Nick Patterson, formerly a senior statistician at Rentech, was on a podcast saying that their most important tool was a simple regression.
I suppose he could’ve been lying, but the believable accounts suggest that the majority of their magic is simply having smart people use simple tools very, very well.
I think this is right - my experience as JS and similar places is that really understanding your data-generating process and the nitty-gritty assumptions of your data/simple model makes a huge difference
E.g knowing what's causing missing values in your data and what implications various fixes on that might have on bias in your linear regressor is probably way way more valuable than fitting some shiny non-linear toy
I suppose he could’ve been lying, but the believable accounts suggest that the majority of their magic is simply having smart people use simple tools very, very well.