{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp06_four_proposal_generalization60_20260803","cell_id":"bounded_rivalry_governance__film_media_production","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_id":"brg-fmp-competing-assembly-pilot-001","proposal_index":1,"version":0,"title":"Bounded Competing Assemblies for Documentary Final-Cut Selection","problem":"A documentary producer commissions several editors to make alternative assemblies from the same footage for one scarce final-cut assignment, but the comparison lacks a governed arena. Editors can improve their odds by reserving useful footage, monopolizing assistants or edit facilities, exceeding nominal labor limits, tailoring cuts to disclosed test-audience reactions, or disparaging rival cuts. The selection can therefore reward resource capture, hidden overwork, and test manipulation rather than the strongest workable narrative treatment.","actors":["Documentary director","Executive producer or commissioning organization","Eligible picture editors and their paid assistant editors","Independent screening and craft-review panel","Production counsel or rights administrator","People depicted in the documentary whose consent or safety could be affected by a cut","Test viewers","Production workers supplying shared footage, transcripts, translations, sound, and edit facilities"],"observable_state":"Two or more editors pursue one final-cut assignment using materially unequal access to footage, labor, facilities, or audience feedback; work logs and media-access records show resource disparities or off-book effort; competing assemblies contain uncleared or unusually harmful material; judging criteria change after cuts are viewed; or participants dispute eligibility, evidence, conflicts, or access without a defined appeal path.","consequence":"The producer may select a cut whose apparent advantage came from privileged inputs, uncompensated effort, manipulation, or exported participant harm. Rivalry also encourages duplicated labor and damaged working relationships while yielding weak information about which editor can responsibly complete the film.","affected_objective":"Use controlled variation among editorial approaches to identify a narratively coherent, technically feasible, ethically supportable final-cut direction while preserving fair labor, shared-source integrity, participant protections, and the legitimacy of the selection.","intervention":"Run a paid, time-boxed competing-assembly round for one explicitly defined final-cut stewardship assignment. Before entrants are known, freeze a rulebook specifying eligibility, an identical read-only footage and metadata package, permitted communications, a capped number of paid editor and assistant hours, prohibited conduct, judging evidence, tie-breaks, conflicts, and appeal procedure. Each entrant produces the same-duration assembly and an editorial rationale. A panel that did not write or produce any entry first evaluates anonymized cuts against a preregistered multi-criterion rubric covering narrative intelligibility, evidentiary fidelity, feasibility of completion, and documented consent or safety constraints; it then conducts a non-anonymous feasibility review. Access and labor logs are audited, foul allegations are reviewed independently, and the producer records a post-round review before assigning final-cut stewardship. The pilot does not transfer underlying rights or displace any contractually protected final-cut authority.","structural_mapping":[{"archetype_element":"Rivalry purpose statement","domain_realization":"The comparison exists to reveal how distinct editorial structures organize the same documentary evidence, not to rank editors generally or intensify workplace pressure."},{"archetype_element":"Scarce prize or selection constraint","domain_realization":"One compensated final-cut stewardship assignment for the documentary, with its scope, term, credit, and fee stated before the round."},{"archetype_element":"Competitor eligibility boundary","domain_realization":"Editors qualify through disclosed craft, availability, conflict, confidentiality, and safeguarding requirements; every entrant receives equal compensation for the assembly round and may decline without employment retaliation."},{"archetype_element":"Contest arena boundary","domain_realization":"Entrants may select and arrange material from the common package and request logged clarifications. They may not reserve footage, contact depicted participants or test viewers, use undisclosed labor, obtain rival timelines, alter source records, or campaign against competitors."},{"archetype_element":"Performance metric and scoring basis","domain_realization":"A frozen multi-criterion rubric assesses narrative intelligibility, fidelity to recorded evidence, completion feasibility, and handling of consent and safety constraints; no audience-response number alone determines the result."},{"archetype_element":"Fair process and due process layer","domain_realization":"Cuts are anonymized for the first review, panel conflicts are disclosed, scores require written reasons, entrants receive a debrief, and material access, arithmetic, or rule-application errors can be appealed within a fixed window."},{"archetype_element":"Anti-sabotage and anti-collusion guardrail","domain_realization":"Immutable source access, audit logs, confidentiality rules, and published foul procedures deter footage withholding, interference, coordinated sham entries, and exchanges intended to prearrange the winner."},{"archetype_element":"Externality and spillover boundary","domain_realization":"Paid-hour limits, overtime prohibition, common safeguarding constraints, and review of risky portrayals keep competitive costs from being shifted to assistants, depicted people, or later post-production workers."},{"archetype_element":"Escalation and arms-race damper","domain_realization":"Identical cut duration, facility access, software baseline, paid labor allowance, and deadline prevent entrants from winning through an off-book hours or production-value arms race."},{"archetype_element":"Winner power and lock-in review","domain_realization":"Victory grants a bounded editing assignment, not ownership of footage, control over future hiring rules, or automatic authority over distribution and participant-safety decisions."},{"archetype_element":"Learning and recalibration loop","domain_realization":"After the round, the producer compares judging results with completion feasibility, labor records, disputes, and stakeholder harms, then revises or retires the format before any reuse."}],"mechanism_mapping":[{"mechanism_slug":"contest_rulebook","role":"Freezes eligibility, common inputs, permitted moves, scoring, conflicts, tie-breaks, and appeals before the competing assemblies begin.","counterfactual_removal":"Without the rulebook, the producer or panel could change expectations after seeing preferred cuts, and entrants could not distinguish legitimate experimentation from access abuse or challenge serious process errors."},{"mechanism_slug":"spending_cap_or_resource_cap","role":"Caps paid editor and assistant hours, cut length, facility access, and eligible production resources, with common definitions and auditable logs.","counterfactual_removal":"Without the cap, entrants with more money, unpaid labor, or facility access could create a polish and endurance arms race unrelated to editorial judgment."},{"mechanism_slug":"sabotage_or_foul_penalty_schedule","role":"Defines graduated responses to source withholding, unauthorized participant contact, undisclosed labor, rival interference, confidentiality breaches, and collusive prearrangement.","counterfactual_removal":"Without a published and consistently applied schedule, prohibited tactics would remain cheap when detected and enforcement could be selectively improvised."},{"mechanism_slug":"anti_collusion_monitoring","role":"Applies the same cross-entry screens to access logs, declared staffing, file histories, and suspiciously coordinated submissions, referring anomalies to an independent inquiry rather than treating them as proof.","counterfactual_removal":"Without monitoring, nominally rival entries could coordinate a result or exchange prohibited information while appearing independently compliant; without the referral limit, innocent similarities could be treated as verdicts."},{"mechanism_slug":"post_contest_impact_review","role":"Tests whether the selected assembly predicted responsible completion and whether the contest generated labor, participant, relationship, or lock-in harms, feeding changes into any later round.","counterfactual_removal":"Without ex-post review, the producer would learn only who won the rubric, not whether the arena selected the intended contribution or exported costs."}],"causal_chain":["Alternative documentary structures may reveal narrative possibilities that a single sequential edit would not expose.","A scarce final-cut assignment makes editors strategically interdependent rather than merely independent creators.","Freezing eligibility, common source access, scoring, conflicts, and appeals limits discretionary favoritism and input asymmetry.","Paid resource caps make additional hidden labor and production polish unavailable as reliable winning strategies.","Immutable access records, foul rules, and anomaly referral raise the detectability and expected cost of withholding, interference, undisclosed labor, and coordinated sham competition.","Anonymized first-stage review reduces identity effects while the later feasibility review restores information needed to judge completion risk.","Multi-criterion scoring keeps narrative strength coupled to evidentiary fidelity, feasibility, and participant constraints rather than to a single audience-response proxy.","Bounded assignment terms prevent the selected editor from converting one win into control of footage, future entry, or non-editorial rights.","Post-round comparison of the winning judgment with labor, feasibility, dispute, and harm records reveals whether the governed rivalry served its stated purpose and supplies a stop-or-revise decision."],"baseline":"The producer hires one editor through reputation and interviews, or informally asks several editors for differently resourced sample cuts and chooses through an undocumented discussion. Inputs, labor budgets, evaluation criteria, conflicts, and rights of challenge are not held constant.","nearest_rivals":["A single editor develops successive cuts with director notes; this avoids rivalrous duplication but provides less independent structural variation and does not test competing approaches under comparable conditions.","A collaborative edit workshop lets several editors jointly explore alternatives; it may preserve knowledge sharing but can converge early, obscure attribution, and does not govern selection for one scarce stewardship assignment.","A conventional paid edit test compares samples; it governs a hiring exercise but may omit common-source controls, externality protections, anti-sabotage rules, appeals, bounded winner power, and post-contest recalibration.","Audience testing of one or more cuts supplies viewer-response evidence but does not govern entrant eligibility, resource escalation, source access, labor spillovers, or the authority attached to winning."],"remaining_contrastive_claim":"The proposal's distinctive claim is limited to the structural combination: when independent editorial variation is desired for one scarce assignment, the comparison should be treated as a governed rivalry whose common inputs, legal moves, judging, labor externalities, contestability, winner authority, and redesign trigger are specified together. It is not a claim that competition is superior to sequential editing or collaboration.","authority_safety":{"decision_authority":"The rights-holding producer may authorize the paid assembly pilot and select an editor only within existing financing, employment, collective-bargaining, consent, privacy, credit, and final-cut agreements. An independent reviewer decides procedural appeals; qualified safeguarding or legal personnel retain authority over non-waivable participant and rights constraints.","authorized_first_step":"With voluntary written agreement from participants, the producer may run one reversible dry run using previously cleared footage, two compensated editors, synthetic or staff-only viewer feedback, no public release, and no automatic production appointment.","excluded_actions":["Using unpaid, speculative, or coerced editorial labor","Changing eligibility, scoring weights, source access, or resource caps after entrant identities or work are known","Exposing one entrant's timeline, rationale, or confidential communications to another","Contacting or filming depicted people for competitive advantage without separate authorization and consent","Treating an anomaly screen as proof of collusion or misconduct","Releasing pilot assemblies publicly or using uncleared material","Overriding contractual final-cut, labor, credit, privacy, accessibility, or safeguarding obligations","Allowing the winner to control future eligibility, judging, source access, or appeals"],"halt_rollback":"Pause judging when common-source integrity, equal access, compensation, confidentiality, panel independence, or participant safety cannot be verified. Preserve logs, pay all authorized work, suspend assignment, let an independent reviewer resolve remediable errors, and either rerun from the last uncontaminated stage or revert to the producer's preexisting noncompetitive selection process. Remove pilot copies according to the agreed retention schedule unless preservation is legally required."},"negative_tests":{"strongest_counterevidence":"Under equal inputs and paid caps, panel rankings remain unstable across repeated blind reviews, while the selected assembly does not predict completion feasibility and the process produces more labor or participant harm than a sequential edit. That pattern would undermine the premise that bounded rivalry yields decision-relevant editorial information here.","problem_falsifier":"Access logs, time records, interviews, and decision documents show that editors are not pursuing the same scarce assignment, cannot strategically affect one another, already receive materially equal inputs and governed review, and exhibit no plausible gaming, sabotage, collusion, escalation, externality, or lock-in pathway. The situation would then be ordinary comparative evaluation rather than the stated unmanaged-rivalry problem.","intervention_falsifier":"In the bounded dry run, rule-compliant reviewers cannot distinguish the assemblies beyond chance or cannot apply the rubric consistently; meaningful differences arise chiefly from uncapped or prohibited resources; safeguard violations increase; or the process cannot produce an appeal-resilient assignment without discretionary rule changes. Any of these results would reject this intervention design for the production.","risks":["Aesthetic judgment may resist reliable rubric-based comparison and produce false precision.","Anonymization may fail because editorial styles or prior footage familiarity identify entrants.","Caps may favor incumbents who already know the footage or own reusable tools and workflows.","Logged oversight may chill legitimate experimentation or expose sensitive creative reasoning.","Competitive framing may damage trust among editors, directors, assistants, and depicted participants.","The producer may use the pilot to obtain valuable speculative work while withholding a genuine assignment.","Safeguarding criteria may be gamed as visible checklist items rather than exercised as judgment.","Collusion screens may falsely flag similar edits caused by the same source material.","A bounded contest may still duplicate labor that a collaborative or sequential process would use more effectively.","The winner's bounded assignment may become de facto lock-in through schedule dependence even without formal rights." ]},"next_evidence_step":"Before any real assignment, conduct one small paid dry run with two volunteer editors using the same previously cleared 20–30 minute source packet to create five-minute assemblies under identical logged hours and tools. Pre-register the rubric and stop conditions; use three conflict-screened reviewers; repeat the blinded scoring after a washout interval; audit access and labor logs; obtain structured reports from editors, assistants, reviewers, and safeguarding staff; and compare score consistency, rule disputes, constraint violations, unplanned labor, and feasibility assessments with one sequential assembly of the same material. The purpose is to test operability and failure modes, not effectiveness at production scale.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"This is the first and only proposal in the sealed cell, so there is no earlier proposal within the cell against which to assess diversity.","revision_record":{"parent_version":null,"progress_targets_addressed":["Complete first-pass specification of the inferred domain problem","Preservation of the archetype's scarce-prize, governed-arena, externality-control, due-process, anti-abuse, and recalibration structure","Operational definition of actors, observable state, authority limits, rollback, falsifiers, and bounded evidence"],"conceptual_changes":[],"operational_changes":[],"evidence_changes":[],"claim_changes":[]}}