Yeah, though that's only four sigmas, and there are 30 teams in the league ...boooeee wrote:...
Another callout - Possessions off of turnovers (1.12 points/poss) are significantly more efficient than possessions off of made shots (1.00 points/poss) or defensive rebounds (points/poss). But the Brooklyn Nets haven't got the memo. Their efficiency off of turnovers is not better than efficiency off of made shots.
Adjusting pace
Re: Adjusting pace
Re: Adjusting pace
Not sure I follow (or maybe I'm missing a statistics joke here).NateTG wrote:Yeah, though that's only four sigmas, and there are 30 teams in the league ...boooeee wrote:...
Another callout - Possessions off of turnovers (1.12 points/poss) are significantly more efficient than possessions off of made shots (1.00 points/poss) or defensive rebounds (points/poss). But the Brooklyn Nets haven't got the memo. Their efficiency off of turnovers is not better than efficiency off of made shots.
Re: Adjusting pace
Let's suppose that every team has a true rate of 1.12 points per turnover possession in the form of 44% 0 points and 56% 2 points, and has had about 1000 turnover possessions this year. Then it wouldn't be that crazy for the 30 teams in the league to range from 1.03 to 1.21 points per possession just from chance.
Re: Adjusting pace
boooeee, a few replies to your very interesting blog post.
(1) I would still recommend breaking out the half-court component from the fast-break component for all possession origins. In "theory", one should expect the efficiency and shot selection profiles of "seven seconds and more" offense to be the same, no matter the origin. (With "four to six seconds" offense more likely to differ in its frequency and perhaps efficiency.)
(2) Further GSW decomposition would make on or more nice blog posts.
(a) As noted, they are a stand-out fast-breaking team, but what is particularly remarkable is their "seven to thirteen second" offense. How is it that they get so many more shots at this point in the clock, relative to other teams? In the half-court shot clock, by definition, there is no man advantage, and early in it, there should be little time for the offense to break the defense down for higher quality shots.
Accordingly, first, it would be interesting to know to what extent their overall margin of offensive superiority is dependent on the different shot clock segments fast-breaks vs. early half court vs. all other times, breaking it down in terms of efficiency and frequency.
(b) And then a comparison with last year's data (assuming you have it). Essentially all of the GSW's improvement has been in terms of offensive efficiency (and what defensive improvement has occurred, likely owes to the more favorable defensive possession origins that the offense has delivered). To what extent is the GSW's offensive improvements explained by the changes in shot selection across the shot clock vs. an upward shift in the time vs. efficiency plot?
(3) Finally, a general, geeky point of great personal curiosity. The plot of "Efficiency by Possession Length" for the League Average is, well... "wrong". From a theoretical perspective, the plot of the efficiency of half-court offenses, in expectation, is concave and monotonically decreasing. Instead, from 26 seconds on (what I eyeball to be about 8% of total possession - what is a non-trivial fraction of the total) efficiency rises! This is very strange. Might it be an artifact of composition across teams? If not, it implies a pretty significant inefficiency in the "average" NBA offense on offensive rebound-originating possessions.
(1) I would still recommend breaking out the half-court component from the fast-break component for all possession origins. In "theory", one should expect the efficiency and shot selection profiles of "seven seconds and more" offense to be the same, no matter the origin. (With "four to six seconds" offense more likely to differ in its frequency and perhaps efficiency.)
(2) Further GSW decomposition would make on or more nice blog posts.
(a) As noted, they are a stand-out fast-breaking team, but what is particularly remarkable is their "seven to thirteen second" offense. How is it that they get so many more shots at this point in the clock, relative to other teams? In the half-court shot clock, by definition, there is no man advantage, and early in it, there should be little time for the offense to break the defense down for higher quality shots.
Accordingly, first, it would be interesting to know to what extent their overall margin of offensive superiority is dependent on the different shot clock segments fast-breaks vs. early half court vs. all other times, breaking it down in terms of efficiency and frequency.
(b) And then a comparison with last year's data (assuming you have it). Essentially all of the GSW's improvement has been in terms of offensive efficiency (and what defensive improvement has occurred, likely owes to the more favorable defensive possession origins that the offense has delivered). To what extent is the GSW's offensive improvements explained by the changes in shot selection across the shot clock vs. an upward shift in the time vs. efficiency plot?
(3) Finally, a general, geeky point of great personal curiosity. The plot of "Efficiency by Possession Length" for the League Average is, well... "wrong". From a theoretical perspective, the plot of the efficiency of half-court offenses, in expectation, is concave and monotonically decreasing. Instead, from 26 seconds on (what I eyeball to be about 8% of total possession - what is a non-trivial fraction of the total) efficiency rises! This is very strange. Might it be an artifact of composition across teams? If not, it implies a pretty significant inefficiency in the "average" NBA offense on offensive rebound-originating possessions.
Re: Adjusting pace
I'm not seeing how you get from SportsVu time of possession to seconds per possession: http://stats.nba.com/tracking/#!/team/p ... TOP&dir=-1italia13calcio wrote:Real easy way to get an estimate of avg. seconds per possession is to just divide teams time of possession per game from the SportVu data by a team's paceCrow wrote:Some of these points have been raised before but some are new and this is perhaps more in sum than I can recall on the topic than anywhere else. Plenty of ways to break it down. At minimum, it would be nice to see some database with offensive and defensive pace in terms of avg. seconds per possessions. Hope to see it, here, at Nylon Calculus or somewhere. An example of something that has been possible for awhile but just hasn't been pushed thru yet.
e.g. Warriors = 17.5 Time of Possession (min)
I assume once you find the offensive time of possession, you are then left with the seconds per possession on defense...