2017-18 team win projection contest

[Crow]

I didn't specify time zone but 5pm was meant to include up to 5 pm Pacific time. 5pm further west might be after tip off and that might be a factor here.

My feeling at moment is to not be too rigid. It is a casual contest with round numbers and decimals. Congratulate who you want to when it is over. Anybody within .05 of the leader is an elite finisher.

[lanqiu]

Sorry, I overlooked the 5 pm deadline and did adjust Hayward's minutes down by 75% in last nights submission.

Here is my corrected projection using no Hayward injury adjustment.
GSW 67.0
HOU 56.8
SAS 56.1
CLE 55.3
OKC 51.4
MIN 50.6
TOR 48.5
WAS 47.7
LAC 47.0
DEN 45.9
MIA 45.5
MIL 44.9
BOS 49.0
UTA 43.3
CHA 42.6
POR 40.2
NOP 38.5
DET 38.2
MEM 37.0
PHI 36.8
DAL 33.9
ORL 33.1
IND 32.4
NYK 30.1
LAL 28.6
BKN 28.1
PHX 25.8
ATL 25.7
SAC 25.0
CHI 25.0

[Crow]

Thanks for this reply. That resolves it in my mind. This entry included in contest.

[knarsu3]

If possible, I wouldn't mind updating my projections with Kevin Pelton's new minutes distribution (which came in a few hours before game time so I wasn't able to do it yesterday) but I understand if it's too late.

[Crow]

Alright by me. I trust you will limit the change to that. If you have a chance to note some of the bigger minute changes, that would possibly be interesting.

[caliban]

Waiting on knarsu3 and which one that is the official projection of the alternatives

With one decimal

[Crow]

Thanks for setting this up and for future tracking (by you and possibly others).

[Mike G]

caliban, that is pretty cool. You're ranking by closest to average, the avg difference being the "error"?
I re-ran my numbers a couple of times since page 1 and before the season.

knarsu, please pick one of your projections for the contest.
. . . .

In this century, the number of seasons that teams have strayed by 3 or more wins from their Pythagorean norms.
Gainers on the left, losers on the right.

Code: Select all

     +3  -3           +3  -3
Chi   6   1      Hou   1   6
LAL   6   1      Min   1   6
Dal   5   1      SAS   2   7
Atl   6   2      Ind   1   4
Brk   7   4      Orl   1   4
Den   3   1      Det   2   5
Por   4   2      Mil   2   5
Cha   5   3      Sac   2   5
Cle   2   1      Bos   2   4
NYK   2   1      GSW   2   4
NOP   4   3      Phx   2   3
OKC   4   3      Mia   3   3
Phl   6   5      Tor   4   4
LAC   1   1      Uta   4   4
Mem   6   6      Was   4   4

The Nets have been notably off the mark in 11 of 18 seasons; the Clipps just twice.
Do small-market teams tend to get the shaft?

[eminence]

Uh-oh, always worrying to be furthest from the average, but I suppose someone has to be

Seems to be driven by the Utah/Toronto win totals, but I feel pretty good about those, so we'll see how it goes.

[Mike G]

If caliban gets my updated numbers in (or anyone else's) you may see I have edged you out for "most unique" in RMSE

Code: Select all

.      RMSE  avErr
.avg	0.00	0.00
RyRi	0.94	0.74
sbs.	1.04	0.84
lnqi	1.42	1.21
cali	1.56	1.33
ATCt	1.83	1.40
shad	2.12	1.63
sndi	2.34	1.73
Crow	2.38	2.08
538_	2.41	2.05
kmed	2.53	2.22
Josh	2.62	2.05
GK5.	2.67	2.12
gold	2.72	2.07
AJBk	2.79	2.07
ncs.	2.79	2.15
Nath	2.82	2.36
emin	3.46	2.83
Mike	3.46	2.72

The first 3 of us to enter have the 3 lowest projections for the Celtics. I'm also low on the Bulls.

Biggest squared departures from average:

Code: Select all

sq         tm  avg.  pro
61  AJBk  Atl  26.8   35
59  emin  Tor  48.4   56
58  Nath  LAC  47.6   40
50  emin  Uta  42.9   50
46  Mike  Hou  58.2   65

41  Mike  Dal  32.6   39
41  Mike  NOP  41.6   48
40  Josh  Sac  26.7   33
39  Mike  SAS  53.2   47
39  sndi  SAS  53.2   47

37  GK5   Phl  35.6   30
36  ncs   LAL  28.0   22
35  AJBk  LAC  47.6   54
33  Josh  Bos  47.7   42
33  ncs   Bos  47.7   42

33  gold  Sac  26.7   21
32  Josh  Mil  45.6   40
31  ncs   Cle  54.5   60
30  AJBk  OKC  51.1   46

[eminence]

Well, good to know I've got somebody out here on a limb with me at least

[caliban]

Edit1: Updated with Mike G's, shadows and knarsu3's edited #'s. If anyone spots an error in their line send me a PM. I won't be doing high frequent updates but I'll post one occasionally.

for fun, 85% avg, 15% netrtg from first round + BOS -3 & BRK -1

[knarsu3]

Sorry for the late reply but I'll go with the adjPM-PTPM as my official projections. It's the 2nd column in that table.

Would've liked to test out the SHACPM based ones but I think it's way too far off on DEN/DAL (who seem like they should be swapped, maybe there's a coding error that I didn't catch).

[Crow]

If there is a data entry error or a flip, I'd say ok to fix but do it (and state it for the tracking table) by Monday?

[eminence]

Whoops, appear to have entered on the wintotals site that the Kings will win 3 games this season, lol. Not sure if that can be adjusted or not.