2015-16 Team win projections

[Mike G]

Season win projections based on current wins + pythagorean W% over remaining games:

Code: Select all

tm    W        tm   W        tm   W
Uta   74      NYK   57      Por   31
GSW   73      Chi   54      Cha   26
Cle   70      Phx   53      Den   23
Tor   68      Dal   50      LAL   19
SAS   65      Sac   45      Ind   13
Okl   65      Atl   45      NOP   10
Min   64      Was   41      Mil    9
LAC   64      Bos   40      Brk    7
Det   61      Orl   32      Hou    5
Mia   58      Mem   32      Phl    3

These don't consider strength of opponent or remaining schedule; just point-diff in 2 to 4 games.

And our large-ish differences with these projections:

Code: Select all

rsm  15.8      km   16.3      Dan  16.6
Cal  16.0      KF   16.4      Crow 16.9
DF   16.0      AJ   16.5      yoop 16.9
DrP  16.2      bbs  16.5      tzu  17.0
itca 16.2      MG   16.5      BD   17.1
snd  16.3      taco 16.6      fpli 17.2
nr   16.3      DSM  16.6      EZ   18.7

Now we can all look forward to improving these errors.
The season is 3.7% completed.

[BasketDork]

Noticing Utah and Golden State's projections topping 72 wins (I think we're all well aware Utah will get nowhere close and Golden State seems to be a consensus pick to win 60+), It stirred up a simple, albeit interesting pondering of mine. Does anyone think ANY team this year has the capabilities, pieces and concrete history of avoiding the injury bug to make a run at that magic "72-10" mark set by Chicago, that turns 20 years old this season? I was an early teen back then, but I still recall how complete of a team they were, how well they dealt with distractions (ahhem..mainly The Worm's antics), all the breaks that seemingly went their way, and how laser-focused that group was night in and night out. Anyone see ANY team making a run at it this year or within the next 2-3 seasons? Just curious to hear some folk's responses..

[BasketDork]

Seeing some projections with Golden State ending up with 67+ W's for this year sparked my curiosity. They are a pretty complete team, but still have some holes.

[Mike G]

I shudder to think what a 50 point win over Mem last night does to the GSW projections.
Last year, they were just about injury free. In playoffs, none of their top guys seemed to be slowed much by any ailment. I don't know what should be expected -- maybe one starter unavailable, on avg?

The minutes I and some others used are pretty generous -- both Livingston and Barbosa getting more than last year -- and my estimate for their season came in right around last season; unsurprisingly, since they've hardly changed. But every team came in a few wins high, when you average them. So I have to knock them down a few. It seemed quite possible there is no improving for Curry, Thompson, or perhaps any of them.

The franchise champs in 1975, in fact, had a bunch of guys having their best seasons in '75 and '76, and then never much else of note.
This team has 5 guys expected to go 2000+ minutes (again). None of these is very replaceable from the bench.

[BasketDork]

I think Curry isn't even at his full potential yet, as evidenced by his ferocious start. I think D.Green's offense can still improve, although we may only see that on some nights or if one of their top guns get injured, his defense is great, if there's room for improvement, he's a lock for a couple DPOY's, between him and Kawhi Leonard. Klay Thompson, I don't know what you think, but I think he's peaked. I think he's still in his peak, but I think last year will probably be his best offensive season over the course of his career, even if only because Curry's usage rate can afford to go up a few ticks. Those extra possessions will have to come from somewhere, right? They have a pretty solid bench, too. I think the Clippers have a deeper bench this year though. I have them as a 60+ win team, hell, correct me if I'm wrong, we all do I believe? Thanks again for letting me participate in the win projections. I was kicking myself for not getting them up sooner after I missed last years altogether.

[Statman]

My very late entry - I got sidetracked with lineup optimization stuff & just didn't get around to just plain finishing the work. Obviously this ignores any games already played - I will be posting ALL projected player WARs (using Kevin's projected minutes) at my site later today, I'm just too tired to format anything right now.

I forgot to post the link to all the Player WAR projections (before start of season using Pelton's minute projections):

http://hoopsnerd.com/?p=885

Quick list of top:

Code: Select all

+------+-------------------+---------------+------------+-------+------+--------------------+------+-----+---------------------+-------+--------+
| WAR  |                   | Less Data=low |            |       |      | Pelton Projections |      |     | Projections 2015-16 |       |        |
+------+-------------------+---------------+------------+-------+------+--------------------+------+-----+---------------------+-------+--------+
| Rank | Player            | Confidence    | Bday       | Age   | Team | G                  | Min  | MPG | HN/48               | WAR   | WAR/48 |
| 1    | Kevin Durant      | 56.85%        | 9/29/1988  | 26.84 | okc  | 73                 | 2628 | 36  | 177                 | 19.33 | 28.95  |
| 2    | Anthony Davis     | 85.12%        | 3/11/1993  | 22.39 | nop  | 74                 | 2664 | 36  | 187                 | 18.57 | 27.44  |
| 3    | James Harden      | 99.45%        | 8/26/1989  | 25.93 | hou  | 78                 | 2808 | 36  | 164                 | 18.27 | 25.61  |
| 4    | Russell Westbrook | 69.85%        | 11/12/1988 | 26.72 | okc  | 74                 | 2516 | 34  | 175                 | 17.45 | 27.29  |
| 5    | Stephen Curry     | 92.89%        | 3/14/1988  | 27.38 | gsw  | 78                 | 2574 | 33  | 170                 | 17.30 | 26.45  |
| 6    | LeBron James      | 87.05%        | 12/30/1984 | 30.58 | cle  | 77                 | 2695 | 35  | 157                 | 16.45 | 24.03  |
| 7    | Chris Paul        | 90.60%        | 5/6/1985   | 30.24 | lac  | 76                 | 2660 | 35  | 157                 | 15.74 | 23.29  |
| 8    | DeMarcus Cousins  | 73.04%        | 8/13/1990  | 24.97 | sac  | 76                 | 2584 | 34  | 168                 | 14.55 | 22.16  |
| 9    | Paul George       | 38.08%        | 5/2/1990   | 25.25 | ind  | 68                 | 2448 | 36  | 146                 | 13.23 | 21.27  |
| 10   | Blake Griffin     | 88.60%        | 3/16/1989  | 26.38 | lac  | 77                 | 2695 | 35  | 140                 | 12.89 | 18.82  |
| 11   | Hassan Whiteside  | 29.59%        | 6/13/1989  | 26.13 | mia  | 78                 | 2184 | 28  | 169                 | 12.66 | 22.81  |
| 12   | Damian Lillard    | 100.00%       | 7/15/1990  | 25.05 | por  | 78                 | 2808 | 36  | 134                 | 11.91 | 16.69  |
| 13   | Kawhi Leonard     | 68.39%        | 6/29/1991  | 24.09 | sas  | 72                 | 2304 | 32  | 141                 | 11.05 | 18.88  |
| 14   | Kyle Lowry        | 88.23%        | 3/25/1986  | 29.35 | tor  | 77                 | 2618 | 34  | 133                 | 10.78 | 16.21  |
| 15   | Kevin Love        | 86.72%        | 9/7/1988   | 26.90 | cle  | 69                 | 2277 | 33  | 138                 | 10.64 | 18.40  |
| 16   | LaMarcus Aldridge | 85.78%        | 7/19/1985  | 30.03 | sas  | 76                 | 2432 | 32  | 134                 | 10.57 | 17.11  |
| 17   | DeAndre Jordan    | 98.38%        | 7/21/1988  | 27.03 | lac  | 78                 | 2730 | 35  | 125                 | 10.42 | 15.02  |
| 18   | Andre Drummond    | 91.51%        | 8/10/1993  | 21.97 | det  | 77                 | 2464 | 32  | 148                 | 10.41 | 16.63  |
| 19   | Nikola Vucevic    | 80.93%        | 10/24/1990 | 24.77 | orl  | 76                 | 2584 | 34  | 140                 | 10.20 | 15.54  |
| 20   | Rudy Gobert       | 59.68%        | 6/26/1992  | 23.10 | uta  | 78                 | 2496 | 32  | 138                 | 10.15 | 16.00  |
+------+-------------------+---------------+------------+-------+------+--------------------+------+-----+---------------------+-------+--------+

[Mike G]

Happy to report everyone is looking better than anyone was, last time.

Code: Select all

DrP   11.6      AJ    12.4      BD    13.0
bbs   12.0      km    12.4      yoop  13.1
DSM   12.0      tzu   12.4      fpli  13.1
Cal   12.0      Dan   12.6      nr    13.1
snd   12.1      rsm   12.7      taco  13.2
KF    12.2      Crow  12.8      itca  13.3
DF    12.2      MG    12.8      EZ    13.9

If last year's PythWins is an entry, its avg error is 13.9 -- relative to that, nobody is doing great.
Dr Pos in the lead at 5.6% of the season. He looks worst on Mia, NYK, Phl, and Orl (tie); best on Hou, Mil, Tor, Uta, and Por (t)

Again not accounting for SOS past or future, projected wins:

Code: Select all

West  W       East  W
GSW   75      Cle   65
Uta   60      Tor   65
SAS   58      Atl   51
LAC   56      Mia   50
Phx   54      Det   48
Min   53      Cha   45
Okl   52      Was   42
Por   45      NYK   41
Dal   40      Bos   38
Den   34      Chi   38
Sac   26      Orl   36
Mem   19      Ind   27
LAL   18      Mil   21
Hou   17      Phl   10
NOP   11      Brk    9

[Statman]

Happy to report everyone is looking better than anyone was, last time.

Code: Select all

DrP   11.6      AJ    12.4      BD    13.0
bbs   12.0      km    12.4      yoop  13.1
DSM   12.0      tzu   12.4      fpli  13.1
Cal   12.0      Dan   12.6      nr    13.1
snd   12.1      rsm   12.7      taco  13.2
KF    12.2      Crow  12.8      itca  13.3
DF    12.2      MG    12.8      EZ    13.9

Is that Dan me?

[Mike G]

Yep. Middly Diddly Dan.

[Dr Positivity]

APBR board looking pretty smart having Portland mean prediction at 35 Ws compared to 26.5-27 for Vegas

[Statman]

APBR board looking pretty smart having Portland mean prediction at 35 Ws compared to 26.5-27 for Vegas

Wow, I had 43 wins for them. That would have been one of my very first over bets if I was a gambling man.

Had to look at it - my projections like Davis', Plumlee's, & Leonard's per minute relative production quite a bit - and Pelton had them playing a combined 29% of total team minutes.

Of course, Lillard is a stud - projected 12th in the NBA in WAR before season start.

[bondom343]

I didn't do an actual prediction, but the lines are made for gambling, not actual predictions. I figured PDX would be an over just by Lillard alone, now looking at McCollum's play (thought he'd improve but he's really impressive so far) they look to be a mid 30s win team easily.

[Mike G]

Quick update:

Code: Select all

Dan  11.3      DrP   11.7      fpli  12.6
Cal  11.4      snd   11.8      nr    12.6
km   11.4      tzu   11.8      itca  12.6
KF   11.4      AJ    11.8      yoop  12.7
bbs  11.5      rsm   12.1      taco  13.0
DF   11.6      MG    12.3      BD    13.5
DSM  11.7      Crow  12.4      EZ    13.9

Statman Dan had the biggest variations from the mean, and several of these outliers are looking good. He has the best looking guesses on Brk, Dal, NOP, and Was -- all of which he was lowest on; Cha and Por, where he was highest.

Code: Select all

west  w    west  w     east  w    east  w
GSW  76    Dal  33     Cle  66    Orl  40
Uta  62    Sac  26     Tor  61    NYK  38
SAS  59    Den  25     Atl  54    Was  33
LAC  57    Hou  25     Cha  52    Ind  32
Por  54    LAL  25     Det  52    Mil  29
Phx  49    Mem  17     Mia  51    Phl  12
Okl  48    NOP  14     Bos  49    Brk  11
Min  42                Chi  42      

If and when b-r.com gets its Playoff Probabilities Forecast running, I'd be inclined to use those. Above, it's just MOV regardless of SOS.

[kmedved]

Using Justin Kubatko's method, and Neil's dummy games value from the comments here, http://statitudes.com/blog/2013/11/12/h ... using-srs/, using SRS to project wins for the rest of the schedule, generates this:



These results are obviously very "flat" right now, due to the adding of 7 dummy games to regress the current SRS to the mean.

[Crow]

I should have left Vegas out of my blend. Low weight; but a mistake, it appears. Almost pulled it. Blended fewer components than in past but left Vegas in. Maybe should have added others. Just did a real quick, small blend and tweak. Maybe too much on the small spur of the moment tweaks.


But it is very early.