Hacker News new | ask | show | jobs
by perl4ever 2351 days ago
You seem to be confusing "known unknowns" with "unknown unknowns". I find this unaccountably irritating because everybody intuitively knows the difference, and that it's huge, except when they pretend to be "rational".

What's the probability of something happening when the probability you are given is probably wrong?
1 comments
tim333 2351 days ago
I'm more just dividing the historical statistics eg.

>commercial airlines in the United States between 2000 and 2010 was about 0.2 deaths per 10 billion passenger-miles

and

>390 motorcyclist deaths per billion vehicle miles

dividing gives 19500x per mile. Though I do more miles by air. It's all a bit approximate.

link
asenk 2350 days ago
1. You can't claim your risk of dying using a particular method of transportation is above another method just by looking at per mile death rates unless you travel a similar distance in both, or say it in a context of making a decision of using one or the other to travel from place to place (although to be precise the length of the trip depends somewhat on the used method).

2. The commercial airlines figure which you've used is not from using the particular method of transportation, 737 max, which the use of you are justifying with the comparison.

3. If you wear a helmet, don't ride drunk, don't speed and otherwise follow traffic rules it will make riding a motorcycle a lot safer than average.

link
tim333 2350 days ago
Well, yeah it's complicated. From the perspective of an all knowing god your chances of dying in some way are probably either 1 or 0 depending on what fate will bring so practical probability estimates are guesses based on limited information. Based on my limited information I'd be happy enough to fly the max. It'll probably be the world's most scrutinised aircraft by the time it flies again anyway.
link
perl4ever 2349 days ago
Guesses based on limited information need to include the information you do have, and not include information you don't have. If a probability doesn't reflect what you know and don't know, it's not valid, period, limited information in general not withstanding.
link