Judgment Framework
Epistemically examine the principles. Do not inherit or assume them. Understand where they came from, what reasons support them, what assumptions they contain, how they hold up against counterexamples, and how they relate to competing principles/opinions/positions. Which principles need this scrutiny is usually clear only once the problem is separated. Expect to move between the two.
Separate the problems. Narrow in and distinguish the primary problem from the secondary problems a solution may create. Do not assume that solving one requires accepting or reversing another. Separating them well often depends on which principle you are reading the situation through — expect to move between the two.
Define success. Specify the primary and backstop outcomes that matter and the evidence that would show they were achieved — quantitative where honest measures exist, qualitative where they don't. Name what matters here that no available metric captures, so the proxy does not silently replace the goal.
Set criteria and constraints before making the tradeoff. Decide what alternatives will be judged against and what boundaries must be respected before choosing among them. Principles guide these choices but do not predetermine the answer.
Make the tradeoff. Choose among the available alternatives explicitly, including what is gained, what is sacrificed, who bears the costs, and what uncertainty remains. Before acting, write down what you expect to observe, by when, and which specific observations would mean the decision was wrong.
Evaluate the result against reality. Compare what actually happened with those tripwires, the success measures, the assumptions, the expected benefits, and the anticipated harms — not with a story assembled in hindsight. If reality contradicts the decision, locate the failure: execution, the tradeoff, the criteria, or the principle itself. Fix at the shallowest level the evidence supports, and no shallower.
(Examine principles ⇄ Separate problems) → Define success → Set criteria & constraints → Make the tradeoff (+ commit tripwires) → Test against reality
The steps are iterative, not a strict sequence — except for two orderings that must hold: criteria before choosing, and tripwires before observing.