APBRmetrics

(Note: revised after I recognized I had gone too far, too quickly)


I checked the correlation of team mid-range shot frequency (as % of all shots) and other stats for this regular season:


Age -0.13
Proj Team Wins -0.35
ORtg -0.57
DRtg -0.06
Pace -0.42
eFG% -0.58
TOV% -0.14
ORB% 0.09
FT/FGA -0.34

Age is simple team roster average and thus not properly weighted for playing time but this could be re-run if anyone has a quick and accurate minutes weighted average team age dataset handy. [see correction below]

Projected team wins, offensive rating, eFG% and FT/FGA all decline on average as the % of team shots taken from mid-range shot frequency increases and vice versa and some of the correlations are what I would call fairly strong (at least strong enough to be worth being aware of). Pace declines with more mid-range shots as well. TOV% is very modestly negatively correlated. Defensive rating and OR% get slightly "better" as mid-range shot frequency increases.

I need to compare this information to several other correlations (to inside shots and 3 point shots) but this much seems like an overall at least mild argument for less rather than more frequency of team mid-range shots. Of the 18-20 teams I know enough about to call (generously perhaps) "analytics"-equipped, about two thirds have higher than league average team mid-range shot frequency. Maybe they have a justifiable rationale for this based on other data and analysis that I am not currently aware of. I will look into it more.

b-r.com has teams' Age in the miscellaneous stats table, on season summary page.

Mike G wrote:b-r.com has teams' Age in the miscellaneous stats table, on season summary page.

That is the source of the data I used, but I assumed it was simple age unweighted by minutes played, as that is what I commonly see around. I just checked it for two teams and it appears in fact to be the minutes weighted age. Good to get that clarified.

Inside shots are fairly modestly correlated (almost .4) with pace and offensive rebounding (compared to the correlations of the other stats considered). The rest are very low.

The 3 pt shooting correlations are pretty close to the mirror image of the mid-range correlations.

The string of correlations for inside shots are generally much milder than for these other two distances. The mid-range - 3 pt decision making of teams is only one of many factors affecting these roll-up stats but the general strategic lesson seems to favor more 3 point shots and less mid-rangers. Maybe there is a supportable rationale for the about 2/3rds of teams with analytic staff (to my knowledge) which have mid-range frequencies above the league average based on other / perhaps deeper analysis (such as what works against strong team playoff defense?). I currently wonder about the wisdom of this basic pattern but will try to look for more understanding.