Page 12 of 13

Re: 2023-24 team win projection contest

Posted: Thu Feb 29, 2024 9:05 pm
by Crow
I didn't have a specific plan for or expectation of low yr to yr change but I guess it happened to a degree of 2nd least.

Re: 2023-24 team win projection contest

Posted: Fri Mar 01, 2024 1:59 pm
by Mike G
Steamlined contest: number of best guesses, incl. exact ties.

Code: Select all

11 medi  Bos Cha Cle Orl Tor Den GSW Hou OKC Por Sac
5  emin  Det Tor Was LAL Min
5  eWin  Atl Brk Chi Ind Phx
4  DQin  Mia Mil LAC NOP
2  vzro  Mem SAS
2  dtka  Phl Dal
1  trzu  Uta
1  ncs.  NYK
1  Crow  Phx 
Gotta say I much prefer this ranking.

Re: 2023-24 team win projection contest

Posted: Fri Mar 01, 2024 7:19 pm
by v-zero
:lol:

Re: 2023-24 team win projection contest

Posted: Wed Mar 06, 2024 3:49 pm
by Mike G
Sixers without Embiid this season are 9-18 and -5.5 ppg. This is better than only the terrible 5 and equal to the Grizzlies.
The b-r.com forecast has them going 12-9 the rest of the way -- Memphis 7-13 with a lighter schedule -- to finish with 47 wins, 6th or 7th in the East.
With 5 fewer wins, they'd still be #8. Embiid may return, of course; but perhaps a bit shaky, or a lot.

Re: 2023-24 team win projection contest

Posted: Mon Mar 11, 2024 1:02 pm
by Mike G
https://www.basketball-reference.com/fr ... _prob.html

Code: Select all

.  avg err   rmse   r^2      .  avg err   rmse   r^2
medi   5.03   6.3   .77      EBPI   7.19   8.5   .60
ChKl   5.80   7.7   .64      vzro   7.22   8.6   .54
EExp   5.82   7.1   .71      KPel   7.70   9.4   .48
vegas  5.95   7.3   .71      DRKO   7.80   9.2   .50
IanL   6.06   7.2   .68      DQin   7.92   9.5   .44
avgA   6.60   8.1   .65      ncs.   7.97   9.7   .42
trzu   6.67   8.0   .66      LEBR   8.08   9.7   .46
emin   6.74   8.3   .57      eWin   8.13  10.0   .50
Crow   6.95   8.4   .59      NuFi   8.27  10.0   .40
dtka   7.07   8.5   .59      23re   8.28  10.0   .45
At least momentarily, all are better than the dummy.

Re: 2023-24 team win projection contest

Posted: Sun Mar 17, 2024 2:11 pm
by Mike G
As the over-reaching teams refuse to lose, and the underachievers can't even beat one another, we've matched our mid January futility level, averaging 7.41 MAE

Code: Select all

.  avg err   rmse   r^2      .  avg err   rmse   r^2
medi   5.31   6.6   .76      vzro   7.47   9.0   .52
EExp   6.19   7.4   .70      EBPI   7.65   8.9   .57
ChKl   6.20   7.9   .64      KPel   8.10   9.8   .47
IanL   6.30   7.4   .68      DRKO   8.14   9.6   .49
vegas  6.36   7.6   .70      DQin   8.35  10.0   .41
avgA   6.94   8.4   .63      ncs.   8.36  10.0   .40
emin   7.02   8.5   .56      LEBR   8.42  10.0   .44
trzu   7.03   8.4   .64      eWin   8.53  10.3   .48
Crow   7.19   8.7   .58      NuFi   8.62  10.3   .38
dtka   7.40   8.8   .57      23re   8.69  10.4   .42
At this time last year, the avg MAE was 6.06; and Mediocre's current err would rank 4th.
Changes since the beginning of March, ranked least to worst:

Code: Select all

err+        3/01  3/17      err+        3/01  3/17
.22   emin   6.8   7.0      .44   medi   4.9   5.3
.29   eWin   8.2   8.5      .45   KPel   7.7   8.1
.30   Crow   6.9   7.2      .45   ncs.   7.9   8.4
.30   IanL   6.0   6.3      .45   ChKl   5.8   6.2
.32   NuFi   8.3   8.6      .47  vegas   5.9   6.4
.36   DQin   8.0   8.3      .47   EExp   5.7   6.2
.39   23re   8.3   8.7      .48   trzu   6.6   7.0
.39   vzro   7.1   7.5      .48   avgA   6.5   6.9
.40   LEBR   8.0   8.4      .50   DRKO   7.6   8.1
.44   dtka   7.0   7.4      .56   EBPI   7.1   7.7
Our worse half has worsened less than the better half; and the bottom-top diff is down almost a full point from early Feb.

Re: 2023-24 team win projection contest

Posted: Wed Mar 20, 2024 11:54 pm
by Mike G
On Dec. 8, the Nets had the 7th best record in the East, along with the 5th best point differential.
https://www.basketball-reference.com/fr ... &lg_id=NBA
They had a few players in early editions of the 'most improved' list. Since that time, their fortunes have changed significantly.

Dividing their 69 games played into thirds (of 23 games each), here are some team rates:

Code: Select all

W   L    ORtg    DRtg    TS%   TRb%   Stl%   Blk%     eFG%   TO%   ORb%  FT/fga
13 10   119.3   116.2   .583   52.0   6.3   10.4     +.026   +.9   +3.2   -.015
6  17   113.8   118.6   .550   49.1   7.7    9.9     -.035   -.7   +1.6   -.035
7  16   111.2   117.4   .556   47.4   7.6    8.0     -.031   -.5   -2.0   -.033
https://www.basketball-reference.com/te ... -advanced/
These last 4 columns -- the Four Factors -- are represented by team-opponent differences. They've cut down on turnovers!
Their early TotalReb% would be top 5 if they'd maintained it. Their latest 1/3 TRb% would be next to last.

ORtg-DRtg has gone from +3.1 to -4.8 to -6.2
Every one of their rotation players has generally declined. Top 8 minutes guys in the same 23-game periods:

Code: Select all

M Bridges    TS%   ORb%  DRb%   Ast%   Stl%  Blk%   TO%   Usg%   ORt  DRt   GmSc   Gm36
23   35.0   .599   3.2   14.1   17.2   1.3   1.3   11.2   26.3   119  117   17.3   17.8
23   35.1   .548   1.7   12.5   16.3   1.3   1.1   10.3   26.2   107  120   13.9   14.3
23   35.6   .527   2.5   11.0   16.5   1.5    .5    9.0   23.0   106  119   12.3   12.4

N Claxton    TS%   ORb%  DRb%   Ast%   Stl%  Blk%   TO%   Usg%   ORt  DRt   GmSc   Gm36
14   27.5   .653  12.0   22.8    7.2    .9   7.9    6.6   14.9   138  111   13.5   17.7
22   31.1   .629   9.8   30.7   10.6   1.4   5.9   13.3   16.2   122  111   14.0   16.2
23   29.1   .630   7.1   28.7   13.0    .8   5.8   11.3   15.9   123  113   12.3   15.2

D F-Smith    TS%   ORb%  DRb%   Ast%   Stl%  Blk%   TO%   Usg%   ORt  DRt   GmSc   Gm36
22   30.5   .631   6.8   11.4    5.8   1.0   2.0    9.5   13.6   123  120    9.2   10.9
20   27.0   .445   5.1   12.7    8.5   1.6   1.9    8.2   13.8    97  118    5.2    6.9
18   27.7   .561   6.2   12.3    9.0   1.9   1.8   11.5   13.3   113  115    7.3    9.5

Cam Thomas   TS%   ORb%  DRb%   Ast%   Stl%  Blk%   TO%   Usg%   ORt  DRt   GmSc   Gm36
14   31.2   .539    .7    9.9   12.0   1.0   1.6    7.3   31.8   109  119   14.1   16.3
23   27.4   .558    .9    9.3   14.5   1.0    .6    9.4   30.6   107  122   11.7   15.4
17   33.5   .546   3.0   11.4   18.9   1.6    .0    7.9   29.6   110  121   14.8   15.9

Cam Johnson  TS%   ORb%  DRb%   Ast%   Stl%  Blk%   TO%   Usg%   ORt  DRt   GmSc   Gm36
16   29.5   .586   6.3   13.7   13.6   1.5    .5    9.1   20.0   123  118   11.9   14.5
21   27.5   .574   2.4   14.1   13.8   1.1    .5    4.7   18.9   118  122    9.9   13.0
16   25.9   .602   2.7   14.1   12.5   2.2   1.5   11.2   21.1   113  116   10.0   13.9

De Smith     TS%   ORb%  DRb%   Ast%   Stl%  Blk%   TO%   Usg%   ORt  DRt   GmSc   Gm36
13   17.6   .505   5.1   12.4   27.8   2.4   1.3   12.2   18.4   113   113   6.3   12.9
18   21.1   .505   5.3   13.5   30.6   3.4   1.0   16.5   18.6   109   115   7.5   12.8
22   18.8   .489   5.4   11.9   25.5   3.2   1.0   15.4   16.9   104   115   5.6   10.7

L Walker     TS%   ORb%  DRb%   Ast%   Stl%  Blk%   TO%   Usg%   ORt  DRt   GmSc   Gm36
17   21.3   .633   1.0   12.0   12.7   1.4   1.6   11.3   26.7   115  118   10.1   17.1
11   14.1   .563   2.1    8.9   11.1   2.6   1.9    7.6   22.6   101  113    5.0   12.8
20   16.6   .498   1.0   13.5   10.7   1.7   1.2    7.6   24.8    95  118    4.7   10.2

D'R Sharpe   TS%   ORb%  DRb%   Ast%   Stl%  Blk%   TO%   Usg%   ORt  DRt   GmSc   Gm36
23   15.6   .581  16.3   28.6    8.4   1.2   4.3   16.4   18.9   116  110    6.6   15.2
14   16.7   .620  19.4   28.0   22.1   1.7   5.6   10.0   19.9   134  116   10.4   22.4
15   14.9   .627  16.8   26.9   18.6   1.3   2.9   21.4   19.7   116  112    6.6   15.9
OK, Day'Ron has picked it up; just missed a lot of games.
First 2 columns are Games and Min/G
Last column is GameScore*36/mpg

Averages for the whole troop:

Code: Select all

games   GmSc   Gm36
01-23   11.1   15.3
24-46    9.7   14.2
47-69    9.2   13.0

Re: 2023-24 team win projection contest

Posted: Sat Mar 23, 2024 3:51 pm
by Mike G
* - All but a few entries with worst errors of the season.

Code: Select all

.  avg err   rmse   r^2       .  avg err   rmse   r^2
medi   5.66   6.9   .75      vzro   7.84    9.3   .50
EExp   6.48   7.7   .69      EBPI   7.96    9.3   .55
IanL * 6.52   7.7   .67      KPel   8.49   10.1   .45
ChKl * 6.54   8.3   .62      DRKO   8.54    9.9   .47
vegas  6.65   7.9   .69      DQin   8.64   10.2   .40
emin * 7.31   8.9   .55      eWin   8.75   10.5   .48
avgA   7.33   8.8   .61      ncs.   8.79   10.4   .38
trzu   7.40   8.7   .63      NuFi * 8.80   10.6   .36
Crow   7.58   9.1   .56      LEBR   8.83   10.3   .42
dtka   7.80   9.2   .56      23re   9.06   10.7   .41
Mediocre is beating the [avg of] rest of us on roughly 2/3 of teams.
Beating us 15-2 by 2.0 or more; 9-0 by 3 or more.

Code: Select all

err  medi  avgA  dif        err  medi  avgA  dif
OKC   10   17    7.2        Phl    6    3    2.4
Hou    2    9    7.0        Ind   13   11    2.3
Por    1    8    6.6        NOP    7    5    1.9
Cha    8   13    5.1        Chi    3    2    1.7
SAS    5   10    4.9        Mia    4    2    1.7
Orl    8   12    4.8        Mil    3    1    1.4
Uta    0    4    3.9        Cle    3    1    1.2
Bos    5    9    3.6        Phx    1    0    0.8
Den    3    6    3.6        LAL    1    1    0.1
Atl    4    6    2.6                    
Tor   10   13    2.4                    
Was   12   14    2.3                    
Sac    2    4    2.2                    
Min    8   10    2.2                    
Det    9   11    2.1                    
LAC    8    9    1.2                    
Mem   16   17    0.8                    
GSW    2    2    0.6                    
Dal    6    7    0.3                    
NYK    2    2    0.1                    
Brk    8    8    0.1                    
Your beating may vary.

Re: 2023-24 team win projection contest

Posted: Tue Mar 26, 2024 11:14 am
by Mike G
There are 13 entries this year who were also in it last season. Year to year has a negative correlation.
Mean absolute errors compared to last year on this date:

Code: Select all

MAE:    2023    2024      from avg
dtka    4.94    7.86    -.98    -.04
vzro    5.45    7.86    -.46    -.04
KPel    5.51    8.53    -.41     .63
LEBR    5.63    8.96    -.28    1.06
DRKO    5.69    8.61    -.23     .72
Crow    5.75    7.64    -.16    -.26
ncs.    5.80    8.89    -.11     .99
emin    5.89    7.15    -.02    -.75
trzu    5.94    7.47     .02    -.43
EBPI    6.00    7.89     .08    -.01
vegas   6.69    6.63     .78   -1.26
EExp    6.77    6.39     .86   -1.51
NuFi    6.81    8.80     .89     .90

avg :   5.91    7.90    corr:  -0.39
The averages shown are for the 13 above and not the whole field. The correlation doesn't change, though.

UPDATE Apr.02
Yesterday, season high MAE were achieved by medi, ChKl, AvgA, trzu, dtka, vzro, KPel, DQin, eWin, and 23Re.
Today it's done by ncs; and again by KP, eW, 23.

Code: Select all

.  avg err   rmse   r^2       .  avg err   rmse   r^2
medi   5.73   7.1   .74      vzro   7.91    9.5   .49
IanL   6.29   7.6   .68      EBPI   7.94    9.3   .55
EExp   6.38   7.7   .69      DRKO   8.59   10.0   .46
ChKl   6.53   8.3   .62      KPel   8.62   10.3   .44
vegas  6.62   7.9   .69      NuFi   8.67   10.6   .37
emin   7.29   8.9   .55      DQin   8.80   10.4   .38
avgA   7.39   8.9   .60      eWin   8.87   10.7   .46
trzu   7.45   8.8   .62      ncs.   8.91   10.5   .38
Crow   7.63   9.3   .54      LEBR   8.94   10.4   .42
dtka   7.88   9.2   .55      23re   9.27   10.9   .39
There are distinct tiers now: the 5's (party of one), 6's (gang of 4), 7's and 8's.
Also, #1's lead has shrunk to smallest since Dec. 1

Re: 2023-24 team win projection contest

Posted: Sat Apr 06, 2024 1:46 pm
by Mike G
Errors have shrunk for all, a lot more for some than for others.
IanL, emin, and NuFi hit their worst marks some time ago; everyone else in the last couple weeks.

Code: Select all

APBR  av err  high  impr     other  av err  high  impr
medi   5.52   5.77   .24      EExp   6.14   6.51   .38
trzu   7.11   7.49   .38      IanL   6.18   6.88   .70
emin   7.20   7.42   .22      ChKl   6.36   6.58   .22
Crow   7.41   7.67   .26      vegas  6.40   6.72   .32
dtka   7.58   7.92   .34      EBPI   7.74   8.05   .31
vzro   7.82   7.94   .11      DRKO   8.30   8.61   .31
DQin   8.52   8.81   .29      KPel   8.38   8.62   .25
ncs.   8.57   8.91   .33      NuFi   8.39   9.03   .64
eWin _ 8.86 _ 8.95 _ .09 ____ LEBR _ 8.61 _ 8.96 _ .35
.avg   7.62   7.87   .25      .avg   7.39   7.77   .39
"We" are now averaging .23 worse than "them".

Re: 2023-24 team win projection contest

Posted: Tue Apr 09, 2024 12:20 am
by Crow
Congrats to medi.

Re: 2023-24 team win projection contest

Posted: Wed Apr 10, 2024 2:16 pm
by mediocre
Thanks! Hoping to avoid the sophomore slump next year

Re: 2023-24 team win projection contest

Posted: Mon Apr 15, 2024 2:43 pm
by Mike G
Final standings:

Code: Select all

.  avg err   rmse   r^2       .  avg err   rmse   r^2
medi   5.57   7.1   .73      vzro   7.87    9.4   .49
EExp   6.33   7.7   .69      EBPI   7.94    9.4   .53
IanL   6.33   7.5   .69      DRKO   8.41    9.9   .47
vegas  6.56   7.9   .69      NuFi   8.46   10.4   .38
ChKl   6.60   8.3   .61      KPel   8.59   10.3   .43
trzu   7.29   8.7   .62      ncs.   8.74   10.4   .39
avgA   7.39   8.9   .60      DQin   8.74   10.3   .38
emin   7.40   8.8   .56      LEBR   8.75   10.3   .43
Crow   7.73   9.3   .53      eWin   9.14   10.7   .44
dtka   7.80   9.2   .55      23re   9.20   10.9   .38
Season high MAE for ChKl, Crow, and eWins

Mediocre wins at exponent from .21 to 3.33

Code: Select all

e = .10           e = .33          e = 1.00 
EExp   2.72      medi   4.29      medi   5.57
IanL   2.88      EExp   4.72      EExp   6.33
Crow   3.36      IanL   4.91      IanL   6.33
ChKl   3.60      ChKl   4.99      vegas  6.56
medi   3.77      vegas  5.33      ChKl   6.60
                     
e = 3.33          e = 10.0         e = 33.3 
medi   8.60      IanL   11.9      IanL   14.5
IanL   8.62      EExp   12.3      medi   15.3
EExp   8.90      medi   12.4      EExp   15.4
vegas  9.21      emin   13.1      emin   15.4
ChKl  10.01      vegas  13.1      EBPI   16.1

Re: 2023-24 team win projection contest

Posted: Tue Apr 16, 2024 12:34 am
by Mike G
mediocre wrote: Wed Apr 10, 2024 2:16 pm Thanks! Hoping to avoid the sophomore slump next year
Congratulations;
and curses.
:)

Re: 2023-24 team win projection contest

Posted: Tue Apr 16, 2024 1:49 pm
by v-zero
Congrats mediocre, hit us all out of the park.

I went back to review my numbers. Very depressing reading.