You rock! This is very helpful, though as you point out it's difficult to separate higher death rates from the fact that certain driver types are drawn to certa

You rock! This is very helpful, though as you point out it's difficult to separate higher death rates from the fact that certain driver types are drawn to certain cars. One notable example: Honda Civic 49 (29-69) 47 (28-65) 2019-20 Honda Civic hatchback 78 (26-130) 20 (0-44) 2019-20 I'd like to know what's happening with the data there that a small tweak in the model with almost no structural changes has a higher death rate. Maybe down to a lower sample (fewer sales of the hatchback) of more dangerous drivers