Why I Measure Everything

My scale reads to 0.1 grams. My thermometer reads to 0.1 degrees. My spreadsheet of household energy consumption goes back 1,847 days. People find this exhausting. I find the alternative exhausting: living inside systems whose behavior you cannot describe.

Here is the argument, in the order I discovered it.

You cannot own what you cannot describe

Ask someone how much they spent on subscriptions last year and watch them guess. The honest answer for most people is "whatever the platforms decided to charge accounts I've forgotten I have." That is not ownership of a budget. That is tenancy in one.

The same is true of attention, of health, of data. If the only party measuring your behavior is the platform profiting from it, then the measurements exist — they are just working for someone else. Every metric you don't collect about your own life is being collected anyway, upstream, by a party with different incentives.

Measurement is how you repatriate that information.

Numbers are a trustless protocol

The deeper reason I measure is that numbers don't require trust. When I say my espresso ratio is 1:2.1, you can replicate it. When I say my server's uptime was 99.4% last quarter, that claim is checkable against a log. Compare that to "the coffee's good" or "the service is reliable" — claims that require you to trust the speaker.

Decentralized systems work for exactly this reason. Bitcoin does not ask you to trust a banker; it asks you to verify a chain of measurements. A mesh network doesn't ask you to trust a carrier; it routes around whatever fails, measurably. Measurement is the primitive that makes trustlessness possible. A man who wants to depend on institutions less must, arithmetically, depend on measurement more.

The calibration habit

One practice I recommend above all others: when you make a prediction, write down a number. Not "I think remote work is here to stay" but "I put 0.85 on the median knowledge worker still being hybrid-or-remote in 2030."

Then check yourself later. Over enough predictions, you learn your own error bars — where you run overconfident, where you sandbag. I have been doing this for six years. My forecasts at 0.90 confidence have come true 87% of the time, which means my 0.90 is honest to within a few points. I know this the way I know my resting heart rate: because I logged it.

What I am not saying

I am not saying quantify your children's hugs. Some things are load-bearing precisely because they are unmeasured, and a man who cannot tell the difference has mistaken the instrument for the life.

But everything you depend on? Every system that can fail you, bill you, deplatform you, or quietly change its terms? Measure it. Get the number. Write it down somewhere that belongs to you.

Because the first step of decentralizing your life is producing your own copy of the truth.