Predictions 2013-14

[Mike G]

As of yesterday morning, our predictions were looking still better.

Code: Select all

Bobb   8.24      Yoop   9.04
jBro   8.30      deep   9.08
416x   8.84      jank   9.47
hDon   8.925     ATC    9.59
eW     8.929     fpli   9.74
ncs    8.99      13Py  10.35

Avg error 9.01, at .241 of season.
The espn team is off by 8.60, so 'they' would rank 3rd here.

[Mike G]

Another team that surprises everyone is Portland. Yoop and I picked them to win 34; they're headed for 59, and none of us picked them for more than 46 (416x).

Code: Select all

Blazers   2012-13 / predicted     projected from 21 G
.         mpg  e484  mpg   eW       e484  mpg   e82
Aldridge  34   1.58   32   8.6      1.88   37  11.7   
Lillard   39   1.08   32   5.8      1.57   36   9.6   
Batum     34    .99   32   5.4      1.31   35   7.9   
Matthews  29    .73   26   3.2      1.13   34   6.5   
Lopez     26   1.11   24   4.5       .90   30   4.6 
  
Wright    22   1.02   20   3.5       .65   14   1.6   
Williams  17    .97   20   3.3       .67   24   2.7   
Leonard   15    .61   16   1.7      1.38    2    .4   
Robinson  10    .55   12   1.1      1.08   12   2.1 
  
Barton    11    .31    8    .4      1.03    1    .2
Watson    10    .28    8    .4       .00    2    .0   
Freeland   6    .43    6    .4       .51   14   1.2   
Claver    10    .10    6    .1      1.34    1    .2  (Crabbe,Allen)

.totals  262         242  38.4            242  48.6

The top 4 are all better AND playing more minutes than expected. Mo Williams is down but a lot better than last year's backup guards.

[Mike G]

Still groping around for underlying causes of east-west imbalance. Here are teams' average age (minutes weighted, via b-r.com)

Code: Select all

West  age       East  age 
Dal   30.6      Mia   30.9
SAS   29.2      Brk   30.0
LAL   28.4      Chi   29.0
Mem   28.4      NYK   29.0
LAC   27.7      Atl   27.3

Den   26.8      Ind   26.9
Min   26.4      Was   26.1
GSW   26.2      Bos   25.7
Por   26.1      Mil   25.6
Phx   26.0      Orl   25.3

Hou   25.9      Tor   25.3
Okl   25.9      Cle   24.8
Sac   25.9      Det   24.7
Uta   25.1      Cha   24.5
NOP   24.6      Phl   23.5

The East has more of the oldest teams and more of the youngest teams.
The West, then, would seem to have more players of prime age. Like twice as many teams aged 26-27.5, or 25.5-28

[AcrossTheCourt]

Older teams are often very good teams because contenders pick up old veterans on cheaper contracts looking for a title. Miami's very old but obviously very good.

As promised, here's my look at Brooklyn:
http://ascreamingcomesacrossthecourt.bl ... oklyn.html

Key stat: allowing 44.9% from above the break three-pointers. League average is near 35% and the next closest team is at 38%. I also don't expect Deron/Pierce/Garnett to continue shooting so poorly, and Kirilenko will help them a lot.

[Crow]

Western teams have 9 of the 10 fastest paces, east 9 of the bottom 10.

Western teams have 10 of top 11 big minute PGs on winshares per 48 minutes.

[Crow]

Eastern teams average spending a few million more than the west. They have 7 of the top 10 payrolls. So overall spending is not the issue holding them back; but GM, coaching and PG leadership all heavily favor the west and shows up in the form of the overall offensive edge and overall wins. Eastern teams have won 37% of the close games they have played- 3pts or less; while the west averages 56%. The west has won about 230% as many games against .500+ win pct teams.



Since no one inside ever talks substantively in public about their use of RAPM (or similar), I cant say that either conference has an edge there or that it has had a net pt differential / win impact for one conference over the other. Among well-known producers of APM the conferences are fairly balanced.

[Mike G]

Western teams have 10 of top 11 big minute PGs on winshares per 48 minutes.

Wow. That's quite amazing. I've got John Wall as the best PG in the East, heading for 10.9 eWins.
Next is Jeff Teague at 7.7

Out west, it's Paul at 14.0, Curry 11.9, Lillard 9.6, Lawson 9.3, Parker 8.8, Conley 8.4, Ellis 8.4, Westbrook 8.3, Thomas 8.1, Holiday 7.8
So yeah, that too is 10 of the top 11 PG.
Irving was supposed to be elite, but this year he's not, yet.

From the current Most Improved list at PG:

Code: Select all

eW+    West Conf.              eW+    East Conf.
.82   Lillard,Damian Por      .50   Wroten,Tony    Phl
.75   Curry,Stephen  GSW      .45   Cole,Norris    Mia
.72   Thomas,Isaiah  Sac      .41   Bradley,Avery  Bos
.60   Lawson,Ty      Den      .39   Stuckey,Rodney Det
.43   Paul,Chris     LAC      .30   Teague,Jeff    Atl
.40   Bledsoe,Eric   Phx            
.39   Blake,Steve    LAL            
.35   Dragic,Goran   Phx            
.32   Brooks,Aaron   Hou            
.31   Lin,Jeremy     Hou            
.30   Jackson,Reggie Okl            
.29   Mills,Patty    SAS            
.27   Ellis,Monta    Dal            
.26   Holiday,Jrue   NOP            

[Crow]

I wonder what management of eastern teams are thinking about their conferences weak performance and in most cases their own middling to weak performance. I wonder how many will seriously consider changing GMs, coaches and/ or PGs. Most will probably ride it out for the season or multiple seasons without making changes in these areas. Many may not even really seriously consider making any of these changes. Many will probably continue to have middling to weak results. In some measure because of these leadership issues.

How much should insiders listen to and consider criticism from the outside? How much do they allow or encourage critique from the inside? Surely it varies and I don't know but I wonder it is generally enough. I wonder how much teams bring in outsiders to offer frank independent assessments. I imagine it probably happens quietly with some past GMs and coaches, maybe with past ties to the organization or not. I'd think this should be a common practice. I wonder if any or many teams bring in outside stat analysts for brief assessment and alternative opinions on key issues. I would guess that a good number of teams fall into having one rigid system of assessment, one ultimate decider, one world view until it get to the breaking point of final unacceptability. But it is just a guess, based on what I see from outside in the stats and on the court.

Leadership turnover may be the only way out of bad results but as suggested previously it can also be the cause of bad results. Got to decide what to stick with and be right.

[Mike G]

We've improved to an avg error of 8.69, with the season .294 finished.

Code: Select all

Bobb    7.66        hDon    8.75
jBro    7.92        Yoop    8.78
416x    8.55         eW     8.83
deep    8.65        ATC     9.09
jank    8.680       13Py    9.73
ncs     8.681       fpli   10.01

We have a clear leader, a clear top 2, a clear trailer, and six of us bunched tightly.
At 8.83, espn is just hanging in there.

[nbacouchside]

We've improved to an avg error of 8.69, with the season .294 finished.

Code: Select all

Bobb    7.66        hDon    8.75
jBro    7.92        Yoop    8.78
416x    8.55         eW     8.83
deep    8.65        ATC     9.09
jank    8.680       13Py    9.73
ncs     8.681       fpli   10.01

We have a clear leader, a clear top 2, a clear trailer, and six of us bunched tightly.
At 8.83, espn is just hanging in there.

Hooray for being middle of the pack!

[Mike G]

While a couple of us have fallen sharply in the last week, our average has improved again, to 8.53

Code: Select all

Bobb   7.35      416x   8.79
jBro   7.60      ATC    8.96
hDon   8.14       eW    9.05
jank   8.25      deep   9.11
ncs    8.40      fpli   9.55
Yoop   8.67      13Py   9.70

Using Pythagorean only (disregarding current W-L), avg error is 8.37

Code: Select all

Bobb	7.21		ATC	8.51
jBro	7.47		deep	8.51
hDon	8.21		Yoop	8.55
jank	8.30		eW	8.92
416x	8.37		fpli	9.63
ncs	8.43		13Py	10.02

[Bobbofitos]

While a couple of us have fallen sharply in the last week, our average has improved again, to 8.53

Code: Select all

Bobb   7.35      416x   8.79
jBro   7.60      ATC    8.96
hDon   8.14       eW    9.05
jank   8.25      deep   9.11
ncs    8.40      fpli   9.55
Yoop   8.67      13Py   9.70

Using Pythagorean only (disregarding current W-L), avg error is 8.37

Code: Select all

Bobb	7.21		ATC	8.51
jBro	7.47		deep	8.51
hDon	8.21		Yoop	8.55
jank	8.30		eW	8.92
416x	8.37		fpli	9.63
ncs	8.43		13Py	10.02

:thumb:

[nbacouchside]

While a couple of us have fallen sharply in the last week, our average has improved again, to 8.53

Code: Select all

Bobb   7.35      416x   8.79
jBro   7.60      ATC    8.96
hDon   8.14       eW    9.05
jank   8.25      deep   9.11
ncs    8.40      fpli   9.55
Yoop   8.67      13Py   9.70

Using Pythagorean only (disregarding current W-L), avg error is 8.37

Code: Select all

Bobb	7.21		ATC	8.51
jBro	7.47		deep	8.51
hDon	8.21		Yoop	8.55
jank	8.30		eW	8.92
416x	8.37		fpli	9.63
ncs	8.43		13Py	10.02

:thumb:

Bob, did you make your predictions based on any sort of metric / system or was it intuition and feel?

[Bobbofitos]

Bob, did you make your predictions based on any sort of metric / system or was it intuition and feel?

Used a metric to determine minute productivity, used basketballmonster's minute projections, and injected an aging curve. Added or subtracted 1 game from each team based on my own disagreements or agreements.

[Mike G]

We've taken a collective turn for the worse, with avg error now at 9.04

Code: Select all

Bobb   7.90      deep   9.11
jBro   8.09      ATC    9.24
jank   8.73      Yoop   9.51
hDon   8.81       eW    9.79
416x   8.96      fpli  10.27
ncs    9.02      13Py  10.34

The ESPN braintrust is also off by 9.04 at this time.