Also some of this stuff is clearly suffering from the tiny sample sizes, such as:
no correlation between 3p% and wins.. I think it's fairly clear a higher 3p% helps us win regardless of correlations this season
ditto the lack of correlation between opponent turnovers and wins
that said, yes we need more paint shots and fewer midrange shots (and maybe fewer threes, but maybe not). If the paint shots are there we should prioritize them. But if we do that and the opponent packs the paint, then we will need to shoot the open threes that would result
What we need to stop doing is hanging around the perimeter for 15 secs before awkward drives, stepback midrange shots, desperation threes at the end of the shot clock, etc