Use caseThink
Think through a decision you keep going round on
A one-page memo in your folder: the options side by side on criteria you chose, what each gives up, the assumptions underneath, and the first thing you are going to do.
- You start with
- A decision you have carried for a while — two offers, rewrite or patch, which direction a project takes — plus any documents that bear on it.
- Time
- About an afternoon.
- Your part
- Most of it. The AI lays out comparisons and attacks them; it does not weigh anything, because the weights are the decision.
What you'll end up with
- The real options on one ruler, against criteria you rewrote yourself.
- What each option gives up, and the assumptions the whole comparison rests on.
- A first action, what comes next, and what is deferred — with the trigger that brings it back.
Plus the part that saves you in three months: a written record of why the rejected options were rejected.
How to think through a hard decision, step by step
Each step applies one mental model to a card on the mind map — picked from a menu rather than recalled, and carried out on your actual options. Expect half the afternoon on steps 2 and 4.
Put the decision in the middle, phrased as a choice.
You"Rewrite the importer or patch it, decide by Friday" works. "Thinking about the importer" does not — a vague centre card produces vague everything downstream, and no method rescues it.
Widen the options before you narrow them.
You ask · MindMax answersMost bad decisions are between two options that arrived together. Brainstorm the space, and run a counterfactual on the constraint you treat as fixed — "what if the deadline moved" often finds the third option.
Put in what only you know.
YouDrag the relevant documents onto the canvas, and hand-write a card with the constraints written down nowhere: the budget, the person who will not go along, the promise you made. Nothing can weigh what it was never told.
Put the options on one ruler — then rewrite the criteria.
You ask · MindMax answersThe options-matrix skill returns a filled-in table, options against criteria; drill into one column to compare on a single criterion. Rewriting the criteria is where the decision gets made — letting a model choose them is delegating it.
Make each option say what it gives up.
You ask · MindMax answersA comparison table flatters whichever option is adequate everywhere. The trade-off skill surfaces the cost you were quietly planning not to pay — the step people skip, and regret skipping.
Attack the comparison, not the options.
You ask · MindMax answersRun the assumption audit on the leading option, then a red-team pass from someone who wants it to fail. If you install the pre-mortem skill from the starter library, it goes here: a year on, it went badly — why?
Turn it into a first action, and save the memo.
You ask · MindMax answersThe action-plan skill splits now, next and explicitly deferred, with the trigger that brings a deferred item back. Convert the branch to a note: it lands in your folder as Markdown, rejected options and reasons attached.
Why it works here
Methods you pick, not prompts you write
Options matrix, trade-offs, assumption audit, red team and action plan are AI Skills on a menu — each carried out on your options, as a table or a card you can drill into.
Thinking models on the canvas →Your constraints are on the board
Documents and hand-written cards sit next to the options, so the comparison runs on the budget and the politics you put there — not on a generic version of your situation.
How the canvas works →The memo is a file you keep
The branch converts to a Markdown note in your folder, rejected options included, and you edit it like any other document.
Markdown editor with inline AI →Also used here: Mental model · Mind map · Your own folder, not a database · Guide: thinking on the canvas
Where it gets hard
- It will not make the call, and should not.
You can end an afternoon with a beautiful canvas and no decision. The missing piece is almost always step 4: you have not said what you are optimising for.
- A weak option set cannot be rescued by comparison.
If the real answer was never on the canvas, a matrix ranks the wrong things accurately. That is what step 2 is for, and it deserves more time than it looks like.
- The red team will annoy you.
Objections to a decision you have privately made read as obstruction. The ones that irritate most are worth reading twice — the reaction tells you how attached you are.
- Anything political or personal is invisible to it.
Who owes whom, what the last reorg did, what you can get approved. It produces a clean recommendation that ignores all of it unless you put it on a card in step 3.
- Scores are not weights.
A filled-in table looks quantitative and is not. Two options that come out close are genuinely close, and the tie-break is a judgement, not a number.
- Wide expansions cost credits.
One skill on the option group is cheap; six critique skills on every option is not, and the third layer rarely changes the answer.