Plane Crash & statistics

has everyone ever noticed that whenever there's a plane crash you get all these "experts" coming on TV saying things like "oh your more likely to be hit by lightning than to die in a plane crash"

This is quite pathetic because its not going to help the people who have just died is it, plus I wonder how often the occupants have been struck by lighting in their lives.

I' ve been struck by lightning twice (and have witnesses so I wasn't dreaming) but never been in a plane crash, ok I came close once but i didn't crash.

So this poses the question, are statistics utter bollocks and if not where do they get their data from.? There's been quite a few plane crashes lately dunno how many lightning strikes.

And example of statistical silliness.

Lets say the chances of being blown up by a bomb on a plane are a million to one, so the chances of there being 2 bombs on the plane is probably a billion to one. So if everyone takes their own bomb on board, safe bombs, then its going to be impossible to be blown up on that plane statistically.

Which is total rubbish, if there's a bomb on board, its going to go off no matter what you do.

Specific examples have nothing to do with accurate statistics- in fact they are completely irrelevant to them. The general extrapolation of answers from statistics- i.e. accidents on planes are very rare hence they are safe- is perfectly good reasoning and I do not understand your aversion to them.

The way to attack statistical information is not to provide specific examples but to attack the methodology. For example, air safety statistics work ion miles travelled, as other transport means do. This is unfairly biasing things in faviur of air travel. This is because every mile travelled in a car is subject to roughly the same danger, whereas in a aeroplane it is really only the take off and landing that is risky, with mid-air failures (like this one) being so rare as to be almost irrelevant.

This means very long air journeys are actually virtually as safe as very short ones, whilst a long car journey is noticably more dangerous than a short one. Which means in turn that judging accidents my miles travelled is unfairly positive to air travel, for which distance travelled is not the issue.

were you the jesus actor in the movie the passion? because he got struck twice while filming his scene while tied to the cross that was staked to a hill in a region with high lightning storms and electrical phenomenon

Originally posted by Bicnarok
has everyone ever noticed that whenever there's a plane crash you get all these "experts" coming on TV saying things like "oh your more likely to be hit by lightning than to die in a plane crash"

This is quite pathetic because its not going to help the people who have just died is it, plus I wonder how often the occupants have been struck by lighting in their lives.

I' ve been struck by lightning twice (and have witnesses so I wasn't dreaming) but never been in a plane crash, ok I came close once but i didn't crash.

So this poses the question, are statistics utter bollocks and if not where do they get their data from.? There's been quite a few plane crashes lately dunno how many lightning strikes.

And example of statistical silliness.

Lets say the chances of being blown up by a bomb on a plane are a million to one, so the chances of there being 2 bombs on the plane is probably a billion to one. So if everyone takes their own bomb on board, safe bombs, then its going to be impossible to be blown up on that plane statistically.

Which is total rubbish, if there's a bomb on board, its going to go off no matter what you do.

Also, this Cracked article is super funny and that.

http://www.cracked.com/article_17416_7-most-bizarrely-unlucky-people-who-ever-lived.html

Also, are you joking or do you really not understand statistics at all?

Originally posted by Bicnarok
has everyone ever noticed that whenever there's a plane crash you get all these "experts" coming on TV saying things like "oh your more likely to be hit by lightning than to die in a plane crash"

This is quite pathetic because its not going to help the people who have just died is it, plus I wonder how often the occupants have been struck by lighting in their lives.

I' ve been struck by lightning twice (and have witnesses so I wasn't dreaming) but never been in a plane crash, ok I came close once but i didn't crash.

statistically, the odds are 1/2 that a coin flip will result in heads. However, I just flipped a coin 5 times, and got heads 4 times.

something being statistically probable does not mean that it must happen as opposed to things that are less probable, nor does the probability of something allow one to make specifically accurate predictions about how things must work.

Originally posted by Bicnarok
So this poses the question, are statistics utter bollocks and if not where do they get their data from.? There's been quite a few plane crashes lately dunno how many lightning strikes.

statistics are essentially averages. So, like in hockey goalies have a goals against average. If Dominik Hasek has a 1.68 goals against avg, it doesn't mean that there are no games where he lets in 7 goals. In fact, he could have a month where each game he played, 5 or more goals were scored against him, though in each other month he could have none.

Plane crashes are also international news. Lightning strikes are not.

Originally posted by Bicnarok
And example of statistical silliness.

Lets say the chances of being blown up by a bomb on a plane are a million to one, so the chances of there being 2 bombs on the plane is probably a billion to one. So if everyone takes their own bomb on board, safe bombs, then its going to be impossible to be blown up on that plane statistically.

Which is total rubbish, if there's a bomb on board, its going to go off no matter what you do.

here is the problem with that scenario. The "million to one" is about being blown up in a terrorist attack, meaning a terrorist makes a bomb, brings it on the plane, and successfully blows it up.

That only refers to the average plane, and the specific odds for any specific plane are going to be different.

There is still a million to one chance that any specific passenger might be the terrorist with the bomb, meaning there is a nearly impossible chance that each passenger might be a terrorist and have a bomb.

I don't understand how you get that everyone bringing a bomb on the plane would reduce the odds of a bombing.

Originally posted by inimalist

I don't understand how you get that everyone bringing a bomb on the plane would reduce the odds of a bombing.

That's what I was wondering, cause I am pretty sure I heard exactly that example as a joke before. Granted, not the funniest joke, still.

Originally posted by Bardock42
That's what I was wondering, cause I am pretty sure I heard exactly that example as a joke before. Granted, not the funniest joke, still.

it might reduce the odds that any single individual may be the one who explodes the plane, as they now have to do it before anyone else does, but on aggregate, it would have to increase the odds.

He's basically got to that point by basing his odds already partway through the scenario, a common error. So, flipping a coin ten times and getting heads every time is unlikely, but if you already have nine heads, the chances of the next one being heads is back to just 50/50 again.

To apply that to his example- the chances of a plane blowing up when everyone on board carrying a bomb are so unlikely as to be close to impossible, so his reasoning is that if everyone on a plane brings a bomb then, statistically speaking, it won't be destroyed.

Point is, if you are ALREADY IN the scenario where everyone has brought a bomb, then the massive part of the unlikeliness has already happened. From that point on, the chances that the plane will be destroyed as opposed to normal (i.e. where NO-ONE has a bomb) are of course increased. Not that it will ever happen.

Furthermore, of course, if you deliberately set out to counter statistics (by deliberately organising it so everyone on a plane has a bomb) then you are playing from a different data from the stats anyway. Stats refer to AVERAGE flights. The chances of any given fight having every passanger with a bomb on it are as close to zero as makes no odds. The chances of a flight where you are specifically trying to make sure that everyone on board has a bomb... those odds are different.

To answer your original question- the point of the statistical annoucements is to keep accidents in proportion and point out how safe air tavel is. As I say, if you want to fight that, attack the methodology. Specific examples bring you nothng (other than odd glances from posters confused as to how you can think statistics work like that).

Originally posted by Bardock42

Also, this Cracked article is super funny and that.

http://www.cracked.com/article_17416_7-most-bizarrely-unlucky-people-who-ever-lived.html

Also, are you joking or do you really not understand statistics at all?

good link that, that Japanese bloke surviving 2 atom bomb sure is a statistic killer.

And I obviously don't understand statistics, as they appear pointless to me as anything can happen any time. I could get shot by a lunatic S.American Indian who got swept over to Europe by a freak tornado, with a blow dart covered in poison from the back of a homosexual poisonous frog... owww what was that what just hi t . m y n e c kkkkkkkkkkkkkk

Originally posted by Bicnarok
"oh your more likely to be hit by lightning than to die in a plane crash"

I guess the irony of it all is when lightening hits a plane and causes it to crash. 😐

Originally posted by Bicnarok
And I obviously don't understand statistics, as they appear pointless to me as anything can happen any time.

The idea of statistics isn't to tell you what will happen, it's to tell you what is likely to happen based on previous things.

Some people need to learn the difference between Conditional Probability and the Memoryless Property of Probability.

Some people need to learn the difference between Conditional Probability and Memo

Originally posted by Symmetric Chaos
The idea of statistics isn't to tell you what will happen, it's to tell you what is likely to happen based on previous things.

but it doesn't work or is very ineffective as events tend happen randomly.

Originally posted by Bicnarok
but it doesn't work or is very ineffective as events tend happen randomly.

Except they don't. Obviously it's very unlikely that people will randomly explode. However as statistics get into the middle ranges they do become less useful outside of pure mathematics and gambling.

Originally posted by Bicnarok
but it doesn't work or is very ineffective as events tend happen randomly.

Yeah, but at a difference chance. Like you said, it's possible that you could get shot by a lunatic S.American Indian who got swept over to Europe by a freak tornado, with a blow dart covered in poison from the back of a homosexual poisonous frog, but it is statistically very, very unlikely...and have you seen it happen lately?

There's not an equal chance of everything happen at any given time.

no, but i have seen life fish fall from the sky. i mean what are the odds of that?

Originally posted by Wild Shadow
no, but i have seen life fish fall from the sky. i mean what are the odds of that?

if you saw it happen, 1.

Well this is a bit odd:

http://news.yahoo.com/s/ap/20090605/ap_on_re_la_am_ca/brazil_plane;_ylt=AmBFABOSNZCxyDvdJ27f2K.CfNdF

statistics are total horse shit.

statistics of any kind are based upon a large sampling of data. Statistics count a multitude of events which each had the same outcome. They do not, however, account for the specific unique variables that occur in each and every person's life that may or may not affect the outcome, the data that is being used. In some scenerios, this is moot. Others, vastly important.

eg. a person could take into account the amount of people living in any city, also take into account the amount of car accidents in that city per year. Based on these two numbers they would give you a statiscal probability of you being in a car accident this year in that city if you're a resident. rubbish. There are so many variables and specifics not accounted for, not the least of which is your own personal driving ability. Some elderly man with poor vision who's hobby is drunk driving would have the same probability of getting into an accident as you. While everybody would indeed have the same probability of being blind sided by another driver, the difference in variables of your life as opposed to his greatly reduces your odds of being in a car accident of your own doing. This, however, would not be shown. Statistics gather data, just an equation. Each person has the same value, 1. It doesn't matter if you're blind, mentally handicapped or Chuck Norris. You're all equal, 1.

Statistics should only ever be used to demonstrate trends in an entire population. They should never be used in respect to a single individual nor a group of individuals significantly smaller than the population used to gather data from.

Let's say only 100 people per year die in plane crashes. There is no probability there in respect to single individuals. Every person who flies on a plane has the same chance of being one of those 100 every time they lift off. For an entire population, only 100 people out of 6.5 billion is pretty good odds for those that comprise it. As an individual, does you no good. You're not concerned about the other 6.5 billion, you're only concerned about 1.......and you have no more/no less of chance of going down in flames than the 100 people who did so last year.