Sensitivity and Scenario Sweep¶
Validation method — instantiates Activation Energy Cost-Benefit Analysis
Tests whether the activation recommendation changes under different assumptions about costs, adoption rate, benefit timing, failure probability, and maintenance burden.
A recommendation is only as trustworthy as the assumptions it rides on, and some of those assumptions are guesses. Sensitivity and Scenario Sweep stress-tests the activation recommendation by systematically varying each uncertain input across its plausible range, and by running a handful of named end-to-end scenarios — best, base, worst — then asking one question: does the decision flip? It does not produce the estimate; it probes how fragile the estimate is and pins down the single assumption the whole recommendation depends on. Where Break-Even Activation Model reports one binding assumption in passing, this mechanism is the full, disciplined exploration of robustness across every uncertain term.
Example¶
A utility is weighing a grid-scale battery-storage investment — a capital-heavy activation whose benefit appears only once regional renewable penetration is high enough to need storage. The base case says invest. The sweep varies each input across its range: carbon price, the pace of renewable adoption, the rate of battery-cost decline, cycle-life degradation, and wholesale-price volatility. A one-way sweep, ranked as a tornado, shows the recommendation is overwhelmingly sensitive to the pace of renewable adoption and nearly indifferent to carbon price. A coherent worst-case scenario — slow adoption together with flat prices — flips the decision from "invest" to "wait." The team stops arguing about carbon price, reframes the entire decision around the one variable that actually decides it, and sets an explicit trigger to revisit once regional renewables reach a defined share.
How it works¶
- One-way sensitivity. Vary each input alone across its plausible range and rank inputs by how far they swing the outcome — the tornado.
- Scenario analysis. Bundle coherent combinations (best / base / worst) rather than jiggling variables independently, because real futures move together.
- Find the flip points. For each decisive input, locate the value at which the recommendation reverses, and check where that value sits in the plausible range.
- Name the decisive assumption(s) — the one or two the whole case rides on — and hand them to staging or probing.
- Feed the result forward. The variable the decision is most sensitive to is exactly what a Pilot Option Probe should resolve first.
Tuning parameters¶
- Input ranges — how wide each variable is swept; too narrow hides fragility, too wide makes everything look decisive.
- One-way vs. joint variation — one-at-a-time sweeps are readable but blind to interactions; scenarios and Monte Carlo capture joint moves at more effort.
- Scenario count and coherence — how many named scenarios, and how internally consistent each is; incoherent worst cases are easy to dismiss.
- Flip-point focus — whether you report full response curves or just the decision-reversal thresholds that actually matter.
- Correlation handling — whether inputs are varied independently or with realistic correlations, which can turn an apparently robust case fragile.
When it helps, and when it misleads¶
Its strength is that it separates a robust recommendation from a knife-edge one, and it pinpoints the assumption worth resolving before commitment — disciplining the overconfidence that a single tidy break-even invites.
Its failure mode follows from its most common form: one-at-a-time sweeps miss interactions and correlated moves, so a case that looks sturdy variable-by-variable can be fragile when several inputs shift together. The classic misuse is sweeping only the comfortable variables within flattering ranges to manufacture a "robust" verdict. The discipline that guards against this is the tornado diagram paired with at least one coherent worst-case scenario, with the ranges fixed before the results are seen so they cannot be quietly narrowed to protect the conclusion.[n1]
How it implements the components¶
Sensitivity and Scenario Sweep realizes the uncertainty-validation side of the archetype — it interrogates the model's inputs rather than producing or acting on them:
probability_and_uncertainty_trace— the sweep is the explicit record of how uncertainty in each input propagates to the recommendation, and which inputs dominate it.post_threshold_benefit_model— benefit assumptions (timing, magnitude, durability) are prime sweep variables, tested for the ranges over which they flip the case.activation_cost_inventory— cost line-items are swept to find which overruns would reverse the decision, revealing where budget discipline actually matters.
The commit / stop rule itself (decision_and_stop_rule) belongs to Break-Even Activation Model and Activation Hurdle-Rate Rule; the post-crossing self-sustainability verification (self_sustainability_condition) belongs to Post-Crossing Feedback Check. This mechanism tests a recommendation — it does not make or confirm one.
Related¶
- Instantiates: Activation Energy Cost-Benefit Analysis — supplies the robustness check that keeps a single optimistic estimate from carrying the whole decision.
- Consumes: Break-Even Activation Model — the valuation whose fragility it probes.
- Sibling mechanisms: Barrier Height Estimation · Break-Even Activation Model · Stage-Gate Activation Review · Pilot Option Probe · Counterfactual Non-Activation Comparison · Post-Crossing Feedback Check · Activation Hurdle-Rate Rule
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Sensitivity and Scenario Sweep operates as an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution because it tests whether the activation recommendation changes under different assumptions about costs, adoption rate, benefit timing, failure probability, and maintenance burden.
Independent corroboration: The frozen evidence defines Sensitivity and Scenario Sweep as 'Tests whether the activation recommendation changes under different assumptions about costs, adoption rate, benefit timing, failure probability, and maintenance burden', so its operative form is Analysis, Modeling & Optimization.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Varying costs, adoption, timing, failure, and maintenance to see whether a recommendation flips is decision analysis under uncertainty.
Related originating lineages:
- Economics & Finance — Discounting, adoption, and expected-failure costs are economic valuation inputs.
- Futurism & Strategic Foresight — Scenario practice packages correlated assumptions into plausible alternative conditions.
- Mathematics — Mathematical modeling, proof, and abstract-structure practice supplies a parallel or contributing lineage for the mechanism's defining operation: tests whether the activation recommendation changes under different assumptions about costs, adoption rate, benefit timing, failure probability, and maintenance burden.
- Organizational & Management Science — Business-case and portfolio reviews test commitment against operational assumptions.
- Statistics & Experimental Design — Statistics, experimental design, and measurement theory supplies a parallel or contributing lineage for the mechanism's defining operation: tests whether the activation recommendation changes under different assumptions about costs, adoption rate, benefit timing, failure probability, and maintenance burden.
Review resolution: The blind reviewers agree that operations_research is the primary origin and differ only on alternate origin disagreement. I preserve every independently explained alternate from both records rather than imposing a numeric cap. I retain cross_disciplinary_synthesis because the combined record shows material contributions from several lineages. The broader reach of multi_domain records portability separately from historical provenance, and encyclopedia_synthesis=true preserves the affirmative synthesis judgment where either reviewer identified one.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] A tornado diagram ranks inputs by how much the outcome swings when each is varied alone across its range, widest bars on top. It is the standard one-way sensitivity display; its known blind spot is interaction effects, which is why it is paired with coherent multi-variable scenarios rather than trusted on its own. ↩