Trunk-or-treat is fucking sad, but necessary in some areas.

Is it bad I assumed the y-axis was like hundreds or thousands?

A textbook example of how to lie with statistics.

In the entire US there are around 10 child pedestrian deaths on an average night. On Halloween there are only around 14 deaths despite thousands more children on the street. This means the actual risk per child is exponentially lower on Halloween night.

So the reality is the opposite of what's being claimed. Halloween is perhaps the least dangerous day of the year for child pedestrians.

Its also the busiest night of the year for pizza delivery and probably any other food delivery too.

Maybe I’m getting old but I think I had this opinion always: I don’t understand why drivers don’t, eh, slow down?

*in the US

Also more pedestrians=more pedestrian accidents, like duh

Base rate fallacy?

Given it is almost certainly the day with the highest concentration of child pedestrians by a wide margin I can only imagine.

Still, sounds like a great reason to be extra special careful on the road on the 31st.

The only day of the year children are allowed to go on foot beyond the end of their home driveway.

I'm surprise it's only a few times the background level. Historically, when more children went out, it was probably higher.

Even without the stats, I would have assumed as much.
The more you have people and cars in the same space, the more they will collide, and the more people will be killed. I expect it's not just kids but the same for adults too.

Double points?