The Metric Is The Measure

One of the more interesting phenomena that one encounters as a student of data is the way that discussions are often framed selectively with regards to certain convenient and self-serving metrics. For example, if one is looking at a conversation of the success of yet another Nicki Minaj single on the sales charts, one will often find fans of that artist emphasizing the success that she has on that particular chart, even if the songs do not get much radio play nor do they remain on the chart for a very long time. Similarly, there are artists who sell a lot of copies of albums but who do not tend to have a lot of big hits, and such fans tend to focus on album sales as the metric of choice when comparing their favorite artists to others. This sort of scenario can be multiplied indefinitely. One can discuss politics where some people will focus on how many popular votes of (possibly deceased or imaginary) morons their party or candidate has won in an election while others talk about how many states their party has won, each of them choosing a metric that supports the argument that they are trying to make.

Let us understand that anything which can be measured can be measured in a great many ways. It is hard to know if any particular metrics are representative of a given thing or whether it is best to include as many different measurements as possible. My own approach is the latter, and I have often found it very interesting to see a bunch of metrics and measurements on the same data, not concerned to make any of them a measure of judgment but rather to see the same things from a variety of different perspectives to see if patterns can be found that provide insight. There is no need to prejudge such a search by only looking for measurements that will support a given conclusion that we wish to make.

What we need to understand is that when we choose what metric we use to judge something, we are not just measuring the thing that we have in mind, but the metric we choose is a measurement of ourselves. How are we seeking to frame what we are measuring? Do we choose measurements that seek to build it up or to tear it down? What we seek to build up and tear down says a lot about us, and we are often rather insensitive to this. We may either seek to promote certain measurements as being legitimate because they tend to be ones which favor us. We may use certain measurements which delegitimize what we are trying to condemn, and similarly those metrics will be cherry-picked. But however we go about it, we are saying far more about ourselves than we are saying about the thing that we are talking about, because we could have chosen any number of ways to discuss something, but chose something very specifically for reasons that served our own rhetoric.

One of the major reasons why people are somewhat skeptical of the use of data in debates and arguments is precisely the way that such metrics are often weaponized to promote certain judgments in a predictable and not always legitimate fashion. To the extent that we can recognize the way that different metrics tell us different parts of the story, we can look at things in a larger point of view by combining a lot of metrics, allowing us to categorize things based on the patterns we find in various metrics and viewing them with other members of their kind. Thus, rather than seeking to delegitimize them or praise them, we are seeking to understand where they belong and what other things are like them, a far more useful task in general. To use data to understand is wise; to manipulate and frame a discussion of a given thing is a sign of the wickedness of the present age.

About nathanalbright

I'm a person with diverse interests who loves to read. If you want to know something about me, just ask.
This entry was posted in Musings. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s