2012 predictions

[EvanZ]

Nets waived Ime Udoka and traded for Mehmet Okur, who will presumably get some of Lopez' minutes. Chuck Hayes playing re-signing his contract with the Kings added about 3 wins to their total in my predictions.

[wilq]

Predictions made or at least tracked here: [...]

Thank you for those links.

Few were in these lists every year but John Hollinger, Vegas and I (using screenname Mountain then Crow) were in the top 3 here twice in 2 tries.

That's why I've tried to widen the picture

[Mike G]

I missed last year, in the middle of relocating, and didn't feel right jumping in late. In fact, after only a game or 2 we can often see surprising things about a team. A nagging injury is clearly major. Some awesome chemistry has been created. ..

So I encourage us to post those predictions! Perhaps no one is really going to predict based purely on a particular system -- minutes are guessed at, averaging and projections are varied, etc. -- but that's OK, or perhaps it's inevitable.

Of course, late is still better than never. And it's Christmas, after all.
Throw your hat in, or be tied for last. You know who you are.

[dudeimstoked]

my first attempt at nba season projections (before the season starts):

Wins Losses
ATL 34.5 31.5
BOS 40 26
CHA 22 44
CHI 48.5 17.5
CLE 19.5 46.5
DAL 41 25
DEN 34 32
DET 23 43
GSW 32.5 33.5
HOU 30.5 35.5
IND 35 31
LAC 42 24
LAL 36 30
MEM 37 29
MIA 53 13
MIL 31 35
MIN 27.5 38.5
NJN 22 44
NOH 24 42
NYK 38 28
OKC 44 22
ORL 41 23
PHI 34 32
PHO 29 37
POR 37 29
SAC 25 41
SAS 37.5 28.5
TOR 21.5 44.5
UTA 28.5 37.5
WAS 20.5 45.5

[Crow]

It appears that the revised projections of Mike G, Evan Z and the new entry of dudeimstoked all have just one team with less than 20 wins. That might be the right strategy (i.e. regression to the mean) for average error reducing but given that 6 teams last season won less than .300, there might be competitive advantage to trying to identify which teams will slip below.300 and below 20 wins. I expect it will be more than one.

Last season 6 teams won over .680, though only 2 were much higher. In the predictions so far only 1-2 are predicted to break that mark. I think that might be closer to being true than at the lower end of things and may be error reducing but may also prove to be a little light compared to the actuals at the end of the season.

[J.E.]

Here goes nothing. These were not done with RAPM, it's just me guessing

West:

SAS 43.5
OKC 43.5
DAL 42.5
DEN 41.5
LAL 41.5
LAC 40
POR 38
MEM 36.5
HOU 34
GSW 28.5
PHO 27
NOH 24.5
UTA 24.5
SAC 20

EAST:

CHI 48.5
MIA 48.5
BOS 42.5
ORL 40
PHI 39
NYK 38
ATL 33
MIL 32
IND 30.5
DET 27
TOR 23.5
NJN 22.5
WAS 21
CLE 18.5
CHA 17.5

[J.E.]

Here's Vegas

Portland 37.5
Atlanta 34.5
Boston 39.5
Charlotte 16.5
Chicago 47.5
Cleveland 16.5
Dallas 43.5
Denver 37.5
Detroit 22.5
GSW 26.5
Houston 31.5
Indiana 36.5
LAC 42.5
Memphis 38.5
Miami 50.5
Milwaukee 31.5
Minnesota 22.5
NOH 24.5
NYK 41.5
OKC 47.5
Orlando 38.5
Philly 35.5
PHX 26.5
Sacramento 20.5
SAS 41.5
Toronto 16.5
Utah 25.5
Washington 20.5

Not sure why there's no New Jersey or Lakers, I'll try to find those tomorrow. Vegas seems to not like Toronto

[Mike G]

It appears that the revised projections of Mike G, Evan Z and the new entry of dudeimstoked all have just one team with less than 20 wins. That might be the right strategy (i.e. regression to the mean) for average error reducing but given that 6 teams last season won less than .300, there might be competitive advantage to trying to identify which teams will slip below.300 and below 20 wins. I expect it will be more than one.
.

No strategy need be invoked in order to predict all teams to have 20+ wins. You could probably retrodict the Cavs of last year to win 20, if everyone is 90-95% healthy.

Every year some teams do much better than expected and some do much worse. Those doing much worse almost always have injury issues that are unpredictable. Those doing much better almost always have unusually good health, chemistry, player improvement, strong rookies -- or several of those.

Even if you get lucky to guess a sharp departure from the norm, it's probably not going to be for the reasons you thought.

[DSMok1]

Here are my predictions (ASPM Projections and newest regression used)

Using ESPN's depth charts and assuming a standard distribution of minutes based on depth at position (I know, that's a pretty rough way of doing minutes)

Code: Select all

Team	Efficiency	Win%	66 Games
MIA	6.39	70%	46.0
LAC	4.81	65%	43.0
SAS	4.33	64%	42.1
CHI	3.83	62%	41.1
OKC	3.81	62%	41.0
ORL	3.50	61%	40.4
LAL	3.25	60%	39.9
DAL	2.24	57%	37.8
MEM	1.85	56%	37.0
BOS	1.77	56%	36.8
POR	1.63	55%	36.5
PHI	0.86	53%	34.9
NYK	0.74	52%	34.6
DEN	0.55	52%	34.2
IND	0.25	51%	33.5
MIL	-0.06	50%	32.9
ATL	-0.22	49%	32.5
GSW	-0.48	48%	32.0
HOU	-0.81	47%	31.3
PHO	-1.97	44%	28.8
SAC	-2.05	43%	28.6
UTA	-2.30	43%	28.1
DET	-2.49	42%	27.7
NJN	-2.61	42%	27.4
TOR	-3.62	38%	25.3
MIN	-3.84	38%	24.9
NOH	-3.86	38%	24.9
WAS	-4.07	37%	24.4
CLE	-5.71	32%	21.3
CHA	-5.75	32%	21.2

[Mike G]

Wins and Losses are rounded to integers, W% is from calculation by eWins.

Code: Select all

West    W%    W   L       East    W%    W   L
SAS   .652   43   23      Chi   .759   50   16
LAC   .638   42   24      Mia   .701   46   20
Dal   .628   41   25      Bos   .634   42   24
LAL   .587   39   27      Orl   .608   40   26
Okl   .587   39   27      Ind   .576   38   28

Por   .578   38   28      NYK   .568   37   29
Mem   .551   36   30      Phi   .504   33   33
Hou   .498   33   33      Atl   .474   31   35
Min   .493   33   33      Mil   .456   30   36
Den   .476   31   35      NJN   .365   24   42

Uta   .456   30   36      Det   .358   24   42
Sac   .445   29   37      Was   .331   22   44
GSW   .441   29   37      Cha   .329   22   44
Phx   .350   23   43      Cle   .323   21   45
NOH   .315   21   45      Tor   .321   21   45

West teams avg 33.9 wins, East 32.1

[Mike G]

Here's what we have so far. I found Kevin Pelton's at http://basketballprospectus.com/article ... cleid=1974
dudeimstoked (dis), your wins seem to total 989, which is 1 short.
J.E., you were 22.5 short of a full load, and also you missed Min, so I gave them 22.5 W on your behalf.

Code: Select all

avg    East   DSM    MG   Vegas    JE     dis    KP
48.7   Mia   46.0   46.3   50.5   48.5   53.0   48.0
47.0   Chi   41.1   50.1   47.5   48.5   48.5   46.0
40.1   Orl   40.4   40.1   38.5   40.0   41.0   40.3
39.4   Bos   36.8   41.8   39.5   42.5   40.0   35.8
37.6   NYK   34.6   37.5   41.5   38.0   38.0   36.0

35.4   Phi   34.9   33.2   35.5   39.0   34.0   35.6
34.3   Ind   33.5   38.0   36.5   30.5   35.0   32.5
32.8   Atl   32.5   31.3   34.5   33.0   34.5   31.1
31.6   Mil   32.9   30.1   31.5   32.0   31.0   31.9
25.0   Det   27.7   23.6   22.5   27.0   23.0   26.1

25.0   NJN   27.4   24.1    na    22.5   22.0   28.8
22.7   Tor   25.3   21.2   16.5   23.5   21.5   28.3
21.3   Was   24.4   21.9   20.5   21.0   20.5   19.5
21.2   Cha   21.2   21.7   16.5   17.5   22.0   28.4
19.9   Cle   21.3   21.3   16.5   18.5   19.5   22.3

avg    West   DSM    MG   Vegas    JE     dis    KP
42.5   Okl   41.0   38.7   47.5   43.5   44.0   40.4
41.6   LAC   43.0   42.1   42.5   40.0   42.0   39.7
41.0   SAS   42.1   43.0   41.5   43.5   37.5   38.3
40.5   Dal   37.8   41.5   43.5   42.5   41.0   36.8
37.9   LAL   39.9   38.7    na    41.5   36.0   33.5

37.4   Por   36.5   38.1   37.5   38.0   37.0   37.5
36.3   Mem   37.0   36.4   38.5   36.5   37.0   32.3
35.6   Den   34.2   31.4   37.5   41.5   34.0   34.7
32.1   Hou   31.3   32.8   31.5   34.0   30.5   32.6
29.0   GSW   32.0   29.1   26.5   28.5   32.5   25.4

27.8   Uta   28.1   30.1   25.5   24.5   28.5   30.1
27.2   Min   24.9   32.6   22.5   22.5   27.5   33.4
27.0   Phx   28.8   23.1   26.5   27.0   29.0   27.8
25.0   NOH   24.9   20.8   24.5   24.5   24.0   31.1
24.9   Sac   28.6   29.4   20.5   20.0   25.0   25.9

Predictions average variance from the mean is 1.89 wins.
Less than 1.00 variance: Por 0.46 (36.5 to 38.1), Orl 0.53 (38.5 to 41.0), Mil 0.70 (30.1 to 32.9)
Greatest variance: Min 3.93 (22.5 to 33.4), Sac 3.10 (20.0 to 29.4), Tor 2.98 (16.5 to 28.3), Cha 2.82 (17.5 to 28.4)

The prognosticator who is closest to the averages is dis, just 1.33 avg per team.
Wildest (most independent) guesser is Kevin P, at +/- 2.44
Everyone else averaging 1.78 to 1.96 from avg.

Vegas is highest on Dal, Mem, NY, Okl; hardest on Cha, Cle, Min, Orl, Tor, Was.

[Crow]

Hollinger

West

1. OKC 45-21
2. Denver 43-23
3. Spurs 42-24
4. Clippers 41-25
5. Lakers 40-26
6. Mavs 39-27
7. Blazers 38-28
8. Grizzlies 37-29
9. Rockets 32-34
10. Wolves 29-37
11. Utah 27-39
12. Golden State 26-40
13. Sacramento 25-41
14. Phoenix 24-42
15.New Orleans 22-44

EAST -Wins
MIA 52
CHI 48
BOS 43
ORL 40
PHI 37
IND 36
NYK 35
MIL 34
ATL 33
NJN 28
DET 22
WAS 22
TOR 19
CLE 18
CHA 13

[Mike G]

Thanks, Crow.
John H is highest on Bos, Den, Mil; lowest on Cha, Det.
The biggest disparity in predictions seems to be JH with 13 wins for the Bobcats vs KP with 28.4

Still thinking EvanZ is going to have an update (or 2).
Someone could snap up the average of these and call it theirs.

[Crow]

I drifted into it again after all. A quick take.

East
Mia 50
Chi 46
Orl 36
Bos 41
NYK 36

Phi 37
Ind 37
Atl 36
Mil 33
Det 23

NJN 24
Tor 21
Was 22
Cha 18
Cle 18

West
Okl 45
LAC 41
SAS 40
Dal 40
LAL 41

Por 39
Mem 40
Den 37
Hou 34
GSW 27

Uta 26
Min 24
Phx 28
NOH 27
Sac 23

[Crow]

I think Evan's projections on page 4 of this thread are the same as what he published here http://www.goldenstateofmind.com/2011/1 ... #storyjump

I checked some but not all and didn't notice any changes.