The more you get to know a concept the more subtle it appears to you and the more approximate and incomplete everyone else’s casual use of the concept starts to appear.

See as an empiricist I tend to think many economists misunderstand what OLS does. On the other hand, I think I have a good understanding of Nash Equilibrium. You, however, as a theorist think that most people don’t properly understand NE but you think they do understand OLS.

What counts as a correct understanding of OLS??

Especially when going from a single binary treatment to multiple or continuous treatments

Never assume that because you know a proposition is true in a simple case (e.g. regression beta = ATE with unconfoundedness & saturated model & constant TE, LATE theorem with binary D and Z, DiD with 2 periods = TWFE…) that it is true in general, when relaxing any of these assumptions.