Comments on: BBR Rankings: Schedule-Adjusted Offensive and Defensive Ratings (November 12, 2010) http://www.basketball-reference.com/blog/?p=8123 NBA & ABA Basketball Statistics & History Mon, 21 Nov 2011 20:56:04 +0000 hourly 1 https://wordpress.org/?v=4.6 By: Neil Paine http://www.basketball-reference.com/blog/?p=8123&cpage=1#comment-36207 Thu, 09 Dec 2010 20:31:29 +0000 http://www.basketball-reference.com/blog/?p=8123#comment-36207 #3 - Good question, I think I'll look at that at some point.

]]> By: Jason J http://www.basketball-reference.com/blog/?p=8123&cpage=1#comment-32359 Fri, 12 Nov 2010 18:53:34 +0000 http://www.basketball-reference.com/blog/?p=8123#comment-32359 Surprised the Hornets and Bucks are so far out in front on the defensive side, even considering we're only a couple weeks in.

]]> By: DSMok1 http://www.basketball-reference.com/blog/?p=8123&cpage=1#comment-32356 Fri, 12 Nov 2010 18:26:55 +0000 http://www.basketball-reference.com/blog/?p=8123#comment-32356 That was quick!

We would then estimate that a weighted average with 7 games of 0 would be the best estimate throughout the season...

Any ideas on how many games of the SRS preseason projection would yield the best estimate? Probably more like 12 or 15...

]]> By: Neil Paine http://www.basketball-reference.com/blog/?p=8123&cpage=1#comment-32355 Fri, 12 Nov 2010 18:19:57 +0000 http://www.basketball-reference.com/blog/?p=8123#comment-32355 I didn't go through this with adjusted efficiency differential per se, but SRS is a quick-n-dirty proxy... Through Nov 11 of last year, you needed to add seven games of zero to get the lowest squared errors when predicting final SRS (postseason included):

team_id final 11-Nov games games of 0 regressed err^2
ATL 3.571043614 5.461515458 8 7.036528279 2.905732152 0.442639342
BOS 4.406563137 10.89290052 9 7.036528279 6.113299772 2.912949942
CHA 1.1603997 -5.996771663 8 7.036528279 -3.190508634 18.93040334
CHI -1.742721136 -1.296426929 8 7.036528279 -0.689748008 1.108752407
CLE 5.798293055 2.93573871 8 7.036528279 1.561923687 17.94682543
DAL 2.763861768 7.07368662 8 7.036528279 3.763467997 0.999212611
DEN 4.139408178 2.932790876 9 7.036528279 1.645937165 6.217397696
DET -5.029896667 1.515624542 8 7.036528279 0.806369403 34.06200164
GSW -3.259376157 -6.815391713 7 7.036528279 -3.398827762 0.01944675
HOU -0.006480229 5.609996361 8 7.036528279 2.98472959 8.947336183
IND -3.11032882 -0.239312367 6 7.036528279 -0.110142376 9.001118698
LAC -6.002380061 -4.478077389 9 7.036528279 -2.513180896 12.17451082
LAL 5.549183748 6.111001031 7 7.036528279 3.047548964 6.258176592
MEM -1.381991474 -10.04628942 9 7.036528279 -5.638165766 18.1150196
MIA 1.665936392 7.869308888 7 7.036528279 3.924415006 5.10072565
MIL 1.26034132 1.460605836 6 7.036528279 0.672236874 0.34586684
MIN -9.053556177 -13.05617783 9 7.036528279 -7.327371513 2.979713493
NJN -8.928493456 -12.35501032 8 7.036528279 -6.573331341 5.54678859
NOH -2.261044175 -3.153380289 9 7.036528279 -1.769736074 0.24138365
NYK -4.009028033 -9.016047932 9 7.036528279 -5.059974951 1.104489424
OKC 3.574587788 5.309555548 8 7.036528279 2.824883749 0.562056146
ORL 7.785182565 2.887539245 9 7.036528279 1.620541102 38.00280438
PHI -3.91639128 -4.316139399 8 7.036528279 -2.296348901 2.624537308
PHO 5.345131816 5.440280464 9 7.036528279 3.053187281 5.253009748
POR 2.642556391 7.70669432 9 7.036528279 4.325141182 2.831091579
SAC -4.049693044 -1.938168918 8 7.036528279 -1.031178944 9.111427376
SAS 4.589121295 3.5950541 7 7.036528279 1.792849214 7.819137547
TOR -1.828240042 -0.061055392 8 7.036528279 -0.032483771 3.224740586
UTA 5.044031124 -2.165981598 8 7.036528279 -1.152383879 38.39555889
WAS -4.716019289 -1.868059274 8 7.036528279 -0.993877969 13.854336
274.1334583
]]> By: DSMok1 http://www.basketball-reference.com/blog/?p=8123&cpage=1#comment-32351 Fri, 12 Nov 2010 17:56:32 +0000 http://www.basketball-reference.com/blog/?p=8123#comment-32351 Any idea of how many games of 0 margin must be added in to get a best estimate of true talent level at this point? I'm guessing something like 5...

]]>