PDA

View Full Version : Why cluster occurrences?



banquo's_bumble_puppy
2012-Apr-02, 04:51 PM
Why do certain events seem to happen in clusters? I know that I have posted on this subject before, but the question nags at me. This year it's cruise ship accidents and people going nuts on airplanes (recent occurrence with a pilot and before that a stewardess). Someone said that we just these happenings are more frequent than we realize and that it is just a trick of the human brain to notice them more or think there's a connection. I believe that that there is more to what we observe/perceive than meets the eye. It's like the infinite improbability drive...but on a small scale. Pay attention...or you'll miss it...

Strange
2012-Apr-02, 05:10 PM
I suspect there are two things here. Firstly, if the events are randomly distributed then there will be clusters (that is what random means). Secondly, if a particularly notable event of a given type occurs then the newspapers will report further events of the same sort, even if they wouldn't normally have been newsworthy. (A sort of media confirmation bias).

For example, there were 388 fatalities in house fires in the UK in 2010. Clearly, not every fatality is reported in the news. If one hits the headlines (because of the number, age, whatever of people involved or because of a "slow news" week) then others in the following days/weeks may also be reported.

Nothing to see here. Move along.

Taeolas
2012-Apr-02, 05:16 PM
I suspect it's as you said, just a false cluster most of the time, coming from something big hitting the news and attuning you to similar events. The media itself is partly to blame for this; when a high profile occurrence happens, they will often start reporting similar but otherwise lesser occurrences more often, leading to the impression that the event is happening more often.

For example, the sinking of a cruise ship in Italy primes the media and general public to cruise ship incidents. Then if another ship runs aground in Alaska, (didn't happen AFAIK, just a made up example) with no deaths or injuries, it will still get world wide headlines due to being another cruise ship disaster, instead of just being a semi-local blurb it would otherwise be.

Of course real clusters could happen too, which is why the incidents are investigated; it could be there is some common part that is flawed and fails after a certain amount of wear and tear, so finding that is important to preventing future incidents. But those tend to be discovered long after the media cluster has faded to the next cluster.

NEOWatcher
2012-Apr-02, 05:35 PM
I suspect it's as you said, just a false cluster most of the time, coming from something big hitting the news and attuning you to similar events. The media itself is partly to blame for this; when a high profile occurrence happens, they will often start reporting similar but otherwise lesser occurrences more often, leading to the impression that the event is happening more often.
I think that has a lot to do with it.
I've heard of plenty of "sick" ships, ship fires, power outages, ferries sinking, and other ship issues on and off for quite some time.
It's extremely rare that a big one like this that sunk in this manner. (especially when it coincides with the 100th anniversary of another example).
So; every fire, flu, and other thing that happens on a ship gets the news as a big story rather than in passing like they would before.

As far as people going crazy on airplanes, it happens all the time (at least it seems that way to me). This time it was a pilot, not long ago it was a flight attendent. Incidently, I know a flight attendent who recently had to quit because the doctor said so. The hours required were causing health issues. So; I'm sure there's some added stress going on.

Swift
2012-Apr-02, 05:43 PM
I suspect there are two things here. Firstly, if the events are randomly distributed then there will be clusters (that is what random means).
This Design of Experiments software I used to use (E-Chip) had the following on their opening screen: "Randomness comes in bunches". I always liked that expression.

Secondly, if a particularly notable event of a given type occurs then the newspapers will report further events of the same sort, even if they wouldn't normally have been newsworthy. (A sort of media confirmation bias).
And there is a third related thing, humans generally look for patterns, even when there is no pattern (search on pareidolia). So even if the news reports everything equally, our brain latches on to apparent patterns. You remember the three cruise accidents, but you don't remember all the single instances of other equally random events.

Tensor
2012-Apr-02, 06:06 PM
I always had the feeling that suspensions came in clusters. So, I went back through the Banned thread and yeah, there is some clustering. But, the reason for that, at least as far as I can see, is that we get some people who burn through their suspensions to bans rather quickly due to their posting styles or the way they ignore Mod's warnings, rules, etc. Such behaviors tend to cluster the suspensions. Other than those type posters, suspensions are pretty much random.

Strange
2012-Apr-02, 06:16 PM
Other than those type posters, suspensions are pretty much random.

You mean .... any of us could be suspended any moment?!

Strange
2012-Apr-02, 06:19 PM
As far as people going crazy on airplanes, it happens all the time (at least it seems that way to me).

I was on a flight from Japan and heard several of the flight attendants chatting at the back - a well-known footballer turned actor (1) had got drunk and started threatening the (mainly female) cabin crew. I don't think it made the news.

(1) So well known, I can't remember the name.

Rhaedas
2012-Apr-02, 06:21 PM
Successful landings and uneventful cruises cluster too, but somehow we don't notice them.

Tensor
2012-Apr-02, 06:25 PM
You mean .... any of us could be suspended any moment?!

Yeah, you know those mods and their dart board. :whistle:

Swift
2012-Apr-02, 06:25 PM
You mean .... any of us could be suspended any moment?!
<evil laugh> http://files.myopera.com/debplatt/smiley/funny-bwahaha.gif

Gillianren
2012-Apr-02, 06:42 PM
Someone said that we just these happenings are more frequent than we realize and that it is just a trick of the human brain to notice them more or think there's a connection. I believe that that there is more to what we observe/perceive than meets the eye. It's like the infinite improbability drive...but on a small scale. Pay attention...or you'll miss it...

A, I think you're referring to the Somebody Else's Problem field.

B, there's a perfectly reasonable explanation (several, in fact), but you choose to ignore that? Why?

Ara Pacis
2012-Apr-02, 07:02 PM
Are we assuming randomness instead of correlation or causation for clusters in this thread? As someone else mentioned, there's a selection bias that involves reporting media as well. There was a documentary recently that talked about the bird death's in AR and elsewhere last year and how random events can seem to cluster, I think it was on NatGeo.

Also, the Titanic wasn't a cruise ship.

pzkpfw
2012-Apr-02, 07:23 PM
If I repeatedly flipped a coin, I'd be very surprised if the results were exactly even: heads, tails, heads, tails, heads, tails, ...

There are going to be cases where it's heads, tails, tails, tails, heads, ...

Were those three tails in a row a mystical bunching?

------

An example I recall from a book (whose name I can't recall) is two rectangles with dots sprinkled on them.

The first rectangle has the dots randomly placed. The second uses an algorithm that doesn't allow a dot to be placed "too close" to previously placed dots.

The "natural" clumping of the first (some dots just end up near each other, with gaps in other parts of the rectangle) compared to the more "even" spread of dots on the second, tends to make people think the second is the more random rectangle; when it's really the least random.

NEOWatcher
2012-Apr-02, 07:33 PM
Also, the Titanic wasn't a cruise ship.
Where did I say cruise ship?
Besides, the concept is the same... a big ship that carries lots of passengers.
Granted, maybe one has more entertainment than the other. But; there wasn't really much in the way of "cruise" ships back then, and the reverse is true today.

Ara Pacis
2012-Apr-02, 09:00 PM
Where did I say cruise ship?Context.

Besides, the concept is the same... a big ship that carries lots of passengers.Fair enough.

Granted, maybe one has more entertainment than the other. But; there wasn't really much in the way of "cruise" ships back then, and the reverse is true today.Kinda the point. But it's also not just about entertainment, but also operation. the first dedicated cruise ship was constructed a decade before the Titanic, and the concept of a dedicated pleasure craft used for cruising can be traced back to antiquity.

Luckmeister
2012-Apr-02, 09:22 PM
What gets me is how often I hear people comment that things happen in "threes." Two similar dire things happen to someone (like car accidents) so they say, "Watch out, everything happens in threes." I ask them what proof they have of that.... to no avail. :rolleyes:

publiusr
2012-Apr-02, 10:08 PM
Oh well, Maybe that superstition will make them a bit more careful. With more and more artificial implements moving a lot of foce, life does seem to resemble the Final Destination movies, 1000 ways to die, Darwin awards, etc. Just more opportunities to go wrong.

That's the skeptical answer, and the proper one, even if its cooler to suggest that all our guardian angels have turned drunks...


EDIT: Would you believe that I just had a bad coughing fit as I finished the above statement? Somebody took offense ;)

Strange
2012-Apr-03, 07:17 AM
What gets me is how often I hear people comment that things happen in "threes." Two similar dire things happen to someone (like car accidents) so they say, "Watch out, everything happens in threes." I ask them what proof they have of that.... to no avail. :rolleyes:

The proof is that (eventually) the third shoe will drop.

DonM435
2012-Apr-03, 03:11 PM
On one of the old 1960s sitcom shows, a mysterious fortune-teller tells the lead player that he would have three misfortunes the next day.

Sure enough, he (the typical paternal doofus) injures himself twice before breakfast.

The helpful neighbor suggests that he go outside for safety's sake, because "They say that two-thirds of accidents happen in the home."

"Oh," says the wife, "but he's already had his two-thirds, so he's safer staying home."

(I think it was The Donna Reed Show.)

Ara Pacis
2012-Apr-03, 05:39 PM
Or people say that most accidents happen within a mile of home, so you drive a mile down the road and realize you're in the slums and wonder if you're about to be an outlier.

Gillianren
2012-Apr-03, 06:55 PM
I don't think true accidents are more likely to happen in the slums than anywhere else.

HenrikOlsen
2012-Apr-03, 06:59 PM
Exactly, getting mugged because you were stupid enough to enter the wrong street at the wrong time is not an accident.

publiusr
2012-Apr-13, 06:54 PM
Friday the 13th seems to work. Had to go to the doctor and my toilets--both of them--are out of commission. I can't win.

More card debt...yay!

grapes
2012-Apr-14, 06:28 PM
(I think it was The Donna Reed Show.)To me, sounds like Dick Van Dyke. :)

I often promulgate the occurs-in-threes rule, but I also stand eggs on the equinox, and avoid pennies with heads up. If someone wants to pick those up, more fun for me.

Gillianren
2012-Apr-14, 07:16 PM
That's funny--around here and back in LA, it was tails-up pennies that were bad luck!

DonM435
2012-Apr-15, 01:54 PM
To me, sounds like Dick Van Dyke. :)

I often promulgate the occurs-in-threes rule, but I also stand eggs on the equinox, and avoid pennies with heads up. If someone wants to pick those up, more fun for me.

The doofus father (i.e., pater doofii ) was interchangeable.