Decision Making
Multi-Criteria Decision Analysis
Master structured decision frameworks for high-stakes choices involving multiple competing objectives, uncertain information, and diverse stakeholders. These methods turn overwhelming complexity into manageable, defensible comparisons that you can articulate to others.
Context
Why this exercise
Real high-stakes decisions almost never reduce to a single criterion. Hiring involves skills, cultural fit, growth potential, salary, and risk; strategic investments balance expected return, downside survival, strategic fit, and reputational impact; policy choices weigh aggregate welfare, distributional fairness, political feasibility, and implementation cost. This exercise drills the structured frameworks that turn this overwhelming complexity into manageable, defensible comparisons — weighted scoring matrices, the Analytic Hierarchy Process, dominance analysis, and the disciplined separation of values (what you care about) from facts (how the options actually perform).
Before you start
Multi-criteria decision analysis emerged from operations research in the 1960s and was formalized by Ralph Keeney and Howard Raiffa in their 1976 book 'Decisions with Multiple Objectives.' The core insight is that decision quality depends on cleanly separating three steps that everyday thinking conflates: identifying the criteria that matter (the value side), measuring how each option performs on each criterion (the fact side), and combining the two through explicit weights to produce a comparable score. When these steps stay merged, the decision-maker is effectively reweighting criteria on the fly to match an emotional preference, which produces conclusions that cannot be defended to others and cannot be revisited when circumstances change. Thomas Saaty's Analytic Hierarchy Process, developed in the 1970s, added a structured method for deriving weights through pairwise comparisons, which surfaces inconsistencies in stated preferences and forces them to be resolved.
Several advanced patterns recur often enough to be worth memorizing. Dominance analysis identifies options that are worse than another option on every criterion — these can be eliminated immediately, narrowing the comparison to genuinely competing alternatives. Sensitivity analysis asks how much the weights would have to change to flip the conclusion, which reveals when a decision is robust versus when it sits on a knife edge. Pareto frontier analysis maps the tradeoff between two or more criteria so the decision-maker can see what they would have to give up on one dimension to gain on another. And the disciplined use of expected value combined with downside-survival constraints — a single number for ranking, plus a separate veto on options whose worst case is unacceptable — handles most high-stakes decisions under uncertainty.
The wrong-answer options in this exercise are designed to look like reasonable simplifications — pick the option with the highest single score, defer to the most experienced person in the room, use only quantitative factors and ignore the qualitative ones. Each of these is a known failure mode that structured analysis is designed to prevent. The deeper lesson, articulated by Keeney in 'Value-Focused Thinking,' is that the choice of criteria and weights is itself the most important part of the decision — get those right, and the comparison is mostly arithmetic; get them wrong, and no amount of careful scoring will produce a good outcome. As you work the scenarios, practice naming the criteria before evaluating any option, assigning weights before scoring, and running sensitivity checks on your conclusion. For more on structured choice under uncertainty, see Decision Making.