Screen options for dominance before debating preferences
Remove only options that are no better on any declared objective, then retain genuine tradeoffs for the decision owner.
Kiguri Blog
Inspect a failure-condition tree without inventing probabilities
Use explicit AND and OR conditions to inspect an invented failure model, then separate its logic from unverified real-world causes.
Review condition combinations with a pair-coverage register
Check a small synthetic fixture set against every factor pair, while keeping untested three-factor combinations visible.
Model a workflow with paper tokens before changing the real routine
Replay one fixed arrival sequence under two queue rules and track every token without treating the model as a forecast.
Test a reusable document template with contrasting examples
Expand an invented invitation twice to expose stale names, missing substitutions, and contradictory conditional text.
Review a calculation chain in a fictional spreadsheet
Trace source cells, copied ranges, and rounding stages before trusting an apparently reasonable total.
Check whether issue reports describe the same failure
Compare conditions and reproduction paths before linking similar reports to one investigation.
Slice proposed work into complete small outcomes
Choose a narrow end-to-end result instead of splitting work into partial departmental outputs that nobody can use yet.
Review an artifact change from both producer and consumer sides
Compare explicit version pairs so a new artifact does not silently acquire a different meaning for an older reader.
Review a state model for reachable and contradictory behavior
Trace event sequences, inspect overlapping guards, and test persistent rules across an invented process model.
Check cross-references before a shared document is released
Match internal pointers to unique intended targets, then recheck the whole document after moving or renumbering sections.
Review a synthetic example before sharing it more widely
Map identifying clues across an invented example and check that removing unnecessary detail preserves its explanation.