I think the jury is still out since these cars are deployed in very specific, limited environments. In my opinion you need to compare only the areas they are dr

I think the jury is still out since these cars are deployed in very specific, limited environments. In my opinion you need to compare only the areas they are driving in vs the accidents in only those areas. If they don't use freeways or specific streets, or drive in the rain, than those needed to be excluded in the data. There are other factors like if they make unintended stops, cause traffic, etc. that don't get marked as accidents. I would love it if they all worked and saved lives but I'm not yet sold on this.