News for hackers. Stuff that matters.
Re:Test Can Tell How Likely a Young Driver Is to Crash by noqc on Saturday January 13, 2024 @10:31AM (#38980816)
You cannot determine how likely something is to occur in a universal sense. To be accurate, the headline should read "young drivers who fail test 20% more likely to be in crash than those who pass". As it stands, this reads like an advertisement for a test you can give your teen to decide whether or not they should be allowed to drive. A 20% improvement in predictive power is not worth the money, since it probably shouldn't change your decision.
Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Saturday January 13, 2024 @12:48PM (#38982495)
One possible difference if this is deployed in real life: teens who score well may become overconfident, or at least not as cautious as they should be.

I don’t think the converse is as likely to manifest differently than in the original research, because teens who did poorly would likely already know this (from having hit or nearly hit things). Drivers who did well would not know how close their score was to the ceiling; merely not hitting stuff wouldn’t necessarily lead to overconfidence.

Re:Test Can Tell How Likely a Young Driver Is to Crash by swader999 on Saturday January 13, 2024 @10:40AM (#38980911)
Kids receiving nice cars as gifts from parents on their sixteenth birthday was a reliable crash predictor for my classmates growing up. The ones that worked hard to contribute to the cost and often bought second hand seemed to have a lot less accidents.
Re:Test Can Tell How Likely a Young Driver Is to Crash by kjkjadksj on Saturday January 13, 2024 @10:45AM (#38980966)
The kids driving a lease would not only total their cars routinely but get a new one immediately after. No consequences.
Re:Test Can Tell How Likely a Young Driver Is to Crash by analog31 on Saturday January 13, 2024 @11:05AM (#38981172)
My mom taught high school. She noticed that the kids with cars also tended to have lower grades.
Re:Test Can Tell How Likely a Young Driver Is to Crash by rr808 on Saturday January 13, 2024 @11:45AM (#38981667)
I'd love to know if long term earnings are affected. I know a bunch of people who can't drive and it really seems to restrict what you can do in life.
Re:Test Can Tell How Likely a Young Driver Is to Crash by bossyTeacher on Saturday January 13, 2024 @12:10PM (#38981975)
> I know a bunch of people who can't drive and it really seems to restrict what you can do in life

I would imagine that you are a US citizen. Car dependency is more of an American problem imo. The fact that you depend (yes, dependency not power or freedom) on an expensive piece of metal to go anywhere is crazy. Imagine if medieval peasants had to have a horse to go anywhere from work to the food market

Re:Test Can Tell How Likely a Young Driver Is to Crash by rr808 on Sunday January 14, 2024 @02:10PM (#38993408)
I am both American and EU citizen. Yes US is more car dependent but its the same everywhere. My mother can't drive and lives rurally, she rarely gets to visit her relatives unless someone drives her.

> Imagine if medieval peasants had to have a horse to go anywhere from work to the food market

Not sure where you're going with this. 2 of my grandfathers were farmers and they had horses to work the farm and deliver goods.

Re:Test Can Tell How Likely a Young Driver Is to Crash by bossyTeacher on Sunday January 14, 2024 @07:39PM (#38996027)
If you live in a rural area, it's understandable but if you live in a city it's not great. Re: where I am going to this, I already stated my point. Dependency on your own piece of metal for transportation purposes is shocking. Ideally, driving would be a choice and you would always have public transportation as a backup. But if as a society, you reach a point where not owning a piece of metal restricts your activities in your area, it sounds like an awful issue with serious socioeconomic implications.
Re:Test Can Tell How Likely a Young Driver Is to Crash by Symbiote on Saturday January 13, 2024 @01:46PM (#38983127)
It's not necessary to drive at age 16 though, learning by the end of college should be more-or-less equivalent.
Re:Test Can Tell How Likely a Young Driver Is to Crash by bossyTeacher on Saturday January 13, 2024 @12:07PM (#38981930)
Is this in US? Any thoughts on why is this the case?
Re:Test Can Tell How Likely a Young Driver Is to Crash by analog31 on Saturday January 13, 2024 @01:53PM (#38983212)
My mom thought it was a distraction -- driving around as a social activity was a thing. Also, supporting a car meant getting a part time job, which took time and energy away from studying.

It may depend on whether having a car is a means to an end, or an end unto itself. For instance, the college bound kids may have had aspirations other than having a car.

Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Saturday January 13, 2024 @12:51PM (#38982535)
Just my experience, but this was not at all what I saw growing up. It wasn’t even true among the kids with fancy/new cars where I went to school (public magnet school in CA). Were there some spoiled rich kids who didn’t work hard? Sure. But most of the wealthy kids were not slackers, and there wasn’t even a strong correlation.
Re:Test Can Tell How Likely a Young Driver Is to Crash by analog31 on Saturday January 13, 2024 @01:51PM (#38983186)
For reference, my high school was mostly blue collar.
Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Saturday January 13, 2024 @02:18PM (#38983495)
Thanks for the context. Mine was a mixed school, part magnet and part blue collar/immigrant.
Re:Test Can Tell How Likely a Young Driver Is to Crash by gruez on Saturday January 13, 2024 @10:22AM (#38980723)
>The kids who drove best ended up having a 10% lower-than-average crash risk. The ones who experienced major issues had an 11% higher-than-average chance of getting into a wreck.

That seems pretty weak in terms of predictive power.

Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Saturday January 13, 2024 @12:45PM (#38982452)
I would be interested to see the absolute numbers here, since the vast majority of drivers (even teens) don’t get into a crash.
Re:Test Can Tell How Likely a Young Driver Is to Crash by Symbiote on Saturday January 13, 2024 @10:22AM (#38980734)
A little towards this is included in the British driving test, like the example of the boy missing the car behind the trees.

The "hazard perception test" is a video showing driving, and the student must identify (by clicking) the likely hazards.

https://m.youtube.com/watch?v=SdQRkmdhwJs (I think, I can't watch it with sound right now.)

Re:Test Can Tell How Likely a Young Driver Is to Crash by hank808 on Saturday January 13, 2024 @11:54AM (#38981774)
If you want to share this with a friend or someone, copy and paste this into a message [thought it might save people a minute]:

Original Article: https://www.wsj.com/tech/personal-tech/this-test-can-tell-ho...

Archived Article: https://archive.is/FsmyG

Discussion: https://news.ycombinator.com/item?id=38980501

Re:Test Can Tell How Likely a Young Driver Is to Crash by whatever1 on Saturday January 13, 2024 @11:14AM (#38981276)
Being overly cautious can also cause crashes.

There are cases that if you hesitate (for example stopping because you think you cannot clear the intersection on time), you surely gonna cause accident.

My sibling was a much more careful and law abiding young driver compared to me. The crash tally says otherwise.

Re:Test Can Tell How Likely a Young Driver Is to Crash by Retric on Saturday January 13, 2024 @11:23AM (#38981397)
Not all crashes are equivalent. Being rear ended for stopping too soon has different risks than being T-Boned for running a read light.

Several driving behaviors reduce your risk of death while increase the risks you end up in an accident.

Re:Test Can Tell How Likely a Young Driver Is to Crash by jl6 on Saturday January 13, 2024 @11:35AM (#38981546)
A sudden stop is something the driver behind should be prepared to handle, so if there’s a crash, it’s caused by a lack of caution on the part of the driver behind, not overcaution of the braking driver.
Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Saturday January 13, 2024 @12:55PM (#38982584)
It is generally presumed that the driver who rear ends a stopping car is the one who is at fault. But this is a rebuttable presumption: the following driver can shift the responsibility if he can show that he was following at a safe distance and that the driver in front braked too aggressively/for no reason. For example, if you’re doing 70 on the freeway and the car in front of you screeches to a halt to avoid hitting a squirrel, you could prove this is unreasonable braking. If it’s a deer in the road, you’re going to have a harder time. It depends on the circumstances.
Re:Test Can Tell How Likely a Young Driver Is to Crash by mindslight on Saturday January 13, 2024 @02:18PM (#38983492)
> the car in front of you screeches to a halt to avoid hitting a squirrel, you could prove this is unreasonable braking

Uh, what? I agree with the rest of your comment, but avoiding killing a squirrel is a legitimate reason to brake hard.

I've rarely seen squirrels on controlled access roads where going 70 is the norm though (proper landscaping plus road noise keeps them away), so I'd say we're really talking about 30~50mph situations.

Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Saturday January 13, 2024 @02:48PM (#38983799)
The point was the speed. Slamming on the brakes at 70 to avoid killing a squirrel is unreasonable, according to most people, if the likely outcome is causing a huge pileup. People don’t slam on the brakes on the freeway unless it’s to avoid serious injury or damage.

Even at 50 I’d think that slamming on the brakes could be unreasonable, since it is pretty rare and would be unexpected by other drivers.

Re:Test Can Tell How Likely a Young Driver Is to Crash by Ferret7446 on Sunday January 14, 2024 @03:42AM (#38988623)
By definition, a safe following distance means you can brake in time if the car in front effectively turned into a giant concrete wall (e.g., if they swerved right in front of, or crashed into, a large immobile obstruction).

If the driver in front "braked too aggressively", then you are not at a safe following distance, sorry.

The safe following distance is much longer than most drivers realize.

Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Sunday January 14, 2024 @12:14PM (#38992201)
In many places, it is impossible to maintain this follow distance. If you leave this distance when driving on the freeway in California, for example, cars will repeatedly cut you off, shrinking your follow distance. It is literally not possible to maintain what you describe.
Re:Test Can Tell How Likely a Young Driver Is to Crash by Ferret7446 on Sunday January 14, 2024 @07:17PM (#38995883)
That is not what "impossible" means. It is perfectly possible to maintain this distance, which is based on your speed. You just have to be comfortable with getting cut off or slowing down.

(Note that I am not dictating what you should be doing, I am simply commenting that

1. if you maintain a safe follow distance, sudden braking isn't an issue

2. if sudden braking is an issue, it is your fault

3. it is perfectly possible to maintain safe follow distance, even if it irritates you.

Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Sunday January 14, 2024 @09:03PM (#38996612)
If there is a regular amount of traffic, it is actually impossible. I've tried. You can achieve it instantaneously, but you cannot maintain it because you will keep getting cut off.

It is also dangerous — especially for you — because it causes a lot of merges to happen right in front of you. Cars will end up passing you on both sides, and some fraction of them will try to then cut you off. If this happens from both sides, you're in big trouble.

Also, from a legal perspective, you are incorrect that if you rear-end someone, you are at fault (IAAL).

Re:Test Can Tell How Likely a Young Driver Is to Crash by genocidicbunny on Saturday January 13, 2024 @12:01PM (#38981857)
True, but it doesn't really change the calculus at all. Good defensive driving includes considering the human factors, which also means anticipating when other drivers don't drive well or suddenly decide to do something stupid.

So yes, the person behind you that rear-ended you might be in the wrong, but you're still going to be left with a life-long spinal injury. As they say -- you can be dead right, but you're also dead.

Re:Test Can Tell How Likely a Young Driver Is to Crash by Ferret7446 on Sunday January 14, 2024 @03:38AM (#38988605)
Is that causing crashes, or are other people causing crashes that involve you?

If everyone else is driving aggressively, then a crash may occur if you drive cautiously, but the law in most places I know would be on your side.

Re:Test Can Tell How Likely a Young Driver Is to Crash by myself248 on Saturday January 13, 2024 @10:14AM (#38980655)
Re:Test Can Tell How Likely a Young Driver Is to Crash by linusg789 on Saturday January 13, 2024 @11:01AM (#38981136)
Re:Test Can Tell How Likely a Young Driver Is to Crash by jsjohnst on Saturday January 13, 2024 @10:34AM (#38980840)
Here’s the company behind this and a description of the experience: http://www.diagnosticdriving.com/how-readyassess-works
Re:Test Can Tell How Likely a Young Driver Is to Crash by xnx on Saturday January 13, 2024 @10:19AM (#38980698)
What's the current curriculum for drivers Ed look like? I would hope it would be at least a hundred hours in a high fidelity simulator like this with detailed scoring.
Re:Test Can Tell How Likely a Young Driver Is to Crash by swader999 on Saturday January 13, 2024 @10:43AM (#38980937)
In Canada, there's young drivers that teaches a lot of defensive driving tactics. Then there's others that teach how to pass the test. I've saved myself several times over the years with the brake and avoid and once with the head on collision techniques.
Re:Test Can Tell How Likely a Young Driver Is to Crash by henriquez on Saturday January 13, 2024 @11:00AM (#38981130)
We need to be teaching kids offensive driving tactics. Too many drivers are passive and have no idea about the physical capabilities of their vehicles. This leads to more accidents than needed.

For example, merging into the right traffic lane of a freeway from an on-ramp. Most drivers think the correct way to do this is to accelerate to 50mph and use their left turn signal. Wrong. The speed of traffic on the freeway is 85mph and inadequate acceleration produces a huge hazard when other drivers have to make a sudden lane change to avoid your stupid Honda CRV. The correct action would have been to accelerate to 98mph while merging to minimize the probability of a collision with existing drivers on the road, no turn signal necessary.

There are numerous other examples, such as four Honda CRVs pulling up to a 4-way stop simultaneously, but the general solution to all roadway problems is aggressive acceleration and disregard for other drivers, since they’re probably dumb anyway.

Source: I learned to drive in California.

Re:Test Can Tell How Likely a Young Driver Is to Crash by seadan83 on Saturday January 13, 2024 @12:46PM (#38982463)
Your point has validity to it, I would not agree regarding: 85mph vs 98mph

(1) The right hand lane of any free-way should not be going that fast! The fastest speed limit I've seen I think was 80 in the middle of Kansas.

(2) 98mph is unsafe for a number of cars, it's a lot of momentum. You need to have decent tires, and be a bit lucky nothing goes wrong.

(3) Merging at +15mph compared to traffic flow is not going to be that smooth. You'll be running too fast and will need to brake or merge left too quickly. Either way, it's not being part of the flow of traffic.

(4) You mentioned, "accelerate to 98 while merging". I think that is also wrong. When changing lanes it is a good idea to accelerate by about +3mph. +15mph is overkill. Further, a merging car should be at speed before it even starts to move left. The acceleration ramp, the on-ramp, that is where you match traffic speed and then merge in.

Though, if I had to pick between merging at 15mph too slow vs too fast, I'd probably pick to merge too fast.

Re:Test Can Tell How Likely a Young Driver Is to Crash by kjkjadksj on Saturday January 13, 2024 @11:24AM (#38981414)
People in CA are some of the worst drivers and I’ve drove in probably 30 states. No respect for the speed limit either way because no cops in socal run radar, so you get people ripping 100 miles an hour on the same highway as someone going 40mph with ten feet of metal scrap pilled in their pickup truck. No respect for posted signage or even the directionality of the lane; today on the 101 an accident caused traffic and I saw people driving on both shoulders as well as exiting the freeway through the onramp, jumping a curb in the process. All the guardrails on the highways are beat and scraped up, almost every square inch has had an impact since it was built it seems. I’ve also seen probably a half dozen cars by now lodged into a storefront, like it must not even be that rare if I’ve encountered that six times already, and I’ve never seen that anywhere else. Also whenever it rains even a little, people devolve even more and go about 35mph on the freeway with no one in front of them.
Re:Test Can Tell How Likely a Young Driver Is to Crash by henriquez on Saturday January 13, 2024 @11:34AM (#38981539)
> Also whenever it rains even a little, people devolve even more and go about 35mph on the freeway with no one in front of them.

And God forbid it snows. Even with a light dusting people just stop on HWY 17 halting traffic for hours until emergency services is able to tow everyone off the mountain.

Re:Test Can Tell How Likely a Young Driver Is to Crash by diggan on Saturday January 13, 2024 @11:06AM (#38981185)
> no turn signal necessary

You're a BMW owner by any chance?

If you're planning to turn to the left/right, use your turn signal, even if someone would deem it unnecessary. Better to use the turn signal too much than too little.

Overall, highway on-ramps looks wildly different all across the world, and what you say would be impossible in lots of places around me as an example, where you need to be able to basically stop at the on-ramp if there is too much traffic, to find the right gap, otherwise you'd run straight into others.

> We need to be teaching kids offensive driving tactics

Hard disagree :) I think Spain where I live has a huge problem with too offensive driving, where even if the rules/laws says something, people act on the "first come, first served" basis when it comes to preference.

Re:Test Can Tell How Likely a Young Driver Is to Crash by henriquez on Saturday January 13, 2024 @11:10AM (#38981226)
Not a BMW driver but I never keep my blinker fluid topped up. In California, turn signals just telegraph your intent to make a move, giving other drivers valuable milliseconds to block your approach. The element of surprise is crucial to an offensive driving regimen.
Re:Test Can Tell How Likely a Young Driver Is to Crash by seadan83 on Saturday January 13, 2024 @12:51PM (#38982527)
One thing I do like about California drivers - they do generally let you in. Hard disagree to not use turn signal in CA.

> The element of surprise is crucial to an offensive driving regimen.

Really is fun and games in the abstract. Unfortunately we lose the idea that the roadways are a communal space, instead everyone else is just an obstacle. I love that fore-sight, the ability to predict a mere _30_ seconds into the future disappears for so many drivers. Just observe what people will do to gain 1 or 2 car lengths of ground, as if they could not predict that they would do effectively nothing swerving around to the lane that happened to gain 10 feet of open concrete.

Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Saturday January 13, 2024 @12:58PM (#38982613)
There’s a joke that in LA, a turn signal is used to tell other drivers what you just did.

In other parts of CA, it’s pretty different. Even in LA, I used my blinkers like normal (but yes, sometimes you do have to start moving in your lane for others to take you seriously).

Re:Test Can Tell How Likely a Young Driver Is to Crash by mindslight on Saturday January 13, 2024 @02:24PM (#38983559)
> turn signals just telegraph your intent to make a move, giving other drivers valuable milliseconds to block your approach

You're thinking of turn signals wrong. Generally you shouldn't be using them as a request, but rather a declaration of "this is happening". If another driver can act to thwart your action in "milliseconds", you didn't have the margin to do it in the first place.

Re:Test Can Tell How Likely a Young Driver Is to Crash by henriquez on Saturday January 13, 2024 @02:49PM (#38983807)
That's just ignorant. I've been playing Mario Kart for decades and I know how to drive. It's not my fault if other drivers are terrified by my sudden maneuvers.
Re:Test Can Tell How Likely a Young Driver Is to Crash by mminer237 on Saturday January 13, 2024 @10:43AM (#38980947)
It depends on the state, but typically you have like 30 hours of classroom learning, then six hours driving with an instructor, then like 50 hours of driving with a parent. They also have to pass a test, although that is often in the behind-the-wheel now. Nobody is buying a simulator.
Re:Test Can Tell How Likely a Young Driver Is to Crash by dgrin91 on Saturday January 13, 2024 @11:29AM (#38981465)
I'm not sure what state you're talking about, but most I know are way less. presuming the learner is over 18 the requirements tend to be something like

1. A written exam, takes ~30 mins to 1 hour at the DMV 2. A 5 hour class (when I took this the 5 hour class clearly hadn't been updated for decades) 3. A driving exam that took ~30 minutes end to end

My understand is that most states are like this, though it gets a bit more stringent if you are under 18

Re:Test Can Tell How Likely a Young Driver Is to Crash by caddemon on Saturday January 13, 2024 @12:00PM (#38981843)
I think it's pretty normal to get a learner's permit like a year before your license, so you can theoretically do the 50 hours with a parent. But I've never seen that part formally enforced, you could just do nothing until you're of age for the license driving test. Mostly depends on your parents.
Re:Test Can Tell How Likely a Young Driver Is to Crash by BenjiWiebe on Saturday January 13, 2024 @01:04PM (#38982669)
In Kansas there is no class, and you have to be 16.5 to get a full driver's license. Written test and driving test.
Re:Test Can Tell How Likely a Young Driver Is to Crash by mminer237 on Saturday January 13, 2024 @05:19PM (#38985200)
I'm in Illinois. Of course, you used to be able to just wait until you were 18 and take the test if you didn't care about driving without an adult and not have to do the 50 hours or behind-the-wheel course, but I think they've upped the requirements since I took it.
Re:Test Can Tell How Likely a Young Driver Is to Crash by gnicholas on Saturday January 13, 2024 @01:01PM (#38982634)
> Nobody is buying a simulator.

I think some parents would if offered the chance. It gives the opportunity to see how your kid responds in situations (like red light running) that would not likely come up during the drivers test, or even during 50 hours of driving practice. You can throw things at a kid in a simulator that he won’t have seen in real life, and that you wouldn’t throw at him in real life testing.

Re:Test Can Tell How Likely a Young Driver Is to Crash by jpsouth on Saturday January 13, 2024 @10:26AM (#38980766)
In the UK at least you can take your test pretty much whenever, typically your instructor would say you’re ready to try and put you forward for it but there’s no requirement that I’m aware of to do a minimum number of hours.

We also don’t use simulators, UK driving students are on standard roads with other road users.

Re:Test Can Tell How Likely a Young Driver Is to Crash by Thedarkb on Saturday January 13, 2024 @11:08AM (#38981203)
In Ireland you need to have held your provisional license for at least six months and done 12 lessons with a certified instructor before they allow you to take the test.
Re:Test Can Tell How Likely a Young Driver Is to Crash by jpsouth on Saturday January 13, 2024 @09:48PM (#38986942)
It seems like we’ve no rules in the UK around it, as long as you’ve passed your theory (which you can book before 17 so it’s ready for your birthday) you can book your practical.

You’ll struggle to book one less than 6 months in advance I believe, but in theory you could be driving at 17 years and a couple weeks.

Re:Test Can Tell How Likely a Young Driver Is to Crash by kjkjadksj on Saturday January 13, 2024 @10:32AM (#38980822)
In the US its not required after a certain age and more drivers are waiting and skipping it
Re:Test Can Tell How Likely a Young Driver Is to Crash by Broken_Hippo on Saturday January 13, 2024 @10:55AM (#38981062)
And to be fair, it isn't that much wait in some places.

In Indiana, for example, drivers education lets you get it about 6 months earlier than if a relative teaches you. Without drivers ed, you can get a learner permit at 16 as long as you pass the "knowledge test".

Re:Test Can Tell How Likely a Young Driver Is to Crash by fortran77 on Saturday January 13, 2024 @09:59AM (#38980511)
They should give this test to old drivers, too! (And I say that speaking as a senior citizen.)
Re:Test Can Tell How Likely a Young Driver Is to Crash by voakbasda on Saturday January 13, 2024 @11:47AM (#38981696)
While I agree, this would be career suicide for any politician brave or foolish enough to suggest it.
Re:Test Can Tell How Likely a Young Driver Is to Crash by ACV001 on Saturday January 13, 2024 @10:58AM (#38981096)
Bullshit for the insurance companies to charge more