TWO QUESTIONS, TWO WORKSPACES_
Decision Analysis asks: which available option best satisfies what we must have, what we want, and the risks we can accept?
Problem Analysis asks: what distinguishes where, when and how the deviation occurs from comparable situations where it does not?
CAPYBARA RULE: Do not use a decision score to diagnose a cause, and do not use a causal hypothesis as a substitute for explicit decision criteria.
DECISION ANALYSIS_
PROBLEM ANALYSIS_
Describe the deviation across WHAT, WHERE, WHEN and EXTENT. For each dimension, compare IS with a useful IS NOT. Then look for distinctions and changes that could explain the boundary.
Possible causes should be tested against those distinctions. A cause that explains one detail but conflicts with several others should lose confidence. The best current fit is still a hypothesis until verification.
CAPYBARA CASE #004_
In The Pond Protocol, Decision Analysis compares pond options using MUSTs, weighted WANTs and risk. In the teaching case, Quiet Marsh / Pond B best satisfies that defined decision structure.
Problem Analysis tackles a different question: why is the current pond condition abnormal? The feeder near the north inlet becomes the best-fitting hypothesis because it matches the observed distinctions. The next move is verification — for example, reducing feeder output and observing whether the deviation changes.
COMMON FAILURE MODES_
Scoring before defining the rules
If weights and criteria move after the preferred option becomes visible, the matrix becomes justification rather than analysis.
Letting MUSTs become soft preferences
A MUST that can be traded away is probably a WANT. Be explicit.
Treating the highest score as an automatic answer
The matrix structures judgment. It does not absorb every uncertainty, political constraint or implementation risk.
Skipping IS / IS NOT
Problem Analysis gets its power from boundaries. “What is different here?” is usually more diagnostic than generating a long list of generic causes.