Hacker News Reader

Comment by balderdash | Hacker News Reader

This reminds me of the Micromort - which I thought was supposed ch great idea, but never seemed to go anywhere…

https://en.m.wikipedia.org/wiki/Micromort

I dunno, I use it all the time. Putting the danger of Covid-19 into micromort terms was a useful exercise for instance.

Useful for what? Peace of mind? Or as a comparison, where you calculate the micromorts for taking a vaccine, or social distancing, etc?

In terms of deciding how careful to be about it. An interesting property of Covid is that getting it for the first time without vaccination is about the equivalent of a typical person's full actuarial Micromort budget for a year regardless of their age. That's a serious level of danger but not a terrifying one and worth some a good number of lifestyle changes until I got vaccinated but not every potential change.

Wonderful, yeah, great way to put it into perspective. I hope everyone knows by now, but the media feeds off unbounded, hysterical-style risk. Putting a cap on the worst case (or average worst case) frees you from the fear of the unknown.

> A micromort (from micro- and mortality) is a unit of risk defined as a one-in-a-million chance of death.

Ah... so it happens nine times out of ten.

> The micromort concept was introduced by Ronald A. Howard

I was genuinely surprised that not by Terry Pratchett.

A challenge I have with understanding and communicating risk is that some risk is an instant and some is across distance/time/some quantity.

If you skydive your chance of death is X.

If you drive your chance of death is Y per hour (or km).

Any advice from anyone on the best way to reconcile this? We could add a unit time or distance to the first kind, but I feel that betrays the fact that it’s mostly about “does your parachute deploy?” and is disconnected from how long or far you skydove that day.

(these examples are easy to debate but I hope you get what I’m trying to communicate)

Aren’t they already the same? You’re just not applying the same metric in both cases.

The chance of dying in a car accident isn’t the number of miles driven, it’s the chance of a failure happening and killing you — the distance driven is simply more rolls of the dice.

The same with parachute-failure — the number of chutes opened is the number of dice rolls. You should be saying skydiving death rate is X per chute

What is the driving equivalent to one skydive?

I think that question genuinely illustrates their point that some risk is best considered as a rate whereas some is as a discrete event but I am very interested if you can elaborate a perspective that reconciles that

The link gives both of those, actually. Skydiving is 8 micromorts per jump, car travel is one micromort per ~250 miles. So the equivalent chance of death for one skydive is driving 2000 miles.

But how to combine them, i.e., micromorts per skydiving jump via a car that travels 250 miles in the air before driving it out of the back of a cargo plane?

I think we should just add them together and call it 9 micromorts, assuming the car has an appropriately sized parachute.

That seems to further solidify the original observation.

I do appreciate you finding their break even point though!

I think the answer is that it’s always a rate; it’s just that the skydiving is such a rare event we can implicitly drop the “per chute” because we’re assuming it’s always going to be “1 chute”.

If you took the skydiver instructor, doing hundreds of rounds, then suddenly skydiving should no longer be considered a “discrete event” — your determination of how to treat it isn’t based on the event-type itself, but the object undergoing the event.

There is also the microlife which denotes reduction in life expectancy that is more applicable for lifestyle and environmental factors (everyday you live in Mexico City vs London takes 30min off your life) vs acute risks (sky diving)

https://en.m.wikipedia.org/wiki/Microlife

I actually use it with friends now and then and find it useful.

Reply on news.ycombinator.com