> We are actually taught stopping distances (inc. reaction times) in the UK, for our theory tests.
That's admirable, but I think it would be better to know why those distances are true, i.e. the velocity-squared rule. The idea is very simple, certainly simpler than memorizing a bunch of distances or even noticing the overall relationship between speed and distance.
> They know the facts and choose to ignore them.
There's a substantial literature about how the teenage brain is wired differently than the brains of those both older and younger. Giving them a car only exacerbates the problem. The irony is that a 10-year-old, and a 25-year-old, would both be likely safer drivers than a 15-year-old.
It would be interesting to see a graph of driver safety versus age -- ascending safety as age progresses from zero to about 12 (hypothetically, of course), then a really big dip in safety centered on 16, then another ascent, a long level spell between 25 and 60, then a tailing off after age 60, finally a terrible and irreversible decline after about 75.