[scarmig]

In fairness to the authors, legal fault is not sufficient to encompass how different vehicle types change accident risk. E.g., if someone breaks abruptly for no reason and that results in rear ending, the breaker wouldn't be "at fault," but they still share part of the actual blame. And I'm sure at least half of the accidents involving taxi cab drivers had them similarly wrong place, wrong time.

Where I object to the authors is the statistical spelunking and secret data sets that aren't publicly available but provided as a favor from NY government as a favor to the author. Are cabbies really 14x as safe as the average driver? It's impossible to say; we know nothing about the data they used to get that estimate. And some things strain credibility: what model do they suggest for Waymos being much safer than cabbies for all accidents (as supported by their own reported data), but suddenly getting much worse for serious injuries and fatal accidents, especially seeing as how all of the three accidents involving the Waymos they use involved a stationary or low speed Waymo where the Waymo is pretty clearly not at fault?