Prompt library
Build a decision matrix that resists gaming
Constraint/criteria separation and the weight-sensitivity flip test are what keep a matrix from being decorated intuition. The disappointment check operationalizes a real decision-science trick: your reaction to the result reveals preferences the criteria list missed.
Last reviewed July 17, 2026
The prompt
Build a decision matrix for this decision: {{decision}}
Options: {{options}}
What matters and any hard constraints: {{criteria}}
Method:
1. Separate constraints from criteria: constraints eliminate options outright (show which die and why); criteria rank the survivors. Mixing these is the classic matrix error.
2. Weight the criteria (sum 100) and justify each weight in one line against my stated priorities.
3. Score survivors 1-5 per criterion, with one line of evidence per cell. No evidence, no score: mark [UNKNOWN] instead.
4. Compute totals, then stress-test: which single weight change flips the winner? If a small plausible change flips it, say the matrix is not decisive and name the missing information that would settle it.
5. Gut check: state which option the matrix picked and ask me whether I am disappointed. Disappointment is data about a criterion I did not list.
Show the full table in markdown.Run in idaptOpens a new chat with the prompt prefilled. Nothing sends until you press send.
Fill in the variables
| Variable | What it is | Example |
|---|---|---|
| {{decision}} | The decision | which city to open our second office in |
| {{options}} | The options on the table | Lisbon, Austin, Toronto |
| {{criteria}} | What matters, plus hard constraints | talent pool, cost, timezone overlap with Berlin; constraint: must allow hiring within 6 months |