The Advice That Works for Everyone Works for No One
Why the advice that worked for someone else may be the wrong map for your life
There is a quote that has been attributed to various sources across the years, and its precise origin is less important than its precision: “All generalizations are wrong, including this one.”
It sounds clever. It is also, in the context of wellness advice, the most important thing you could know.
The Problem With “90% of Your Problems”
You have seen the posts.
“Fix these seven things and 90% of your problems disappear.”
“Most suffering comes from three things: comparison, resistance, and attachment.”
“The biggest skill you can develop is the ability to reset fast.”
Each of these statements sounds true. Some of them contain real insight. And yet, applied to any specific person in any specific situation, they may be entirely wrong — not because the observation is false in general, but because generalization applied to the particular fails by design.
This isn’t a problem with bad advice. It’s a problem with how advice works.
What Machine Learning Already Knows
In machine learning, there is a well-understood failure mode called overfitting.
A model is trained on a dataset. It learns the patterns in that dataset so well that it performs perfectly on the training data — and fails completely when applied to new cases it hasn’t seen before. It has learned the specific rather than the general. It mistakes the map for the territory.
The solution is a validation dataset — a separate set of cases the model was never trained on. If the model generalizes correctly, it performs well on both. If it overfits, it performs well on the training data and poorly on everything else.
Wellness advice has no validation dataset.
A recommendation emerges from observation — clinical experience, a study, a researcher’s own life, a pattern noticed across a specific group of people in specific circumstances. That observation gets packaged as a principle. The principle gets repeated until it feels like fact. Nobody asks: does this generalize? For whom does it work? For whom does it fail?
This isn’t a theoretical concern. Science has documented it at scale. When the Open Science Collaboration systematically retested published psychology findings, only 36% replicated successfully — and effect sizes that did replicate were roughly half as large as originally reported. Researchers have named the deeper problem the “generalizability crisis” — the systematic failure of findings to apply beyond the specific populations, conditions, and contexts in which they were originally produced. What worked in the study doesn’t work for you — not because you’re unusual, but because you were never in the training data.
The authority that makes wellness advice feel trustworthy — the PhD, the peer-reviewed citation, the expert panel — is a separate problem explored in an earlier piece: Science Tells You How. Nobody Told You That Wasn’t Enough. What this article addresses is what happens after the authority is accepted: the generalization fails the specific person, quietly, without explanation.
The “90% of your problems” list may accurately describe what worked for the specific person who wrote it, in their specific life, at a specific moment. Applied universally to everyone reading it, it is an overfitted model presented as a universal law.
And this is much bigger than wellness. The same structure runs through self-help (”copy these habits of successful people”), business (”follow this leadership framework”), finance (”this investment strategy changed my life”), parenting, productivity, and relationships. The “become a millionaire in 60 days” book is the same fallacy — the training dataset is usually one person who succeeded in specific conditions at a specific historical moment, packaged as a universal method. The validation failure is measurable: most people who follow the method don’t get the promised result. But those people don’t write books about it. The failures disappear. The success story gets published, amplified, and repeated until it feels like a proven path.
The Tolstoy Principle
Leo Tolstoy opened Anna Karenina with an observation that has proven more durable than most clinical research: “All happy families are alike; each unhappy family is unhappy in its own way.”
He was pointing at something that wellness advice consistently ignores: the problem is personal. Not in the sense that it is private or sensitive, but in the precise sense that the pattern generating the suffering is specific to that person’s history, that person’s installed rules, that person’s particular place where the map stopped matching the territory.
A generic solution addresses a generic problem. But the problems that actually persist — the ones that don’t respond to the obvious advice — are not generic. They are specific. And a model trained on everyone is a model optimized for no one in particular.
This is why the person who has tried every item on the list and still feels stuck is not failing to apply the advice correctly. They are experiencing the validation failure of an overfitted model.
The Questions Nobody Asks
When you encounter a wellness recommendation, the questions worth asking are precise:
What was the training data? A study of two hundred university students in one country is not humanity. A researcher’s personal experience is a dataset of one. Clinical observation across thousands of clients is more robust — but still filtered through the therapist’s framework, which is itself a generalization.
Was it validated on cases outside the original dataset? Most wellness advice isn’t. It is generated from observation, packaged as principle, and distributed before anyone asks whether it holds in different populations, different contexts, different life situations.
What is the success rate — and who does it fail? A recommendation that works for sixty percent of people sounds promising until you are in the forty percent. The failure rate is rarely mentioned, because the person giving the advice is usually someone for whom it worked. And the people for whom it didn’t work quietly disappear from the conversation.
Is this a rule or a correlation? “Successful people wake up early” may be a genuine observation about a specific population. It tells you nothing about whether waking up early causes success, or whether the recommendation applies to you.
The Personal Diagnostic
The alternative to generic rules is not more sophisticated generic rules. It is a different question entirely.
Not: what does the advice say I should do?
But: what in my specific life is actually working — and what isn’t?
What’s working shows up as aliveness — the desire to move, to act, to engage, to do more of the same. Energy that replenishes rather than depletes. A sense of rightness that doesn’t require forcing. What’s not working shows up as friction, resistance, recurring patterns that keep producing the same results despite repeated effort, and a tiredness that sleep doesn’t fix.
One final point worth making: the people who followed the generic rules and found them insufficient are not at fault. They were given a map built for someone else and told it was universal. The failure isn’t in the person. It’s in the map.
The next time a piece of advice tells you what works for everyone: ask what the training data was. Then ask what’s actually working in your own life right now — and what the evidence of your experience is telling you needs to change.