Mentavo
Two things dropped in one month, and the sentence is there: "I always start and never follow through."
Sentences like that form quickly, sound like self-knowledge, and are often useless – not because they're harsh, but because they rest on a sample nobody counted.
Meanwhile, genuine repetitions go unnoticed for years. The reason is the same: your head is not a reliable counter.
What separates a pattern from a cluster comes down to four points.
Why memory doesn't count as evidence
Memory favors the striking. Two abandoned projects stick; seven finished ones dissolve into ordinary life – especially if they went unspectacularly.
There's a second effect on top: once you have a suspicion, you find support for it. After the sentence "I never follow through", the cases that come to mind are mostly the ones that fit.
This isn't a personal weakness, it's a general feature of remembering. It can't be switched off, but it can be routed around: with notes written before the suspicion existed.
The number that's almost always missing
"Three times makes a pattern" is a popular rule of thumb and worthless without a second number: how many opportunities were there in total?
Three abandoned out of four is an entirely different thing from three abandoned out of fifteen. In the first case dropping things is the rule, in the second it's the exception – and the sentences you tell yourself should differ accordingly.
In practice: before naming a pattern, add the opportunities.
| Observation | Without opportunities | With opportunities |
|---|---|---|
| Three projects dropped | "I don't see things through" | 3 of 4 or 3 of 15 – two different statements |
| Two arguments after family visits | "Family visits escalate" | 2 of 3 visits or 2 of 11 |
| Unproductive on Mondays | "Mondays are lost" | 4 of 8 Mondays, 3 of them after short nights |
Four criteria
Once the numbers are there, four questions decide whether the word pattern is earned.
1. Ratio, not count
Does the behavior occur in a substantial share of the opportunities – or in a small share that simply gets remembered well?
2. Predictive power
A usable pattern allows a forecast. If you suspect that conversations with a particular person regularly drain you, write that down beforehand and check afterwards.
The order is what matters. An explanation after the fact almost always fits; a prediction doesn't.
3. Independence from the setting
Does it show up across different situations or only in one? Being quiet in a single team isn't a personality trait, it's a statement about that team.
This is one of the most common mix-ups: context read as character. From the other direction, the piece on how your environment shapes your habits covers how strongly surroundings steer behavior.
4. Where the evidence comes from
Do the examples come from notes or from memory? Only the first is a sample. Everything else is a selection your suspicion helped assemble.
The counter-check
The single most effective step is an uncomfortable one: go looking for counterexamples on purpose.
Concretely, spend five minutes collecting cases where the suspected pattern did not occur. When did something get finished? When was Monday fine? When did the visit stay calm?
With a real pattern you'll find few counterexamples, and they often have something in common – which delivers the most useful part of the insight: the condition under which it doesn't happen.
With a cluster that has no system, counterexamples are everywhere. In that case it wasn't a pattern, it was a phase.
Why notes beat thinking here
The decisive advantage of recorded observations is that they were made before anybody went looking for anything.
Note briefly over several weeks what happened – without interpretation, as described in the piece on getting started with journaling – and you'll later have a basis that doesn't bend to your current mood.
That's exactly what longer-range overviews are for: in Mentavo, the Mentavo Map shows which parts of everyday life tend to occur together, and makes visible what each recognized connection is based on. It doesn't replace your own judgment – but it shifts the conversation from "my impression is" to "on these days it went like this".
How far that could eventually go is covered in the piece on how AI could recognize your blind spots.
A pattern explains nothing
Even with all four criteria met, a limit remains: repetition tells you that things occur together, not why.
"After long meetings I eat differently in the evening" might be about exhaustion, the hour, a skipped lunch, or the commute home. Take the first plausible explanation and you've found a pattern and invented a cause.
The cautious version is more useful: name the pattern, write two or three possible explanations beside it, and pay attention to the difference next time it occurs.
When the search does harm
There's an overdone version that reliably feels unpleasant: check every situation for patterns and you live inside a running evaluation of your own behavior.
One sign is that explanations start mattering more than experiences. What that tipping point looks like is described in the piece on self-reflection turning into self-monitoring.
A workable limit: one suspicion at a time, and the checking doesn't happen in the moment but in hindsight – in the weekly review, for instance.
How many opportunities are enough?
A fair follow-up question: at what point is it enough?
There's no universal number, but there is a workable rule of thumb. The rarer the behavior, the more opportunities you need before a ratio says anything.
- For something that could happen weekly, six to eight weeks is a reasonable starting point.
- For something that occurs a few times a year – moves, job applications, family conflicts – a statement about frequency barely holds; there, only a precise description of the individual cases carries weight.
- Two out of two is never a pattern, however clear-cut it feels.
It also helps to note your expectation in advance: "I'd guess this happens in roughly every second opportunity." If the actual rate is far below that, it was memory rather than a rule.
What this article doesn't claim
Patterns in your own behavior are not diagnoses. They describe frequencies, not disorders, and they're not suited to classifying yourself as a person.
Nor does the piece claim that only countable observations count. Some recurring things can't be divided into opportunities – there the assessment stays provisional, and that's fine.
Conclusion: opportunities first, pattern second
The difference between a repetition and an impression rarely lies in attention. It lies in the question of how often it could have happened.
Add that one number and you'll regularly reach different conclusions than your first suspicion did – sometimes gentler ones, sometimes sharper ones.
And in both cases, a sentence that can be checked.
