Feasibility Sensitivity Probe¶
Robustness check — instantiates Feasible-Alternative Comparator Calibration
Perturbs the feasibility and cost assumptions to see whether the winning option and its dominance ranking survive — or hang on one fragile estimate.
A comparison can be honest, symmetric, and correctly decided — and still be brittle, if the verdict flips the moment one shaky assumption moves. Feasibility Sensitivity Probe is the stress test that finds out. It takes the decided field and systematically varies the feasibility and cost estimates each option rests on — a longer permitting timeline here, a higher adoption cost there — then re-checks whether the same option still wins and whether the dominance ordering holds. Its defining act is perturbation of the assumptions to test the durability of the ranking, not the making of the ranking itself. It reports which option is robustly best, which comparisons are dominated across the whole plausible range, and which verdicts hang on a single fragile number. It neither builds the option set nor renders the original decision; it tells you how much to trust the one already made.
Example¶
An agriculture ministry has tentatively chosen an input-subsidy scheme — discounting fertilizer and seed — over two alternatives: a cash-transfer scheme and an insurance-premium subsidy. On the base-case numbers, the input subsidy wins. The probe now shakes the assumptions. It varies the projected uptake rate (farmers may enroll at half the assumed pace), the world fertilizer price (which drives the scheme's cost), the leakage rate to non-target recipients, and the rainfall scenario that governs the payoff. Under each perturbation it re-runs the comparison and watches the ranking.
The findings sharpen the decision. Across almost the whole plausible range, cash transfers dominate the insurance subsidy — that comparison is settled and needs no more debate. But the input subsidy's lead over cash transfers is fragile: it wins only if fertilizer prices stay low and uptake stays high; let either slip and cash transfers overtake it. The probe's output is a dominance-and-sensitivity map: some pairings robust, one verdict balanced on two correlated assumptions. The ministry now knows exactly where to spend its remaining diligence — pinning down the fertilizer-price and uptake estimates — rather than re-arguing the whole choice.
How it works¶
- Identify the load-bearing assumptions. Find the feasibility and cost estimates the ranking actually depends on, especially the softest ones.
- Perturb across plausible ranges. Move each assumption through its credible span — singly and, where they interact, jointly — rather than trusting a single point estimate.
- Re-check dominance and the winner. For each perturbation, re-evaluate which option is best and which comparisons stay dominated, mapping where the ordering holds and where it flips.
- Report robust versus fragile verdicts. Separate the settled comparisons from the ones balanced on a single or correlated assumption, and point diligence at the fragile ones.
Tuning parameters¶
- Perturbation range — how wide a span each assumption is swung through. Narrow ranges understate fragility; over-wide ones manufacture flips that no one believes.
- One-at-a-time vs. joint — whether assumptions move singly or in combination. Joint moves catch correlated failures a one-at-a-time sweep misses, at the cost of a much larger space.
- Flip threshold — how close a re-ranking must come before a verdict is called "fragile." A sensitive threshold surfaces more brittleness but can cry wolf.
- Assumption prioritization — whether to probe every input or only the load-bearing few. Focusing on the decisive assumptions is efficient but risks missing a sleeper.
When it helps, and when it misleads¶
Its strength is that it distinguishes a robust choice from a lucky one: it shows which parts of the decision are settled across the plausible range and which rest on a number no one has actually pinned down — the core value of sensitivity analysis in any decision model.[n1] By reporting dominance, it also retires the comparisons that need no further argument, concentrating attention where it matters.
Its failure mode is false reassurance or manufactured doubt. Too narrow a perturbation range makes a fragile decision look solid; too wide a range, or cherry-picked scenarios, can make any option look untrustworthy and stall action indefinitely — analysis as a delay tactic. It can also lull with a tidy map while missing an unmodeled assumption entirely. The guarding discipline is to set perturbation ranges from evidence before seeing the results, probe assumptions jointly where they plausibly move together, and treat a robust verdict as "durable within what we modeled," not "certain."
How it implements the components¶
Feasibility Sensitivity Probe fills the robustness components — the ones that test how much the verdict can be trusted:
dominance_and_sensitivity_matrix— its core artifact: the map of which comparisons stay dominated and which rankings flip as assumptions are perturbed.best_feasible_comparator— it re-tests whether the identified best feasible option remains best across the plausible range, or wins only under a fragile assumption set.
It does not select the winner in the first place — that is Best Feasible Alternative Test's job of applying the decision rule — nor gather each option's underlying feasibility evidence (feasibility_evidence_packet, Shared-Burden Comparison Table). It pressures a decision already made; it does not make it.
Related¶
- Instantiates: Feasible-Alternative Comparator Calibration — the probe reports how robust the feasible-comparator verdict is to shifting assumptions.
- Consumes: Best Feasible Alternative Test supplies the verdict and operative comparator the probe stress-tests.
- Sibling mechanisms: Feasible Option Inventory · Ideal Comparator Filter · Status Quo Costing Sheet · Shared-Burden Comparison Table · Best Feasible Alternative Test · Defect/Disqualifier Sort · Comparison-Class Review Memo
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: The mechanism varies feasibility and cost assumptions within a model and calculates whether the winning option and dominance ranking remain robust.
Nearest alternative: Experiment, Test & Rehearsal — Inputs are perturbed, but this is offline sensitivity analysis of a model rather than exposure of an operational target to generate evidence.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Convergent development
Present-day reach: Multi-domain
Rationale: Sensitivity analysis of a preferred feasible solution is canonical in operations research and optimization.
Related originating lineages:
- Economics & Finance — Economic appraisal independently developed comparative statics and scenario sensitivity for option rankings.
- Statistics & Experimental Design — Statistical uncertainty and robustness analysis materially shape how assumptions are perturbed and conclusions assessed.
Review resolution: Both reviewers agree that operations_research is primary. I retain statistics_experimental_design, economics_finance only as formative origin lineage(s), without treating every later application as an origin. convergent is appropriate because the same operational structure arose through materially independent professional lineages. Reach is multi_domain as a separate applicability judgment: it does not widen or narrow the recorded provenance. Encyclopedia synthesis is false because the artifact is already established enough that encyclopedia-specific synthesis is not required. The secondary differences are reconciled with no unresolved primary-provenance ambiguity.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Sensitivity analysis systematically varies a model's inputs to see how much its conclusions move — separating results that hold across the plausible range from those that hinge on a single uncertain estimate. Applied to a feasible-comparator decision, it distinguishes a durable choice from one that merely won on the base-case numbers. ↩