I find it ridiculous that no more than 3 miles every week produces more problems for riding a bike
The thing is what you actually want to know is what the likely hood of actually being in an collision if I go for a ride down a road. This is a very hard statistic to generate in a meaningful manner. I come up with the time between collisions per length of road surface. But I'm still not convinced that that is any more meaningful a statistic than collisions per mile traveled.The stats will be based on the number of miles travelled per vehicle for a give stretch of road. Therefore the denser the traffic the lower the incident per vehicle/mile over that piece of tarmac.