Skip to content
Go back

Keeping metrics tied to what matters

By Leo Traven

TL;DR

Metrics make progress easier to see

Someone starts exercising because they want to feel better in their body. A scale offers a convenient way to track progress. Its readings are easy to record, and a falling line gives them a clear sense of achievement. There are reasons to want that clarity. Memory is unreliable, and a record can reveal a change that would otherwise go unnoticed. A shared measurement also lets people compare results without having to explain every detail of their circumstances. Metrics can help people coordinate and hold each other accountable to an agreed standard.

In The Score: How to Stop Playing Somebody Else’s Game, philosopher C. Thi Nguyen examines why scoring systems can give activities direction while also drawing us away from what we care about. He explains that the usefulness and limitations of metrics are connected. A measure that people can easily share across contexts has to leave out some of the detail that makes each situation different. That makes a metric a partial view. Weight tells the person exercising something about their body. It cannot, by itself, tell them whether exercise is giving them the life they wanted.

A clear score can quietly replace the purpose

Getting healthier does not have to mean losing weight. Someone might want to sleep better, move more comfortably, or have enough strength to carry their shopping bags. Another person might care about endurance or flexibility. What they want from better health affects what would count as progress for them.

The substitution can happen gradually. The person begins by using their weight as a sign that exercise is helping. Over time, the number becomes the thing they want to improve. They start choosing activities according to their effect on the scale, and a week with no weight loss feels like failure regardless of how they experienced it. Nguyen calls this value capture: a simplified account of what matters takes over the richer values someone started with. He distinguishes being captured from responding to an incentive. Someone can pursue a score for practical reasons while still judging it against other things they care about. Capture occurs when the score becomes their own definition of success.

A person could decide that losing weight serves a particular purpose. The problem arises when that purpose disappears from their reasoning, so they no longer notice when the measure and the value diverge. Even a perfectly accurate measurement can then direct them toward something they never consciously chose.

Choosing a metric still requires a value judgment

The question is how to use that clarity without letting it decide what matters for us. A number can look objective because people agree on how to count it. Choosing what to count still involves a judgment. Treating weight as the definition of health makes a different choice from treating strength, sleep, or everyday comfort as the priority. An exact number cannot tell us which of those things matters most.

The same problem appears when a student chooses a university mainly through its ranking. Suppose the ranking rewards published research while the student cares most about teaching and close guidance. A university can improve its position without becoming better at what that student wants. The ranking’s apparent authority makes it easy to skip the decision about which qualities should matter.

We can also produce false clarity by narrowing our values to whatever is easiest to explain. A person who cares about several competing things may struggle to explain their choices clearly. Someone who judges everything through one score can produce a consistent explanation almost immediately. That consistency may come from having discarded the difficult parts.

As Nguyen puts it:

If we are sincere and respectful of the value density of the world, we will rarely be able to generate quick, coherent, accessible explanations.

A role can make us feel we have no choice. A manager might say that their job requires maximizing a particular target, treating the role as if it had settled every question about worthwhile work. Real constraints may limit their choices, but the target still contains choices about value. Following it does not make those choices disappear.

Keep judging the metric against the purpose

Start by being clear about what the activity is for. Then choose a measure that reflects that purpose as closely as possible, while naming what it leaves out. If someone wants enough strength to carry their shopping bags, observing whether that task becomes easier may be more useful than watching their weight. A grip-strength measurement could provide additional information. Its usefulness depends on whether it helps them understand the capability they care about.

Keep returning to the purpose after adopting the measure. Nguyen describes trying a step counter to see whether it helps with health, then remaining able to decide that it no longer helps. That judgment requires some understanding of health beyond the step count itself. When the number and the purpose disagree, investigate the gap. Perhaps the measure needs changing, perhaps the activity needs changing, or perhaps measurement is adding little. Someone who walks because they enjoy it can pay attention to that experience without turning every walk into a score.

Values can also develop through experience. What someone wants from exercise may change as they discover an activity they enjoy. Keeping track of the underlying value means allowing that reconsideration, rather than fixing a target once and letting it govern every later choice. Metrics are useful when they help us see and pursue what we care about. Preventing value capture requires keeping the underlying value in view, including the parts that remain difficult to count. We need to remain willing to revise or drop a metric when pursuing it draws us away from that value.


Share this post on:

Previous Post
Building self-reorganizing software systems
Next Post
Working across specialties