There's a lot of interesting conclusions that can be found from play-by-play time series data (the NCAA dataset also on BigQuery has data since 2009). Here's a quick new visualization of the Distribution of Basketball Home Team Points at Each Minute of NCAA Games, by Season: https://i.imgur.com/8Ar0J2W.png
nice! would have loved a little more detail in your conclusions section, but it's neat to see how my basketball intuition matches/mismatches your data.
one analysis that would be cool is an effectiveness chart of the mix of shots for each team/game. i suspect made percentage is correlated to the mix of shots (3's vs. layups vs. mid-range for example).
if you're a guard-heavy 3 shooting team, then defenders can adjust to gaurding the 3 to force lower-effectiveness mid-range shots. the question is whether that's an effective strategy, because at some point, the defense will have to start gaurding for the drive too, which loosens up the 3's again.
There's a lot of interesting conclusions that can be found from play-by-play time series data (the NCAA dataset also on BigQuery has data since 2009). Here's a quick new visualization of the Distribution of Basketball Home Team Points at Each Minute of NCAA Games, by Season: https://i.imgur.com/8Ar0J2W.png