We all start thinking so in
some form ever experienced
to have
we wake up one morning and us
it is not so good we have it on
headache us is evil and it will
not better over the day and then
we probably do the dumbest
what we could do
like google in our symptoms and you
know that again runs ten minutes
later one is convinced that one that
it comes to an end
you do not have much longer to live but
is that really is what really
rational other situation
new job first working day
We all learn our employees
know and we would say it was
good if we are such a two from the beginning
people we have to stick to
to whom we can trust
after the first working come to
home and think us how can we
then our employees will appreciate
situation 3 many of us have profound
Conviction about different things
I call that little personal
truth with me for example because that
so while I went to school
am
I was deeply convinced that it was
are two kinds of people, people who
understand mathematics and such the
just were not made around
to understand mathematics and I was up
In any case, the idea for a warehouse is more likely
the mathematics nothing is to then to
second high school and when I graduated
have learned and me a bit
it has been more closely involved
suddenly started making fun and
I understood things I was not
understood and that has of course
this fundamental belief that I
always had somehow shaken and
I did not really know how to do it
deal with it because he situations with
We always encounter them in everyday life
again
We make assumptions about ourselves
yourself about others about the whole
world
and then we do observations
either support these hypotheses
or contradict them
and the question is how do we handle it
and is there a rational way with it
deal and that should be my stoxx
had to be when we talk about it
to deal rationally with hypotheses
then a good starting point maybe
the scientific method in the
you have scientific method in
To hypothesize one either says something is
it is true or or that is wrong
let’s say we take the example
from our first day of work we learn
We want to know Paul and that’s what we want
somehow decide if we trust
can not so is a cool guy
or not if we have one there
then scientifically would answer
we would say okay it is the same
probably that ‘s a cool one
type is and not
we make no assumptions in principle
next step we collect quite a lot
Information about days and weeks
away in all interactions with Paul
think about what that means is a
trusting persons or not if
we all have a lot of information
have collected
We decide either we can
Do not trust him too many negative
had interactions or we can
the trust of many positive
had experiences
There are only a few problems with it
this method in everyday life
For one thing, we have just these
assumptions we make but that is
our humanity for example
or our deepest convictions and
that is absolutely unrealistic that we
go through life and every day
wake up and say maybe we are
we may not leave you dead
more information will gather the next one
problem is that we are usually simple
not so much information to
have available
if we think we are dying
we just do not have time long
wait and see how that looks
we prefer to develop one
bit of panic expired and the third
point is that not every observation
So important is some
Observations tell us more about
our year’s conviction and
and some less and one way out
thomas base offers this situation
Statistician who lived 300 years ago
with Thomas pace was the eponym
probably the most important
theories of statistics for that here
but that is not all
not so bad that we have one now
mathematical formula been with the me
you did not want to shock much anyway
more important is actually the
the intuition behind these theories it
unconditional probability it goes
is about estimating how
probably a hypothesis right
is given that we have a certain
make observation
How does it work?
that we think before we own
observations
how likely is this hypothesis
that was our example
have human knowledge and say if
we know each other know a person
learn and then we ask ourselves like
We do not have that person then we have to
intuitively more but then he does not bring
that could be our basis
his chance we were like
so first he does not do it then
we are watching the second step
we think it suits us
observation to our hypothesis or www
she contradicts our hypothesis and
how can we reconcile this?
and the third step is not
as just described
either to maintain the hypothesis
or to discard but we update
just our hypothesis so the
Probability that we are the other
find a person cool rises or falls
bit but we do not decide
cool person or not cool person
even more clearly, it is perhaps the
this formula as we do not currently
have seen what is not bad
if you do that once you do that
practically on with the in parts
small example we go to the doctor
routine tests on a disease which is about
ten out of every 1,000 people are developing
get a week when calls from
doctor and the test positive
The doctor also tells us that
Probability that this test
accurate is only 90 percent
Now the question is how likely
we have this disease most
People would say 90 percent shit
but it is not that easy
we look at the fact times
numbers
that is 1000 people we know that
ten from that 1000 people
statistically have the disease
these ten people are tested and
to 90 percent comes a real one
Result therefore a positive result
out but with a person that will be
misdiagnosed whole and the whole
then get negative diagnosis so far
of course
but what you can not forget now
is that these 990 healthy people
also be tested
and here too is the
error probability 1 out of 10 so
One in ten people will be wrong
So diagnosed although this
people are healthy are 99 people
get a positive result though
they are not sick at all and that could be
you can now use them in the
mechanical formula and then you would
consider yourself okay what is that
Probably that we are the
illness has given us that one
positive result from the format back
got and it’s only nine
Persons the correct positive
got result and 99 persons
which is falsely a positive one
got results which we called
a positive result our
chance to actually get sick too
his is still 8 3 percent
and that we are actually healthy
as at 91.7 percent yes and pay this
that sounds somehow come up again
the example
but it is actually very close to
example
in the right life and that in mum
graphic studies in women
Breast cancer screening is that
about the same pay mechanical
thinking is based on this formula and on
this simple principle we have one
hypothesis
We have a base probability
we think about it
What does this observation mean to us?
our hypothesis and then we update
our hypothesis and have in principle
nine values just like one
positive result that ours
Probably 9 percent that for
the disease have what is now so
exciting at this theory why
I do why do I do and have
about that has two reasons to
One is this mechanical formula
expression of a whole philosophical
direction base was not just a statistician
He was also a theologian and his first
The fair had to be written like his
God can be almighty at the same time
and is valid and that it does anyway
suffer in the world there and base has
we argue we may own our own
fallibility does not coincide with the fallibility
from god confused he meant we
Can the truth of God in principle
we do not really experience that
can never really understand that
and the people are now mistakes and
we can never do this truth
suffice
that sounds very religious and somehow
for the first time not at all
statistics but fair will be a bit like that
clearer from the example of the essay
was named in which it is exactly this
Probability front he went was
published by price that was a
good friend of base and he has this
essay published after base death and
he described the following example
I imagine a person in front of the
cave lives a cave like plato
at that time, too, for example
has introduced and comes in person that
this cave out and looks to the first
times the sun rise and the people
I do not know if this is an anomaly
or is that a regularity
is and the person survives every day
on the world you turn to the sun again
who the sun rises and the person
is always a bit safer
every day that it is a regularity
is and not an anomaly but i really
100 percent can be sure of this
you person that the sun really
it’s a little bit better
the essence of mechanical we think
Never really 100 percent sure
but we can always be safer
with every observation we become
do as others why base formula
interesting is that she
it’s all over there now
There are so many use cases last
week there was a bad hurricane
Usa hurricane always and first saw it that way
out as if the hurricane heading for miami
turns and then at some point they have
changed and it was called
but rather direction tampa and this
Hurricane predictions are based on
Brazilian bill man
look at what is the situation
just for other observations
we have and because of this
we have to update in
which direction of the hurricane goes if
an airplane crashes more where you can
do not know where to look for him
Been also applied so far
make a hypothesis about it in which
the plane is probably located
then look at things like
currents linde et cetera around then
to further narrow down in which
area that aircraft could be lying
if you use a e-mail program yours
spam filter is basically one too
bit of spockfilter model looks
how likely is one
e-mail a spam email given that
words that are in there right now
for example penis enlargement are there
good indicators for that there is one
So spam emails is on one side
that this philosophical whole whole
philosophical direction in this
little formula you put on the
other side of this huge potential
these huge application possibilities
and we can do that in everyday life
use this formula and a few
Examples and we go these three
steps after
1 the hypothesis is our first example
was
we feel like a sick headache
abdominal pain like google what we have
could
and somehow your first hit is
probably meningitis
then there are symptoms
Dizziness is dizzy, too
but what you do not forget
may is one thing
Meningitis in Western Europe
simply very unlikely
the probability of a hindered
Develop first is 0.0 0 3
percent more for this example at times
that the probability is a percent
is the basis probability of
Next step is we think ourselves
our symptoms are ours
observation how likely are
These symptoms have been caused by
a meningitis or if we
a brain inflammation would have like
we probably would this
yes, that has a lot of symptoms
but probably that classic
be symptoms of lengthen to accept
Once 90 percent then we have to
ask what if we had the disease
do not have this bad illness
what other possibilities are there
that we observe these symptoms
It could be a flu , for example
namely, like to be weather-sensitive
but maybe not
weather sensitivity because of our symptoms
So have not been for a long time
more such headaches have
let’s say 75 percent and now
we do the same again what we
just made
we only compare the cases in
which we observed the symptoms and
we look at what is the share
the red area on the area blue
plus red and are just 1.3 percent
So base probability that we
a meningitis have at 1
percent the symptoms are not
particularly meaningful because of that
does not increase our probability
so wet yes and the probability
that because of other illnesses have the
not that bad at 898
percent observation I have earlier
Not every observation is the same
important and that is also true in the
For example, if we now to
example still very high goal
still would get
photosensitivity would come to it
neck stiffness then that’s in the crowd
Symptoms are very unlikely
somehow of the simple flu
be caused and we have ours
Probably update that it does
probably that’s a bad one
Disease is we should go immediately
go to the doctor
yes we can in the example with
with paul apply to our staff
We meet Paul first impression is quite
good and our human knowledge is too
pretty good actually at three
of four people we meet
we correctly with our first
assessment and that’s why we say
okay 75 percent chance that
the paul a cool guy is first day
first meeting with the boss and powell
stands out because he is our boss
The city praises and now we think
What does that say about Paul?
if Paul’s nice guy would be ours
Chief praise but probably already
is already a very nice gesture but
in paul would be no negative or no
Trustworthy guy is also very
probably that he is the boss and
just to beat them around
to stand there and say that observation
not really a lot of our
Probability remains at 75 percent
next day we have problems with
our computer and we ask paul
help and he actually helps us one
for forty-five minutes this problem too
remedy
we ask ourselves again how
probably that’s good if Paul
type would be that he makes us
would be very likely
small probability still not that
if just really
deal with something else
yes and if he is like no one no person
we should stick to
is unlikely otherwise
So helps very valuable information
now we are 93 percent sure
that we can trust
the third step is the update and there
It is important for us to be continuous
update so we’ll make you one
observations and then decide
but we do a observation
update our chances
make another one
observation and so on and so on yes
in the example of my little mat
dilemma I started sometime
to give statistics and yes many
I met people who said that
I’m not a math type at all
not at all in the end very good
so statistics are written and are
even own statistics on viewers
won and with each one of these
I could basically observe my observation
my mine was reversed and say
so now there is no small one
no type question everyone can in principle
dull men would like that brings me
even at the end always these three
Steps take note of what our base
probability as meaningful
is our observations actually
and how does our customized look
Hypothesis from three financial statements
Maybe some might say now
yes we have that great
subjectively we want somehow
rational is also somehow objective
but that’s not so bad
yes, it is important that we are at all
first make clear what we
Probably determined for me
Events when eating and the formula
still works my second point
It ‘s good to make it’s name
good hypotheses we should set up
do that as often as possible but that
the third point we should always
already one of our hypotheses too
adjust
base formula does not work if the
probability is 100 percent or
0% and as base has already said
it’s probably about the end
not really the truth
Finding or knowing is really
it’s always a little less
to be wrong
Thank you very much
”
Be First to Comment