[PREV - SCATTER_THE_NOISE] [TOP]
PRECISELY_ACCURATE
August 7, 2023
The canonical example is target shooting.
In the ideal case all your shots will group
together at the center, but in reality
they'll often be scattered widely, and the
center of the scatter might not even be at
the center of the target.
We often analyze these results by looking
at how well the average gets near the
bullseye, but also by looking at the size
of the grouping, e.g. by calculating the
"standard deviation".
In the usual jargon, we say that the shots
are "accurate" if the average is near the
center, but we say the shots are "precise" This distinction is very
if the grouping is tight. common, but the jargon is
really remarkably awkward: it
This can be a useful way doesn't map well to colloquial
of analyzing the result: understanding of the words,
which seem like synonyms to
You might, for example, find that anyone without a technical
all your shots are actually background.
grouped tightly together, but not
centered on the target (high Myself, I've been familiar
precision but low accuracy), and with it for many years, but
you can conclude from this that even so I need to stop and
there's nothing wrong with your think sometimes to make sure
aim, but there could be an issue I haven't gotten the terms
with the gun sight that needs to backwards.
be fixed.
But: while the precision/accuracy
distinction can be revealing,
I think it's important to remember
it's an artificial imposition on
the data.
This is an analytical device
which might tell you
something about underlying
phenomena, but there's no
reason it has to.
Treating accuracy and
precision like two knobs
that can always be spun
independently of each SCATTER_THE_NOISE
other is often going to be
a mistake.
The reason I've indulged in
the above rather conventional
exposition is that something
similar is done in:
"Noise" (2021) by
Kahneman, Sibony and Sunstein See the introduction,
"Two Kinds of Error"
SCATTER_THE_NOISE
But I think the conventional exposition
is clearer, because it works with the
individual shooter case. Sunstein et
al have got their eye on the
predictions and judgments made by
multiple experts in an organization,
and so the authors keep trying to talk
about shooting teams, but that fuzzes The entire point of the
things up too much, e.g. it makes it distinction is that it
harder to bring in the issue of can point you at an
individual aim vs the quality of the underlying cause:
gun sights. shooter vs gun sight.
The idea that an entire
shooting team would only
have one gun between them
is a stretch.
And really, if you wanted
to analyze the results of
a shooting team, the place
where you would begin is
an individual analysis of
the data for each shooter.
And rather than work with the admittedly
awkward terms "precision" and "accuracy",
the authors introduce their own favored
terms "noise" and "bias".
Neither are without problems.
To my ear, the word "noise" The word "bias" has
suggests issues with interference severe negative
in communications-- not scatter connotations:
in judgment and prediction. technically it may just
mean any systematic
Further "noise" also strongly error, but colloquially
suggests "randomness" which could it suggests things like
be begging an important question. racial bias.
"Noise" suggests "Bias" suggests
unintended, accidental. hidden agendas,
Neither, I think, is
something you would want
to assume at outset. I'm capable of
intentional
noise, myself:
NOISE
It's a continual irritation to me that
the word "noise" is abused in every
other sentence of this book-- for me,
it's a fingernails-on-blackboard noise. This rhetorical "fingernails"
still feels like a useful
I think they were working too figure of speech to me,
hard at trying to sound edgey. though really half of the
people alive now have never
Not to mention original. even seen a blackboard, let
This bold new discovery they're alone heard one.
talking about isn't all that
new, and their prescription
doesn't strike me as the panacea
they make it out to be.
--------
[NEXT - SCATTER_THE_PAGES]