APBRmetrics

Several articles have suggested that the Warriors and the Warriors in the finals vindicated Mike D'Antoni and 7 seconds or less. Steve Kerr suggested that. But how much did it?

Warriors in finals played at a pace for a bit less than 91 poss./game. That was most for finals for 13 years but still 3 less than reg. season league average and about 7.5 less than reg. season team average.
2006 Suns at their furthest advancement were held to a similar last series pace. 5 less than regular season team average. 06 Suns didn't drop off as much but neither really succeeded at playing fast, just faster than late playoff norm.

2006 Suns shot in first ten seconds 44% of time in regular season. 2015 Warriors 45%. How much in final playoff series? Anyone with play by play query skills want to check? With pace down it is likely this is done too, but would be interesting to see.

On 3pt rate, 06 Suns were at 29% in regular season, just 23% in their final playoff series. 2015 Warriors were at 31% regular season but 37% (corrected) in finals. But a team offensive efficiency of 107 for the Warriors does not seem enough to declare the offense a smashing success to me. The 99 Defensive Efficiency (or weak Cavs offense) seems like the bigger reason.

Will we see faster / higher 3pt rate in finals? Probably. But regularly and as the main reason for victory? I dunno.

Could you compare vs. historical average of NBA finals? That's the more interesting comparison--the finals has an absurd intensity level, and that will always favor the defense.

I misread / computed Warriors' 3pt rate late last night. Corrected post above.


The 9 yr finals avg pace is around 87.5 2015 finals Warriors played about 4% faster.

9yr period selected because I consider the 2007 Spurs the dawn of the current era.

Avg. 3pt rate of titlist is about 29%. 2007 Spurs 27%, 08 Celtics pushed it to 33%. Then in their final series outs the next 3 seasons they fell back to 21-24% each time. Did they lose focus or did the defense deny? From 2009 til now 3pt rate almost always climbing back up (with exception of 2013, Spurs containment?) but only passed 08 Celtics this season. Another defensive contribution / failure in part, but Warriors get credit for pushing about 27% above the 9 yr avg. (and about the same avg. for previous 4 yrs).

During the season, Warriors held opponents to .337 eFG% -- 5th lowest in the league and .013 below league avg.
They also allowed 5th lowest 3FGA/FGA at .248 -- .020 below avg.
Bulls and Pels were better on both.

Cavs' RS/PO numbers are wack, but they weren't really the same team.
It looked to me like the GSW were pushing the pace as much as possible, while the Cavs were slowing it as much as possible. This led to a number of Cle 24-second violations, which led to quite the disparity in team/individual turnovers.

http://www.basketball-reference.com/pla ... riors.html
Cle "won" the TO% battle, in the Four Factors box, 11.1 to 11.9
This is based on the sum of individual TO.

However, b-r.com also informs me that Cle had 87 total TO, vs 78 by GSW: The Cavs apparently had 14 'team TO' in the Finals, to just 2 for GSW.

Using individual TO in the Pace calculation.
Poss = FGA + 0.4*FTA - 1.07*(ORb + OppDRb - TO))*FGX + TO
... we get Pace of 90.9 for GSW and 90.4 for Cle

Using total team TO, we have Cle 92.6 - 91.2 GS (poss/48 min)

In order for both teams to have had roughly equal number of possessions, I'd guess GSW had more of those team rebounds, in which Draymond and Tristan are fighting for the ball, and it goes out of bounds, awarded to one team or the other.

With both TO and Reb being systematically under-counted, the actual Pace is always going to be higher than what shows up by conventional calculation.
With the aforementioned greater intensity of playoff competition, does it not seem likely that contested/uncontrolled team rebounds, and/or team TO, might also increase in frequency? And that part or most of Pace drop-off might be a statistical quirk, rather than a real phenomenon?

With how quarters can unevenly end but evenly start by rule, team possessions don't have to equal. Some small part of the pace decline may involve team turnovers or team rebounds but the biggest share of a 7.5 possession decline on pace for GSW is beyond these factors.

Well, I went through the pbp for Game 6, and I count 100 possessions for each team.
b-r.com says the pace was 96.9
The Cavs had the ball for 26:19, or 15.8 sec/poss
GSW possessions totaled 21:41, or 13.0 per
Each team had a 1-2 sec poss at the end of a quarter.

On avg, a possession was 14.4 sec; but Cle was about 9% slower, while GS was 11% quicker than that.
For the Cavs, I count 5 OReb by team and 3 DReb by team.
For GS, just 1 of each.

Some part of the possession estimation formula must be estimating these 'uncounted' rebounds.