>>2048781>flying is statistically the safest form of travel on a per-mile basis But that doesn't mean anything. Of course it looks good per-mile, since nobody takes a plane to go 100 miles.
The true value, the one which shows actual lethality, is fatalities per incident rate. Even if you dilute the pool by adding all incidents (ex. flap stuck in take-off trim), not just crashes, even if you remove fatalities from non-crash-related causes (ex. people who died from stroke or heart attack while flying), it will still look bad.
The important part isn't really to not have incidents, it's not to fucking die in them. People get t-boned all the time, but they only die if a) their car was old and lacked side airbags; b) the other driver was a drunkard blowing past a red at 70 mph.