APBRmetrics

The discussion of the analysis of basketball through objective evidence, especially basketball statistics.
It is currently Fri Nov 28, 2014 11:56 am

All times are UTC




Post new topic Reply to topic  [ 234 posts ]  Go to page Previous  1 ... 12, 13, 14, 15, 16  Next
Author Message
 Post subject: Re: Power Ranking
PostPosted: Sun Feb 19, 2012 5:23 pm 
Offline

Joined: Fri Apr 15, 2011 2:05 am
Posts: 95
Mike G., can't you do the same thing with Chicago? Using ESPN's expanded standings page:

Games decided by 3 or less:
Chicago 3-0
OKC 4-2

Chicago's played a much softer schedule than OKC, with 20 games vs. below .500 teams instead of just 12 for OKC.

Games vs. .500 or better:
Chicago 7-6
OKC 13-5

Whether OKC's schedule is ranked 9-15th, or Chicago's 25-30th...isn't it clear that OKC has played the tougher schedule...and has better results vs. good teams...and that both could be closer to .500 in close games moving forward than they've been so far?

_________________
Blogging basketball at http://www.statintelligence.blogspot.com/


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Sun Feb 19, 2012 5:25 pm 
Offline

Joined: Fri Apr 15, 2011 2:05 am
Posts: 95
Evan, would you have Philly over SA if they had faced the same schedule in your methodology? That's an interesting test for me because I watched the Spurs beat them in Philly, winning every quarter...during the stretch where Philly was also losing at home to Dallas and the LA Clippers.

_________________
Blogging basketball at http://www.statintelligence.blogspot.com/


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Sun Feb 19, 2012 6:25 pm 
Offline

Joined: Thu Apr 14, 2011 10:41 pm
Posts: 815
Location: Hotlanta
I'm using the same method as Jerry ("adjusted ratings" so to speak), so the schedule is implicitly in the results. It's not something added to the ratings later like SRS+SOS.

_________________
The City: A Golden State Warriors-Centric NBA Blog
NBA WOWY: Find statistics of your favorite team with any arbitrary combination of players on or off the court.


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Sun Feb 19, 2012 6:41 pm 
Offline

Joined: Fri Apr 15, 2011 2:05 am
Posts: 95
Evan, any concerns about the ratings given that Philadelphia is 6-9 against teams at .500 or better, while San Antonio is 15-5...and the Spurs won at Philly, winning every quarter? (Plus, Philly has played a home heavy schedule, while Spurs have more road games...don't have time at the moment to see if/how that influences the .500 or better data).

If you were an oddsmaker, would you trust your ratings enough to make Philly a neutral site favorite over SA in a best of seven matchup?

_________________
Blogging basketball at http://www.statintelligence.blogspot.com/


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Sun Feb 19, 2012 7:53 pm 
Offline

Joined: Thu Apr 14, 2011 10:41 pm
Posts: 815
Location: Hotlanta
Jeff Fogle wrote:
Evan, any concerns about the ratings given that Philadelphia is 6-9 against teams at .500 or better, while San Antonio is 15-5...and the Spurs won at Philly, winning every quarter? (Plus, Philly has played a home heavy schedule, while Spurs have more road games...don't have time at the moment to see if/how that influences the .500 or better data).

If you were an oddsmaker, would you trust your ratings enough to make Philly a neutral site favorite over SA in a best of seven matchup?


Vegas has Philly and SAS essentially a push. I wouldn't bet on that game.

_________________
The City: A Golden State Warriors-Centric NBA Blog
NBA WOWY: Find statistics of your favorite team with any arbitrary combination of players on or off the court.


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Sun Feb 19, 2012 8:47 pm 
Offline

Joined: Fri Apr 15, 2011 2:05 am
Posts: 95
Yeah, but...the market has badly misread those situations. Philly is now 5-10-1 ATS vs. teams at .500 or better (Knicks just moved up to .500, which adds another straight up loss and non-cover to the category for Philly). San Antonio 15-5 ATS vs. teams at .500 or better.

Versus quality...it's a mismatch for the Spurs. Versus crap teams, Philly soars past expectations. Philly won't draw a crap team in the playoffs...and wouldn't be facing a crap team in the example of a head-to-head neutral site series with the Spurs.

Just throwing this stuff out there. I know you guys do extremely in-depth calculations. I was surprised Philly was getting so much stat-type respect when they were just crushing patsies. I figured the recent gauntlet would be a good test to see if they were for real. They lost at home to the Spurs, Mavs, and Clips, and lost pretty big at Orlando. To me that's a flunked test, at least to this point (more challenges are ahead and they could turn things around). Always sample size issues within a season. I don't think if you equalize the schedules or flipped conferences these would be seen as even teams (Philly 1-5-1 ATS vs. West). Metrics that are adjusting for schedule and still show them as equal concern me.

Ginobili's out again, so we won't see SA at full strength star-wise for awhile. They've got a very strong resume even with him missing a lot of games. Philly's resume is extremely bottom heavy in a way that doesn't look like it would translate well if you equalized schedules. They're ONLY better at crushing patsies...and a lot of crushing patsies involves choices the coach makes with a big lead...

Might be fun to do a poll on that if any mod is interested. "Who do you believe is the better team, San Antonio, Philadelphia, neither?"

Would even Doug Collins pick the Sixers?

_________________
Blogging basketball at http://www.statintelligence.blogspot.com/


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Sun Feb 19, 2012 10:10 pm 
Offline

Joined: Thu Apr 14, 2011 10:41 pm
Posts: 815
Location: Hotlanta
Maybe capping point differential would be a way to address it. Maybe something like 10-12 should be the cap on wins.

_________________
The City: A Golden State Warriors-Centric NBA Blog
NBA WOWY: Find statistics of your favorite team with any arbitrary combination of players on or off the court.


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Mon Feb 20, 2012 1:05 am 
Offline

Joined: Thu Apr 21, 2011 8:25 pm
Posts: 188
Location: Boone, NC
Jerry,

Are these adjusted for b2b, etc, at all?

Can these be added in the same manner as coaches? i.e. an extra player on the court would be "b2b" :)


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Mon Feb 20, 2012 3:03 am 
Offline

Joined: Thu Apr 14, 2011 10:41 pm
Posts: 815
Location: Hotlanta
Those kinds of factors can definitely be added. Just a matter of spending the time and effort.

_________________
The City: A Golden State Warriors-Centric NBA Blog
NBA WOWY: Find statistics of your favorite team with any arbitrary combination of players on or off the court.


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Mon Feb 20, 2012 4:38 am 
Offline

Joined: Fri Apr 15, 2011 2:05 am
Posts: 95
Agree with you Evan about margin caps. Maybe there's a creative way to deal with East/West imbalance too. Just checked ESPN's standing page...West is 85-60 vs. the East this year. Miami and Chicago are 13-5 vs. West, everyone else in the East is 47-80. Only four Western teams have losing records vs. the East...

_________________
Blogging basketball at http://www.statintelligence.blogspot.com/


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Mon Feb 20, 2012 11:55 am 
Offline

Joined: Fri Apr 15, 2011 8:28 am
Posts: 660
Jeff Fogle wrote:
JE, why do you believe there's a gap from Miami and Chicago to OKC in your data?

Miami 24-7 vs. soft schedule (21st in Hollinger, 25th in Sagarin)
Chicago 25-8 vs. easy schedule (29th in Hollinger, 27th in Sagarin)
Oklahoma City 23-7 vs. decent schedule (9th in Hollinger and Sagarin)

(Hollinger of ESPN, Sagarin of USA Today for international readers)

Is it because OKC has been less hit by the injury bug, and you're rating everyone at full strength? Just wondering about that...because most places I see have OKC up in a big three rather than gapped out of a big two. Doesn't mean they're right and you're wrong. Was just wondering what you believed caused that separation in your process...
It's because OKC doesn't win the games with as much point differential as MIA/CHI

Quote:
Jerry,

Are these adjusted for b2b, etc, at all?

Can these be added in the same manner as coaches? i.e. an extra player on the court would be "b2b"

Not adjusted for b2b. That would certainly help some, but probably not too much after 30 games. Yes that can be added in the same manner as coaches :)
Would be interesting to do some b2b adjustment on player RAPM level. Some players, especially star players, are asked to play >30m on b2b nights, and that certainly hurts their RAPM

_________________
http://stats-for-the-nba.appspot.com/


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Mon Feb 20, 2012 12:03 pm 
Offline

Joined: Thu Apr 14, 2011 10:41 pm
Posts: 815
Location: Hotlanta
I capped MOV at 12 (or -12) and scaled the ratings by the ratio of the sd of the original ridge ratings to these (~1.6):

Code:
RK   TEAM   CAP12
1   CHI   6.47
2   MIA   6.29
3   OKC   6.21
4   SAS   5.73
5   PHI   4.73
6   LAC   3.44
7   DAL   3.16
8   POR   2.78
9   ORL   2.63
10   DEN   2.32
11   IND   2.24
12   HOU   1.79
13   ATL   1.61
14   MEM   1.52
15   LAL   1.28
16   MIN   0.18
17   UTA   -0.35
18   BOS   -0.70
19   GSW   -0.99
20   MIL   -1.36
21   NYK   -1.78
22   CLE   -1.86
23   PHX   -2.07
24   SAC   -4.25
25   TOR   -4.45
26   NOH   -5.57
27   DET   -6.04
28   NJN   -6.50
29   WAS   -7.17
30   CHA   -9.49

_________________
The City: A Golden State Warriors-Centric NBA Blog
NBA WOWY: Find statistics of your favorite team with any arbitrary combination of players on or off the court.


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Mon Feb 20, 2012 12:45 pm 
Online

Joined: Thu Apr 14, 2011 11:18 pm
Posts: 598
Location: Maine
EvanZ wrote:
Maybe capping point differential would be a way to address it. Maybe something like 10-12 should be the cap on wins.


I've investigated that in the past, and it generally hurts predictive performance. I was looking at college football, though... That said, it would make more sense to use it with college football than the NBA in my opinion.

_________________
APBRmetrics Forum Administrator
GodismyJudgeOK.com/DStats/
Twitter.com/DSMok1


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Mon Feb 20, 2012 3:39 pm 
Offline

Joined: Thu Apr 14, 2011 10:41 pm
Posts: 815
Location: Hotlanta
I certainly don't have a dog in the fight. Just thought Jeff would want to see it.

_________________
The City: A Golden State Warriors-Centric NBA Blog
NBA WOWY: Find statistics of your favorite team with any arbitrary combination of players on or off the court.


Top
 Profile  
 
 Post subject: Re: Power Ranking
PostPosted: Mon Feb 20, 2012 5:52 pm 
Offline

Joined: Fri Apr 15, 2011 2:05 am
Posts: 95
Evan, those struck my eye as the kind of range sharps would have right now...with a few differences here and there on certain teams. Decided to line up my estimates of the market with that range. My estimates are in parenthesis.

Looks like the market (or a misread from me on the market) has the worst teams further down the ladder than the trimmed margin effort does. Maybe trimming blowouts has the hazard of making really bad teams look better than they are at the same time it's helping prevent pollution on the bullying end. Bullies shouldn't be rewarded for bullying, but patsies are still patsies. Will have to think on that a bit.

1 CHI 6.47 (+6)
2 MIA 6.29 (+7)
3 OKC 6.21 (+6)
4 SAS 5.73 (+4)
5 PHI 4.73 (+3)
6 LAC 3.44 (+2)
7 DAL 3.16 (+3)
8 POR 2.78 (+2)
9 ORL 2.63 (+3)
10 DEN 2.32 (0, I had docked them for injuries)
11 IND 2.24 (+2)
12 HOU 1.79 (0)
13 ATL 1.61 (+1)
14 MEM 1.52 (+1)
15 LAL 1.28 (+2)
16 MIN 0.18 (-1)
17 UTA -0.35 (-2)
18 BOS -0.70 (+1)
19 GSW -0.99 (-2)
20 MIL -1.36 (-1)
21 NYK -1.78 (0)
22 CLE -1.86 (-4)
23 PHX -2.07 (-5)
24 SAC -4.25 (-6)
25 TOR -4.45 (-6)
26 NOH -5.57 (-8)
27 DET -6.04 (-6)
28 NJN -6.50 (-7)
29 WAS -7.17 (-8)
30 CHA -9.49 (-11)

Thanks for calculating that for me. Appreciate it very much.

JE: I know Chicago has better margins. The bulk of that has come from running up the score on bad teams within a much softer schedule. If you throw out games against the teams at .333 or worse (a threshold we used in an earlier discussion I believe), you get:

Chicago: +6.1
OKC: +5.5

And, if OKC had 7 home games vs. that level instead of 2 (Chicago has 7, OKC 2), there would have been opportunities to crank up their overall average I'd think (if the coach had chosen to do so). Tough to be confident about what would happen in games that aren't being played. OKC doesn't get a lot of home games against the lousy teams in the East the way the Eastern powers will through the season.

DS Mok: I think in most years that capping probably wouldn't matter much in the NBA. Seems like this year we have some teams that are SO BAD because of poor talent, poor preparation, and a growing attitude across the league that you might as well be horrible if you're not going to reach the playoffs because it improves your draft hopes...that it might become needed within smaller sample sizes. Unbalanced schedules within a 30 game sampling might cause pollution that won't be as prevalent after 66 games, or in a normal 82 games. Philly's recent run may be making a case for that, and may make a further case once their home/road split evens out (19-13 home heavy at the moment)...

_________________
Blogging basketball at http://www.statintelligence.blogspot.com/


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 234 posts ]  Go to page Previous  1 ... 12, 13, 14, 15, 16  Next

All times are UTC


Who is online

Users browsing this forum: DSMok1, Google [Bot], Yahoo [Bot] and 3 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group