The routines nobody wrote down

Every morning routine you have read about comes from someone it worked for. That is the whole problem.

The person who woke at five, wrote for an hour, and built a career on it has a story worth telling and a reason to tell it. The person who tried the same schedule for three weeks, slept badly, and went back to waking at seven wrote nothing down. There was nothing to write. Failure at a habit is not a story, it is an absence, and absences do not get published.

So the advice arrives pre-filtered. Not falsified, not exaggerated, just incomplete in one specific direction: you see the cases where the method and the person happened to fit, and none of the cases where they did not.

The filter runs more than once

Survivorship bias is usually drawn as a single cut: the sample you can see, and the sample removed before you saw it. In self-improvement the cut is unusually deep, because it happens several times over.

• Habits that fail early leave no artefact. Nobody writes up the cold plunge they abandoned on day four.
• Habits that fail slowly get reframed. A routine that quietly stopped working is remembered as a phase rather than a result.
• The people best placed to write about a method are the ones it suited, which is the least representative sample available.
• Distribution finishes the job. A routine that produced an ordinary outcome does not travel, so the surviving accounts skew toward the extremes.

Each pass removes the same kind of evidence: the ordinary, the partial, the abandoned. What is left is a library of methods that all appear to work, written by the people for whom they did.

What it costs you

The damage is not that you try something useless. Most published routines are reasonable, and some of them will suit you.

The damage is to your sense of the base rate. When every account you read ends in success, the implied success rate sits close to one, and a method that fails you starts to look like a personal failing rather than an ordinary outcome. That reading is wrong, and it is expensive. It pushes people to repeat a method harder instead of dropping it, and to treat a bad fit as a defect in themselves.

Advice can tell you a method worked for someone. It cannot tell you how many people it did not work for, because those people are not in the sample.

The correction is not general scepticism. It is remembering what the evidence actually supports: that the method is possible, not that it is likely.

The one record that is not filtered

Your own history is the only unfiltered evidence you have about what works for you. It includes the attempts that failed, which no external source will ever hand you, and it is drawn from a sample of exactly the person you are trying to advise.

That is a much smaller dataset than the internet's, and a much better one for this particular question. It has three properties the published version lacks:

• It contains the failures. You know which routines you abandoned, and roughly when you stopped.
• It is specific to your constraints: your sleep, your job, your commute, the people you live with.
• It was not written to persuade anyone, so nothing has been rounded off.

Keeping it is unglamorous. A date, what you did, whether it held, and the point at which it stopped is enough. The value is not in any single entry. It is in being able to look back across six months and see a pattern that no account of someone else's morning could have given you.

None of this makes the published routines worthless. Read them as a catalogue of what has been tried, not as a ranking of what works, and check the result against the only record that includes your failures.