Comments on: BBR Rankings: Schedule-Adjusted Offensive and Defensive Ratings (December 31, 2010) http://www.basketball-reference.com/blog/?p=8510 NBA & ABA Basketball Statistics & History Mon, 21 Nov 2011 20:56:04 +0000 hourly 1 https://wordpress.org/?v=4.6 By: Greyberger http://www.basketball-reference.com/blog/?p=8510&cpage=1#comment-38205 Sat, 01 Jan 2011 02:34:52 +0000 http://www.basketball-reference.com/blog/?p=8510#comment-38205 In the West, Spurs say, "get behind me, devil." The other storyline is the delayed emergence of a legit fourth place - surely Utah OKC or Denver will move up from the pack and closer to the top teams.

Not too early for me to say since I don't care much about the East - we probably have our four home court teams and the rest of the season is about five through eight and who gets what spot.

]]> By: Neil Paine http://www.basketball-reference.com/blog/?p=8510&cpage=1#comment-38189 Fri, 31 Dec 2010 18:29:02 +0000 http://www.basketball-reference.com/blog/?p=8510#comment-38189 Re: #2 - Steps to determine this:

1. Regress to end-of-season ratings by adding 413 possessions of rating 0.00

2. Convert eff. diff. ratings to SRS by multiplying by 92.53/100

3. Convert SRS to WPct using WPct = 0.51419884 + 0.035212338*SRS + 0.00000026073081*SRS^5 + 0.000052608571*SRS^4 - 0.000039551538*SRS^3 - 0.00000035817527*SRS^6 - 0.0021272644*SRS^2

4. Determine stdev of NBA talent = 0.144 (reference)

5. Estimate each team's true wpct talent using (0.5 / 0.144^2 + wpct / stdev^2)/(1 / 0.144^2 + 1 / stdev^2), where stdev = (SQRT((wins + losses) * wpct * (1 - wpct))) / (wins + losses)

6. Simulate rest of season 1000 times using talent estimates

The results:

team_id Champs Finals
MIA 34.3% 48.3%
BOS 22.5% 34.0%
SAS 15.2% 34.0%
LAL 8.0% 21.3%
DAL 6.5% 16.7%
CHI 3.4% 7.6%
ORL 2.6% 6.3%
UTA 1.7% 8.2%
OKC 1.5% 6.0%
DEN 1.4% 6.2%
ATL 1.1% 1.9%
NOH 0.6% 3.9%
POR 0.5% 1.7%
HOU 0.3% 1.4%
PHI 0.2% 0.7%
NYK 0.1% 1.0%
IND 0.1% 0.1%
MEM 0.0% 0.5%
MIL 0.0% 0.1%
PHO 0.0% 0.1%
NJN 0.0% 0.0%
TOR 0.0% 0.0%
CLE 0.0% 0.0%
DET 0.0% 0.0%
CHA 0.0% 0.0%
WAS 0.0% 0.0%
MIN 0.0% 0.0%
GSW 0.0% 0.0%
LAC 0.0% 0.0%
SAC 0.0% 0.0%
]]> By: Stuart Chuang Matthews http://www.basketball-reference.com/blog/?p=8510&cpage=1#comment-38178 Fri, 31 Dec 2010 16:01:58 +0000 http://www.basketball-reference.com/blog/?p=8510#comment-38178 Come the end of each season, what is the percent chance a team will win a championship based on their position in this ranking?

]]> By: Neil Paine http://www.basketball-reference.com/blog/?p=8510&cpage=1#comment-38173 Fri, 31 Dec 2010 15:00:51 +0000 http://www.basketball-reference.com/blog/?p=8510#comment-38173 Btw, minimizing the margin in the ratings like DSMOk suggested last week actually didn't change them at all:

team_id offense defense overall
MIA 4.42 -5.32 9.74
BOS 2.32 -6.48 8.80
SAS 5.36 -2.42 7.78
LAL 4.53 -1.47 5.99
DAL 1.64 -4.25 5.89
CHI -1.63 -7.09 5.45
ORL 0.42 -4.13 4.55
UTA 3.33 0.20 3.13
DEN 4.72 2.02 2.70
OKC 3.20 1.05 2.15
NOH -2.24 -4.03 1.79
ATL 0.89 -0.61 1.49
HOU 3.78 2.65 1.14
POR 0.16 -0.68 0.85
NYK 3.49 3.32 0.17
MEM -2.40 -2.17 -0.22
PHI -1.31 -0.94 -0.36
IND -3.98 -2.85 -1.13
PHO 5.75 6.96 -1.20
MIL -5.89 -4.54 -1.35
GSW -0.39 3.28 -3.68
TOR 0.52 4.49 -3.97
LAC -2.19 2.14 -4.33
DET -0.60 4.12 -4.72
NJN -4.35 1.23 -5.58
MIN -1.66 4.29 -5.95
CHA -5.25 0.71 -5.97
WAS -3.79 2.20 -5.99
SAC -5.93 2.96 -8.89
CLE -4.98 4.21 -9.20

This was using an exponent of 2... I didn't have time to re-arrange the spreadsheet so that I wasn't raising a negative # to the 1.5 power.

]]>