Build a defensible two-option decision from constraints, criteria, tradeoffs, reversibility, and a clear commitment rule.
A good comparison starts before scoring. First define the decision, remove options that fail real constraints, and decide what evidence would matter. The guides in this collection teach the core mechanics without pretending that every value can be reduced to one number.
Start with the complete method for choosing between two options.
Write the two options in comparable terms, name the decision owner, and set the time horizon. A choice between accepting a job now and waiting indefinitely is not balanced. The second option needs a concrete review date and a description of what waiting is expected to produce.
A criterion earns a place only if a meaningful difference between the options could affect the decision. Separate hard constraints from preferences, avoid counting one benefit twice, and define what a high or low score means before scoring.
Cheap, reversible choices need a short check and a review point. Expensive or difficult-to-reverse choices need verified inputs, downside analysis, and explicit exit costs. The amount of analysis should follow the cost of being wrong.
The method should follow the structure of the decision, not the desire to produce a more impressive score.
| Situation | Best starting method | Why it fits |
|---|---|---|
| One option may be unacceptable | Constraint check | A failed requirement should remove an option before preferences are scored. |
| Several independent tradeoffs matter | Weighted decision matrix | Weights make priorities explicit while sensitivity testing shows whether the lead is stable. |
| The main difference is ease of exit | Reversibility test | Compare the time, money, and harm required to recover from a poor choice. |
Start with constraints and criteria, then add opportunity cost, reversibility, or a matrix only when that extra structure changes the decision.
Remove any option that fails a real constraint, compare the remaining tradeoffs, test how reversible the choice is, and set a deadline for commitment.
A weighted matrix is useful when several criteria matter at once. Its value comes from clear inputs and sensitivity testing, not from the decimal in the final score.
The cost of a choice includes the best realistic alternative you give up. Make that tradeoff visible before comparing features.
Move quickly on choices with cheap exits. Slow down when consequences compound, exit costs are high, or other people depend on the result.
Use criteria that connect to the decision purpose, separate constraints from preferences, and describe one distinct consequence each.
Use a decision tree when the choice unfolds in stages and later actions depend on events that have not happened yet.
A group decision is not a vote on one blended score. Make the affected people, decision rights, non-negotiable needs, and distribution of consequences visible before seeking agreement.
