Page 9 of 13
Re: 2022-23 team win projection contest
Posted: Thu Feb 02, 2023 2:24 am
by Crow
0.15 from 5th.
5th to 11th are quite close.
Re: 2022-23 team win projection contest
Posted: Fri Feb 03, 2023 10:51 am
by Mike G
Code: Select all
. avg err rmse r^2 avg err rmse r^2
KPel 4.72 6.098 .52 22Re 5.91 7.24 .32
dtka 5.03 6.095 .61 538R 5.92 7.42 .50
avgA 5.15 6.27 .58 vegas 6.06 7.83 .48
vzro 5.31 6.66 .50 4141 6.09 8.75
LEBR 5.39 6.42 .55 TmRk 6.15 7.76 .47
EBPI 5.42 6.89 .50 BIPM 6.21 7.61 .64
ncs. 5.43 6.74 .50 MPra 6.24 7.77 .56
DRKO 5.52 6.53 .52 AnBa 6.32 8.06 .50
emin 5.52 7.18 .50 nuFi 6.44 8.05 .45
trzu 5.62 6.96 .56 EExp 6.92 8.42 .43
Crow 5.68 6.92 .55 538E 7.10 8.78 .27
11 teams are projecting within 1.5 wins of 41 -- 8 in the West.
UPDATE Feb. 7 -- KPel lead in MAE = RMSE lead for dtka.
Code: Select all
. avg err rmse r^2 avg err rmse r^2
KPel 4.60 5.99 .55 538R 5.77 7.29 .51
dtka 4.76 5.83 .65 vegas 5.81 7.46 .52
avgA 4.99 5.97 .61 TmRk 5.83 7.41 .51
vzro 5.13 6.50 .53 22Re 5.86 7.22 .34
LEBR 5.22 6.16 .58 MPra 5.92 7.34 .61
ncs. 5.25 6.54 .53 BIPM 5.94 7.25 .67
emin 5.30 6.90 .54 AnBa 6.03 7.76 .54
DRKO 5.39 6.31 .55 nuFi 6.11 7.83 .48
EBPI 5.40 6.87 .51 4141 6.25 8.88
trzu 5.45 6.60 .60 EExp 6.59 8.09 .47
Crow 5.51 6.54 .59 538E 7.00 8.68 .29
Smallest error of the year for dtka, vzro, LEBR, emin, 538R, vegas, TmRk, MPra, AnBa, nuFi, EExp.
Re: 2022-23 team win projection contest
Posted: Fri Feb 03, 2023 3:45 pm
by Crow
Any interest in doing a east / west split?
I lead in another contest of 55 on west rank (not wins). Wondering if my win projection performance is worse in east. It often has been in past but not always. Wondering about others as well.
Re: 2022-23 team win projection contest
Posted: Tue Feb 07, 2023 12:49 pm
by Mike G
The West has been much less predictable than the East:
Code: Select all
west err east err
KPel 5.00 emin 3.80
dtka 5.34 DRKO 4.08
538R 5.56 ncs. 4.10
EBPI 5.72 dtka 4.18
vzro 5.76 KPel 4.20
LEBR 6.06 LEBR 4.38
Crow 6.24 vzro 4.50
trzu 6.26 trzu 4.64
BIPM 6.26 Crow 4.80
ncs. 6.40 nuFi 4.92
vegas 6.62 vegas 4.98
MPra 6.64 TmRk 4.98
TmRk 6.68 EBPI 5.08
DRKO 6.70 MPra 5.20
emin 6.80 AnBa 5.20
AnBa 6.86 BIPM 5.62
EExp 6.92 538E 5.84
nuFi 7.30 538R 5.98
538E 8.16 EExp 6.26
avg. 6.38 avg. 4.88
Emin makes gains in the West with big surges from GSW, LAC, Min, SAS; lesser from Hou, LAL; or from collapse by Sac, Uta, OKC (unlikely).
Also needs continuation of struggle by Dal, NOP; continued overachievement from Por, Den
Re: 2022-23 team win projection contest
Posted: Tue Feb 07, 2023 1:09 pm
by Crow
Thanks.
So I am better in East in raw terms (like everyone) but better in West compared to competition.
Re: 2022-23 team win projection contest
Posted: Tue Feb 07, 2023 1:20 pm
by Mike G
Better than the avg. entry by .14 in the west and by .08 in the east, yeah.
This could swing the other way in a day or 2.
Re: 2022-23 team win projection contest
Posted: Thu Feb 09, 2023 12:09 am
by Crow
MAE lead cut in half in 4 days. Will there be a new leader at next update?
I improved but not fast enough.
Re: 2022-23 team win projection contest
Posted: Sat Feb 11, 2023 1:37 am
by Crow
Pacers 2-13 recently and well behind tonight in 2nd half.
Re: 2022-23 team win projection contest
Posted: Sat Feb 11, 2023 10:10 am
by Mike G
A contest!
Code: Select all
. avg err rmse r^2 avg err rmse r^2
KPel 4.78 6.24 .54 538R 5.73 7.42 .51
dtka 4.85 5.99 .64 MPra 5.90 7.42 .60
avgA 5.14 6.15 .60 BIPM 5.95 7.35 .66
vzro 5.19 6.66 .53 22Re 5.98 7.44 .34
LEBR 5.25 6.30 .58 vegas 6.05 7.60 .51
ncs. 5.42 6.74 .52 TmRk 6.07 7.56 .50
emin 5.45 7.07 .53 AnBa 6.14 7.91 .52
EBPI 5.47 7.04 .50 nuFi 6.25 8.05 .46
trzu 5.49 6.73 .59 4141 6.51 9.17
DRKO 5.50 6.53 .54 EExp 6.68 8.20 .46
Crow 5.53 6.69 .58 538E 7.03 8.80 .29
MPra again with his lowest error of the season.
New low for the top-to-bottom difference (2.24)
I'm curious whether people find this visually useful:
Code: Select all
Atl w Bos w Brk w Cha w Chi w Cle w Det w Ind w
dt 50 vz 58 em 51 KP 41 nc 41 b-r 52 vz 31 KP 38
tr 48 tr 57 b-r 48 vz 39 b-r 40 em 50 nc 27 b-r 36
cr 47 b-r 57 nc 47 dt 33 cr 40 tr 48 em 27 nc 34
KP 46 nc 57 tr 47 nc 32 KP 38 vz 48 KP 26 em 33
nc 45 em 56 KP 46 em 32 dt 38 cr 47 dt 24 vz 32
vz 44 dt 55 dt 44 tr 32 tr 38 dt 46 tr 24 dt 31
em 44 KP 54 cr 44 cr 30 vz 37 KP 43 cr 23 tr 28
b-r 41 cr 53 vz 42 b-r 24 em 35 nc 43 b-r 23 cr 28
Mia w Mil w NYK w Orl w Phl w Tor w Was w
nc 51 em 55 cr 44 b-r 34 tr 54 cr 52 vz 40
tr 50 b-r 54 b-r 43 KP 31 dt 53 tr 50 KP 40
cr 50 dt 52 vz 43 em 28 em 52 nc 49 b-r 39
vz 49 vz 52 nc 43 tr 28 nc 52 dt 48 dt 38
em 48 tr 50 dt 43 cr 28 vz 51 KP 47 nc 37
dt 47 KP 50 tr 42 nc 28 cr 51 vz 44 em 34
KP 46 cr 49 KP 42 dt 26 b-r 50 em 40 tr 33
b-r 44 nc 48 em 36 vz 26 KP 48 b-r 39 cr 30
Dal w Den w GSW w Hou w LAC w LAL w Mem w Min w
nc 47 b-r 54 em 52 nc 27 cr 51 dt 43 cr 52 em 53
tr 47 cr 53 cr 51 cr 27 em 49 vz 43 b-r 51 tr 50
vz 46 em 51 tr 50 KP 27 tr 49 em 42 dt 50 dt 49
dt 46 dt 49 dt 47 em 24 vz 45 nc 39 tr 48 cr 47
cr 46 tr 49 vz 47 tr 23 nc 45 b-r 37 nc 47 nc 47
b-r 45 vz 48 nc 44 b-r 21 dt 45 tr 37 em 47 KP 46
em 44 KP 48 KP 42 dt 20 KP 44 KP 37 vz 47 vz 45
KP 43 nc 47 b-r 42 vz 18 b-r 42 cr 35 KP 46 b-r 41
NOP w OKC w Phx w Por w Sac w SAS w Uta w
cr 49 b-r 42 nc 54 em 40 b-r 45 vz 33 b-r 41
tr 48 cr 27 tr 52 b-r 40 cr 39 KP 31 vz 39
KP 48 KP 27 vz 51 tr 37 tr 37 em 31 KP 35
nc 46 dt 26 em 50 KP 37 dt 37 nc 31 nc 35
dt 46 vz 26 dt 50 dt 36 nc 37 dt 26 dt 32
vz 44 nc 25 cr 50 vz 33 KP 37 cr 25 cr 29
b-r 43 tr 24 KP 49 cr 33 vz 32 tr 25 tr 29
em 43 em 23 b-r 44 nc 28 em 32 b-r 20 em 28
UPDATE Feb.12
Code: Select all
. avg err rmse r^2 avg err rmse r^2
KPel 4.83 6.32 .54 538R 5.78 7.52 .50
dtka 4.84 5.97 .64 MPra 5.88 7.38 .61
avgA 5.13 6.17 .60 BIPM 6.04 7.37 .65
vzro 5.15 6.68 .53 vegas 6.10 7.60 .51
LEBR 5.25 6.30 .58 22Re 6.10 7.57 .33
emin 5.43 7.12 .53 AnBa 6.13 7.90 .52
ncs. 5.44 6.78 .52 TmRk 6.16 7.57 .50
DRKO 5.49 6.59 .54 nuFi 6.33 8.10 .46
EBPI 5.50 7.10 .50 4141 6.59 9.27
trzu 5.53 6.75 .59 EExp 6.72 8.21 .46
Crow 5.55 6.72 .58 538E 7.14 8.93 .28
Lowest errors at select exponents:
Code: Select all
e= .10 e= .27 e= .50 e= 2.00 e= 6.7 e= 15
Crow 2.25 emin 3.63 KPel 4.08 dtka 5.97 LEBR 9.50 Crow 11.98
emin 2.64 vzro 3.68 dtka 4.21 avgA 6.17 dtka 9.74 LEBR 12.05
4141 2.77 KPel 3.74 vzro 4.24 LEBR 6.30 Crow 9.75 dtka 12.68
vzro 2.77 Crow 3.74 emin 4.33 KPel 6.32 DRKO 10.08 DRKO 12.84
538R 2.83 538R 3.87 ncs. 4.54 DRKO 6.59 avgA 10.27 vzro 13.30
Re: 2022-23 team win projection contest
Posted: Sun Feb 12, 2023 3:18 pm
by Crow
I find the by team list very useful. I have used it to compare to BRef and leaders but perhaps I should to the first folks I need to pass as well. I don't know if I'll comment on the specifics right now because I am far back but I see a lot of teams where I am very close to BRef projection and a lot of teams where I am extreme high or low. That could give me a chance in the end, if the latter teams move my way.
Best I can tell myself at the moment is .13 from 6th and slightly closer to leader than a few days ago. I am modestly higher ranked on rmse and r2 fwiw.
I am not clear how to interpret / react to the exponent range with me being a leader at both extremes other than thinking that it would be great for me if either team performances (or scoring) changed dramatically. Would appreciate hearing your interpretation if you have more.
I think it is fairly unlikely Kevin Pelton will be ahead in the end. The projections seemed too regressed to me. That has been very effective to date. I can't see it lasting. But if it does, that is quite the story. It would probably be because of calling the middle better than the extremes.
Was the compression in the win range simply a product of the data & method Kevin or consciously anticipated or predicted prior to the calculaton? Would be interesting to hear.
Right now BRef is projecting 6 50 plus and 4 in 20s. Pelton projected only 1 50 plus and 3 in 20s. Last season it was 8 50 plus and 6 in 20s. I projected 9 and 6 respectively. The distribution has to swing further back to last season's for me to have a chance.
The strong showing of the Apbrmetrics average is a good collective story for the handful that participated.
Re: 2022-23 team win projection contest
Posted: Sun Feb 12, 2023 5:41 pm
by Mike G
The very low exponents favor those with the most 'direct hits' -- within one win of the current projections. Big misses are not much worse than medium sized.
High exponent greatly punishes the big misses, while minimizing the small ones. As long as your worst miss (OKC) is less than anyone else's worst, you win the big end.
It's kind of unusual for anyone to lead at more than one point on the continuum. And the low-exponent leader changes almost daily; if you are right-on with a prediction, a win or a loss are both bad.
Re: 2022-23 team win projection contest
Posted: Sun Feb 12, 2023 6:42 pm
by Crow
Thanks.
Re: 2022-23 team win projection contest
Posted: Sun Feb 12, 2023 6:45 pm
by Crow
Average over last 8 year's is slightly less than 6 teams with 50 plus wins. By using last season's 8 as a guide, I went beyond average a bit more than I might have if I had checked the average beforehand.
Low is 2 but only 1 year less than 5.
Re: 2022-23 team win projection contest
Posted: Tue Feb 14, 2023 10:26 am
by Mike G
Code: Select all
. avg err rmse r^2 avg err rmse r^2
dtka 4.82 5.96 .64 538R 5.78 7.46 .51
KPel 4.84 6.29 .55 MPra 5.81 7.29 .61
avgA 5.13 6.14 .61 BIPM 6.09 7.40 .65
vzro 5.16 6.64 .53 22Re 6.13 7.59 .34
LEBR 5.33 6.26 .59 vegas 6.19 7.61 .51
DRKO 5.47 6.56 .54 AnBa 6.26 7.90 .53
EBPI 5.47 7.08 .50 TmRk 6.27 7.59 .50
trzu 5.48 6.68 .60 nuFi 6.43 8.11 .46
emin 5.49 7.09 .53 4141 6.64 9.31
ncs. 5.53 6.81 .51 EExp 6.86 8.20 .47
Crow 5.60 6.67 .59 538E 7.11 8.89 .29
Re: 2022-23 team win projection contest
Posted: Tue Feb 14, 2023 2:55 pm
by dtkavana
#1 baby. LFG.