Learning guide and study notes
Which option should we choose? Use it when alternatives have uncertain outcomes.
Decision thinking structures high-stakes choices under uncertainty by separating controllable decision nodes from chance events, weighing expected outcomes alongside tail risk, and identifying which missing fact is worth investigating first.
A decision tree makes choices, uncertain branches and their consequences explicit. Probabilities and outcome values are assumptions to inspect, not automatic facts.
Work through these steps in order, applying the tools above. Write down the answer to each step before moving on.
State the decision point, realistic alternatives, key uncertainty and constraints. [1] Source: MIT OpenCourseWare — Decision Analysis Basics (opens in a new tab)
Distinguish decision nodes from chance nodes; attach justified probabilities and outcome measures where possible. [1] Source: MIT OpenCourseWare — Decision Analysis Basics (opens in a new tab)
Compare expected results and downside exposure, then specify what new information would reverse the choice. [1] Source: MIT OpenCourseWare — Decision Analysis Basics (opens in a new tab)
Draw three decision options and two uncertain states for a service change; identify a missing probability or value you must investigate.
Situation: A critical service needs a resilience choice.
How to apply: Compare single-site and replicated designs under several outage and recovery states.
Outcome: Recommend a design with explicit cost, recovery and uncertainty assumptions.
Situation: A release passes functional tests, but load testing revealed an intermittent connection leak under peak traffic.
How to apply: Map the decision tree: cutover now (with chance of peak outage and emergency rollback) vs. 7-day delay (fixed schedule cost, lower outage probability).
Outcome: Ship behind a 10% traffic cap now while fixing the leak, avoiding both a full schedule slip and a full-scale outage.
Situation: A team must connect an internal billing system to a partner API within two months.
How to apply: Compare build vs. buy across delivery probability, ongoing maintenance burden, and API change frequency.
Outcome: Run a two-day spike on the partner API edge cases—the single unknown that determines whether the commercial connector fits.
Illustrative examples showing how the lens changes decisions; not measured outcomes.
Watch out: A precise expected value does not compensate for invented probabilities or ignored tail risk.
Combine this lens with others rather than relying on it alone. Browse all thinkings
A sound decision compares alternatives against explicit objectives and uncertainty, then commits with review conditions.
Use a problem of your own. Write a short answer to each prompt, then return after trying the method.
Answer from memory before opening the suggested answer.
Yes. Uncertainty can lead to an adverse outcome even with a sound process. Judge the process using the information available when the choice was made.
The comparison uses inconsistent assumptions. Apply the same scenarios, horizon, and criteria to each option, then examine uncertainty.
These sources support the method; the examples and exercises are original teaching illustrations.