Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Speeding Slows You Down (By a Lot)

algorithmsoup.wordpress.com

Post reply on HN
Print view View original

11–20 of 133 posts

Re: Speeding Slows You Down (By a Lot)

#11

Post by sshine »

codeflo wrote: post #3

Why would you add the lost life expectancy to the length of the trip? That doesn’t make any mathematical sense to me. Edit (this might be wrong, see Edit 2 below -- I'm leaving this as is because otherwise, some of the responses don't make sense): I noticed that the calculation is wrong even within the article's own logic. Supposedly, this calculates "the expected length of the trip (including dead time) at different…

"Time spent dead" is a lot of time added to your trip.

But only by a fraction corresponding to the probability of dying.

Re: Speeding Slows You Down (By a Lot)

#12

Post by jp57 »

When people say “the probability doubles” they really mean that the odds double, right?

If my probability of dying is .6, it can’t double to 1.2. However if my odds of dying are 3:2, they can double to 6:2 or 3:1 or P=0.75. The odds can continue doubling indefinitely and P will asymptotically approach 1.

Re: Speeding Slows You Down (By a Lot)

#13

Post by thfuran »

tzs wrote: post #10

Note that there are some types of accidents whose probability goes down the faster you go. An example would be getting hit head on by someone who veered out of the opposite lane into yours. The probability of being in that kind of accident goes up as the number of cars that you pass in the opposite direction goes up. Without loss of generality we can assume that there are no entrances or exits on the other lane betwe…

The closing speed between you and the swerver is almost certainly positively correlated with collision rate. Given that the base rate of even seeing someone swerve into opposing traffic is so low, I'd expect that effect to dominate. Even if not, severity of injury definitely strongly positively correlates with collision speed.

Re: Speeding Slows You Down (By a Lot)

#14

Post by photochemsyn »

The safest driving speed on busy roads is the speed everyone else is going. If everyone would drive this way, overtaking accidents (often very serious injury-wise) would be eliminated.

This means slow drivers are just as dangerous as fast drivers, and also points to why self-driving cars would greatly reduce highway accidents. Assuming they'd all be monitoring each other's speeds, they could coordinate like swarms of drones, and thus could drive safely at faster speeds.

To go even faster, the cars could link up into a single line, under coordinated control, and zip along like a high-speed train (which is why trains are the most efficient transport system, at a nice optimum balancing speed and energy consumption). However, this 'train of cars' has some advantages, as you could then just detach from the train and drive on independently to your local destination, avoiding the last mile problem.

Re: Speeding Slows You Down (By a Lot)

#15

Post by crznp »

codeflo wrote: post #3

Why would you add the lost life expectancy to the length of the trip? That doesn’t make any mathematical sense to me. Edit (this might be wrong, see Edit 2 below -- I'm leaving this as is because otherwise, some of the responses don't make sense): I noticed that the calculation is wrong even within the article's own logic. Supposedly, this calculates "the expected length of the trip (including dead time) at different…

The advantage of driving faster is that you convert driving time into free time.

Death converts living time (including free time) into death time.

Subtracting them makes mathematical sense: you have more or less free time. Though it is napkin math. Eg: it treats all free time as equivalent. If speeding gets you to a job interview on time, which gets you the job, that is more important than speeding so that you have time to do a crossword.

Might as well do the crossword in the car, eh?

Re: Speeding Slows You Down (By a Lot)

#16

Post by alexwhb »

While the methods in this post may not be a super accurate depiction of reality.. I did find it entertaining. Additionally I agree generally with the point the post is making. It is shocking to me that we accept so whole heartedly the risk associated with driving when, comparatively, it is shockingly unsafe relative to other modes of transportation. In the US somewhere around 41,000 people die per year from traffic accidents. If we compare that to train travel (especially in Europe where safety standards are significantly better) it’s somewhere in the ballpark of 32x safer. Additionally we accept the liability of a car shockingly easily. Yes we all should have insurance, but that insurance only goes so far. If you accidentally kill somebody on a bike you’re likely to serve prison time for manslaughter. There are several other downsides we accept that I won’t get into, but it’s pretty interesting to me that especially US society has accepted these facts so seemingly easily.

Re: Speeding Slows You Down (By a Lot)

#17

Post by giantg2 »

This is idiotic. The penalty wouldn't apply to the trip if you didn't incur it (die) during the trip.

Also, as they noted, there are a bunch of other factors. Some of those factors are equally or more important as the ones they are looking at, such as alcohol impairment (both about 30% of traffic fatalities).

The biggest fatality risk factor is not being buckled in. About 45% of all occupant fatalities are people who were not buckled. This will greatly skew the stats if not properly accounted for.

Edit: why disagree?

Re: Speeding Slows You Down (By a Lot)

#18

Post by thfuran »

jp57 wrote: post #12

When people say “the probability doubles” they really mean that the odds double, right? If my probability of dying is .6, it can’t double to 1.2. However if my odds of dying are 3:2, they can double to 6:2 or 3:1 or P=0.75. The odds can continue doubling indefinitely and P will asymptotically approach 1.

Likely they're talking about small probabilities and meant what they said. If not, who knows. (Quite probably not even them)

Re: Speeding Slows You Down (By a Lot)

#19

Post by kragen »

jp57 wrote: post #12

When people say “the probability doubles” they really mean that the odds double, right? If my probability of dying is .6, it can’t double to 1.2. However if my odds of dying are 3:2, they can double to 6:2 or 3:1 or P=0.75. The odds can continue doubling indefinitely and P will asymptotically approach 1.

when they're talking about probabilities like the one in 600 thousand of dying on this trip, they are the same

if you have a model where a probability goes outside the (0, 1) range for some input parameter value, that obviously isn't a valid answer, but it may just mean that the parameter value is well outside the model's range of validity. you are surely right that if we extrapolate this probability model of doubling the risk of death per 16 kph speed increase beyond about 400 kph we get invalid probabilities, but i think that probably the model breaks down well before that point

Re: Speeding Slows You Down (By a Lot)

#20

Post by kragen »

alexwhb wrote: post #16

While the methods in this post may not be a super accurate depiction of reality.. I did find it entertaining. Additionally I agree generally with the point the post is making. It is shocking to me that we accept so whole heartedly the risk associated with driving when, comparatively, it is shockingly unsafe relative to other modes of transportation. In the US somewhere around 41,000 people die per year from traffic a…

i just spent three hours on the bus today and i would much rather have spent an hour in a car instead, though my risk of dying would surely be higher

there are a lot of places within 100 km of my house that i'd enjoy visiting but where public transport doesn't go. if i had a car, i could go there easily. if that means shortening my life expectancy by 20 minutes, that seems like a worthwhile tradeoff. the risk i'd be imposing on innocent people around me by driving a car is ethically trickier

Post reply on HN