> You really need to look at incident rates rather than totals when making comparisons across categories like this.

Depends on what you want to figure out. For finding out the comparative danger of an individual doing activity A versus activity B, sure. But if the group doing something exceedingly dangerous is quite small, it might not warrant an intervention as far as public health etc is concerned. Scale matters (sometimes).

On the flip side, you can of course argue that the individuals who just happen to be in proximity of the dangerous activity are being unfairly exposed to risk they haven't consented to (if the activity poses a danger to bystanders, that is). We don't let people do whatever they want just because the people it affects are few, either.

Really I should have just said that the rate of injuries in cars is irrelevant to safety considerations with skateboards. If there are dangers in skateboarding that can be reasonably reduced with better safety, that would be a good thing to do. Likewise with cars. If one is more dangerous than the other, it's not a logical reason to focus solely on that one.

curious how cars have become safer for pedestrians? i've anecdotally absorbed the opposite, but maybe i'm a biased bicyclist. Cars have only become increasingly massive; an arms race to "protect your family!" meaning the passengers.

I believe EU regulations lead the way here. Basically, part of crash testing now involves impacts with pedestrians. My understanding is that the primary effect is on the front-end shape of vehicles. When striking a pedestrian, you want them to roll up onto the hood, into the windshield, to spread out the impact as much as possible (in area and time). More recently, systems like automated emergency braking have helped too.

Here, this wikipedia page gives an overview: https://en.wikipedia.org/wiki/Pedestrian_safety_through_vehi...