5 ms·

If you knew all this, would you have avoided driving over the bridge? Would you have wanted the local government to close it indefinitely awaiting repairs? Let

If you knew all this, would you have avoided driving over the bridge? Would you have wanted the local government to close it indefinitely awaiting repairs? Let's say it would collapse with 100% certainty randomly in the next three years, and you're in the danger area for 2 minutes, with a 20% chance of fatality (in fact, nobody died). That's still around a one in 4 million chance any given trip kills you, about the same as 30 miles of driving for the average American driver. Most people would accept that level of risk. Perhaps not to save a couple of minutes on the journey, but if everyone was redirected to another route at rush hour, it might cost each commuter 10-20 minutes. A handful of newsworthy bridge collapses per decade across the US doesn't seem so bad. Instead of negligence, perhaps that indicates an appropriate level of maintenance and risk tolerance, and an appropriate human price to add to the 500,000 other road deaths over the same period.

arethuza 2y ago
"Most people would accept that level of risk" I wouldn't and I definitely wouldn't if I had other people in the car.
- AdamN 2y ago
  It's likely that you live in an area with similar mortality risk from air pollution. Is there a reason not to move to an area with lower risk from air pollution?
  - arethuza 2y ago
    Yes, I'm being "irrational" by some viewpoint, I know that.
  - jorvi 2y ago
    Loneliness is quite damaging to someone’s health. Why are you irrationally suggesting to him to dump all his social ties and net raise his risk of disease and death? :)
  - 1992spacemovie 2y ago
    I get your point, however, no offense intended, this type of questioning is kinda asinine. One risk is immediate (right now), and to a relative extent, an easy choice to mitigate. The other risk is long term (maybe never), and mitigating the risk is much more costly (uprooting family, moving, finding a new residence... with today's prices).
  - jameshart 2y ago
    Air pollution causes ‘excess deaths’ in the sense that people die of air pollution related diseases at some rate instead of dying of something else, generally a short time later. There’s a very different risk assessment between long term choices that have a remote effect on the likely ultimate cause of your inevitable death, versus the probability of premature death through a sudden violent traumatic event.
  - michaelt 2y ago
    > It's likely that you live in an area with similar mortality risk from air pollution. They're not really equivalent, unless you've got a high-risk condition like asthma. Statistically speaking, a 100% chance of losing one year of lifespan (out of the 76 years an American can expect) is a greater mortality risk than a 1% chance of instant, immediate death. In fact, if you're aged 26, a 1% chance instant death has half the mortality risk - and the older you get, the better it looks. However, if you look at humans' revealed preferences, people would much rather lose one year of lifespan than take a 1% chance of losing their entire remaining lifespan. Mortality risk isn't a very good predictor of human preferences, it turns out.
    - bcrosby95 2y ago
      Let's look at money. Would you rather have a 1% chance of losing a million dollars today, or a 100% chance of losing $20k when you're 76? Statistically speaking, a small chance of losing everything today is actually worse than a guaranteed chance of losing a smaller amount later.
spuz 2y ago
I believe regardless of the actual probability of failure, most people would have refused to drive across the bridge had they seen one of the supports had become completely detached. Another way to put it is if authorities had closed the bridge in 2019, I believe most people wouldn't complain when they were presented the photo of the detached support. Also I don't think it's possible for your risk calculation to be done preemptively. We don't know the final breaking point of a piece of steel until we break it. All the calculations and modelling of the bridge will have been done with error margins and because of that we don't have any choice but to over-engineer things so we always stay outside of the worse-case margin for error. Given all that, I don't think it's unreasonable to expect the risk of a bridge collapse to be a lot lower than the average risk we take on the road.
beAbU 2y ago
Yeah it's exactly this mentality that resulted in the Ford Pinto scandal of the 70s. How does that 1:4M probability change when you commute over the bridge twice a day? Humans are notorious for being bad gauging risk, and we very often make the mistake that past success means lower risk of future failure, which for a deteriorating bridge is exactly the opposite of what's going to happen. In my opinion bridges shouldn't collapse, like ever. Annual inspections for years prior pointed to a severely deteriorating structure, even after the temporary cable stays were put in place. The tweet with a picture of a completely detached member a couple of years before the collapse makes this even more egregious. This is a textbook example of where bureaucracy prevails over common sense. Heads should roll. Thankfully nobody died, but the lack of maintenance and upkeep resulted in a total failure, wasting lots of taxpayer money to replace.
- thaumasiotes 2y ago
  > How does that 1:4M probability change when you commute over the bridge twice a day? You can get a good sense of this from the equivalence listed immediately after the odds: >> That's still around a one in 4 million chance any given trip kills you, about the same as 30 miles of driving So it'd be closely analogous to driving a 30-mile commute twice a day. How much risk do you feel that involves?
  - Panzer04 2y ago
    2x daily 30 mile commute sounds like a (relatively) substantial risk to me? Driving is already more or less the most dangerous activity most people engage in nowadays, tolerated by long acclimatization and sheer utility. Moreover, people generally perceive things in which they (theoretically) have more control over as safer - You can control/mitigate the risk you undertake as a driver to some extent (drive slower, bigger gaps, etc), not so for a random bridge collapse.
- EricE 2y ago
  The Pinto thing was dramatically overblown. And seriously set back the acceptance of small cars in the US. For a supposed environmentalist and consumer advocate, Ralph Nadar did FAR more harm than good by grifting off a bunch of sensationalist poppycock. That he's still looked upon as some sort of folk hero is beyond disgusting and just shows how gullible people continue to be paying attention only to the sensationalistic superficial propaganda and not looking deeper.
rob74 2y ago
A well-maintained (and well designed, but there's no reason to think the design was at fault here) bridge shouldn't collapse. Ever. "Appropriate maintenance" in this case would have been to periodically unclog the drains so the water can run off in a controlled fashion and not pool and corrode the supports. How expensive can that be? Instead, the bridge collapses (and they were lucky that no one was killed in the collapse), and they have to replace it for millions of dollars. Money saved on maintenance is the very definition of the proverb "penny wise, pound foolish" IMHO.
- amanaplanacanal 2y ago
  I don’t know how things work in other countries, but in the US new construction is almost always funded with a big chunk of federal dollars (deficit spending), but maintenance has to be done out of your own state or local budget. The incentives are all fucked up.
  - wazoox 2y ago
    Same in other countries for many things. It's relatively easier to spend big on white elephants but almost impossible to get basic day-to-day operation money. Case in point right now at the France-Italy border : they're building the longest rail tunnel in the world under the Alps for tens of billions. There is a tunnel already (Frejus tunnel), which is used at 25% of its capacity, but public money is raining on this unstoppable, useless project. On the other end, all over Europe and particularly in France and Italy, rail and road infrastructures are in a state of disrepair. Wouldn't this money be better used repairing thousands of bridges, tunnels, etc? Sure, but no politician could campaign on "I've got billions spent in our region!". No multinational civil engineering and construction company such as Eiffage or Bouygues would get big money. All the big boys are in favour of the project, and all the small guys simply don't count. This is all completely fucked-up, frankly. It's a general failure of democracy and a global corporate takeover.
    - em-bee 2y ago
      that's because the frejus route is unsuitable for high speed rail. https://en.wikipedia.org/wiki/Turin%E2%80%93Lyon_high-speed_railway#The_present_high-altitude_line https://en.wikipedia.org/wiki/Turin%E2%80%93Lyon_high-speed_... Additional traffic limitations stem from the impact of excessive train transit on the population living near the line. Some 60,000 people live within 250 m (820 ft) of the historical line, and would object to the noise from late-night transits. In 2007 the conventional line was used for only one-third of this calculated total capacity. This low use level was in part because restrictions such as an unusually low maximum allowable train height and the very steep gradients (26–30‰) and sharp curves in its high valley sections discourage its use. A 2018 analysis, by contrast, found the existing line close to saturation, largely because safety regulations now prohibit passenger and freight trains from crossing in a single-tube tunnel. This very significantly reduces the maximum allowed capacity of the 13.7 km long Fréjus tunnel, which trains of one type must now fully cross before any train of the other type can be allowed in the other direction. The historical line's path through the deep Maurienne valley is also exposed to rockfalls, and a major landslide in August 2023 forced its closure for most likely over one year.
thaumasiotes 2y ago
> Let's say [...] you're in the danger area for 2 minutes, with a 20% chance of fatality > That's still around a one in 4 million chance any given trip kills you That's a one in five chance that any given trip kills you, unless 20% means something very different to you than it does to me.
adrianN 2y ago
The average American driver includes drunk and tired drivers who play on their phone while speeding. The numbers for sober people driving reasonably e.g. on their commute are probably better.
- ncruces 2y ago
  Those drive on the same road, right next to you.
  - adrianN 2y ago
    That doesn’t make my statement wrong.
pintxo 2y ago
Someone else posted the NTSB video on the bridge: https://www.youtube.com/watch?v=J-VnWB4fiFk https://www.youtube.com/watch?v=J-VnWB4fiFk I am fairly certain that a large majority of people, if they'd have had access to those images of the bridges structural members, would have stopped using it.
- kube-system 2y ago
  Most people are unable to accurately diagnose a bridge structure other than "yeah looks rusty". And if that were the case, they wouldn't be driving much of anywhere in that part of the country, which is called the "rust belt" for a reason. Pittsburgh has a very high number of bridges, if you go anywhere in that city and you don't want to cross a rusty bridge, you can't go very far.
  - icegreentea2 2y ago
    There's rusty, and then there's "I can see daylight through the steel". I think most people can accurately diagnose that as an actual, real problem.
    - kube-system 2y ago
      That isn't an uncommon sight around Pittsburgh. Whether or not it is an imminent structural issue depends on where the holes are.
      - pintxo 2y ago
        This one settled it for me: https://youtu.be/J-VnWB4fiFk?si=BHg_ppZIAb2ykU2-&t=301 https://youtu.be/J-VnWB4fiFk?si=BHg_ppZIAb2ykU2-&t=301 It's an important enough member to warrant 10 (I am assuming symmetry here, you cannot see all of them) not quite tiny bolts and nuts, but you can see clearly through the thing.
        kube-system 2y ago
danielhep 2y ago
Something I think you're missing here is that when the bridge collapses, cars can no longer drive across it until a new bridge is built, so the inconvenience is almost certainly much worse than if they'd just closed it for maintenance for a bit in the first place.
- hedora 2y ago
  Also, it probably would have been a lot cheaper to have someone roto-rooter all the drains of all the bridges in Pittsburg once a year than to clean up one collapsed bridge and rebuild it on short notice. I suspect they have other bridges with similar water damage and have to pay to fix those too.
jameshart 2y ago
> Let's say it would collapse with 100% certainty randomly in the next three years, and you're in the danger area for 2 minutes, with a 20% chance of fatality Or how about let's say it would collapse with 100% certainty randomly in the next three years, and a school bus with twenty kids in it drives over the bridge twice a day, with each child facing a 20% chance of fatality. ‘I probably won’t be the one who dies when it collapses’ is a terrible metric for whether or not we should try to mitigate the risk of a bridge collapse.
- quesera 2y ago
  School buses are never a useful way to think about risk. You might not be intending it, but this is a cheap manipulation technique. Bridges sometimes fail catastrophically, with risk to life. Modeling risk is a necessary way to consider the costs of mitigation. Leaving aside the poor helpless babies, what metric would you suggest?
  - jameshart 2y ago
    School buses are an excellent way to think about risk if you are a school transportation planner. I was trying to give you an example of someone who might make a different risk calculation than an individual.
tzs 2y ago
> That's still around a one in 4 million chance any given trip kills you, about the same as 30 miles of driving for the average American driver. For those wondering if that is right, it is. Here's the math. Americans drive about 3.2 trillion miles per year and about 40 000 people are killed. That's one death per 80 000 000 miles. Assuming each mile is equally deadly, that chances you survive a given a mile would be 79 999 999 / 80 000 000. To survive a trip of N miles, you have to survive each individual mile sequentially. The chances of that would be (79 999 999 / 80 000 000)^N. The chance of not surviving that trip would then be 1 - (79 999 999 / 80 000 000)^N. For N = 30 that is 1 / 2 666 667, which is close enough to dmurray's number to count as a match. There's enough fuzziness is in the inputs that all we can hope for is the same ballpark. I've seen others say the rate is one death per 120 000 000 miles, and for N = 30 that does give 1 / 4 000 000, so I'd guess they are using that rate. > Most people would accept that level of risk. Perhaps not to save a couple of minutes on the journey, but if everyone was redirected to another route at rush hour, it might cost each commuter 10-20 minutes One big difference is that with the bridge everyone has the average risk. I cross the bridge, I'm rolling a d4000000 and hoping I don't get a 1. With a car I can take steps to make the chances of dying on my particular trip much lower than average. With the car I can often time my trip so as to go at times of day or during weather conditions or during traffic conditions when accident rates are lower.
simple10 2y ago
Are we going to start seeing Google Maps offer alternate routes to avoid sketchy bridges? Seems possible. Airline booking sites already offer filters to exclude 737 Max planes.

arethuza 2y ago

"Most people would accept that level of risk" I wouldn't and I definitely wouldn't if I had other people in the car.

AdamN 2y ago
It's likely that you live in an area with similar mortality risk from air pollution. Is there a reason not to move to an area with lower risk from air pollution?
- arethuza 2y ago
  Yes, I'm being "irrational" by some viewpoint, I know that.
- jorvi 2y ago
  Loneliness is quite damaging to someone’s health. Why are you irrationally suggesting to him to dump all his social ties and net raise his risk of disease and death? :)
- 1992spacemovie 2y ago
  I get your point, however, no offense intended, this type of questioning is kinda asinine. One risk is immediate (right now), and to a relative extent, an easy choice to mitigate. The other risk is long term (maybe never), and mitigating the risk is much more costly (uprooting family, moving, finding a new residence... with today's prices).
- jameshart 2y ago
  Air pollution causes ‘excess deaths’ in the sense that people die of air pollution related diseases at some rate instead of dying of something else, generally a short time later. There’s a very different risk assessment between long term choices that have a remote effect on the likely ultimate cause of your inevitable death, versus the probability of premature death through a sudden violent traumatic event.
- michaelt 2y ago
  > It's likely that you live in an area with similar mortality risk from air pollution. They're not really equivalent, unless you've got a high-risk condition like asthma. Statistically speaking, a 100% chance of losing one year of lifespan (out of the 76 years an American can expect) is a greater mortality risk than a 1% chance of instant, immediate death. In fact, if you're aged 26, a 1% chance instant death has half the mortality risk - and the older you get, the better it looks. However, if you look at humans' revealed preferences, people would much rather lose one year of lifespan than take a 1% chance of losing their entire remaining lifespan. Mortality risk isn't a very good predictor of human preferences, it turns out.
  - bcrosby95 2y ago
    Let's look at money. Would you rather have a 1% chance of losing a million dollars today, or a 100% chance of losing $20k when you're 76? Statistically speaking, a small chance of losing everything today is actually worse than a guaranteed chance of losing a smaller amount later.

spuz 2y ago

I believe regardless of the actual probability of failure, most people would have refused to drive across the bridge had they seen one of the supports had become completely detached. Another way to put it is if authorities had closed the bridge in 2019, I believe most people wouldn't complain when they were presented the photo of the detached support. Also I don't think it's possible for your risk calculation to be done preemptively. We don't know the final breaking point of a piece of steel until we break it. All the calculations and modelling of the bridge will have been done with error margins and because of that we don't have any choice but to over-engineer things so we always stay outside of the worse-case margin for error. Given all that, I don't think it's unreasonable to expect the risk of a bridge collapse to be a lot lower than the average risk we take on the road.

beAbU 2y ago

Yeah it's exactly this mentality that resulted in the Ford Pinto scandal of the 70s. How does that 1:4M probability change when you commute over the bridge twice a day? Humans are notorious for being bad gauging risk, and we very often make the mistake that past success means lower risk of future failure, which for a deteriorating bridge is exactly the opposite of what's going to happen. In my opinion bridges shouldn't collapse, like ever. Annual inspections for years prior pointed to a severely deteriorating structure, even after the temporary cable stays were put in place. The tweet with a picture of a completely detached member a couple of years before the collapse makes this even more egregious. This is a textbook example of where bureaucracy prevails over common sense. Heads should roll. Thankfully nobody died, but the lack of maintenance and upkeep resulted in a total failure, wasting lots of taxpayer money to replace.

thaumasiotes 2y ago
> How does that 1:4M probability change when you commute over the bridge twice a day? You can get a good sense of this from the equivalence listed immediately after the odds: >> That's still around a one in 4 million chance any given trip kills you, about the same as 30 miles of driving So it'd be closely analogous to driving a 30-mile commute twice a day. How much risk do you feel that involves?
- Panzer04 2y ago
  2x daily 30 mile commute sounds like a (relatively) substantial risk to me? Driving is already more or less the most dangerous activity most people engage in nowadays, tolerated by long acclimatization and sheer utility. Moreover, people generally perceive things in which they (theoretically) have more control over as safer - You can control/mitigate the risk you undertake as a driver to some extent (drive slower, bigger gaps, etc), not so for a random bridge collapse.
EricE 2y ago
The Pinto thing was dramatically overblown. And seriously set back the acceptance of small cars in the US. For a supposed environmentalist and consumer advocate, Ralph Nadar did FAR more harm than good by grifting off a bunch of sensationalist poppycock. That he's still looked upon as some sort of folk hero is beyond disgusting and just shows how gullible people continue to be paying attention only to the sensationalistic superficial propaganda and not looking deeper.

rob74 2y ago

A well-maintained (and well designed, but there's no reason to think the design was at fault here) bridge shouldn't collapse. Ever. "Appropriate maintenance" in this case would have been to periodically unclog the drains so the water can run off in a controlled fashion and not pool and corrode the supports. How expensive can that be? Instead, the bridge collapses (and they were lucky that no one was killed in the collapse), and they have to replace it for millions of dollars. Money saved on maintenance is the very definition of the proverb "penny wise, pound foolish" IMHO.

amanaplanacanal 2y ago
I don’t know how things work in other countries, but in the US new construction is almost always funded with a big chunk of federal dollars (deficit spending), but maintenance has to be done out of your own state or local budget. The incentives are all fucked up.
- wazoox 2y ago
  Same in other countries for many things. It's relatively easier to spend big on white elephants but almost impossible to get basic day-to-day operation money. Case in point right now at the France-Italy border : they're building the longest rail tunnel in the world under the Alps for tens of billions. There is a tunnel already (Frejus tunnel), which is used at 25% of its capacity, but public money is raining on this unstoppable, useless project. On the other end, all over Europe and particularly in France and Italy, rail and road infrastructures are in a state of disrepair. Wouldn't this money be better used repairing thousands of bridges, tunnels, etc? Sure, but no politician could campaign on "I've got billions spent in our region!". No multinational civil engineering and construction company such as Eiffage or Bouygues would get big money. All the big boys are in favour of the project, and all the small guys simply don't count. This is all completely fucked-up, frankly. It's a general failure of democracy and a global corporate takeover.
  - em-bee 2y ago
    that's because the frejus route is unsuitable for high speed rail. https://en.wikipedia.org/wiki/Turin%E2%80%93Lyon_high-speed_railway#The_present_high-altitude_line https://en.wikipedia.org/wiki/Turin%E2%80%93Lyon_high-speed_... Additional traffic limitations stem from the impact of excessive train transit on the population living near the line. Some 60,000 people live within 250 m (820 ft) of the historical line, and would object to the noise from late-night transits. In 2007 the conventional line was used for only one-third of this calculated total capacity. This low use level was in part because restrictions such as an unusually low maximum allowable train height and the very steep gradients (26–30‰) and sharp curves in its high valley sections discourage its use. A 2018 analysis, by contrast, found the existing line close to saturation, largely because safety regulations now prohibit passenger and freight trains from crossing in a single-tube tunnel. This very significantly reduces the maximum allowed capacity of the 13.7 km long Fréjus tunnel, which trains of one type must now fully cross before any train of the other type can be allowed in the other direction. The historical line's path through the deep Maurienne valley is also exposed to rockfalls, and a major landslide in August 2023 forced its closure for most likely over one year.

thaumasiotes 2y ago

> Let's say [...] you're in the danger area for 2 minutes, with a 20% chance of fatality > That's still around a one in 4 million chance any given trip kills you That's a one in five chance that any given trip kills you, unless 20% means something very different to you than it does to me.

adrianN 2y ago

The average American driver includes drunk and tired drivers who play on their phone while speeding. The numbers for sober people driving reasonably e.g. on their commute are probably better.

ncruces 2y ago
Those drive on the same road, right next to you.
- adrianN 2y ago
  That doesn’t make my statement wrong.

pintxo 2y ago

Someone else posted the NTSB video on the bridge: https://www.youtube.com/watch?v=J-VnWB4fiFk https://www.youtube.com/watch?v=J-VnWB4fiFk I am fairly certain that a large majority of people, if they'd have had access to those images of the bridges structural members, would have stopped using it.

kube-system 2y ago
Most people are unable to accurately diagnose a bridge structure other than "yeah looks rusty". And if that were the case, they wouldn't be driving much of anywhere in that part of the country, which is called the "rust belt" for a reason. Pittsburgh has a very high number of bridges, if you go anywhere in that city and you don't want to cross a rusty bridge, you can't go very far.
- icegreentea2 2y ago
  There's rusty, and then there's "I can see daylight through the steel". I think most people can accurately diagnose that as an actual, real problem.
  - kube-system 2y ago
    That isn't an uncommon sight around Pittsburgh. Whether or not it is an imminent structural issue depends on where the holes are.
    - pintxo 2y ago
      This one settled it for me: https://youtu.be/J-VnWB4fiFk?si=BHg_ppZIAb2ykU2-&t=301 https://youtu.be/J-VnWB4fiFk?si=BHg_ppZIAb2ykU2-&t=301 It's an important enough member to warrant 10 (I am assuming symmetry here, you cannot see all of them) not quite tiny bolts and nuts, but you can see clearly through the thing.
      - kube-system 2y ago

danielhep 2y ago

Something I think you're missing here is that when the bridge collapses, cars can no longer drive across it until a new bridge is built, so the inconvenience is almost certainly much worse than if they'd just closed it for maintenance for a bit in the first place.

hedora 2y ago
Also, it probably would have been a lot cheaper to have someone roto-rooter all the drains of all the bridges in Pittsburg once a year than to clean up one collapsed bridge and rebuild it on short notice. I suspect they have other bridges with similar water damage and have to pay to fix those too.

jameshart 2y ago

> Let's say it would collapse with 100% certainty randomly in the next three years, and you're in the danger area for 2 minutes, with a 20% chance of fatality Or how about let's say it would collapse with 100% certainty randomly in the next three years, and a school bus with twenty kids in it drives over the bridge twice a day, with each child facing a 20% chance of fatality. ‘I probably won’t be the one who dies when it collapses’ is a terrible metric for whether or not we should try to mitigate the risk of a bridge collapse.

quesera 2y ago
School buses are never a useful way to think about risk. You might not be intending it, but this is a cheap manipulation technique. Bridges sometimes fail catastrophically, with risk to life. Modeling risk is a necessary way to consider the costs of mitigation. Leaving aside the poor helpless babies, what metric would you suggest?
- jameshart 2y ago
  School buses are an excellent way to think about risk if you are a school transportation planner. I was trying to give you an example of someone who might make a different risk calculation than an individual.

tzs 2y ago

> That's still around a one in 4 million chance any given trip kills you, about the same as 30 miles of driving for the average American driver. For those wondering if that is right, it is. Here's the math. Americans drive about 3.2 trillion miles per year and about 40 000 people are killed. That's one death per 80 000 000 miles. Assuming each mile is equally deadly, that chances you survive a given a mile would be 79 999 999 / 80 000 000. To survive a trip of N miles, you have to survive each individual mile sequentially. The chances of that would be (79 999 999 / 80 000 000)^N. The chance of not surviving that trip would then be 1 - (79 999 999 / 80 000 000)^N. For N = 30 that is 1 / 2 666 667, which is close enough to dmurray's number to count as a match. There's enough fuzziness is in the inputs that all we can hope for is the same ballpark. I've seen others say the rate is one death per 120 000 000 miles, and for N = 30 that does give 1 / 4 000 000, so I'd guess they are using that rate. > Most people would accept that level of risk. Perhaps not to save a couple of minutes on the journey, but if everyone was redirected to another route at rush hour, it might cost each commuter 10-20 minutes One big difference is that with the bridge everyone has the average risk. I cross the bridge, I'm rolling a d4000000 and hoping I don't get a 1. With a car I can take steps to make the chances of dying on my particular trip much lower than average. With the car I can often time my trip so as to go at times of day or during weather conditions or during traffic conditions when accident rates are lower.

simple10 2y ago

Are we going to start seeing Google Maps offer alternate routes to avoid sketchy bridges? Seems possible. Airline booking sites already offer filters to exclude 737 Max planes.