train crashes and base rate fallacies

derailing the derailment narrative

“oh my gawd, another train crash!”

“it this some kind of sabotage?”

this seems to be quite the furor on social media right now as the newshound junior reporting team is suddenly amplifying every ill outcome.

but let’s catch our breath for a moment.

the US has about 1000 train derailments a year. that’s roughly 3 per day.

so is the rate actually up, or are we just having another “summer of the shark/food factory fire” media moment where an event that is actually no more common than usual suddenly seems so because it’s being breathlessly reported at high frequency all of a sudden?

the internet (and humans in general) seem to really struggle with the idea of "base rate" likely because you can make nearly anything look like some crazy clustering or wild trend by just starting to report on it all the time.

we generate ideas about outcome rates based on what we see and hear about.

and this can seriously warp perspective.

it’s an evergreen tactic.

whatever the babblesphere becomes monomaniacally fixated upon suddenly looks like a crisis.

it's easy to pass off "this happens all the time but you never bothered to notice" as "scary new crisis."

it's also attention garnering and profitable.

lots of people have lots of incentive to inflame rather than inform.

the cubscout with the scariest story about disembodied hands winds up in journalism school.

but it's not a lens through which you want to see the world, so don't trust any of these claims unless they reference a base rate.

and no, this is also not some “decaying infrastructure” story. as always, context is key.

Bachman @ElonBachman

Tracks (purple line) have gotten a lot safer. Humans (orange line) have maybe gotten a bit worse on a per mile basis.

it’s also worth noting that chemical spills and releases happen all the time. here’s some useful insight from bloomberg in 2014 (back when they were a seriously solid news source)

according to “el google”:

so, while yeah, the whole thing is basically meme entrapment

perhaps some caution to be sure one has a sense of facts and magnitudes may be warranted here.

bloomberg continues:

obviously, this is a super scary image. no one sees this and thinks “oh, cool, no bigge!” it’s atavistic.

some fish and a fox seem to have died. i’m sure you did not want to stand too close to this and obviously this is not a good thing, but just how big a crisis is this really?

i have no particular expertise here, so as far as the palestine crash and its starkly photogenic disposal methods, i can only go on what folks who i know to be sober, serious experts in the space and have good records of being right are saying, to take it for what it’s worth:

i have spoken to two good friends about the ohio spill. both work in industry, have deep familiarity with chemicals, chemical engineering, chemical manufacturing, spills, and vinyl chloride.

neither has a dog in this specific fight.

neither seems terribly concerned and both say this is being vastly overblown. both also agree that this was 100% the correct disposal method. they doubt it has any large or even intermediate scale implications and that all these scare maps about poisoned farmland are wild misunderstandings about quantity, potency, and concentration.

lots of things from open heart surgery to giving birth are terrifying if you’ve never seen them before.

and obviously, the images here are striking.

but lemme tell you, so is this:

and in terms of putting raw carcinogens into the sky that affect people 3 states away, what just happened in ohio is not even a rounding error compared to the forestry management calamities of california.

it’s not my intent to wade into the “is any of this OK/tolerable/desirable” debate here because it’s inherently subjective and reasonable people can have reasonable differences on what kinds of pollution to tolerate for the trappings of modernity.

i’m just speaking to whether or not what just happened is unusual and to what extent it moves the baseline.

and best i can tell the answer is “not discernably.”

for good or for ill, this seems to happen all the time.

per EPA, we release 3-4 billion pounds of toxics release inventory chemicals a year. (this is almost certainly under-reported).

as i was about to press send, gatopal™ doomberg published this which i think is excellent and worth reading.

Railroaded

so, while what happened in palestine was unfortunate, it’s neither terribly rare nor terrible dispositive vs what else is going on.

it seems like we’re going to be heading for “the spring of the toxic traincrash” but when you see such claims, always consider the base rate.

otherwise when some locomotive hits a henhouse, it’s going to break the internet…