Power Ranking

[EvanZ]

Yes, Prediction Tracker (a guy named Todd) for some reason uses the opposite convention to everyone else. It kind of annoys me, but he's doing a good service, so tough to complain.

This is what I'm using for lines:

http://www.wagertracker.com/Odds.aspx?d ... &sport=NBA

Just replace the date with whatever date you want.

[Crow]

Alright, thanks for the clarification and tip. I'll try to sort it out. Sorry for my confusion.

[EvanZ]

Oh, it's not your fault. I just happened to make the mistake first, so I can correct the next person.

[Crow]

So I guess for the 3 "best" picks by this method went 1-2 this time with two close calls.

[Jeff Fogle]

Will be interesting to see how the markets and various ratings systems deal with Philadelphia. They're showing a clear "well-suited to blow out patsies" tendency that the markets are reacting to vs. bad teams. But, the schedule has been so soft thus far for the Sixers that they have few games vs. winning teams.

lost 107-103 vs. Portland (currently 7-4)
lost 102-99 at Utah (currently 6-4)
won 96-86 vs. Indiana (currently 8-3)
lost 85-79 at NY (currently 6-5)

Not included yet: Miami, Chicago, Orlando, Atlanta, Oklahoma City, San Antonio, LAL, LAC, Denver, Dallas.

Some good teams are coming up finally:
Wednesday: vs. Denver
Friday: vs. Atlanta
Saturday: at Miami

Then another batch down the road:
1/30: at Orlando
2/1: vs. Chicago
2/3: vs. Miami
2/4: at Atlanta
2/6: vs. LAL
2/8: vs. SA
2/10: vs. LAC

Interested to see pricing strategies this week, particularly if Wade is back for Miami by Saturday. There's a lot to like about the Sixers, but they're just 1-3 in the "Super League" of teams with winning records, and they've missed most of the serious championship threats...

[J.E.]

You can try Rotoworld, not sure it's better though:

http://www.rotoworld.com/teams/injuries/nba/all/

Those are pretty good, thanks

They even give you additional info on return date, other than "day to day" etc., when you hover over the info button

Maybe I'll write a parser for the info on the "!" button. RMSE for the RAPM derived rating can definitely be reduced when we actually know who's playing instead of guessing. Especially in a season like this, with players missing games left and right
(The above moved from a post moved to new 76ers thread.)

On team specific HCA, here's a ten year difference in home point differential vs away point differential. I think if there is an effect, it should be the same every year*. Thus you can use more years to reduce noise.

*unless it is somehow player dependent. If it is, the effect is probably minor, I would believe

Oklahoma;2.34
Clippers;3.01
Boston;2.06
Orlando;3.35
Memphis;3.12
Minnesota;2.81
Jersey;3.12
Antonio;3.02
Houston;2.64
Chicago;3.36
Philadelphia;2.23
Detroit;2.301
Lakers;3.565
Toronto;3.091
Denver;4.679
Orleans;2.951
Cleveland;4.182
Dallas;3.180
Portland;3.414
Milwaukee;3.59
Washington;3.589
Miami;2.698
York;2.25
Charlotte;3.761
Golden;4.334
Phoenix;3.437
Utah;4.471
Sacramento;4.109
Atlanta;3.696
Indiana;3.9

[DSMok1]

Excellent!

I would say, though, that it should be city-dependent, so split out Seattle from OKC (and whatever other franchises moved).

[Jeff Fogle]

Interesting that Lakers are 3.56 and Clippers are 3.01 on the same floor even after the noise reduction of a 10-year study. Evidence perhaps for some player related influences that are meaningful...at least for something like generally high quality players for one team and lower quality for the other (more respect from officials in terms of drawing fouls?). So much in common in terms of what visitors are dealing with (travel, games at sea level, an assumption for a similar number of b2b's falling on night two), yet there's still a half a point over 10 years. Thanks for posting those J.E.

[J.E.]

I would say, though, that it should be city-dependent, so split out Seattle from OKC (and whatever other franchises moved).

Yeah I did that. I also separated the Charlotte Bobcats from the Charlotte Hornets.

Jeff, I think part of it is atmosphere/fans. I don't know the situation in LA, but the Lakers seem to have had a higher attendance!? Then again, they did probably also draw higher attendance at opponents. Hm. Maybe it's the floor color? The dancers?

[Mike G]

Here's what J.E. reported, sorted:

Code: Select all

HCA  Western Conference             Eastern Conference
4.68   Den      3.18   Dal       4.18   Cle      3.12   NJN
4.47   Uta      3.12   Mem       3.90   Ind      3.09   Tor
4.33   GSW      3.02   SAS       3.76   Cha      2.70   Mia
4.11   Sac      3.01   LAC       3.70   Atl      2.30   Det
3.57   LAL      2.95   NOH       3.59   Mil      2.25   NYK
3.44   Phx      2.81   Min       3.59   Was      2.23   Phl
3.41   Por      2.64   Hou       3.36   Chi      2.06   Bos
3.41  West      2.34   Okl       3.35   Orl      3.15  East

The 3 biggest HCA are in the West. The 4 smallest are in the East; in 4 of the most venerable franchise locales.
Are visiting teams "excited" to be playing in these tradition-steeped places?
Are their teams jaded, or regularly booed at home when they aren't doing well?

[EvanZ]

Denver and Utah are at altitude. And California fans are the best!

[Mike G]

Denver and Utah are at altitude. ..

Den has the largest HCA over the last few years, at 4.68; Utah is 2nd with 4.47 .
If altitude is a big part of this, would we expect their HCA against one another to be somewhat less than it is against the league in general? Maybe close to the norm of 3.3 ?

Well ... since 1985-86, they've played one another 59 times in each home venue.
In Denver, the Nuggets are 31-28, avg +.90 differential.
In SLC, they're 9-50, avg -10.36

The difference is 11.26 PPG.

[EvanZ]

Denver and Utah are at altitude. ..

Den has the largest HCA over the last few years, at 4.68; Utah is 2nd with 4.47 .
If altitude is a big part of this, would we expect their HCA against one another to be somewhat less than it is against the league in general? Maybe close to the norm of 3.3 ?

Well ... since 1985-86, they've played one another 59 times in each home venue.
In Denver, the Nuggets are 31-28, avg +.90 differential.
In SLC, they're 9-50, avg -10.36

The difference is 11.26 PPG.

That is puzzling.

[EvanZ]

I've started calculating VPR (Vegas Power Ratings) for the prior week:

http://thecity2.com/vegas-power-ratings/

Here are last week's ratings:

Code: Select all

RK	TEAM	LONG	LVWK
1	MIA	Miami Heat	9.30
2	OKC	Oklahoma City Thunder	6.52
3	CHI	Chicago Bulls	6.45
4	ORL	Orlando Magic	4.60
5	POR	Portland Trailblazers	4.29
6	DEN	Denver Nuggets	3.87
7	SAS	San Antonio Spurs	3.33
8	DAL	Dallas Mavericks	3.09
9	PHI	Philadelphia 76ers	3.08
10	LAL	Los Angeles Lakers	3.02
11	LAC	Los Angeles Clippers	2.30
12	BOS	Boston Celtics	2.14
13	IND	Indiana Pacers	1.67
14	ATL	Atlanta Hawks	1.62
15	MEM	Memphis Grizzlies	0.79
16	HOU	Houston Rockets	0.77
17	NYK	New York Knicks	0.59
18	MIL	Milwaukee Bucks	-1.61
19	PHX	Phoenix Suns	-1.69
20	UTA	Utah Jazz	-1.86
21	MIN	Minnesota Timberwolves	-2.01
22	GSW	Golden State Warriors	-2.48
23	TOR	Toronto Raptors	-4.30
24	CLE	Cleveland Cavaliers	-4.56
25	DET	Detroit Pistons	-4.77
26	SAC	Sacramento Kings	-4.88
27	NOH	New Orleans Hornets	-5.30
28	NJN	New Jersey Nets	-5.42
29	CHA	Charlotte Bobcats	-5.55
30	WAS	Washington Wizards	-6.03

It seems that even with just one week of games, it's enough to get a good order, at least.

[DSMok1]

This begs a Bayesian treatment, with previous games weighted at a^n, where a < 1.0 and n = number of days ago. a could be found via cross-validation.