Discussion about this post

User's avatar
Samuel Garfield's avatar

Been working in the field of analytics myself for over a decade.

The worst misapplication of it is folks who think that everything can essentially be boiled down to rational statistical analysis or that a good model applies in every situation. To use a model correctly you have to understand what the model accounts for and use your human judgment to decide when to ignore it (or tweak its advice).

If you don’t you risk running things like a D&D character with 20 INT and 1 WIS: the kind of person who analyzes a vampire using medical knowledge and recommends a blood transfusion to cure its “severe anemia” or some shit

PNWRider's avatar

This very much the old Green Bay model. Draft a QB late they may pop. They either become a starter. ( Love or Hassleback ) then you can either keep them or try them

I don’t think Darnold is going anywhere anytime soon.

43 more comments...

No posts

Ready for more?