P2P No. 31 — Single measure, complex thing
On the pitfalls of GDP, citations, and five-star reviews.
Peter Medawar, in his essays on science, critiques science for trying to describe complex things, such as intelligence or economic welfare, by a single value. This is impossible, he argues, since the nuance gets lost. This point struck a cord: I believe that comparison can be harmful, and now I see more clearly why:
A single measure of success, such as the number of scientific papers, does not tell you the complex backstory.
Unfortunately, sometimes these proxies matter; more precisely, people often use them for comparison—because it makes comparison easy. However,
when there is a metric, there will be a way to trick it.
I understand the human desire to distill complex phenomena into a single truth, though this cannot be done without losing details. However, when only the big picture matters, then details can be distracting. Evaluating candidates does not fall into that category, though.
Even looking for the most delicious ice cream is too complex to decide only based on the average review score: if an ice cream vendor gets an average of three stars out of five, you avoid that place. However, if reading the reviews tells you that the ice cream is the best in the city, but the queue is always very long, and you happen not to care about that, then you would be deceived by only looking at the average score.
Going nerdy, I could say that life is a multiobjective optimization task. This means that you want to get to “the optimum” by reaching multiple goals simultaneously: in your profession, hobbies, and relationships. The caveat of single metrics is, exactly as in multiobjective optimization, that detail is lost. I could get the same life satisfaction (think: same average review score) by being successful professionally and not having time for my hobbies, or vice versa—assuming that I value both the same (which I do not).
More importantly, single metrics get rid of the story, which can hurt when that metric determines whether you succeed. For job applications, the cover letter can provide the nuance and boundary conditions to show your evaluators (including yourself, when self-doubt nags at you) that from the cards you got dealt, this is the best anyone could have gotten out.
Actionable advice
When making a comparison, someone might (read: you will always find someone) be better than you w.r.t. a single metric, but that does not mean that you should do the same. Remember, the subject of your comparison is playing from different cards and probably has a different life goal.