We blow more leads because we shoot the most 3s. 1-18 from 3 yesterday after starting 10-20. If they shoot a paltry 5-18 we aren’t having this conversation.
My guess is that from an expected points per possession perspective, shooting lots of good 3s maximizes that. But there is probably more variance in the outcome - so you are going to have stinkers.
I think this is why most Cs fans criticize Joe when we blow leads because when you are up big you should change your approach to minimize that variance - eg go away from so many 3s. Drive and get to the line. But they don’t seem to change their system.
I've posted this before, but it bears repeating that variance works oppositely to this. It's a reasonable mistake to make, because lots of basketball writers and sports-talk types will use 'variance' as a synonym for 'low-percentage shot, when actually it means the exact opposite.
For sake of simplicity:
Suppose you have a team that shoots midrange shots on every possession and hits them 50% of the time.
And you have another team that shoots threes on every possession and hits them 33% of the time.
Even though both teams are scoring the same points per possession, the team shooting the midrange shots has a
higher variance, because the variance we're actually measuring is the expected points per possession compared to the total shot attempts. In other words, you don't get more points for a higher FG%, but you do get more points for shooting a three.
I used this example last time and it's still pretty good: check out
the game we played against the Wizards at the beginning of February - we absolutely outplayed them on both ends of the floor, but it was 'only' a four-point game because they shot (and hit) way more threes.