Risk review.
Two scenarios on the board.
One:
"The payments provider
has an outage this year."
Two:
"The payments provider
has an outage this year,
during our busiest weekend,
because of a bad certificate rotation,
and their status page says nothing for an hour."
The room spent forty minutes
on the second one.
It felt real.
You could picture it.
Somebody had seen
something just like it.
The first one
felt like filler.
But the second
is a special case of the first.
Every outage
on the busiest weekend
caused by a certificate
is also an outage.
It cannot be more likely.
It can only be less.
In 1983
Amos Tversky and Daniel Kahneman
described a woman called Linda.
Clever,
outspoken,
concerned with social justice.
Then they asked
which was more probable.
That Linda is a bank teller.
Or that Linda is a bank teller
and active in the feminist movement.
Most people picked the second.
It fits the description better,
so it feels more likely.
They called it
the conjunction fallacy.
In the same paper
professional forecasters
did it too,
rating a detailed chain of events
as more probable
than one link of it alone.
Here is why it happens.
The mind does not compute probability.
It checks how well the story hangs together.
Each detail
makes the story
easier to believe,
and every detail
also has to happen.
Coherence goes up.
Probability goes down.
You feel only the first.
So watch for it
wherever stories are told.
In risk registers.
In design reviews
where the scary scenario
gets the hour.
In incident reviews,
where the most complete explanation
wins the room
before the evidence is in.
Write the bare version
next to the detailed one.
Ask what each extra detail
needs to be true.
Plan for the broad case
first.
The vivid story
is easy to prepare for
and rarely the one
that arrives.
– Serguey Asael Shinder
Top comments (0)