Satisficing Threshold Design¶

Decide what “good enough” means before endless comparison, accept an option that clears protected floors and the aspiration threshold, and stop searching with an auditable path to reopen.

Essence¶

Satisficing Threshold Design turns “good enough” from an excuse into a governed decision-closure rule. It defines the decision and affected needs, separates noncompensable safety, legality, rights, feasibility, accessibility, and interoperability floors from compensable preferences, states an aspiration level and evidence burden, governs the candidate population and search order, and accepts a sufficiently supported option when further comparison no longer justifies delay and opportunity cost. Closure transfers resources and responsibility and remains conditional on real performance, changed context, fairness, and explicit reopen triggers.

Compression statement¶

Satisficing Threshold Design converts bounded rationality from an excuse into a governed decision rule. It names the decision and needs; separates mandatory safety, rights, legality, and feasibility floors from compensable preferences; defines measurable sufficiency thresholds and confidence; specifies the candidate population and search order; budgets time, information, and review; records each candidate consistently; manages uncertainty, multiple qualifying options, and exceptions; accepts and commits when the rule is met; and monitors real performance. Search can continue when the cost of a false acceptance is high, a better option is plausibly near, the sample is biased, or evidence is weak. It stops when a qualifying choice exists and expected improvement no longer justifies opportunity cost. Thresholds are reviewed after outcomes, distributional effects, strategic gaming, or context change. The archetype does not equate cheap, familiar, first, average, minimally compliant, or politically convenient with sufficient.

Canonical formula: accept candidate c when every mandatory_floor_i© is satisfied, aggregate_or_vector sufficiency© meets declared aspiration_threshold, confidence© meets evidence_requirement, unresolved_exception_risk is within authority, and expected_value_of_further_search <= remaining search_cost + delay_cost + opportunity_cost; reopen when performance, context, evidence, fairness, or assumption triggers invalidate that conclusion.

When to Use This Archetype¶

Use this archetype when optimization is unknowable, unstable, disputed, or too expensive; candidates arrive sequentially; comparison consumes material time or attention; or a group repeatedly oscillates between endless search and opportunistic first-fit closure. It is most valuable when stakeholders can state needs and protected floors but cannot justify pursuing a global optimum.

option comparison expands while likely decision improvement becomes small
no feasible global optimum is knowable within available time information or computation
candidates arrive sequentially and delaying choice has material cost
stakeholders use “good enough” without a shared threshold or protected floor
requirements continue growing after viable options appear
teams repeatedly revisit a decision because closure and reopen rules are absent
a first-fit rule creates order dependence or incumbent advantage
optimizing a proxy degrades robustness fairness adaptability or human value
emergency or capacity limits require proportionate bounded choice

Do not use it to lower a duty of care, normalize scarcity, or bypass legitimate review. It is also not the same as defining the smallest artifact, pruning impossible search branches, bounding the option universe, or stopping marginal effort. Those interventions can compose with satisficing at different decision stages.

Structural Problem¶

Without a governed threshold, criteria remain vague or move after a favorite appears; a single score compensates for protected harm; incumbent access determines which candidates appear first; time pressure is asserted without a budget; point estimates near the threshold are treated as certain; and acceptance produces no funded handoff or monitoring. The system then alternates between analysis paralysis and arbitrary convenience. “Good enough” hides whose need counts, who bears residual risk, how long sufficiency must last, and what evidence would reopen the choice.

The root tension is search and evidence have real costs, but closure becomes arbitrary or harmful unless “good enough” is tied to legitimate need, protected floors, evidence, fair access, and a reversible commitment.

analysis paralysis and missed opportunity windows
escalating information and comparison cost
preference drift and criteria manipulation
incumbent first-order and presentation bias
acceptance of options that violate noncompensable needs
hidden inequity and inaccessible qualification paths
fragile commitments and repeated reopening
overoptimization complexity and loss of adaptability

Intervention Logic¶

The intervention separates two gates that are often conflated. The acceptance gate asks whether a candidate clears every floor, meets the aspiration threshold, carries adequate evidence, and uses authorized exceptions. The search-stop gate asks whether additional comparison has enough expected value to justify its cost, delay, and opportunity cost. A candidate can qualify while more comparison remains warranted because the order is biased or evidence weak; a search budget can expire with no qualifier, which triggers scope, capacity, threshold, or authority review rather than silent floor reduction.

define the decision the commitment horizon the cost of delay and who bears consequences
inventory needs obligations constraints and preferences with affected-party input
classify noncompensable safety rights legality feasibility and interoperability floors
define good-enough thresholds for remaining decision dimensions and state aggregation rules
justify aspiration levels using need evidence history benchmarks capacity and consequence
define evidence quality uncertainty margins and what counts as threshold satisfaction
define candidate eligibility search scope order channel and inclusion safeguards
allocate time information computation and review budgets plus an emergency fallback
evaluate each candidate through a consistent record without hiding missing or disputed evidence
test whether search order access or presentation systematically changes who can qualify
accept only when floors threshold confidence and exception authority are satisfied
if several options qualify apply the declared tie or preference rule rather than reopening optimization by stealth
compare expected benefit of more search with delay attention and opportunity cost
record acceptance rationale residual uncertainty rejected exceptions dissent and reopen triggers
transfer decision resources obligations monitoring and fallback to an accountable implementer
review realized sufficiency distributional effects gaming and threshold drift

The control loop observes candidate evidence, uncertainty, search coverage, delay, burden, and emerging alternatives; interprets whether apparent passage reflects real sufficiency or gaming, bias, missing evidence, or a moving standard; decides to continue, accept, apply a legitimate tie rule, escalate a conflict, or reopen; commits through a funded handoff; and verifies sufficiency against real outcomes rather than paperwork alone.

Key Components¶

The sixteen components cover semantics, values, evidence, search design, closure, legitimacy, implementation, and learning. Floors and aspiration thresholds are distinct. Candidate access and order are distinct from candidate evaluation. Acceptance and search stopping are coupled but not identical. Exception authority does not erase fairness review. Commitment is not complete until ownership, resources, safeguards, fallback, monitoring, and reopening travel with the decision.

Component	Description
Decision and Commitment Boundary ↗	Define what choice is being closed, for whom, over what horizon, and which downstream commitments follow. Prevents one threshold from being reused across materially different decisions or time scales. The exact choice to be closed, affected population, use context, commitment duration, dependencies, and downstream actions. Required fields: decision statement, affected parties, decision owner, use context, commitment horizon, delay cost, reversibility, dependency map. Validation questions: What becomes binding when the option is accepted? Who bears consequences without controlling the choice? When may the decision be revisited? Does delay change the feasible option set? Failure signals: scope ambiguity, hidden downstream commitment, wrong horizon, unowned consequence. Safeguards: decision charter, stakeholder map, horizon statement, dependency review.
Need and Outcome Set ↗	State the needs, outcomes, obligations, and affected-party interests the accepted option must address. Distinguishes actual sufficiency from convenient feature counts or stakeholder power. The real needs, outcomes, obligations, service conditions, and stakeholder interests that make an option sufficient or insufficient. Required fields: core need, affected party need, required outcome, time requirement, quality requirement, burden limit, evidence source, need owner. Validation questions: Which need is essential and which is preference? Can the option meet the need over the full horizon? Are less powerful affected parties represented? Has scarcity lowered aspiration below legitimate need? Failure signals: feature proxy for need, omitted party, temporary fit claimed permanent, deprivation adaptation. Safeguards: need trace, affected party review, outcome measure, periodic revalidation.
Noncompensable Floor Set ↗	Identify safety, legality, rights, feasibility, accessibility, and interoperability conditions that cannot be traded away. A high aggregate score cannot compensate for a floor breach. Safety, legality, rights, consent, feasibility, accessibility, interoperability, and other conditions that no compensatory score may override. Required fields: floor name, authority basis, test, evidence, pass rule, exception authority, remedy, review trigger. Validation questions: Why is this floor noncompensable? Is the pass test construct-valid and accessible? Can an exception be lawful and proportionate? Does one floor conflict with another? Failure signals: weighted offset of harm, hidden exception, weak floor measure, contradictory floor. Safeguards: conjunctive gate, independent review, exception record, conflict escalation.
Aspiration and Sufficiency Threshold ↗	Define the good-enough level for each compensable dimension and any valid aggregation or conjunctive rule. Records whose needs and evidence justify the aspiration level. The good-enough level for each compensable outcome and the declared rule for combining dimensions. Required fields: dimension, threshold, unit, direction, basis, aggregation rule, owner, effective period. Validation questions: What evidence supports the aspiration level? Is the threshold need-based or incumbent-based? Can dimensions validly compensate one another? Is a different threshold needed across contexts? Failure signals: post hoc threshold, arbitrary number, hidden weight, moving goalpost. Safeguards: precommitment, benchmark review, versioning, sensitivity analysis.
Evidence and Measurement Protocol ↗	Specify measures, sources, uncertainty, missingness, validation, and the rule for deciding threshold passage. Prevents false precision and evidence standards that change after seeing candidates. Measures, sources, sampling, validation, uncertainty, missingness, freshness, and the rule for determining threshold passage. Required fields: measure, construct, source, collection method, uncertainty, missingness rule, freshness, validation. Validation questions: Does the measure represent the need or only an easy proxy? Are sources comparable across candidates? How are missing and contested values treated? Can evidence be independently checked? Failure signals: proxy failure, selective evidence, stale measure, candidate specific standard. Safeguards: construct review, common evidence template, provenance, independent sample.
Candidate Population and Search Order ↗	Define eligibility, discovery channels, ordering, sampling, and inclusion checks for options considered. Makes first-fit order effects and access barriers visible. Eligibility, discovery channels, sampling, ordering, presentation, and inclusion rules for options considered. Required fields: eligibility, candidate sources, coverage goal, order rule, presentation standard, accommodation, exclusion reason, reopening. Validation questions: Which plausible option classes are absent? Does order correlate with power or familiarity? Is minimum comparison necessary before first-fit acceptance? Can excluded candidates challenge an error? Failure signals: incumbent first, channel exclusion, presentation bias, nonrepresentative first fit. Safeguards: multi channel search, rotation or randomization, minimum set, access audit.
Search Resource and Time Budget ↗	Bound time, attention, information, computation, money, and delay while preserving high-consequence review. A budget is a decision input, not permission to ignore protected floors. The planned limits on time, attention, information, computation, money, and delay for search and evaluation. Required fields: start, deadline, effort budget, information budget, computation budget, delay cost, reserve, extension authority. Validation questions: Is the budget proportionate to consequence and reversibility? Who bears the cost of delay and review? Can urgent constraints be independently verified? Is extension triggered by evidence rather than indecision? Failure signals: analysis paralysis, manufactured urgency, unequal review budget, unbounded extension. Safeguards: timebox, consequence tier, extension rule, emergency review.
Candidate Evaluation Record ↗	Apply criteria consistently and preserve evidence, missing data, exceptions, conflicts, and evaluator rationale. Supports audit and prevents invisible shifting of standards. A comparable record of criteria, evidence, missingness, conflicts, floor tests, threshold results, exceptions, and rationale for each candidate. Required fields: candidate, evaluator, criteria version, evidence, floor results, threshold results, uncertainty, conflict, rationale. Validation questions: Were all candidates tested under the same current criteria? Are exceptions and missing evidence visible? Can a reviewer reconstruct the result? Were corrections preserved rather than overwritten? Failure signals: selective documentation, criteria version mismatch, hidden conflict, irreproducible result. Safeguards: common template, immutable revision, conflict disclosure, audit sample.
Uncertainty and Confidence Margin ↗	Define how close calls, noisy measures, model risk, and contested evidence affect acceptance. Can require conservative margins, probes, sensitivity checks, or independent verification. The rule for incorporating measurement error, model uncertainty, missing evidence, variability, and disputed interpretation near a threshold. Required fields: uncertainty source, interval or grade, confidence requirement, conservative margin, probe rule, high consequence rule, reviewer, update. Validation questions: What uncertainty could reverse passage? Is a conservative bound or expected value appropriate? Can a cheap probe materially improve confidence? Does uncertainty differ systematically by candidate group? Failure signals: point estimate pass, asymmetric missingness, false precision, ignored model risk. Safeguards: confidence bound, sensitivity test, bounded probe, uncertainty disclosure.
Acceptance and Search-Stop Rule ↗	State when a qualifying option may be accepted and when expected improvement no longer justifies continued search. Separates threshold passage from the optional value of more comparison. The conjunctive conditions for accepting a candidate and the comparison of expected further-search value with search, delay, and opportunity cost. Required fields: floor pass, threshold pass, confidence pass, exception status, additional search value, cost, decision authority, timestamp. Validation questions: Did the candidate clear every required test? What plausible gain remains from more search? Are costs measured from all affected perspectives? Is closure reversible and monitored? Failure signals: qualifier without stop basis, stop without qualifier, hidden opportunity cost, status driven closure. Safeguards: conjunctive rule, marginal search review, decision record, reopen trigger.
Multiple-Qualifier and Tie Rule ↗	Choose among several sufficient options using a declared secondary preference, rotation, lottery, or other legitimate rule. Prevents optimization from returning through an undisclosed tie breaker. A declared legitimate method for choosing among several candidates that all meet the sufficiency standard. Required fields: qualifier set, secondary dimension, tie method, conflict rule, randomization or rotation, rationale, authority, review. Validation questions: Is the secondary rule consistent with satisficing rather than a hidden full optimization? Does it create structural advantage? Could rotation or lottery improve legitimacy? Are conflicts disclosed? Failure signals: secret tiebreaker, prestige default, optimization by stealth, arbitrary lottery. Safeguards: predeclared rule, conflict review, fairness test, rationale.
Exception and Escalation Rule ↗	Route infeasible, urgent, disputed, or high-consequence threshold conflicts to authorized review. Records temporary deviations, authority, expiry, and remedy. The authority and process for infeasible, urgent, disputed, novel, or high-consequence cases that cannot follow ordinary thresholds. Required fields: trigger, failed condition, necessity, authority, temporary rule, safeguards, expiry, review, remedy. Validation questions: Is the exception necessary and proportionate? Which protected floor is implicated? When does ordinary governance return? Can affected parties challenge the decision? Failure signals: exception as norm, urgency laundering, no expiry, inaccessible appeal. Safeguards: independent approval, time limit, retrospective review, public or bounded record.
Bias, Fairness, and Affected-Party Check ↗	Test threshold, search access, order, measures, and burden for systematic exclusion or capture. Includes accessible challenge and non-retaliation where the decision affects rights or livelihood. Assessment of whether thresholds, measures, search access, order, evidence burden, exceptions, and outcomes systematically disadvantage actors or groups. Required fields: affected group, access path, pass rate, evidence burden, outcome burden, accommodation, challenge, corrective action. Validation questions: Who helped define good enough? Does access to qualification differ? Are historical proxies perpetuating exclusion? Is appeal safe and usable? Failure signals: disparate exclusion, inaccessible evidence, retaliation, stakeholder omission. Safeguards: impact test, accommodation, ombuds channel, nonretaliation, corrective threshold review.
Commitment and Implementation Handoff ↗	Transfer the accepted decision, resources, obligations, safeguards, monitoring, and fallback to an accountable owner. Satisficing is incomplete if a decision closes but never becomes usable action. The transfer of the accepted option, authority, resources, requirements, safeguards, residual uncertainty, monitoring, and fallback to an accountable owner. Required fields: accepted option, owner, authority, resources, obligations, transition, monitoring, fallback, communication. Validation questions: Can the owner implement under the assumed conditions? Are safeguards and exceptions understood? What happens if readiness is absent? Are affected parties notified appropriately? Failure signals: unfunded choice, owner gap, paper acceptance, lost safeguard. Safeguards: readiness gate, funded plan, acknowledgment, fallback.
Performance Monitor and Reopen Trigger ↗	Observe whether the accepted option remains sufficient and activate fallback or renewed search when declared conditions fail. Protects against lock-in, threshold gaming, context change, and delayed harm. Outcome, floor, burden, context, and evidence signals that test continuing sufficiency and initiate correction, fallback, or renewed search. Required fields: indicator, baseline, threshold, cadence, owner, trigger, immediate action, reopen authority, fallback. Validation questions: Does monitoring measure actual need rather than only contract compliance? Can delayed or distributional harm surface? Is reopening timely and funded? Are enhancements separated from true insufficiency? Failure signals: monitoring absent, trigger ignored, perpetual reopening, sunk cost lock in. Safeguards: outcome dashboard, protected signal, materiality rule, termination or change path.
Threshold Review and Learning Cadence ↗	Reassess aspiration, floors, evidence, search process, outcomes, and exception history at scheduled and event-based reviews. Prevents temporary scarcity or low expectations from becoming a permanent low standard. Scheduled and event-based review of needs, floors, aspiration levels, measures, search order, budgets, exceptions, outcomes, and distribution. Required fields: cadence, trigger, evidence, participants, authority, option set, version, transition, next review. Validation questions: Did outcomes validate the original sufficiency claim? Have capacity or needs changed? Are exceptions showing a bad threshold or a bad process? Who is missing from review? Failure signals: downward threshold drift, obsolete measure, exception accumulation, temporary rule permanent. Safeguards: versioned review, affected party participation, benchmark, sunset, post change validation.

Common Mechanisms¶

Checklists, matrices, confidence bounds, timeboxes, procurement protocols, and reopen reviews are mechanisms. They implement replaceable parts of the lifecycle and should never be mistaken for the parent or used alone to claim a defensible satisficing process.

Mechanism	Description
Acceptance-Criteria Checklist ↗	Pins a single deliverable to a short list of pass/fail conditions written before work starts, so 'finished' means every box is ticked and nothing beyond them is owed.
Aspiration-Level Decision Rule ↗	Accept the first adequately evidenced candidate meeting the declared aspiration subject to order safeguards. It is a sequential choice, not the archetype itself. Selection and control: sequential_arrival costly_comparison reversible_choice measurable_sufficiency Minimum configuration: aspiration threshold and protected floors, candidate order rule, evidence confidence requirement, minimum search or coverage guard, acceptance authority, reopen condition. Implementation: set aspiration using need consequence and evidence diversify or randomize initial candidate access where order bias matters evaluate consistently record the first qualifying point and any continued-search rationale monitor accepted performance Validate with: bounded search duration, qualifying performance, controlled order effect, lower revisit rate. Misuse risks: first convenient wins, low aspiration, incumbent advantage, false certainty; safeguards: minimum set, confidence margin, fairness audit, threshold review.
Minimum Service or Quality Floor ↗	Establish noncompensable performance accessibility safety or rights conditions. It is a protected threshold, not the archetype itself. Selection and control: public_service regulated_choice safety_critical_option vulnerable_population Minimum configuration: floor definition and authority, valid test and evidence, affected-party and accessibility check, noncompensation rule, exception authority expiry and remedy, breach response. Implementation: identify nontradeable obligations validate the test against actual outcomes publish material requirements separate ordinary preference scores audit outcomes and exceptions Validate with: zero unapproved floor bypass, usable access, timely breach response, stable outcomes. Misuse risks: minimum becomes ceiling, inaccessible standard, captured exception, paper compliance; safeguards: outcome test, independent review, appeal, improvement backlog.
Timeboxed Search with Stop Gate ↗	Allocate a comparison period then apply a governed accept continue or escalate gate. It is a bounded search, not the archetype itself. Selection and control: delay_cost expanding_comparison decision_fatigue adequate_candidate_pipeline Minimum configuration: search start deadline and budget, candidate channels and coverage target, criteria and evidence standard, stop-gate decision options, extension authority and limit, emergency fallback. Implementation: estimate delay and additional-search value set the timebox before search begins monitor candidate and evidence coverage hold the gate with conflicts disclosed record any extension or reset rationale Validate with: fewer open ended searches, timely commitment, adequate coverage, bounded extensions. Misuse risks: manufactured deadline, rushed low quality search, automatic extension, sunk search cost; safeguards: independent urgency check, consequence tiering, one extension rule, fallback.
Qualified Shortlist Then First Fit ↗	Create a minimally representative eligible set before applying a sequential acceptance rule. It is a order bias control, not the archetype itself. Selection and control: biased_arrival_order incumbent_market multiple_channels moderate_screening_cost Minimum configuration: eligibility and outreach channels, minimum shortlist size or diversity, common evidence package, order rotation or randomization, first-fit rule and tie handling, exclusion and appeal record. Implementation: map candidate access barriers source through independent channels verify eligibility without ranking prematurely apply the declared order rule audit pass and exclusion by channel and group Validate with: broader access, lower incumbent bias, stable quality, transparent exclusion. Misuse risks: token shortlist, arbitrary minimum, hidden preference order, excessive screen; safeguards: coverage metric, channel audit, randomization, affected party review.
Threshold with Confidence Bound ↗	Require the conservative evidence bound rather than the point estimate to clear the threshold. It is a uncertainty control, not the archetype itself. Selection and control: measurement_noise model_uncertainty close_calls high_false_acceptance_cost Minimum configuration: outcome estimate and uncertainty model, required confidence or evidence grade, conservative direction, missingness and model-risk note, probe or additional-evidence rule, independent review threshold. Implementation: define consequence-specific confidence verify comparable evidence quality calculate or grade the conservative bound probe close cases where information value is high preserve residual uncertainty after acceptance Validate with: fewer false passes, consistent close calls, explicit uncertainty, efficient probes. Misuse risks: false precision, asymmetric evidence burden, overconservatism, model misspecification; safeguards: sensitivity analysis, qualitative review, fairness test, model validation.
Good-Enough Option Matrix ↗	Display floors, thresholds, evidence gaps, exceptions, and qualifying options without pretending to optimize. It is a transparent comparison, not the archetype itself. Selection and control: multiattribute_choice stakeholder_review noncompensable_criteria several_qualifiers Minimum configuration: rows for candidates and columns for criteria, floor and aspiration labels, evidence uncertainty and missingness, pass fail close and exception states, tie rule and rationale, version and decision record. Implementation: define dimensions and noncompensation normalize evidence presentation without erasing context review close calls and missingness identify the qualifier set apply the declared tie and stop rule Validate with: visible tradeoffs, lower hidden weighting, reproducible qualifiers, understandable rationale. Misuse risks: color coding false certainty, covert weighting, information overload, cherry picked columns; safeguards: plain language notes, audit, affected party review, raw evidence link.
Satisficing Procurement or Hiring Protocol ↗	Close a selection when qualification, fairness, confidence, and bounded-comparison rules are satisfied. It is a institutional selection, not the archetype itself. Selection and control: high_comparison_cost many_acceptable_options delay_cost recurring_selection Minimum configuration: role or need definition, legal and protected floors, job or service-valid thresholds, outreach and candidate-order guard, structured evidence record, conflict recusal challenge and monitoring. Implementation: validate requirements against actual need support accessible candidate participation create a minimum comparison pool apply structured evidence and confidence hand off onboarding or implementation with review Validate with: shorter cycle, sustained performance, fair access, fewer criteria changes. Misuse risks: credential inflation, discriminatory proxy, preferred vendor first, minimum quality ceiling; safeguards: validation, accommodation, conflict review, outcome and equity monitor.
Emergency Temporary Threshold ↗	Permit a proportionate temporary choice under urgency with expiry fallback and retrospective review. It is a time limited exception, not the archetype itself. Selection and control: material_time_pressure temporary_capacity_loss reversible_interim_option explicit_authority Minimum configuration: emergency trigger and evidence, nonwaivable floor set, temporary aspiration and scope, authority duration and expiry, fallback monitoring and remedy, retrospective review and restoration. Implementation: verify urgency independently where feasible choose the least harmful reversible qualifier record deviations and notify affected parties monitor closely and maintain alternatives expire and review without automatic renewal Validate with: timely safe action, no floor breach, completed expiry, restored ordinary threshold. Misuse risks: emergency ratchet, manufactured urgency, permanent low standard, rights bypass; safeguards: sunset, renewal evidence, independent review, remedy.
Performance-Triggered Reopen Review ↗	Reopen search when realized performance, context, equity, or protected floors invalidate acceptance. It is a adaptive governance, not the archetype itself. Selection and control: long_commitment uncertain_field_performance reversible_contract changing_context Minimum configuration: outcome and protected-floor signals, trigger materiality and cadence, monitoring owner and independent channel, immediate containment or fallback, reopen authority and option path, closure and learning record. Implementation: instrument real need and burden outcomes protect adverse-signal reporting evaluate triggers without sunk-cost bias activate fallback and bounded renewed search update thresholds and search design from evidence Validate with: timely reopen, reduced repeat failure, preserved fallback, threshold learning. Misuse risks: ignored trigger, perpetual reopening, punitive reporting, no alternative capacity; safeguards: independent signal, materiality rule, termination right, learning review.

Parameter / Tuning Dimensions¶

Threshold design is a coupled choice among aspiration, evidence, search breadth, delay, consequence, and reversibility. Tightening one dimension can shift burden or exclusion into another.

Aspiration level¶

Anchor the threshold in actual need, obligation, outcome, and horizon. Compare benchmarks, historical performance, capacity, and affected-party evidence, but do not let an incumbent baseline or deprivation define legitimacy by itself.

Noncompensation rule¶

Declare which dimensions are conjunctive floors and which preferences may trade. Weighted scores are acceptable only where compensation is legitimate; safety, rights, legality, consent, accessibility, or core feasibility often require separate gates.

Evidence and confidence¶

Tune measurement validity, provenance, freshness, missingness, close-call treatment, and independent verification to consequence and reversibility. A conservative margin can reduce false acceptance but may unfairly exclude candidates with less evidence access.

Candidate coverage and order¶

Specify eligible classes, discovery channels, sampling, presentation, accommodation, minimum pool, and waiting period. Sequential first-fit rules save delay only when order and access do not systematically privilege incumbents or well-connected options.

Search budget and stop sensitivity¶

Budget time, money, computation, attention, information, and delay. Compare the expected value of another search increment with those costs, but never interpret budget exhaustion as automatic permission to lower a protected floor.

Tie and multiple-qualifier logic¶

When several options qualify, apply a declared secondary preference, rotation, lottery, diversification, or portfolio rule. Prevent optimization from returning through an undisclosed tie breaker after sufficiency has already been established.

Exception authority and urgency¶

State who may authorize deviation, why it is necessary and proportionate, which floors remain nonwaivable, how long it lasts, what safeguards apply, and who receives remedy. Emergency thresholds require expiry and ordinary-threshold restoration.

Commitment horizon and reopen sensitivity¶

Short horizons and reversible choices permit lighter evidence and faster reopening. Irreversible or catastrophic decisions require stronger review, fallback, and challenge. Reopen only for material trigger failure, not every new preference.

Invariants to Preserve¶

Every valid implementation preserves these conditions even when domains and mechanisms change:

Every accepted option clears all noncompensable floors.
Aspiration thresholds remain tied to actual needs, obligations, affected parties, and commitment horizon.
Evidence and uncertainty rules are declared and comparably applied before candidate preference hardens.
Candidate access, discovery channels, ordering, and presentation remain reviewable for bias.
Acceptance and search stopping remain separate, explicit tests.
Exceptions have legitimate authority, reasons, safeguards, expiry, review, and remedy.
Commitment transfers resources, ownership, monitoring, fallback, and reopen triggers.
Thresholds do not drift below protected need through scarcity, gaming, automation, or repeated emergency.

If closure becomes faster only because floors, candidate access, evidence, or challenge were weakened, the rule is convenience rather than satisficing.

Target Outcomes¶

The target is proportionate closure: less search and delay without an increase in protected failure, exclusion, lock-in, or untested sufficiency.

criteria floors thresholds and search budgets documented before selection
reduced comparison time and candidate churn without more floor breaches
consistent evidence treatment and fewer post-hoc criteria changes
accessible exception and challenge use without retaliation
explicit tie handling and lower hidden optimization after qualification
accepted options meet declared needs over the commitment horizon
decision delay opportunity cost and review burden decline proportionately
no increase in safety rights legality accessibility or feasibility failures
distributional outcomes remain within protected bounds
reopen and fallback actions occur when triggers are met
thresholds improve from outcome evidence rather than drift downward silently

Realized outcomes, distributional effects, exceptions, threshold drift, missed alternatives, and reopen behavior test the rule. Contract passage alone is not outcome evidence.

Tradeoffs¶

A lower threshold and narrower search can accelerate commitment but increase false acceptance and inequity. A higher threshold and broader search can improve assurance but produce paralysis, exclusion, and opportunity loss.

Lower thresholds and narrower search reduce delay but increase false acceptance and the risk of normalizing low expectations.
Higher thresholds and broader search increase assurance but can create paralysis, exclusion, and missed opportunity windows.
Conservative confidence margins reduce close-call error but can impose unequal evidence burdens.
Sequential first-fit selection saves comparison cost but amplifies discovery-channel and order bias.
A single aggregate threshold simplifies choice but may hide noncompensable deficits and incommensurable needs.
More affected-party input improves legitimacy but requires time, access support, and conflict handling.
Frequent reopening improves adaptation but weakens commitment and can revive optimization by stealth.
Automation improves consistency and scale but can reproduce proxy bias and remove accessible human challenge.

The appropriate setting depends on consequence, reversibility, evidence availability, option arrival, delay cost, and who bears residual risk. High-stakes decisions require specialist review and do not outsource legal, clinical, or ethical duties to a generic heuristic.

Failure Modes¶

Most failures come from a threshold that is illegitimate, a search process that is biased, evidence that is incomparable, or commitment that lacks monitoring and reopening.

Threshold Set Too Low¶

Cause. Scarcity convenience or power defines aspiration below actual need. Consequence: Chronic underperformance becomes formally acceptable.

Mitigation. Need trace protected floors outcome review affected-party standing and upward revision trigger.

Threshold Set Too High¶

Cause. Perfectionism or incumbent specifications make qualification impractical. Consequence: Search paralysis cost and exclusion persist.

Mitigation. Evidence-based aspiration feasibility review cost-of-delay accounting and scoped requirement challenge.

First Fit Order Bias¶

Cause. Candidate order reflects access visibility or incumbent position. Consequence: A sufficient but systematically advantaged option always wins.

Mitigation. Minimum comparison set randomized or rotated order channel diversity and order-effect audit.

Compensatory Floor Bypass¶

Cause. A weighted score offsets a safety rights or legality failure. Consequence: Serious harm is laundered through aggregate performance.

Mitigation. Conjunctive noncompensable floors exception authority independent review and remedy.

Vague Good Enough Language¶

Cause. Criteria remain interpretive until a favored candidate appears. Consequence: Selection becomes unreviewable.

Mitigation. Predeclared measures examples evidence and close-call rules plus immutable revision history.

Metric Gaming¶

Cause. Candidates optimize visible thresholds while degrading unmeasured needs. Consequence: Formal passage diverges from real sufficiency.

Mitigation. Countermetrics qualitative validation random audit and post-acceptance monitoring.

Urgency Manufacture¶

Cause. Decision-makers create or exaggerate time pressure to narrow search and review. Consequence: Preferred options bypass comparison and safeguards.

Mitigation. Independent urgency basis emergency expiry retrospective review and conflict disclosure.

Unsupported Uncertainty¶

Cause. Point estimates near the threshold are treated as certain. Consequence: False acceptance and unstable closure.

Mitigation. Confidence margin probe sensitivity analysis and conservative high-consequence rule.

Aspiration Adaptation To Deprivation¶

Cause. Repeated scarcity lowers expectations rather than correcting capacity. Consequence: Inequity and poor service become self-reinforcing.

Mitigation. Benchmark need rights and peer outcomes; separately record capacity gap and investment obligation.

Search Budget Capture¶

Cause. Powerful actors receive more time evidence and channels than others. Consequence: Qualification access becomes unequal.

Mitigation. Common process accommodation support transparent budgets and distributional audit.

No Implementation Handoff¶

Cause. Decision closure is celebrated without resources ownership or safeguards. Consequence: The selected option fails for reasons unrelated to sufficiency.

Mitigation. Readiness gate funded handoff accountable owner fallback and monitoring.

Lock In After Acceptance¶

Cause. Sunk cost status or contract blocks warranted reopening. Consequence: An option remains after it stops being sufficient.

Mitigation. Reopen triggers termination rights review cadence and alternative readiness.

Perpetual Reopening¶

Cause. New preferences are mislabeled threshold failures. Consequence: Closure never creates stable action.

Mitigation. Commitment horizon materiality rule change control and separate enhancement backlog.

Automated Threshold Discrimination¶

Cause. Historical measures or proxies encode unequal opportunity. Consequence: Automated closure scales exclusion.

Mitigation. Construct validity fairness testing appeal human review drift monitoring and audit access.

High Stakes Satisficing Without Safeguards¶

Cause. Bounded rationality is invoked to shortcut irreversible safety rights or clinical decisions. Consequence: Avoidable severe harm.

Mitigation. Strong evidence independent review second check proportionality and nonwaivable standards.

Neighbor Distinctions¶

Neighbor boundaries are stable by decision stage. Several may appear in one program, but each owns a different intervention object.

Minimum Sufficient Solution¶

Implements the smallest scoped solution meeting a requirement after or around choice; this archetype governs decision acceptance and search closure.

Marginal Stop Rule¶

Stops the next increment when marginal benefit no longer justifies cost; this archetype accepts a candidate that clears an aspiration level.

Optimal Stopping Rule¶

Models expected reward of stopping versus future observations; this archetype supplies a governed sufficiency threshold even when probabilities are unavailable.

Bounded Search Pruning¶

Excludes branches using proof or bounds; it does not define good-enough acceptance or commitment.

Solution Space Bounding¶

Makes a large search space tractable; it does not say when a found option is sufficient.

Overoptimization Guardrail¶

Prevents further optimization from damaging protected values; it may trigger use of a satisficing rule but does not own candidate acceptance.

Heuristic Vs Algorithm Tradeoff And Selection¶

Chooses how to decide; this archetype is one specific closure rule that either a heuristic or algorithm may implement.

Minimum Viable Learning Release¶

Releases a minimum artifact to learn from real feedback; this archetype can select such a release but does not define its learning lifecycle.

Good Enough Option Design¶

Provisional option-presentation direction that should route to this parent after acceptance rather than create a competing full archetype.

Cross-Domain Examples¶

Cross-domain transfer depends on the same logic: needs and floors are explicit, candidate access is governed, passage is evidenced, closure is authorized, and real performance can reopen the decision.

Procurement¶

Define mandatory security legality interoperability and service floors, then good-enough cost and usability thresholds; search a fair supplier pool; accept a qualifier under the tie rule; preserve conflicts implementation readiness performance and termination rights. It preserves the parent lifecycle across procurement constraints.

Hiring And Admission¶

Use role-valid requirements accommodations structured evidence and a representative candidate channel; avoid pedigree proxies first-fit order bias moving standards and aggregate scores that offset essential or protected concerns. It preserves the parent lifecycle across hiring and admission constraints.

Incident Response¶

Select the first safe workable containment or restoration option meeting time-bound floors, with explicit temporary authority monitoring fallback and post-incident restoration of normal criteria. It preserves the parent lifecycle across incident response constraints.

Healthcare¶

Choose an acceptable option under clinical evidence safety patient values access and time, without equating cost convenience or population average with individual sufficiency. Preserve consent, escalation, second review, monitoring, and change of course. It preserves the parent lifecycle across healthcare constraints.

Public Service¶

Set entitlement safety accessibility and timeliness floors, distinguish them from service enhancements, record capacity gaps, and prevent temporary resource scarcity from redefining rights. It preserves the parent lifecycle across public service constraints.

Software And Product¶

Close a design or release decision when required user outcomes reliability security accessibility and evidence thresholds are met; keep enhancements separate and reopen on field failure or changed need. It preserves the parent lifecycle across software and product constraints.

Automated Search And Recommendation¶

Use explicit acceptability constraints, confidence, diversity and fairness guards before early stopping; log the reason and allow correction when a later option reveals systematic order bias. It preserves the parent lifecycle across automated search and recommendation constraints.

Extended example¶

A procurement team defines a two-year service decision, mandatory security, legality, accessibility, interoperability, and continuity floors, plus aspiration thresholds for reliability, support, transition effort, and total cost. It opens several discovery channels, standardizes evidence, audits order effects, and budgets six weeks of search. Two suppliers qualify. A declared tie rule selects the lower transition risk rather than reopening unrestricted optimization. The handoff funds implementation and fallback; field performance and a security breach trigger can reopen the choice. Annual review compares actual service with needs and checks whether incumbent familiarity lowered the threshold.

Non-Examples¶

These cases use “good enough” language or bounded work but do not satisfy the complete archetype:

first option is good enough without the parent lifecycle.
lowest cost equals sufficient without the parent lifecycle.
minimum compliance equals real need without the parent lifecycle.
weighted score offsets every harm without the parent lifecycle.
define threshold after seeing favorite without the parent lifecycle.
use urgency to skip authority and review without the parent lifecycle.
reduce search without checking candidate access without the parent lifecycle.
keep searching because more data is possible without the parent lifecycle.
accept without implementation capacity without the parent lifecycle.
treat satisficing as permanent low expectation without the parent lifecycle.
automate thresholds without construct or fairness review without the parent lifecycle.

Minimum viable products, shortlists, screening tools, timeboxes, confidence bounds, and option matrices become satisficing only when they sit inside the whole governed decision-closure lifecycle.

Abstractions this archetype builds on — directly (a source ingredient) or as a related pattern. Links follow the typed catalog namespace.

Built directly on (1)

Satisficing: Accept good-enough solution.

Also references 10 related abstractions

Bounded Rationality: Limited decision capacity.
Decision Fatigue: Reduced decision quality over time.
Heuristic: Mental shortcuts.
Opportunity Cost: Value of best alternative.
Optimal Stopping Rule: A rule maps a sequence of observations to a halt decision, trading the cost of stopping too early against stopping too late.
Scope Creep: A perimeter ratchets outward through small additions each judged against a drifting current baseline rather than the original charter, with no owner of the trajectory.
Screening: Inducing self-revelation.
Search and Retrieval: Locate and extract information.
Threshold: Safe vs harmful levels.
Uncertainty: Incomplete knowledge.

Variants¶

Narrower or domain-specific specializations that share this archetype's core structure. Recognized variants are established; candidate variants are provisional.

First Qualifying Option · implementation variant · recognized

Candidates arrive sequentially and the first one that clears every protected floor, the aspiration threshold, the evidence requirement, and any minimum-comparison safeguard can be accepted. The variant saves delay and comparison cost but makes arrival order, access, and presentation especially consequential.

Distinct from parent: This variant changes one bounded dimension of the parent's decision-closure contract while retaining threshold legitimacy, evidence, commitment, monitoring, and reopening.
Use when: Candidates genuinely arrive over time and reserving an acceptable option has value; Delay has a material and documented cost; Eligibility and discovery channels do not systematically hide plausible qualifiers; Rejection and acceptance evidence can be recorded consistently.
Typical domains: procurement, hiring, sequential offers
Common mechanisms: aspiration level decision rule, timeboxed search with stop gate, satisficing procurement or hiring protocol

Confidence-Bounded Satisficing · risk or failure variant · recognized

Acceptance depends on a conservative confidence bound, robust scenario result, or explicit probability of threshold passage rather than a point estimate. The variant is appropriate when measurement noise, model risk, sparse evidence, or close calls could reverse qualification.

Distinct from parent: This variant changes one bounded dimension of the parent's decision-closure contract while retaining threshold legitimacy, evidence, commitment, monitoring, and reopening.
Use when: Evidence is noisy or uncertain enough to change pass or fail status; False acceptance has greater consequence than bounded additional evidence collection; A meaningful uncertainty model, sensitivity range, or qualitative confidence rubric exists; A small probe or independent check can improve the decision before closure.
Typical domains: clinical decision, vendor selection, model governance
Common mechanisms: threshold with confidence bound, good enough option matrix, performance triggered reopen review

Emergency Temporary Satisficing · temporal variant · recognized

Verified urgency permits a temporary, reversible, narrowly scoped qualifying option under an explicitly time-limited aspiration rule. Nonwaivable floors remain protected, deviations are recorded, monitoring is intensified, and ordinary thresholds automatically return at expiry.

Distinct from parent: This variant changes one bounded dimension of the parent's decision-closure contract while retaining threshold legitimacy, evidence, commitment, monitoring, and reopening.
Use when: Delay creates immediate material harm and an ordinary search cannot finish in time; A reversible or containable interim option exists; Emergency authority, scope, expiry, fallback, and remedy can be defined; Ordinary selection or service capacity can continue or restart in parallel.
Typical domains: incident response, emergency procurement, service continuity
Common mechanisms: emergency temporary threshold, acceptance criteria checklist, performance triggered reopen review

Portfolio or Multiattribute Satisficing · governance variant · recognized

A candidate or small portfolio must clear a vector of floors and aspiration levels across several attributes, affected groups, scenarios, or complementary roles. It avoids collapsing all dimensions into one compensatory score and can select multiple sufficient options when no single option legitimately covers the whole need.

Distinct from parent: This variant changes one bounded dimension of the parent's decision-closure contract while retaining threshold legitimacy, evidence, commitment, monitoring, and reopening.
Use when: Several dimensions are legitimate and not fully commensurable; Coverage across groups, roles, regions, or scenarios matters more than a single winner; Complementary options can jointly satisfy the need within a bounded resource envelope; Noncompensation and aggregation rules can be stated before comparison.
Typical domains: portfolio selection, public service coverage, multi role hiring
Common mechanisms: good enough option matrix, minimum service or quality floor, qualified shortlist then first fit

Near names: Good-Enough Decision Threshold, Aspiration-Level Decision Rule, Sufficiency-Based Choice Closure, Bounded Search Acceptance Rule, Acceptable-Option Stop Design.