Personal Dev Decision Making¶
14 cards — 🟢 4 easy | 🟡 6 medium | 🔴 4 hard
🟢 Easy (4)¶
1. What is a decision journal and why does it separate process from outcome?
Show answer
A decision journal records the situation, options, reasoning, assumptions, confidence level, and success/failure criteria before you act. It separates process from outcome because good decisions can have bad outcomes (bad luck) and bad decisions can have good outcomes (good luck). Without a journal, outcome bias rewrites your memory of the reasoning. The journal preserves what you actually knew and thought at the time.2. What is the difference between reversible and irreversible decisions, and how should you treat each?
Show answer
Reversible decisions (Type 2, two-way doors) can be undone at low cost — feature flags, A/B tests, trying a new CI tool. These deserve bias toward action: decide and iterate. Irreversible decisions (Type 1, one-way doors) are costly or impossible to reverse — deleting production data, public disclosure, firing someone. These deserve careful analysis, broad input, and a pre-mortem. The trap is treating every decision like Type 1, causing analysis paralysis.3. Why should decisions be evaluated based on future costs and benefits only, ignoring past investment?
Show answer
Past investment is gone regardless of what you decide next. The sunk cost fallacy tricks you into continuing a failing path because of prior time or money spent. The correct frame: "If I were starting fresh today with no prior investment, would I choose this path?" If no, the prior investment is irrelevant. The emotional pull of sunk costs is strong, so this question must be asked explicitly and honestly.4. What is the Eisenhower Matrix and why do most people spend too much time in the "urgent but not important" quadrant?
Show answer
The Eisenhower Matrix sorts tasks by urgency and importance into four quadrants: Q1 (urgent + important) — do now, Q2 (not urgent + important) — schedule and protect, Q3 (urgent + not important) — delegate or batch, Q4 (not urgent + not important) — eliminate. Most people live in Q3 because urgency creates a false sense of importance — responding to every Slack ping, attending every meeting, fighting every small fire. Q2 (strategic work, learning, relationship building, system improvements) is where long-term career leverage lives, but it never feels urgent enough to prioritize.🟡 Medium (6)¶
1. What belongs in a pre-mortem and how does it surface risks that optimism suppresses?
Show answer
A pre-mortem assumes the decision has already failed: "It is six months from now. This project failed completely. Write the failure story." List 3-5 failure modes, identify the most likely one, note early warning signs, and define mitigations you can implement now. It works because asking "what could go wrong?" invites optimism, while "tell me why it failed" gives social permission to voice doubts honestly.2. What does "externalizes" mean in tradeoff analysis and why is it the most dangerous column?
Show answer
Externalizes is the cost you push onto someone else: a different team, future engineers, users, the on-call rotation. It is the most dangerous column because externalized costs are invisible to the decision-maker but very real to the person who absorbs them. Skipping integration tests externalizes debugging cost to on-call. Deferring tech debt externalizes complexity to future engineers. Always ask: who bears the cost of this choice?3. What is the difference between satisficing and maximizing, and when is satisficing the better strategy?
Show answer
Maximizing means seeking the absolute best option by evaluating all alternatives. Satisficing means defining "good enough" criteria and choosing the first option that meets them. Satisficing is better for reversible decisions, time-constrained choices, and decisions with diminishing returns from further research. Maximizers report more decision fatigue and less satisfaction. For most engineering decisions, a "good enough" choice made quickly outperforms a "perfect" choice made too late.4. What is the regret minimization framework and when should you apply it?
Show answer
Ask: "When I am 80 years old, which choice will I regret more — doing this or not doing this?" It works best for Type 1 (irreversible) life decisions where analytical frameworks produce ambiguous results: career changes, starting something new, taking risks. It cuts through spreadsheet analysis by accessing your deeper values. It is less useful for routine operational decisions where data-driven analysis is more appropriate.5. What is confirmation bias and what are practical techniques to counteract it in technical decisions?
Show answer
Confirmation bias is the tendency to seek, interpret, and remember information that confirms your existing beliefs while ignoring contradictory evidence. In engineering: if you believe the database is the bottleneck, you will find evidence supporting that and miss network latency data. Techniques: 1) Assign a "red team" member to argue the opposite position. 2) Write down your hypothesis and what evidence would disprove it before investigating. 3) Seek disconfirming evidence first. 4) Ask "what would have to be true for the alternative to be correct?" 5) Use structured decision matrices instead of gut feeling.6. What is anchoring bias and how does it affect engineering estimates and incident response?
Show answer
Anchoring bias causes over-reliance on the first piece of information encountered. In estimation: if someone says "this should take two weeks," subsequent estimates cluster around that number regardless of actual complexity. In incidents: the first hypothesis anchors investigation, and the team tunnels on it even when evidence points elsewhere. Countermeasures: 1) Have each person estimate independently before sharing. 2) During incidents, explicitly ask "what if our current theory is completely wrong?" every 15 minutes. 3) Use reference class forecasting (how long did similar past projects actually take?) instead of intuition.🔴 Hard (4)¶
1. What is outcome bias and how does it distort your ability to learn from decisions?
Show answer
Outcome bias is judging a decision's quality by its outcome rather than by the reasoning at the time. A good decision with available information can have a bad outcome (unlucky), and a poor decision can succeed (lucky). If you judge only by outcomes, you learn to be lucky rather than thoughtful. To counter it: use a decision journal that records reasoning before outcomes are known, then review whether the process was sound regardless of what happened.2. What are second-order effects and why do most ethical surprises in engineering hide there?
Show answer
First-order effects are what happens directly; second-order effects are what happens because of what happened. Adding detailed logging (first order: better debugging; second order: data retention liability and potential surveillance). Automating a manual process (first order: efficiency; second order: knowledge holder leaves and edge case docs are never written). Surface them by asking: "And then what?" for each affected group, and "What happens when this scales 10x?"3. What is the minimum viable content of a decision record, and why does undocumented decision-making become folklore?
Show answer
Minimum: date, decision, context, options considered (including "do nothing"), rationale, tradeoffs accepted, risks acknowledged, review date, and decision-makers. Without documentation, decisions become folklore that mutates over time. Six months later, nobody remembers why PostgreSQL was chosen over DynamoDB, and the new architect proposes the same migration that was already rejected. A 10-minute decision record saves hours of re-litigation.4. What is groupthink, what conditions produce it, and how do you prevent it in engineering teams?