Maximin / Satisficing Rule¶
Selection procedure — instantiates Robust Solution Selection
Chooses an option that maximizes the minimum acceptable performance or clears a defined performance floor across scenarios.
The Maximin / Satisficing Rule is a selection procedure that works entirely on the gain side of the decision. In its maximin form it ranks candidates by their worst-case performance and picks the one whose worst case is best — the option that lifts the floor highest. In its satisficing form it drops ranking altogether and admits any candidate that clears a pre-declared acceptable floor in every scenario, choosing among survivors on a secondary criterion like cost. Either way, the governing idea is a guaranteed minimum you can live with: the rule asks "how good is this option when things go worst?" and rewards the answer that keeps the guaranteed level high. That framing — maximize the minimum benefit — is precisely what separates it from a rule that instead minimizes the maximum loss.
Example¶
A smallholder cooperative in a semi-arid district must choose one drought-tolerance level of maize seed to distribute to every member for the coming season, before anyone knows what the rains will do. Three varieties are on offer: a high-yield hybrid that thrives in a wet year but collapses in drought, a hardy landrace that yields modestly whatever happens, and a mid-tier variety in between. The cooperative treats the season as four scenarios — good rains, average, late-onset, and drought — and looks only at the guaranteed floor.
The agronomist tabulates each variety's worst yield across the four. The high-yield hybrid's worst case (a failed drought crop) is catastrophic; the mid-tier's worst case is thin but survivable; the landrace's worst case is the highest of the three. Under a maximin reading the landrace wins outright — its guaranteed minimum beats every rival's. The cooperative then applies a satisficing sanity check: it had pre-declared a floor of "enough grain to cover member subsistence in any scenario," and only the landrace and mid-tier clear it in all four. Because feeding families is the binding concern, they distribute the landrace and accept that a wet year will leave yield on the table.
How it works¶
The procedure has three moves and no optimization search. First, declare the acceptable floor in advance — the minimum outcome the decision can tolerate in any scenario — so the rule cannot be relaxed after seeing results. Second, for each candidate, read its worst performance across the scenario set; that worst value is the candidate's robustness score. Third, apply the rule: maximin selects the candidate with the highest worst-value; satisficing keeps every candidate whose worst-value clears the floor and defers to a secondary criterion among them. The two forms differ only in whether the floor is derived (the best achievable minimum) or fixed (a stated line), and the choice between them is itself a declared parameter.
Tuning parameters¶
- Floor height — where the acceptable line sits. A high floor rejects fragile options but can disqualify everything; a low floor lets a barely-adequate option pass as robust.
- Maximin vs. satisficing — optimize the minimum, or merely require it to clear a fixed bar. Optimizing squeezes out the safest option; satisficing preserves other goals once "safe enough" is met.
- Scenario inclusion bar — how implausible a scenario must be before it stops driving the worst case. Admitting extreme tails makes the rule conservative to the point of paralysis.
- Tie-break criterion — what decides among satisficing survivors (cost, simplicity, speed). This is where efficiency re-enters after safety is secured.
When it helps, and when it misleads¶
Its strength is a hard guarantee under genuine ignorance: when you cannot assign probabilities to scenarios and cannot afford a catastrophic outcome, choosing by the best worst-case is a defensible, legible commitment — the classic logic of satisficing when optimizing is infeasible or unwise.[n1] It is fast, needs no distribution over scenarios, and produces a decision anyone can audit against the floor.
Its failure mode is over-conservatism. Because only the single worst cell decides, one lurid, low-probability scenario can dominate the choice and push the cooperative toward a timid option that wastes value in every likely future — the "maximin is the rule of the pessimist" critique. The classic misuse is smuggling an incredible worst case into the scenario set to justify a preferred safe option. The guarding discipline is to hold the scenario-inclusion bar and the floor as pre-declared parameters, and to report what the rule gives up in the better scenarios so the conservatism is a visible choice rather than a hidden default.
How it implements the components¶
acceptable_performance_threshold— the pre-declared floor is this component made explicit; the rule is built around it.robust_selection_rule— the maximin-or-satisficing decision procedure is the selection rule, chosen before results are seen.robustness_metric— each candidate's worst-across-scenarios value is the transparent robustness score the rule ranks on.
This rule works on gains, not losses: it does not minimize a worst_case_bound on downside — that's Minimax Decision Rule, which caps the ceiling of losses where this rule raises the floor of benefits.
Related¶
- Instantiates: Robust Solution Selection — supplies the acceptability-floor selection rule the archetype applies once candidates are profiled.
- Consumes: Scenario Robustness Check supplies the per-candidate scenario performance the floor is read against.
- Sibling mechanisms: Minimax Decision Rule · Decision Matrix Under Uncertainty · Monte Carlo Robustness Screen · Regret Analysis · Robust Optimization Model · Robust Policy Design Review · Scenario Robustness Check · Stress-Tested Plan Review
Editorial Notes¶
Form Classification¶
Form family: Decision, Gate & Allocation
Rationale: Maximin / Satisficing Rule operates as a case-specific gate, selection, routing, prioritization, or resource disposition because it chooses an option that maximizes the minimum acceptable performance or clears a defined performance floor across scenarios.
Independent corroboration: The frozen evidence defines Maximin / Satisficing Rule as 'Chooses an option that maximizes the minimum acceptable performance or clears a defined performance floor across scenarios', so its operative form is Decision, Gate & Allocation.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Worst-case robust choice rules were formalized in decision analysis and operations research.
Related originating lineages:
- Economics & Finance — Game theory and decision under uncertainty shaped maximin reasoning.
- Organizational & Management Science — Simon's satisficing tradition supplied the acceptable-floor component.
Review resolution: Both independent reviews place the primary provenance in operations_research. The queued differences (reported_ambiguity, origin_mode_disagreement, domain_reach_disagreement, encyclopedia_synthesis_disagreement) concern secondary metadata, not primary lineage. The final retains economics_finance, organizational_management only where a reviewer supplied a formative-lineage rationale; downstream use or broad applicability by itself is not treated as origin. origin_mode=cross_disciplinary_synthesis because the supplied rationales identify formative contributions that are composed in the mechanism's present form. domain_reach=universal records established application breadth separately from provenance. confidence=medium preserves the more cautious evidence assessment. encyclopedia_synthesis=true records whether either reviewer identified deliberate corpus-level composition.
Attribution caveat: The mechanism intentionally combines maximin and satisficing lineages.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; medium confidence.
Notes¶
[n1] Herbert Simon's satisficing — choosing an option that clears an aspiration level rather than searching for the global optimum — is the intellectual root of the floor-clearing form of this rule, and its justification when full optimization is infeasible or the stakes reward a guaranteed-good-enough outcome. ↩