Evaluation Criteria Suspension During Divergence

During a protected divergent phase, deliberately defer ordinary evaluative filters so more varied options can be generated, then restore those filters through a governed convergence step.

Version

v1 · 2026-08-24 · History

Solution archetype #

412

Problem family

Decision, Search & Optimization Failure

Problem subfamily

Exploration, Exploitation & Variation Balance

Essence

This archetype protects a divergent phase from the premature return of the very criteria that will later be needed for convergence. It does not abolish evaluation. It changes the timing of evaluation: selected criteria are named, parked, and then reintroduced after the option space has expanded.

The pattern is useful when teams need novelty, alternatives, reframing, or exploratory hypotheses, but ordinary filters such as cost, feasibility, precedent, polish, evidence burden, or institutional acceptability appear too early. Early critique can make the process look disciplined while quietly eliminating the exploratory material needed for better design.

Compression statement

Evaluation criteria are necessary for selection, but when applied too early they shrink the option space, reward familiar frames, and turn divergence into incremental convergence. This archetype creates a bounded suspension: name which filters are deferred, keep non-suspendable safety and ethical floors active, generate and capture diverse possibilities, prevent premature critique, then reintroduce evaluation criteria in a staged way so novelty is preserved long enough to be examined.

Canonical formula: productive_divergence = bounded_suspension(criteria_subset) × protected_generation × idea_capture × staged_reintroduction(criteria)

When This Archetype Applies

Partial catalog groundingSome structural conditions are represented by existing abstractions, but no sufficient condition set is fully represented.

Diagnostic problem

A group or decision system needs divergent exploration, but ordinary evaluation criteria enter too early. Feasibility, cost, evidence burden, precedent, status expectations, or implementation constraints dominate the generative phase, causing premature convergence, local-optimum bias, self-censorship, narrow option sets, and over-selection of familiar proposals.

Applicability expression4 distinct conditions

Premature idea evaluationandAnticipatory self-censorshipandRecurring familiar optionsandConvergence criteria constrain discovery

Algebraic1∧2∧3∧4

3=a∨(b∧b′)

′ context guard? connective not recorded∅ no catalog witness yet

groundedpartly groundedopen

4 conditions, all required.

4Required in every casenumbered 1–4

These hold no matter which pattern applies.

1

Premature idea evaluation · open

A nominally divergent process evaluates ideas as they appear. more

and — also required

2

Anticipatory self-censorship · open

Participants self-censor in anticipation of ridicule, feasibility, budget, or expert rejection. more

and — also required

3

Recurring familiar options · grounded · any one of 2

The same familiar options recur despite new problems or opportunities. more

a

domainEinstellung Effect— Diagnose a skilled solver's failure on a near-surface variant not as missing knowledge but as retrieval-driven blocking — the first-cued familiar method is retrieved, judged adequate, and halts search before a simpler dominating alternative is ever compared.

or — any one of these establishes this case

b

domainGolden Hammer Anti-Pattern— Diagnose a team applying its familiar technology to a poorly-fitting problem because the acquisition cost of an alternative is visible while the misfit cost is diffuse and downstream, so fluency reshapes what counts as the right tool before fit is ever asked.

AND both parts required

context guardThe incumbent technology was selected in at least one earlier tool-choice session and recurs in the current new-problem session.

suppliesThe recurring options are the same familiar options across the compared sessions. · The recurrence occurs across multiple sessions.

How this was matched — 3 shared + 3 branches

The same familiar option set recurs across sessions despite a materially new circumstance.

All of

• comparisonThe recurring options are the same familiar options across the compared sessions.
• quantifierThe recurrence occurs across multiple sessions.
• relationThe familiar-option recurrence persists despite the applicable new circumstance.

…and any one of

• branchA new problem is present in the later session or sessions.
• branchA new context is present in the later session or sessions.
• branchA new opportunity is present in the later session or sessions.

and — also required

4

Convergence criteria constrain discovery · open

Criteria valid for convergence are too narrow for discovering new frames or solution families. more

Other requirements and context (3)

Why these sit outside the expression

Goal — a goal states an intended outcome or evaluation criterion, not a pre-existing situation that independently summons the archetype.

Supporting context — it may accompany or help interpret the situation, but it is not a load-bearing condition in a sufficient diagnostic set.

• GoalTeams need more option variety before making a high-commitment decision.

• Supporting contextEarly-stage ideas are too underdeveloped to be fairly assessed by implementation-stage criteria. more

• Supporting contextStatus hierarchies, expert norms, or risk aversion suppress weak signals and unconventional contributions.

Coverage

1 of 4 conditions grounded · 3 open.

Read the methodology · Download the trigger-logic data

Structural sequence

1. Define the divergent phase boundary.
2. List which criteria are suspended and which guardrails remain active.
3. Generate and capture ideas without immediate critique or scoring.
4. Check whether the output is diverse enough to justify convergence.
5. Restore criteria through a staged convergence gate.
6. Record what was learned about the process and the option set.

The suspension is temporary, scoped, and reversible. That is what distinguishes the archetype from irresponsibility or unbounded speculation.

Why it works

Many early ideas are ambiguous: they may look impractical until combined with another idea, translated to a different scale, or reframed around a different user. If feasibility critique arrives immediately, weak-form ideas never mature. By holding evaluation criteria in reserve, the system gives exploration enough room to produce material worth evaluating.

The key is the phrase during divergence. The archetype is not anti-evaluation. It is anti-premature-evaluation.

Components in practice

This archetype protects a divergent phase from the premature return of the criteria that convergence will later need, changing the timing of evaluation rather than abolishing it, and its components work as a bounded, reversible suspension with both a floor and an exit. The divergence phase contract frames the whole episode, telling participants when evaluation is suspended and when it will return so deferral is understood as method rather than license. The suspended criteria register removes confusion by naming exactly which filters — cost, feasibility, precedent, polish, evidence burden, institutional acceptability — are deferred, while the non-suspendable guardrail set keeps safety, legality, consent, and dignity active so suspension never shades into irresponsibility. These three components define the scope and limits of the protected window.

The remaining components protect generation inside that window and ensure the process eventually closes. The judgment deferral norm blocks not only explicit critique but the subtle status signals that quietly collapse exploration, and the idea capture buffer preserves rough, weak-form material long enough for it to combine, mutate, and reveal latent value before any scoring. Finally, the convergence reentry gate ensures the suspension is temporary, returning the process to responsible evaluation through a staged sequence — typically beginning with potential, learning value, and frame diversity before cost and feasibility — so the harshest filters do not rebound first and erase the value of divergence. Together these components give exploration room to produce material worth evaluating while guaranteeing that disciplined selection still arrives.

Component	Description
divergence phase contract ↗	tells participants when evaluation is suspended and when it will return.
suspended criteria register ↗	prevents confusion by naming deferred filters.
non-suspendable guardrail set ↗	keeps safety, legality, consent, and dignity active.
judgment deferral norm ↗	prevents both explicit critique and subtle status signals from collapsing exploration.
idea capture buffer ↗	preserves rough material long enough for combination.
convergence reentry gate ↗	ensures that the process eventually returns to responsible evaluation.

Common mechanisms

Simple mechanisms include no-critique rules, yes-and extension, deferred scoring queues, wild-card quotas, and timeboxed blue-sky sprints. More structured mechanisms include morphological matrices, anonymous idea capture, separate generator/evaluator roles, and two-pass evaluation. These are mechanisms, not the archetype itself. The archetype is the governed phase design that suspends selected criteria and later restores them.

Boundary notes

This archetype is closest to False Convergence Prevention, Option Preservation, Local Optimum Escape, and Progressive Narrowing. It differs because its central move is neither testing convergence, preserving existing options, perturbing search, nor filtering candidates. Its central move is controlling the timing of evaluation criteria across the divergence-to-convergence boundary.

It should collapse into a mechanism only when the artifact is merely a facilitation rule such as “no criticism during brainstorming.” It remains a full archetype when the process includes criteria registration, guardrails, capture, coverage checking, and staged reintroduction.

Example

A strategy team wants to identify new growth options. In prior meetings, every idea is immediately rejected as too expensive, too unfamiliar, or too hard to explain to leadership. The team creates a divergent phase where cost, current staffing, and near-term executive approval are deferred, while legal and reputational guardrails remain active. Ideas are captured without ranking. After the phase, the team reintroduces criteria in stages: strategic potential, learning value, risk, cost, and implementation sequence. Most ideas still fail, but the process reveals two hybrid options that would not have survived immediate feasibility critique.

Failure modes

The most common failure is indefinite suspension: the team never returns to convergence. Another is hidden evaluation, where people obey the no-critique rule verbally while punishing unusual ideas socially. A third is guardrail loss, where harmful ideas are protected as creativity. A fourth is criteria rebound, where the harshest filters return first and erase the value of divergence. Each failure is controlled by the same discipline: scoped suspension, active guardrails, careful capture, and staged reentry.

Quality and review notes

The pre-draft disposition check found no exact accepted archetype, pilot archetype, variant addition, alias-map entry, duplicate-merge entry, or previous batch output matching this candidate. The draft is merge-sensitive because judgment deferral can appear as a mechanism, but the broader structure fills the related-only gap for divergence_convergence_in_the_design_process as a direct source-prime archetype.

Common Mechanisms

8 documented mechanisms across 6 implementation forms.

The grouping reflects forms represented among the mechanisms currently documented for this archetype; an absent form is not necessarily an impossible implementation.

Analysis, Modeling & Optimization · 1 mechanism

• Morphological Matrix — Lays the problem out as a grid of independent parameters and their options, then generates ideas by systematically combining across cells.

Assessment, Review & Assurance · 1 mechanism

• Two-Pass Evaluation — Evaluates in two ordered passes — a permissive first pass on potential and learning, then a strict second pass on cost and feasibility — so rough-but-promising ideas are not killed by the harshest filter first.

Communication, Facilitation & Learning · 2 mechanisms

• Timeboxed Blue-Sky Sprint — Fences divergence inside a short, fixed timebox with criteria suspended, then hard-stops into convergence when the clock runs out.
• Yes-And Extension — Requires each contribution to be accepted and extended rather than judged, turning would-be blocks into additive building.

Organization, Role & Governance · 1 mechanism

• Separate Generator and Evaluator Roles — Splits generation and evaluation across different people so no one critiques what they are still trying to produce.

Protocol, Workflow & Routine · 1 mechanism

• Deferred Scoring Queue — Routes every raw idea into a holding queue that is scored later in one batch, so capture is never interrupted by evaluation.

Rule, Policy & Commitment · 2 mechanisms

• No-Critique Rule — Bars spoken criticism during the divergent phase so fragile ideas survive long enough to be built on — while safety and legality objections stay allowed.
• Wild-Card Quota — Mandates a minimum number of deliberately extreme, rule-breaking ideas so the option set is forced past the safe and familiar.

Related Abstractions

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

Built directly on (1)

• Divergence-Convergence in the Design Process: Expand then refine ideas.

Also references 23 related abstractions

• Ambidexterity (Exploit vs. Explore): Balance exploit vs explore.
• Boundary: Defines system limits.
• Bounded Rationality: Limited decision capacity.
• Closure: Ensures operations remain within a set.
• Constraint: Limits possibilities to guide outcomes.
• Convergence: Movement toward stable state.
• Decision: Committing to one alternative from a set under uncertainty and trade-off, collapsing open deliberation into a chosen path and foreclosing the others.
• Design Prototyping: Early models for testing.
• Diversity: Maintaining functionally distinct types within a system so that variation provides resilience and coverage that uniformity cannot.
• Experimental Design: Structuring an investigation through deliberate intervention, controlled assignment, and measurement so that causation can be distinguished from mere correlation and confounding.

▸ Show 13 more

Variants

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

Brainstorming Judgment Deferral · implementation variant · recognized

A session-level version that prohibits critique while ideas are being generated.

• Distinct from parent: Narrower than the parent because it focuses on workshop behavior rather than full divergence/convergence process design.
• Use when: A group ideation session is vulnerable to premature criticism or status dismissal.
• Typical domains: brainstorming, design studios, team retrospectives
• Common mechanisms: no critique rule, yes and extension, deferred scoring queue

Blue-Sky Research Sandbox · domain variant · recognized

A research-planning form that temporarily defers grantability, publication certainty, and immediate method feasibility.

• Distinct from parent: Domain-specific to scientific or scholarly agenda formation.
• Use when: A research group needs to explore speculative questions before narrowing to fundable or testable projects.
• Typical domains: basic research, R&D portfolio design, doctoral seminar planning
• Common mechanisms: timeboxed blue sky sprint, two pass evaluation

Feasibility Embargo for Concept Generation · implementation variant · recognized

A variant that specifically delays engineering, cost, schedule, or implementation feasibility objections until concepts are generated.

• Distinct from parent: Narrower than the parent, which can suspend many forms of evaluation beyond feasibility.
• Use when: Feasibility critique is valid later but prevents target-state imagination or concept variety.
• Typical domains: product design, software architecture, urban planning
• Common mechanisms: deferred scoring queue, two pass evaluation

Criteria Reintroduction Ladder · temporal variant · candidate

A staged convergence variant that restores criteria from broad potential to hard constraints in a deliberate sequence.

• Distinct from parent: It is the convergence-side complement of the parent archetype.
• Use when: The first convergence pass would otherwise eliminate novel ideas too harshly.
• Typical domains: innovation portfolio selection, policy option review, creative concept evaluation
• Common mechanisms: two pass evaluation, deferred scoring queue

Near names: Judgment Deferral During Ideation, No-Critique Divergence, Blue-Sky Sandboxing, Feasibility Embargo.

Editorial Notes

Problem Classification

Classification: Decision, Search & Optimization Failure → Exploration, Exploitation & Variation Balance

Problem kernel: evaluation criteria suppress variation before options form

Rationale: Feasibility, precedent, status, and evidence burdens enter the divergent phase early enough to prevent meaningfully different candidates.

Independent corroboration: The earliest necessary condition in the frozen evidence is: A group or decision system needs divergent exploration, but ordinary evaluation criteria enter too early. That is a exploration exploitation and variation balance problem because A system generates too little, too much, or poorly protected variation and cannot transition deliberately between novelty, evaluation, exploitation, consolidation, and retirement.

Review outcome: Independent reviewer agreement; high confidence.