2023-24 team win projection contest

[Crow]

For those inclined.

[Mike G]

Appealing to anyone who knows of anyone who has spent the time conjecturing minutes allocation to players, listed by team; and who can direct to or actually post these numbers.

[DarkStar48]

Appealing to anyone who knows of anyone who has spent the time conjecturing minutes allocation to players, listed by team; and who can direct to or actually post these numbers.

Of the publicly available minutes projections, I have often gotten the best results with ones from Josh Lloyd (@redrock_bball) and BasketballMonster (paywall: basketballmonster.com).

[Crow]

Minute estimates matter... but how much? If systems matter and synergies matter, one isn't going to get to the exact final answer with average individual impact estimates x minute estimates.

Unless one has a way of estimating system impacts (likely somewhat revised and unstated or under-surmised) and synergies (RAPM pushed hard may help some but not perfectly), the result could easily be 10-20% plus or minus the simple product states above. Even best guesses on these are not that reliable. Guesses on fine minute assignments especially outside the top 4-6-8 are just one of at least 5 kinds of guesses.

[DarkStar48]

Kevin Pelton released his win projections for 2023-24.

Link: https://www.espn.com/nba/insider/inside ... l-30-teams

[DarkStar48]

Andy Bailey released Bleacher Report’s 2023-24 win predictions.

Link: https://bleacherreport.com/articles/100 ... le-release

[DarkStar48]

ESPN Expert Forecast

Link (East Conf.): https://www.espn.com/nba/insider/inside ... rence-team

Link (West Conf.): https://www.espn.com/nba/insider/inside ... rence-team

[Mike G]

A lot of smart people seem very high on the Bucks this year, but I don't know what to think. Everyone on the team will have to play more minutes than they did last season.
Here's my current layout of minutes:

Code: Select all

Bucks'24   tm23  Min23   age  Min?   e484    PER   WS/48   BPM
Giannis     Mil   2024   28   2400   3.02   29.0   .204    8.5
Lillard     Por   2107   32   2300   2.17   26.7   .205    7.1
Beasley     LAL   2093   26   2100    .69   10.9   .032   -2.0
B Lopez     Mil   2373   34   2100   1.37   18.4   .161    2.1
Middleton   Mil    801   31   2000   1.41   17.4   .117    0.8

Portis      Mil   1818   27   1900   1.61   17.8   .139    0.7
Connaughton Mil   1443   30   1500    .52   10.0   .085   -1.0
Payne       Phx    968   28   1200    .93   13.2   .053   -1.2
Crowder     Mil    340   32   1100    .92   14.6   .165    1.2
Beauchamp   Mil    701   22   1000    .30    8.2   .027   -4.8

Bucks'24   tm23  Min23   age  Min?   e484    PER   WS/48   BPM
Washington  Hou    433   21    700    .18    7.5  -.016   -6.3
AJ Green    Mil    345   23    700    .67   11.5   .111   -0.9
R Lopez     Cle    299   34    400    .46   11.6   .094   -3.0
A Jackson   #36      0    R    200   (.22   11.1   .022   -4.)
Thanasis    Mil    206   30    200    .02    6.7   .022   -7.6

totals           15951       19800   1.34   17.1   .120    1.3
Livingston  #58                     
M Bolden     z                     
J Gortman                        

Their #2, 4, 6, 8, and 9 in minutes last year are gone. Only 3 of the new guys ran even 500 last year.
Assigning no minutes to those bottom 3, as historically they are in a stratum that plays little.
The aggregate needs to pick up almost another 4000 minutes to reach 48+ per game.

Giannis hasn't hit 2400 min. in 5 years. Dame avg's just over 2000 in last 4 -- and he's 32.
Lopez in perfect health last year managed just 30 mpg. Portis has peaked at 26 mpg.

EDIT: Sixers seem similarly stretched, and with the Harden mess on top of it.

[DQuinn1575]

so you have the 10 at 17,600 minutes. I looked at last year, and had only these teams above that amount for their top 10 guys
ATL
BOS
CHI
CLE
MIA
NYK
SAC

with the average being 16,671 - now I didnt factor in traded guys (like Bridges), but still I think with the age/injury concerns, you probably want to drop the minutes for MIL for their top 10 guys.

Looking at this, CHI & SAC were 1-2 with 18,451 & 18,361, with BOS & CLE were 18.042 and 18.083 and again the average under 17k. Most would say that CHI & SAC were lucky and have a few more missed minutes, maybe making 18k the top, but the average projection should maybe be 17k ?

[DarkStar48]

John Hollinger’s 2023-24 win projections

Link (East Conf.): https://theathletic.com/4974382/2023/10 ... onference/

Link (West Conf.): https://theathletic.com/4965360/2023/10 ... onference/

[v-zero]

I need to sense check these so they may be altered before opening tip, but I'll get the ball rolling.

Edit: These have been edited after I spotted a few issues during the sanity check, most teams have changed by about half a win but a few notable bigger changes are:

Knicks get another 2.2 wins, Milwaukee lose 1.4, Phoenix gain 1.4, Golden State gain 6.2 (I had made a pretty huge error with Golden State), Dallas gain 3, Portland lose 2, San Antonio lose 1.7, Toronto lose 2.2, Boston lose 1.2, Chicago lose 1.2.

Edit2: The loss of Steven Adams for the season drops Memphis just over 3 wins and pushes everybody else in the league up by about 0.1 wins.

Code: Select all

BOS	58.5
NYK	55.7
MIL	52.2
PHI	51.7
CLE	49.3
DEN	49.2
PHO	47.4
GSW	47.0
MIN	46.0
NOP	44.9
SAC	43.4
ATL	43.1
LAC	42.9
MEM	42.8
MIA	42.5
BRK	41.0
TOR	40.8
DAL	40.0
LAL	39.3
CHI	39.0
UTA	39.0
OKC	37.3
IND	34.3
CHO	33.4
WAS	32.8
ORL	32.7
HOU	30.5
POR	28.1
SAS	23.6
DET	21.6

[Crow]

Thunder, 21st in immediately above projection list.

[v-zero]

Thunder, 21st in immediately above projection list.

22nd now after my sanity check. Honestly I personally expect them to be better than that, but this is a purely numerical forecast.

The one I am most bemused by is the Knicks, who actually gained in the sanity check. I can't really imagine them getting to 56 wins. The explanation is simple enough: Thibodeau loves to ride his best players and recently the Knicks squad has been very healthy and developing well. My model suggests they have nine players at or above league average who will make up almost all of their minutes this season, so their amount of court time for players around replacement level is far lower than any other team in the league. Realistically there's a good chance they will have to play more replacement level minutes than the model thinks.

[Crow]

My eyes did not see the Knicks projection first time through. I should look into that one.


Probably ends up with ten at 10+ min. used per game for season with rest, injuries and players having weak stretches. 11 last season. But, yeah; big fall off after that.

Not much roster change in off-season. I'd be looking for a consolidation trade or some sort of upgrade by trade. Probably should have been already imo; but there is time and they did make an in-season move last spring.

[DQuinn1575]

First time ever trying this, and still adjusting my formulas, but figured I'd better post something if I'm going to do it. My aging formulas don't like GSW, and I didn't give anybody healthy going in less than 65 games. If I get to it I will tweak these, but if not here goes:

BOS 56.8
PHI 54.3
LAL 52.2
CLE 51.6
MIL 50.9
DEN 48.3
NOP 48.1
PHO 47.5
ATL 46.1
LAC 45.9
MEM 45.3
MIA 44.3
NYK 44.1
BRK 42.5
CHI 42.2
TOR 42.1
UTA 41.7
MIN 41.4
ORL 39.5
SAC 38.7
GSW 37.0
DAL 35.8
OKC 34.2
WAS 33.2
CHO 32.0
POR 32.0
IND 31.9
SAS 27.4
DET 23.3
HOU 21.1