Why big goals resist being broken down
Ask someone to break down "get fit" and you get: join a gym, go three times a week, eat better, sleep more. That is not a breakdown. It is the same goal restated at a smaller font size, and it fails the same way, because none of those items tells you what to do on a Tuesday when you are tired and the gym is shut.
The problem is that most breakdowns split by topic when the useful split is by uncertainty. Topics are easy to generate and tell you nothing. Uncertainty tells you where the goal will actually break.
Split by what you do not yet know
Take the goal and ask: what would have to be true for this to happen, and which of those things am I least sure about?
For "get fit," the honest answer for most people is not the exercise. It is: whether any routine survives a bad week, whether the thing is enjoyable enough to repeat, and whether the schedule it assumes actually exists. Those are the branches. "Go three times a week" is a guess at the answer to the second one, presented as if it were a plan.
For "change careers," the uncertainties are usually: whether you would like the new work in practice, whether anyone would hire you without the standard background, and whether you can survive the income gap. Everything else — CV, portfolio, courses — is downstream of those three and cheap to sequence once they are resolved.
Splitting this way produces four to six branches instead of twenty tasks, and each one is a question you can actually attack.
Make the first move small enough to be embarrassing
For each branch, name the smallest action that would produce real information. Not the first step of the eventual plan — the cheapest possible probe.
- "Would I like this work?" → one conversation with someone doing it.
- "Would anyone hire me?" → read ten real job ads and mark which requirements you genuinely fail.
- "Does the schedule exist?" → track next week honestly before designing anything.
Each takes under an hour and each can invalidate a plan you might otherwise spend four months on. That asymmetry is the entire argument for working this way.
Stop when the branch can fail on its own
A common question is how deep to go. The answer is: stop splitting when a branch could fail for its own reasons rather than as a symptom of its parent.
"Build a portfolio" can fail independently — you might build the wrong thing, or nothing. "Open the editor" cannot; it fails only because the portfolio branch failed. So the portfolio is a branch and opening the editor is a Todo inside it.
This is also where structure stops being free. Every level you add is a level you have to read past forever. In Minty Path, parent branches are structural — they take no manual status and complete automatically once every end branch beneath them is Complete — which keeps the cost of an extra layer honest. You never end up hand-maintaining progress on an organisational heading.
Only some of it is active
Once the map exists, the temptation is to start everywhere. Do not. Choose one to three branches as the current Focus and let the rest sit visible and inactive.
The unfocused branches are doing work by existing. They are the reason you can stop thinking about nutrition while you find out whether the schedule is real — the map is holding it, so you do not have to. A plan whose whole value is that everything is written down loses that value the moment everything is also active.
Define the finish before you start
For each branch, decide what would exist afterwards that does not exist now. A conversation that happened. A file. A number you measured. A page someone else can open.
Minty Path will not let a branch reach Complete without an Evidence entry attached — a file, link, screenshot, result, or a written note explaining what you applied. That sounds like friction and behaves like clarity: you cannot quietly declare a branch finished because you got bored of it. Either evidence exists or the branch is still open, and knowing which is most of what a plan is for.
Re-split when the map stops matching reality
The first breakdown is a guess made with the least information you will ever have. Expect two or three restructures.
The signal is usually the same: one branch keeps growing while the others sit still. That normally means it was never one branch. Split it, and often move it out entirely — in Minty Path, Copy, Move, and Link all turn a branch into its own Growth Path, and a branch that has quietly become a project deserves that promotion rather than continuing to distort the map it started in.
Where breakdowns usually go wrong
- Splitting into a checklist. Twenty sequential tasks is not a breakdown; it is the same goal with more surface area to feel behind on.
- Going three levels deep on day one. Detail invented before contact with reality is fiction, and it is expensive fiction because you will defend it.
- Splitting by topic instead of uncertainty. Nutrition/exercise/sleep looks organised and tells you nothing about which one will sink the goal.
- Leaving everything active. If all six branches are current, you have not broken down the goal, you have multiplied it.
- Refusing to delete. A branch you have skipped past in three consecutive reviews is not waiting. It is finished, or it was never real.