shinyapps.io/DARKO/

then click current player skill projections and read down the list.

[Mike G]

Darko certainly looks interesting by weighing recent games more heavily. Everyone else seems to just take the season as a whole. Currently, Luka Doncic has just about the highest rating of the year, and for his whole career. So it's not just a few recent weak games that gives him very underwhelming numbers and ranking. Alongside some other stats:

Code: Select all

Luka    DPM    RPM    BPM    WS/48    PER
2019    2.2    1.1    3.9    .101    19.6
2020    3.2    4.3    8.4    .207    27.6
2021    2.7    3.0    6.8    .163    25.3
2022    3.1    6.5    8.2    .159    25.1
2023    3.5    9.0   10.8    .249    30.8
23rank   18     1      2       3      3
#1    Embiid  Luka   Jokic   Jokic  Jokic
#       6.0    9.0   12.1    .300    32.5

The 3 plus-minus stats basically agree in quality -- up, down, up, up -- but not in quantity.

I can rather closely counterfeit his Darko scores from either his RPM or his BPM:
D = (BPM + 0.63)^(0.52)
D = (RPM + 3.9)^(0.49)

Code: Select all

yr    DPM    bpmD    rpmD
19    2.2    2.19    2.20
20    3.2    3.14    2.80
21    2.7    2.84    2.58
22    3.1    3.10    3.15
23    3.5    3.55    3.50

The DPM figures are from holding a straight-edge over the player career graph, from the endpoint of each season.

[Crow]

May be an extreme case but it is a reminder that checking multiple metrics can provide added perspective, even though it might increase uncertainty. I'll usually do that but might not in a hurry.

[DarkStar48]

Jokic, Giannis A. and Tatum in top 4 on DARKO projection (with 2nd place Embid). Doncic 17th.

DARKO has never really liked Dončić, even after some methodological changes. (See: https://twitter.com/kmedved/status/1255902457554755588)

Also, maybe the DARKO ratings had changed, but as of the Dec. 19 update, Jokić is 1st in DPM with 6.1 — not Embiid.

[Mike G]

Jokic, Giannis A. and Tatum in top 4 on DARKO projection (with 2nd place Embid). Doncic 17th.

DARKO has never really liked Dončić, even after some methodological changes. (See: https://twitter.com/kmedved/status/1255902457554755588)

Also, maybe the DARKO ratings had changed, but as of the Dec. 19 update, Jokić is 1st in DPM with 6.1 — not Embiid.

A 40-27-10 game might account for that.
The BPM and RPM conversions I posted above may apply only to Luka. For him to expect 6.0 in DARKO, he'd have to be at 30+ in both.

[Crow]

RPM here is from ESPN? I hadn't check that yet this year. I guess I will now... maybe just once.

I assume still no explanation of past method changes or recent ones? Anyone have any new opinions on its quality?

Anyone else have their own RAPM to share?

[Mike G]

It will be amusing to read and watch the reasons that the Joker should not be MVP again, as he daily puts up the superhuman numbers.

Code: Select all

xyzW    Player          Tm    Min   perW    WS    bpmW
16.2   Nikola Jokić    Den    972   15.0   15.7   17.9
15.8   Luka Dončić     Dal   1136   15.4   14.3   17.6
13.7   Kevin Durant    Brk   1165   13.8   12.8   14.3
12.9  Domantas Sabonis Sac   1041   11.2   14.1   13.4
12.6  Shai G-Alexander OKC   1077   13.0   11.8   13.1

12.5   Jayson Tatum    Bos   1186   12.7   11.9   12.9
12.0   Anthony Davis   LAL    836   12.3   12.1   11.6
11.9  Donovan Mitchell Cle   1091   10.9   11.5   13.5
11.8   Joel Embiid     Phl    859   12.1   11.1   12.2
11.8 Tyrese Haliburton Ind   1041   10.6   11.3   13.4

xyzW    Player          Tm    Min   perW    WS    bpmW
11.0   Giannis         Mil    931   11.7    9.4   11.9
11.0   Stephen Curry   GSW    894   10.2    9.9   12.9
9.8    DeMar DeRozan   Chi   1138   10.7   10.0    8.8
9.6    Julius Randle   NYK   1162    9.3    9.2   10.2
9.5   Zion Williamson  NOP    835    9.5    9.1    9.8

9.3   Lauri Markkanen  Uta   1080    9.1    9.5    9.3
9.0   Pascal Siakam    Tor    842    8.8    8.2   10.0
9.0   Jimmy Butler     Mia    731    8.2    9.4    9.4
8.8   Devin Booker     Phx   1002    9.1    8.3    9.1
8.6   Trae Young       Atl   1069    9.7    6.9    9.3

These alleged 'wins' are extrapolations to the full season, all unofficial.
The current b-r.com tracker:

Code: Select all

Prob%   per game:        W%    Pts    Reb   Ast   Stl   Blk   TO
.508   Nikola Jokić     .66   25.4   11.0   9.4   1.5   .7   3.6
.136   Luka Dončić      .53   32.8    8.3   8.7   1.7   .5   3.7
.117   Joel Embiid      .63   33.1    9.7   4.5   1.1  1.6   3.8
.058   Giannis          .67   31.2   11.2   5.1    .9   .9   3.8
.058   Jayson Tatum     .71   30.9    8.1   4.1   1.2   .9   2.7

.041   Kevin Durant     .64   29.9    6.6   5.3    .8  1.6   3.5
.030  Domantas Sabonis  .55   17.9   12.4   6.7    .8   .5   2.5
.021  Donovan Mitchell  .65   28.9    3.8   4.5   1.4   .3   2.9
.017   Anthony Davis    .39   27.4   12.1   2.6   1.3  2.1   1.9
.015   Stephen Curry    .47   30.0    6.6   6.8   1.0   .3   3.1

[bchaikin]

It will be amusing to read and watch the reasons that the Joker should not be MVP again

you mean like the fact that he's not much of a defender, that denver is currently the 4th worst team in the league defensively, that they allow a league 5th highest 2pt FG%? like those reasons?

[Crow]

I see Jokic has a higher DBPM than any Heat player (almost double Butler). Higher OBPM too of course (about 60% above Butler).

On D-EPM, Butler has a small fractional lead. On O-EPM, it is Jokic by almost 80%.

One player is not a team defense. No Nuggets player other than Jokic has a positive D-BPM. Butler has 9 such positive teammates.

Completely different team designs & results.


Voters probably won't be completely rationale. The MVP regular race numbers race is mildly interesting to me, but I'll be more interested in playoff leadership and progress. Other candidates have more on that part of their resumes. Butler ahead of Jokic on that so far. Jokic with one conference finals and not a close one. In the long-run, he'll need more there or it will be a significant argument diminishing his career greatness some.

[Mike G]

Denver opponents shoot .025 better eFG% when Jokic is on the floor, compared to when he's not.
Yet somehow they score 4.4 pts/100 poss less vs Den w Jokic, than w DeAndre Jordan or other.

With Jokic playing, the Nuggets' D-Rtg is 113.7 -- just off the NBA avg 113.3
When he's out, their defense rivals the Pistons or Spurs (2 worst).
Their ORtg w Jokic is 124.1, almost 6 pts/100 better than #1 Celtics.

To more than make up for the opponent shooting, Den enjoys a 80.1 DReb% with Jokic on the floor. This is 1.3 better than the #1 Bulls.
They also have had a 3.9 Turnover/100 advantage with Jokic playing. This flips to a 2.9 disadvantage when he sits -- 6.8 TO difference in an avg game.
https://www.basketball-reference.com/te ... 23/on-off/

[DarkStar48]

I think defensive metrics overvalue his defensive impact towards the NBA top tier. In addition, his rim protection has regressed from last season.

However, Jokić is no longer a “bad defender,” nor is he uniquely vulnerable for someone of his size.

[Mike G]

It's truly a 2-man race. For a while it was 40-something to 30-something.
Unless both guys get injured, I don't see how anyone else makes a run now. Giannis and Embiid keep missing games, and Durant is out a few weeks now.

Code: Select all

Prob%  player   tm   W%    Pts   Reb   Ast   Stl  Blk   TO
.570   Jokic   Den  .67   25.3  10.8   9.5   1.4   .6   3.4
.214   Doncic  Dal  .56   34.0   8.8   8.7   1.6   .6   3.7
.053   Durant  Brk  .68   29.7   6.7   5.3    .8  1.5   3.5
.047  Giannis  Mil  .64   32.0  11.8   5.2    .8   .9   4.1
.041   Tatum   Bos  .70   30.8   8.1   4.1   1.1   .8   2.7

.025  Sabonis  Sac  .53   18.9  12.4   6.7    .8   .5   2.6
.016   Morant  Mem  .67   27.2   5.9   8.0   1.1   .2   3.6
.015  Mitchell Cle  .63   28.8   3.9   4.8   1.4   .4   3.1
.011 Haliburton Ind .56   20.3   4.0  10.3   1.8   .4   2.6
.009   Zion    NOP  .60   26.0   7.0   4.6   1.1   .6   3.4

I wonder how the Prob% would shift if b-r.com would use their very own Projected team W% instead of the current W%.

[v-zero]

you mean like the fact that he's not much of a defender, that denver is currently the 4th worst team in the league defensively, that they allow a league 5th highest 2pt FG%? like those reasons?

And yet his main competition at this point in time is even less of a defender. It's a pretty empty argument in terms of MVP voting, Jokic won't win it because of voter fatigue, and that's about it.

[DarkStar48]

you mean like the fact that he's not much of a defender, that denver is currently the 4th worst team in the league defensively, that they allow a league 5th highest 2pt FG%? like those reasons?

And yet his main competition at this point in time is even less of a defender. It's a pretty empty argument in terms of MVP voting, Jokic won't win it because of voter fatigue, and that's about it.

Jokić DefRtg: 113.615 pts/100poss.
Jokić raw Def. On/Off: -4.60 pts/100poss.
Jokić FiveThirtyEight “adjusted” Def. On/Off: -2.62 pts/100poss.
Jokić D-RAPM rank: 102

Dončić DefRtg: 115.758
Dončić raw Def. On/Off: +3.65
Dončić FiveThirtyEight “adjusted” Def. On/Off: +1.33
Dončić D-RAPM rank: 404