Do you know what is the most dangerous place
When I say dangerous, I mean
a place where it is most probable for you
What is the most dangerous place in France?
the most dangerous in France, the one where you are most likely
to die, is actually a bed.
Indeed, about 70% of deaths are happening in retirement homes
or in hospitals, and for an important majority of them
in a bed.
Besides, Mark Twain himself once said:
“Bed is the most dangerous place in the world,since 99% of people die there.”
In fact, all of this comes from a mistake, which is very classical in sciences,
and outright common in statistics,
and is named the “causalation effect” (also called “stork effect” literally translated from French)
– Sounds a bit dummy, for a name.
– in latin, it is called “Post hoc ergo propter hoc”; is it better for you?
ERGO PROPTER HOCwhich literally means
“after this, and then because of this”
It is indeed a common mistake
of confusing correlation and causation. You know, when someone tells you you can
make statistics say whatever you want, what’s hiding behind
is this : a confusion between correlation and causation.
When there is causation, there is correlation: when it rains, the ground gets wet.
Sometimes, obviously, the connection is slightly more subtle than this:
When it rains, the ground gets wet.When it rains, the ground gets wet. When it rains, the ground gets wet, cars go more slowly.
Therefore, when it rains, cars go more slowly.
It’s still doing quite well.
But if I tell you that from my window I see a wind turbine;
and that the more the wind turbine turns, the windier;
the wind turbine creates wind.
Then here, this an obvious mistake, but it is exactly
this kind of mistake.
and the problem is that we rely
so much on figures nowadays
that we no more seek
to understand if what we are saying makes sense or not.
A study showed that if we fall asleep with our shoes, we increase our chances
to wake up with a headache.
It is also a classical mistake whist consists in not considering
that both of the correlation elements have a common cause.I say it again (in english this time): if I get massively drunk
all night-long, I increase my chances to get asleep
with my shoes,as well as the odds to wake up with a headache.
and examples like this one, there are plenty of them.
I will give you one which is a little more subversive:
if a high school student smokes marijuana, he/she increases his chances to get
poor marks. Is marijuana responsible for
his/her poor marks?Does he/she start smoking marijuana because he/she has poor marks ?
or is he/she in such environment that he/she at the same time, has poor marks,
and starts smoking marijuana? I’m going to take another example
which will be meaningful for everyone
— because you could not possibly miss it these last days —
Since Neymar Jr is injured…
(sigh) no, this is George Michael…
…there you go…
so, since Neymar Jr is injured,
Brasil’s soccer team can’t play anymore…
…1-7 against Germany.QED!
there you go.
Of course, the problem with this causalation, is that in generalit’s completely preposterous, but, in general, it is not as preposterous as it seems…
It is often very difficult to measure
effectively the difference there is between a simple correlation
and a causation; as it often involves a lot of parameters.
Because that’s what life is about: a lot of parameters.
It drives us CRAZY!!
By the way, it exists a parody of religion which is called pastafarianism.Yes. Pastafarianism.
Which gospel states that global warming
is due to pirates’ decline.
For yes, there is effectively a correlationbetween pirates’ decline and the average temperature on the Earth’s surface.
There is a website called “Spurious correlations”
— obviously, it is American — which takes statistical
data that are fully accessible, and tries to find
correlations, in order to show that, sometimes, we find something, and that, sometimes,
we find anything…
The more we eat ice cream, the more we feel like to watch a movie featuring Nicolas Cage.
This website also enables one to notice that the less we arrest youngsters smoking
marijuana, the more the bees produce honey.Anyway, I could spread out many, there are hundreds and hundreds of them,
the link is in the description below,
some are more interesting than others, some more peculiar than others,
what matters is that you fully understand this:
one can make figures say anything.
It is essential
to keep a clear mind on what we are told; moreover and for
your information, I allow myself to tell you how scientists knowhow to not fall in such a trap. An experienced scientist is
a scientist who doesn’t try to prove that what he thinks is true;
an experienced scientist seeks to prove that what he thinks is wrong.
And it is only when
he can’t prove it on the long run,
that he will convince himself
that what he thinks is right.
Take that, Aristotle!
Many thanks to all of you for being
more and more numerous to be willing to take some time to e-penser. You are now more than 50,000
to be willing to take some time to e-penser.
It’s crazy!It’s just crazy!
Today, we could fill a quite big soccer stadium
No clue on why we would do that, but we could do it! There you go!
Do not hesitate, of course, to share
this episode if you liked it
oh f*** it!
Share this episode whatever you liked it or not!
— why would I care after all —
Yes, that’s the way I am.
I will see you soon for a new episode, and till then, as usual,
keep being curious, and take some time to e-penser.
Subtitles by Matthieu Audras