2023 Playoff misc.

[Crow]

So 3 of the 4 metrics are pretty exactly averaged at their expected average for first 2 rounds. In the 3rd round, 2 become slightly elevated. e484 is exactly centered the whole way (by the division by opponent performance formula). BPM is higher than "expected" (at least by my simple expectation of zero) in all rounds.

How to interpret these other patterns?

[Mike G]

Total "wins" thru 3 rounds, which is just the avg of 4 wins estimates.

Code: Select all

wins    playoffs      tm   Min    PER   WS/48   BPM   e480     perW    WS    bpmW   eWin
3.72   Nikola Jokić  Den   583   31.6   .303   13.4   3.27     3.44   3.55   3.91   3.97
3.09   Jimmy Butler  Mia   669   25.8   .217    7.2   2.51     3.05   2.92   2.91   3.50
3.06   Jayson Tatum  Bos   799   22.5   .167    5.3   2.17     3.04   2.68   2.89   3.61
2.51  Anthony Davis  LAL   608   23.9   .202    5.7   2.18     2.51   2.47   2.29   2.77
2.49   LeBron James  LAL   619   22.5   .170    7.6   2.09     2.36   2.12   2.78   2.69

2.31   Jamal Murray  Den   587   23.0   .193    5.5   2.04     2.30   2.28   2.17   2.50
2.30   Devin Booker  Phx   459   28.6   .236   10.7   2.11     2.39   2.18   2.61   2.02
1.94  Jalen Brunson  NYK   443   23.0   .227    7.7   2.16     1.74   2.02   2.01   1.99
1.72  Stephen Curry  GSW   493   20.5   .132    6.5   1.86     1.65   1.31   2.01   1.91
1.62   Caleb Martin  Mia   552   17.4   .178    3.8   1.19     1.45   1.97   1.68   1.37

wins    playoffs      tm   Min    PER   WS/48   BPM   e480     perW    WS    bpmW   eWin
1.46  Derrick White  Bos   593   16.1   .151    3.0    .83     1.38   1.80   1.63   1.03
1.46   Kevin Durant  Phx   466   21.1   .114    4.5   1.65     1.62   1.07   1.54   1.60
1.43   Marcus Smart  Bos   679   15.3   .112    0.8   1.01     1.46   1.53   1.29   1.44
1.35  Austin Reaves  LAL   579   15.4   .135    2.2    .97     1.26   1.57   1.41   1.17
1.31   Bam Adebayo   Mia   641   15.7   .089   -0.9   1.39     1.44   1.15    .81   1.86

1.31   Jaylen Brown  Bos   751   15.5   .050   -0.9   1.21     1.65    .75    .94   1.90
1.29 Robert Williams Bos   417   19.5   .208    3.4   1.04     1.30   1.74   1.21    .91
1.27   Al Horford    Bos   617   12.7   .119    2.8    .78      .96   1.48   1.65   1.00
1.25   James Harden  Phl   427   18.6   .112    4.1   1.63     1.24    .96   1.35   1.45
1.23  Michael Porter Den   507   15.1   .144    1.5   1.21     1.07   1.47   1.10   1.28

wins    playoffs      tm   Min    PER   WS/48   BPM   e480     perW    WS    bpmW   eWin
1.03   Kevon Looney  GSW   325   18.9   .206    3.1   1.35      .97   1.35    .90    .91
1.02   Tyrese Maxey  Phl   423   16.9   .117    2.2   1.12     1.06    .99   1.03    .99
.97  Anthony Edwards Min   199   27.3   .190   10.4   2.51      .98    .76   1.11   1.04
.96   Joel Embiid    Phl   336   20.4   .091    1.9   1.87     1.12    .61    .78   1.31
.93   Bruce Brown    Den   396   16.2   .158    1.4    .82      .93   1.26    .84    .68

.91   Tobias Harris  Phl   392   16.2   .131    1.4   1.04      .92   1.03    .84    .85
.87   Rui Hachimura  LAL   388   16.0   .146    2.2    .64      .90   1.14    .95    .52
.87   Aaron Gordon   Den   534   13.1   .121   -1.3    .63      .88   1.30    .59    .71
.86  Draymond Green  GSW   367   14.6   .126    2.8   1.08      .73    .93    .98    .82
.83   Kyle Lowry     Mia   454   13.1   .092    1.4    .82      .75    .84    .97    .78
.83   RJ Barrett     NYK   377   14.7   .105    1.3   1.25      .76    .80    .79    .98

Each 'wins' column sums to 86.4
EDIT: "... sums to 79.4"

[Crow]

52 games actually won.

So how does 86.4 "wins" match up?

[Mike G]

Sorry about that.
32+16+8 = 56 games won by series winners.
Losing teams have totaled 23 wins. So that is 79.

Somehow I included the Lakers twice -- after 2 rounds and after 3. Then I manually deleted their duplicitous lines from the text file, but not from the spreadsheet.
So their 8 wins from rd1 & rd2 are the difference.

There are some fractional differences due to overtime games; but then maybe players have "done more" in 53 minutes than in the usual 48.

[Crow]

Ok.

I forgot losers won some games. Oof. Got off-track.

[Mike G]

So 3 of the 4 metrics are pretty exactly averaged at their expected average for first 2 rounds. In the 3rd round, 2 become slightly elevated. e484 is exactly centered the whole way (by the division by opponent performance formula). BPM is higher than "expected" (at least by my simple expectation of zero) in all rounds.

How to interpret these other patterns?

Sorry I didn't see this post earlier.
Here's the table again, somewhat refined. Teams are listed next to their opponents, winners over losers. They're ranked by combined BPM, which could be called 'strength of competition' for the series.
I also took the liberty of scaling Win Shares to average .100 WS/48 min.

Code: Select all

rd1   PER   ws/48   BPM   e480      rd2   PER   ws/48   BPM   e480
NYK   14.8   .133   2.3   1.18      Bos   16.8   .146   2.9   1.25
Cle   12.0   .067   1.0    .82      Phl   14.5   .054   0.7    .75
LAL   15.6   .142   2.4   1.21      Den   17.8   .158   2.9   1.25
Mem   12.1   .058   0.7    .78      Phx   14.1   .042   0.6    .75
Phx   18.3   .143   2.3   1.17      LAL   16.1   .120   1.6   1.09
LAC   14.8   .057   0.7    .83      GSW   13.6   .080   0.9    .92
Bos   17.5   .136   1.9   1.14      Mia   14.2   .116   1.3   1.13
Atl   15.4   .064   0.7    .85      NYK   13.0   .084   0.3    .87
Den   16.0   .149   2.1   1.25                  
Min   13.8   .051  -0.1    .76      rd2   15.0   .100   1.4   1.00
GSW   14.9   .113   1.2   1.06                  
Sac   13.7   .087   0.7    .94      rd3   PER   ws/48   BPM   e480
Mia   16.6   .126   1.3   1.14      Den   17.3   .137   2.7   1.17
Mil   14.9   .074   0.1    .86      LAL   15.4   .063   1.1    .83
Phl   16.3   .174   2.2   1.32      Mia   15.7   .121   1.5   1.12
Brk   12.2   .026  -0.8    .68      Bos   14.5   .079   0.6    .88
                              
rd1   14.9   .100   1.2   1.00      rd3   15.8   .100   1.5   1.00

PER is mostly 'offense', so the most talent-rich 1st round series (Knicks-Cavs) was a defensive struggle and looks 'worst' by PER.
BPM is the one that 'does it right', or sincerely tries to. It doesn't account for 'playoff Jimmy' or 'playoff Murray' or 'playoff LeBron'; nor does it reckon whether Durant after 40 minutes is equal to Durant before 35. Still, it should be close to right.

I'm squeamish about assigning more 'wins' than games; so I have stuck to 1 total player-win per game-played -- that is 1.00 eW/480+ minutes. Adjusting relative to BPM might give a more 'honest' total, like 'equivalent wins vs average team'; but it's still an estimate.

I don't know why WS/48 for a series runs high. Teams avg .102-.106 in a series. [For this post, it's adjusted to .100 per series.]
In close series, WS/48 is Pythagorean -- note that (ignoring the decimal) it is just twice the difference from avg, as is eW/480. In lopsided series (Phl/Brk) it departs from the asymptotic range (zero to .200) that constrains Pyth (and eW).
If you lose by more than about 15 ppg, you can have negative WS.

[Mike G]

The Joker's first Finals game (27-10-14, .781 TS) does not make the top100 playoff games; but it ranks pretty well among this year's entries in BPM*Min/48:

Code: Select all

bpmMin  player   tm   opp
27.5   Harden   Phl   Bos
24.6   Tatum    Bos   Phl
23.3   Booker   Phx   Den
22.3   Morant   Mem   LAL
20.0   LeBron   LAL   Den
18.5 * Jokic *  Den * Mia
17.9   Butler   Mia   Mil
16.6   Jokic    Den   LAL
15.8   Booker   Phx   LAC
15.2   Jokic    Den   Phx
14.0   Booker   Phx   LAC

This actually seems a more valid ranking, since GmSc says nothing about Defense. Dominating a low-scoring game vs thriving in a high-scoring one.
I see ESPN has an article on top Finals games, according to GmSc.

[Mike G]

Through 20 playoff games, the Heat have shot .524 eFG% when Jimmy Butler is on the court and .622 when he is off.
Opponents have shot .505 when he's not in the game vs .542 when he's out there.
So despite his leading the team in assists (by a lot) and his defensive skills, and his genius at both ends, his presence has been accompanied by a whopping -.135 deficit in team-opponent shooting%

Miami rebounding% is also worse with Jimmy, .478 vs .505 without.
Turnovers are essentially the same: a wee bit more for Heat and opponents when Butler is in.
https://www.basketball-reference.com/te ... 23/on-off/

Per 100 on-off net rates so far:

Code: Select all

On-off     Pts    eFG%    TO%   TRb%
Martin    10.5    .060     .2    5.6
Vincent    9.0    .018    5.6   -7.2
Robinson   8.1    .075    -.9    5.6
Adebayo    8.0   -.025    3.7    4.2
Love      -1.8   -.028     .2    1.4
Zeller    -4.3    .036   -2.5   -4.6
Lowry     -4.9   -.002   -1.8   -4.0
Highsmith -8.3    .030   -5.6    5.6
Strus    -10.4   -.081     .8   -4.8
Butler   -15.9   -.135     .2   -5.4

Numbers refer to team-opponent advantage; so positive TO% on-off is good.

Nuggets:

Code: Select all

on-off   min.    Pts    eFG%    TO%   TRb%
Gordon   608    14.7    .045    6.1     .2
Brown    444     6.9    .028   -1.3    9.0
Porter   578     5.7    .021    -.1     .2
Murray   670     3.5    .006    6.8   -5.2
Jokic    666     2.2   -.039     .2    2.4
Braun    189    -1.4    .022   -2.3    4.0
Green    307    -9.4   -.007   -2.7     .4
KCP      573   -10.2   -.080    1.4   -7.8

[Crow]

The top 8 Butler playoff lineups are +118 in about 530 minutes,l for +10.7 / 48 minutes.

The problem, the super-massive problem, must be in Spoelstra selected dink lineups. Especially with Lowry, Zeller and Highsmith.


The other super-massive issue is how good the Heat are without him. Butler on is far different but still +0.8 / 100p overall.

But those 55 lineups beyond the biggest 8 Butler lineups are apparently -130 in 217 minutes for an average of -28.9 / 48 minutes. Total garbage and at least somewhat reducable.

NBA teams show insufficient awareness and concern or at the very least are taking insufficient action in response to their weak lineup choices. It is astonishing. The winner is often who is less chaotic and sloppy. Very unimpressed by large parts of NBA lineup management. Can't always get good results but probably should shut down more of the worst performing faster and favor the strong far more. There is randomness but there is signal in the results too.

[Crow]

Raw Butler on / off is probably worse than adjusted for opponent and teammate on / off.

[Mike G]

I had thought Miami was winning the coaching battle by employing more players. Unless Denver was able to hammer them to death in another sweep, they're going to lose to exhaustion.

Meanwhile, as a coach it must be tempting to throw out many combinations of players. All your guys have played together a lot in practices and RS games, while the opponent may have never seen some lineups. By the time they've adjusted -- lineup change or coverage -- you may have gained something, however briefly.

Anyway, without Murray or without Gordon, the Nuggets' TO rate skyrockets. What is that about?

[Mike G]

Other than Max Strus, every player in the Miami rotation has shot 3-pointers better in these Finals AND throughout the playoffs, than they did in the regular season.
https://www.basketball-reference.com/te ... /2023.html
Ranked by 2-game Finals 3-pt attempts, showing previous 3FG%, expected 3fg% (= avg of RS and PO), and points above or below that expectation.

Code: Select all

Heat    3fga 3fg   RS%    PO%   exp%   xPts  diff
Strus    19   4   .350   .336   .343   19.6  -7.6
Vincent  16   9   .334   .413   .374   17.9   9.1
Lowry     9   5   .345   .380   .363    9.8   5.2
Robinson  8   3   .328   .440   .384    9.2   -.2
Butler    7   3   .350   .364   .357    7.5   1.5
Love      6   2   .297   .365   .331    6.0    .0
Martin    4   2   .356   .441   .399    4.8   1.2
Highsmith 4   2   .339   .500   .420    5.0   1.0
Adebayo   1   0   .330   .000   .165     .5   -.5
.                                             9.8

Miami totals 10 pts more than expected from the arc. Denver has gone the other way:

Code: Select all

Nuggs   3fga 3fg   RS%    PO%   exp%   xPts   diff
Porter   17   3   .414   .375   .395   20.1  -11.1
Murray   15   5   .398   .391   .395   17.8   -2.8
Jokic     7   3   .383   .469   .426    8.9     .1
Brown     6   3   .358   .318   .338    6.1    2.9
KCP       6   2   .423   .405   .414    7.5   -1.5
Gordon    3   2   .347   .375   .361    3.2    2.8
Green     1   1   .288   .296   .292     .9    2.1
.                                             -7.5

Only the Joker has been significantly better from the arc in PO.

[Mike G]

Finalists Jokic and Butler are #1-2 in both playoff Win Shares and VORP.

Code: Select all

VORP                Win Shares
2.8  Jokic   Den    4.6  Jokic   Den
1.7  Butler  Mia    3.2  Butler  Mia
1.5  Booker  Phx    2.9  Murray  Den
1.5  LeBron  LAL    2.8  Tatum   Bos
1.5  Tatum   Bos    2.6  Davis   LAL

It's not especially close, but is it possible that Butler's big games have had more impact, or been more timely?
I don't discern any such pattern here. Ranked by single game vorp = BPM*Min/48:

Code: Select all

Jimmy Butler                      Nikola Jokic
vorp  Rd G   min   BPM   win     vorp  Rd  G  min   BPM   win
17.9  1  4   41   21.0    5      18.6   4  1   41   22.0   11
11.1  2  4   42   12.7    8      18.5   2  6   38   23.6   25
9.1   2  5   43   10.2   -9      18.1   1  4   39   22.2   -6
8.9   1  3   28   15.1   22      16.8   3  1   42   19.0    6
7.9   2  3   36   10.7   19      15.4   2  4   40   18.7   -5
7.6   3  1   43    8.4    7      15.2   2  2   41   17.7   10
7.3   3  4   39    9.0  -17      14.4   2  3   42   16.5   -7
5.7   1  5   46    5.9    2      13.8   4  3   44   15.1   15
5.6   3  7   43    6.2   19      12.5   2  5   38   16.0   16
5.5   3  2   41    6.5    6      12.5   3  4   45   13.2    2
4.7   1  2   28    8.0  -16       9.7   1  5   42   11.1    3
3.6   3  6   47    3.7   -1       8.1   1  3   38   10.2    9
2.7   4  3   40    3.2  -15       7.4   3  2   42    8.4    5
2.7   1  1   43    3.0   13       4.7   1  2   37    6.0    9
2.5   2  1   44    2.8    7       3.8   1  1   29    6.4   29
2.5   4  1   38    3.1  -11       3.3   4  2   42    3.8   -3
1.2   3  5   34    1.7  -13       3.3   2  1   33    4.8   18
0.6   4  2   40    0.7    3       2.2   3  3   38    2.7   11
-.4   2  6   41   -0.5    4                     
-.7   3  3   31   -1.1   26                     
105.7                            198.0   

Game 2 of these Finals is Joker's only weak game that came in a loss (negative in 'win' column).
The sum of these (bottom line) is almost a 2:1 ratio.

[Mike G]

Jokic leads all players in this postseason in points (by 3 over Butler), rebounds (37 ahead of Bam) and assists (60 more than Murray).
Top 250's alltime are listed by b-r.com

Code: Select all

Pts   points    tm    rk
549   Jokic    Den    48
546   Butler   Mia    51
543   Tatum    Bos    52
493   Murray   Den    87
453   Brown    Bos   131
396   Curry    GSW   232
392   LeBron   LAL   238
         
Reb  rebounds   tm    rk
241   Jokic    Den    75
225   Davis    LAL    98
210   Tatum    Bos   131
204   Adebayo  Mia   142
170   Looney   GSW   250
         
Ast  assists    tm    rk
182   Jokic    Den    22
122   Murray   Den   120
117   Butler   Mia   136
105   Tatum    Bos   200
104   LeBron   LAL   207
101   Smart    Bos   228
         
Stl   steals    tm    rk
36    Butler   Mia    75
28    Murray   Den   190
         
Blk   blocks    tm    rk
50    Davis    LAL    26
34    Horford  Bos   102
25    Embiid   Phl   220
25   RWilliams Bos   220

https://www.basketball-reference.com/leaders/

[Mike G]

As of last night, Joker heads the list of players who have had postseasons now among the top 250 (ever) in total points, rebounds, AND assists.

Code: Select all

Playoffs        Year   Pts   Reb   Ast rk: pts   reb   ast
Nikola Jokić    2023   572   253   186      38    56    20
George McGinnis 1975*  581   286   148      36    36    58
Larry Bird      1984   632   252   136      17    60    79
Larry Bird      1987   622   231   165      18    87    33
LeBron James    2015   601   226   169      24    95    32
LeBron James    2020   580   226   184      37    96    20
LeBron James    2012   697   224   129       7   101    98
Tim Duncan      2003   593   369   127      28     9   103
LeBron James    2016   552   200   160      47   147    42
LeBron James    2018   748   200   198       2   148    15
Scottie Pippen  1992   428   193   147     170   172    63
LeBron James    2013   596   193   152      26   175    52
Giannis A       2021   634   269   108      15    44   182
Nikola Jokić    2020   464   187   109     116   195   177
Jayson Tatum    2023   543   210   105      52   132   204
Larry Bird      1985   520   182   115      68   205   145
Charles Barkley 1993   638   326   102      14    17   222
Paul George     2021   511   183   102      75   204   225
LeBron James    2011   497   176   123      85   226   117
John Havlicek   1969   458   179   100     124   213   230
Khris Middleton 2021   542   174   118      54   234   134
Hakeem Olajuwon 1994   664   254    98      12    57   250
Hakeem Olajuwon 1995   725   227    98       4    94   250

Ranked by best worst category -- rankings on the right.
* - McGinnis' was ABA