?

Log in

No account? Create an account
click opera
February 2010
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
 
 
 
 
 
 
Tue, Jun. 2nd, 2009 08:50 am
Fear of flying

74CommentReplyFlag

krskrft
krskrft
Tue, Jun. 2nd, 2009 01:43 pm (UTC)

This is the major misconception which leads people to use statistics as support for things that are completely outside the bounds of numerical reason.

Airplanes don't have random statistical probabilities of malfunction. They are either prone to malfunction or they aren't. The pilots are either prone to fuck up and crash the plane, or they aren't. Etc, etc, etc.

It's not like, every time you get on a plane, an ethereal entity rolls the dice to determine whether you are the one in 9 million--whatever statistic is being floated--who will die in a burst of flames and twisted metal.

I'm not saying that people should all be irrationally afraid of flying, just that statistics are too commonly used to support the rarity of non-random events.


ReplyThread Parent
oracolodeifont
oracolodeifont
ed.
Tue, Jun. 2nd, 2009 03:12 pm (UTC)

in conclusion, we should be mentioning probabilities ONLY when there is actually somebody rolling a dice, like when I play backgammon?


ReplyThread Parent
mrobot
mrobot
Ben
Wed, Jun. 3rd, 2009 12:09 am (UTC)

this is ridiculous! statistics don't imply randomness, but they are facts about patterns.

we don't need to posit an ethereal dice roller to say that flying is less dangerous than driving, or that x number of flights are probably going to crash this year.

it sounds like you're going to the opposite extreme and saying it's some kind of deterministic world where past events have absolutely no relevance to future ones.


ReplyThread Parent
krskrft
krskrft
Wed, Jun. 3rd, 2009 02:34 am (UTC)

My point is that you don't actually have 1 in 9,000,000 "chance" of dying in a plane crash, or whatever the statistic is.

These types of statistics can tell us important things about discrepancies in the number of incidents that occur, between say passenger cars and commercial jets. But they can't tell us the real "chance" of somebody dying in either one, because there are so many non-random factors that play into whether an incident will or will not occur.

1 in 9,000,000, for example, tells gives us a proportional idea of how many plane crashes occur. The statistic, on the other hand, doesn't tell us whether a plane will or will not crash. So the difference is that it can tell us plenty about what has happened already--i.e. the information we have up to the present moment--but should not be incorrectly applied to what will happen in the future.

The problem is that these types of statistics are almost always used, in popular media, to urge us toward some specific understanding of the future, when in actuality, these statistics almost never govern random events, and therefore cannot tell us with any certainty what will or will not occur in a specific situation.


ReplyThread Parent