Adjusted Impact Rating- Top down metric, just noticed

[j_oxford]

Cavs played like a +10 team pretty much the whole time Mozgov was there. Kyrie and Love played a ton of minutes at the beginning of the season (Lebron as well, but he at least got a little break in there, which means his rating wasn't as deflated as Irving/Love by the beginning of the season), which slightly inflated Mozgov's Rating compared to if he had played the whole season there in the teammate adjustment portion.
That's pretty decent correlation for Jerebko/Kobe

[Crow]

Carmelo
AIR plus 4.7
BPM plus 1.4
RPM plus 2.8
raw on/off -13

[j_oxford]

Carmelo only playing half the season helped his AIR.

[j_oxford]

Realized I never posted this formula for win% online:
Win% = .5 +.03(FFR)

where FFR = .004+1.47(eFG%-LG.eFG%)+1.52(LG.TOV%-TOV%)+.43(OREB%-LG.OREB%)+.32(FT/FGA-LG.FT/FGA)+1.42(LG.eFG%-O.eFG%)+1.4(O.TOV%-LGTOV%)+.44(DREB%-LG.DREB%)+.3(LG.FT/FGA%-O.FT/FGA)

The linear regression I ran has FFR as an input and win% as the output. Adjusted R-squared is .946

[j_oxford]

Although I find it interesting that the equation is different depending on the era. I ran a regression from 76-83 and the .03 x coefficient actually gets bumped from .0305 in current era to .035 (.04) for late 70's - early 80's. This prompted me to look at the standard deviation for both sets of data tested, and in the 70-80's data the standard deviation is 3.56, while the more current data is 4.61 for 2000-2005 and 4.92 from 2010-2014. To me this is proof that during the regular season, there is much less parity now than back in the 70-80's. Or another way to look at it is that the elite teams are more dominant now than they were back then.

1991 seems like the year when the transition to elite regular season teams begins. The most dominant team from 1974-1990 was the 1987 Lakers at +9 FFR, and that rating is bested in 91, 92, 94, 96, 97 (by two teams), 98, 2000, 01, 05 (by two teams), 08, 09, 2013 (by two teams), and 2015.

Any explanations for the lack of dominant regular season teams from 74-90?

[Crow]

Possible influences:

Last rounds of league expansion may be a factor.

David Stern may be a big factor. Between late 80s and early nineties lots changed in NBA averages.

90-91, first time 3ptas over 7 per game. Up to 15 in 5 yrs.
86-87, first time 3pt fg% over 30%. Up to 36% in a few yrs (rim tightness/ uniformity?)
89-90, last time teams avg'd 28 ftas.
93-94, last time teams avg'd 16 TOs.
93-94, last time teams avg'd more than 8.5 steals.
91-92, last time teams avg'd 14 ORs.
92-93, last time avg. pace was above 94.

Stern thru rules and management of officiating had impact on some or many of these trends.
Changes in coaching and athletes play a role too.

Did home court win% change? Yes, big time. http://espn.go.com/nba/story/_/id/12241 ... ge-decline

When did teams go to charter jets generally?

More active medical staffs?

Superstar calls?

When was the first modern era salary cap? 84-85. 89-90 big jump to over 10 mil.

Expanded TV coverage? (did best / big market teans win more?) Local cable revenue not shared?
Sports center? The beginning of Jordan era NBA culture? Mega mega stars? Change in frequency of chuckers? (More than twice as many guys over 15, 18 fgas per game per team in 1978-9 than last season. Median efg% about the same but distribution shape may vary beyond that measure.)

Just throwing out possibilities.

74-90, more player partying and / or saving it for the playoffs?

Bigger & bigger bucks coaxing more of the talent edge out of players hungry for the rapidly expanding available dollars?

More and better video work and the slow rise of analytics by best teams?

Exlansion in underclassmen? Expanded roster sizes?

Rise in big salaries for top coaches?

More emphasis on defense by top teams?

Less white players?

Change in pay level / quality of officiating?

Change in rate of game fixing influence?

http://www.sloansportsconference.com/?p=12728
Lots more parity literature to check for clues.

Shorter shot clock?

[Crow]

1 point of AIR on average yields how much FFR?
How much does playing stat "style" / type affect this conversion rate??? Avg. cost per AIR pt and by type?

If you look at WARP player clusters (Bradford Doolittle's recent ESPN articles) what are their avg conversion rates, value ratios? How much does AIR vary from WARP in player ratings / tiering? Which explains recent season results better? Throw in RPM / RAPM, other metrics.

[j_oxford]

Superstar calls?

When was the first modern era salary cap? 84-85. 89-90 big jump to over 10 mil.

74-90, more player partying and / or saving it for the playoffs?

Bigger & bigger bucks coaxing more of the talent edge out of players hungry for the rapidly expanding available dollars?

More emphasis on defense by top teams?

These all seem like possibilities.

Foul rates have gone down, but limiting foul calls for non-superstars artificially creates superstar calls.

Salary Cap jump seems like a big part of it. Players see just how big the "pie" is getting, and they are willing to play hard every night to get a bigger slice of the pie.