And also while you're there, if no car ever in the history of your app goes down the road at the speed limit ever it's a good indication you'll never be able to ever do it at that speed. e.g. small narrow single lane country roads which are only theoretically 60mph roads.

▲

xp84 5 days ago | parent [-]

Imagine the data they must have on the speeds people actually drive on every mile of every road, they’d easily be able to warn you not that you’re “over the speed limit” as in driving 70 on the freeway, but more usefully, if you’re in the top X percentile of speeds usually or even currently driven, which actually is a decent measure of unsafe was and would also be a great predictor of likelihood to get a ticket.

▲

yonatan8070 5 days ago | parent | next [-]

I saw this video recently where the author set up a camera to record sections of highways and measure the speeds of drivers, and make cool graphs out of it.

https://youtu.be/TYTaNsnBjcw

▲

sebastiennight 4 days ago | parent | prev | next [-]

With the data you're mentioning, it's probably just as easy to build an accident predictor model as well.

	▲	hsbauauvhabzb 4 days ago \| parent [-]
		Or sell the data to third parties instead because that wouldn’t bring profit.

▲

Flimm 4 days ago | parent | prev [-]

I share your intuition that your likelihood of getting a ticket is related to the speed of other vehicles. Presumably, police choose to prioritise ticketing the worst offenders when there are too many offenders to handle.

But I don't share your intuition that safety is also relative in that way. If you're driving dangerously (too fast, or while drunk), you're driving dangerously, even if everyone else is driving dangerously too. If you're in a country where nobody wears a seatbelt, it's still prudent to wear a seatbelt, just as much as in a country where that is the norm. I don't think Google Maps should encourage people to drive as dangerously as everyone else. Quite the opposite!

	▲	xp84 a day ago \| parent \| next [-]
		"dangerously" and "fast" are not synonyms (while drunkess obviously is dangerous regardless). The 280 in San Mateo County is designed for 80MPH, with banked turns, gentle curves, etc. The speed limit is of course 65. If most people are driving 80-85, and you enter that highway and drive 60, which is a perfectly reasonable speed according to the speed limit, you are much more likely to cause an accident than if you just drive the speed of the people around you. It's fine to point out that many people are terrible drivers and that a given crash that happens is more dangerous at a higher speed and if everybody were to drive under 60 at all times we'd all be safer, but clearly that will never happen unless we install a totalitarian government, put governors on all cars and give prison time for disabling them, with enforcers stationed everywhere to monitor. But no democracy would vote for that, so I don't think it's worth spending much mental energy on such hypotheticals. > encourage I think just the opposite: My feature would encourage people like me, who drive a "fast car" and can occasionally accidentally go too fast if a road is especially uncongested, to slow down to the speed that is customary or that others are driving, by reminding me that I could get a ticket and that driving faster than everyone else is dangerous.
	▲	hsbauauvhabzb 4 days ago \| parent \| prev [-]
		Relative speed is important though. Where I live it’s completely legal to drive half the speed limit, and people often drive above it. People driving a consistent speed would reduce lane changes, breakaway traffic etc. imo speed consistency is what’s valuable, not preventing upper outliers.