Hacker News Reader

Comment by dsfyu404ed | Hacker News Reader

>ings have gotten much safer the past couple decades so that number is somewhere around 0.75% or lower now.

On a statistical level, sure. On an individual level you can improve your person chance for survival greatly by not driving drunk and putting down your phone, the shitposting can wait. Also if you've made it past your teens your odds improve greatly.

>I want to say being injured in a car accident is somewhere around 20-30% lifetime chance.

That sounds absurdly high. 25% is 1 in 4. Assuming odds were equal or higher in the past then the average person has one grandparent that's been injured in a car crash.

Whatever the rates for injury/death are they're probably not evenly distributed among the population.

View on HN

Actually, OP's figures are pretty much backed up by the figures you get when googling this question.

Chances to die of a car crash/ any kind of vehicle crash is about 1 in 100 according to here [0].

I can't find something similar straight away for injuries, but car accidents are ubiquitous and 2m people are injured every year (vs 30k fatalities) [1; cached]. So something above >20% seems plausible.

[0]: https://www.iii.org/fact-statistic/facts-statistics-mortalit... [1]: https://webcache.googleusercontent.com/search?q=cache:dnqo-l...

Some UK figures here:

https://www.theguardian.com/news/datablog/2011/oct/28/mortal...

mortality due to transport accidents / total deaths = ~2000 / ~500000 ~= 1/250 or 0.4%

> On an individual level you can improve your person chance for survival greatly by not driving drunk and putting down your phone

Let me just add sleep-deprived driving to the mix

Wow so 4.6 million people seriously injured in car accidents in the US in 2016. Given an 80 year average lifespan you are statistically at 110% risk of serious lifetime injury using the most basic statistics.

4.6M / 330M = yearly risk 1.39% x80 years = 111.5% chance of serious injury

I'm sure that includes things like whiplash though.

Probably more correct to do one minus the chance of no injuries.

So the chance of no injury is (1 - 0.0139) ^ 80 = 32.6%, making the chance of serious injury 67.4% over 80 years.

That’s not how probability works - the probability of getting heads when you flip a coin twice is not 2x50 = 100%.

The rates really are that high.

Funny that you mention grandparents: if you graph the crash rate by age, you get a deep U-shape with the very youngest and oldest having a much higher rate.

Doesn't pretty much any crash at a decent speed (even if the airbag doesn't go off) cause an injury like a bruise from the seatbelt or temporary sore muscles?

Yeah, but so does closing a door on your hand. I would hope that whatever we're considering an "injury" is something severe enough to usually require medical treatment (kind of like when someone say "mass shooting" you don't usually think of a double murder suicide). Bruising except in the most extreme cases does not require treatment. Similarly for whiplash, if it's less severe than having a stiff neck in the morning like many people get from sleeping on a crappy air mattress or couch bed I wouldn't count it I think that including such minor injuries is misleading.

Whatever we count as an "injury" should probably be more severe than the little mishaps we have with our bodies from going through life.

>Whatever we count as an "injury" should probably be more severe than the little mishaps we have with our bodies from going through life.

The fact that this is apparently an unacceptable opinion here is mind boggling to me

If we're basing out stats on unimportant injuries (minor bruises and stiff necks that go away in a day) then we're measuring something that we don't care about, injuries that do not meaningfully affect people. What we should care about is injuries that have appreciable recovery time (serious bruising, having a sore neck for multiple days, etc, etc) because those negatively affect people enough to be worth caring about. If we're including injuries we don't really care about in our numbers then something is wrong with our numbers.

I dare someone to actually tell me why you think we should not distinguish injuries that are so minor as to not appreciably affect people from those that do seriously affect people. Tracking minor and significant injuries in the same category misleads anyone who has to reason about or make decisions based on those statistics.

What about not driving late when bars are open or closing and people get behind the wheel tired and buzzed?

This is true but there are also less people on the road.. That being said there are drunk drivers at anytime on the road.

Wondering if those numbers contain a bit of an ambulance chaser skew in there?

I doubt it. Among people I know, I've known 3 people injured as pedestrians, one person who was killed as a pedestrian, and a couple of months ago I was injured as a pedestrian.

Cars are ridiculously dangerous. Something needs to be done about it. Good thing there's Waymo and other companies like them.

Holy crap, where do you live, just outside Thunderdome?

One injury & the friend who passed away were in Sydney. Another friend was injured crossing the road in Bangkok. I was hit by a taxi whilst crossing the road (legally, with a walk light) in Singapore.

In my experience, humans on average are just quite bad at paying attention. The only reason autonomous vehicles didn't surpass our driving abilities years ago is because we tend to make our roads unpredictable (road work, human behaviours, etc.) and it turns out that humans are able to quickly comprehend those situations.

So, oddly enough you do sort of live outside the Thunderdome presuming you're in Sydney.

Reply on news.ycombinator.com