{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp06_four_proposal_generalization60_20260803","cell_id":"bounded_rivalry_governance__criminology_forensic","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_id":"brg_parallel_hypothesis_resource_gate_p3_v0","proposal_index":3,"version":0,"title":"Parallel-Hypothesis Gate for Cold-Case Resources","problem":"A major-case review unit has several plausible explanations for an unresolved case but only limited specialist analysis, laboratory testing, and investigative-review capacity. Informal rivalry develops over which hypothesis receives those resources. When the incumbent case team controls the chronology, evidence requests, and presentations to decision-makers, a hypothesis can prevail through early ownership, selective testing, evidence hoarding, or attacks on alternatives rather than through discriminating predictions. Unbounded competition can also produce duplicative analysis, escalating requests, and intrusive proposals directed at people associated with the case.","actors":["Incumbent case investigators","Independent hypothesis teams","Major-case review supervisor","Evidence custodian","Forensic laboratory and specialist analysts","Independent scoring and appeal panel","Prosecution and defense disclosure officials","People named in investigative hypotheses","Legal authorities responsible for searches, testing, charging, and disclosure"],"observable_state":"The unit can observe multiple plausible hypotheses competing for a fixed pool of specialist hours; one team controlling the master chronology or briefing sequence; repeated requests designed mainly to confirm an incumbent explanation; alternative-theory memoranda based on incomplete evidence access; uncited or differently coded claims in competing presentations; duplicated work; and proposals whose privacy, evidence-consumption, or third-party burdens are not compared alongside their expected informational value.","consequence":"Scarce investigative resources can follow organizational position or persuasive presentation rather than the hypothesis offering the clearest lawful test. A prematurely dominant theory may narrow subsequent evidence collection, while rival teams may escalate accusations, duplicate work, or propose burdensome actions merely to remain competitive. The resulting ranking can be mistaken for a finding about guilt even though it only reflects a resource-allocation process.","affected_objective":"Allocate bounded next-stage analytical resources toward hypotheses that make distinct, falsifiable, evidence-grounded predictions while preserving disclosure duties, evidentiary integrity, individual rights, and the ability of later evidence to reopen the field.","intervention":"When an authorized cold-case review identifies at least two materially distinct and minimally supported hypotheses, create separated hypothesis teams with equal access to a read-only, citation-addressable copy of the review record. Before receiving additional results, each team registers its hypothesis, contradicting evidence, discriminating predictions, proposed tests, uncertainty, and expected privacy, evidence-consumption, and third-party burdens. A frozen rulebook prohibits field contact, public accusation, alteration of evidence, undisclosed outside searches, withholding record items, and personal attacks. Equal analyst-hour and request-page caps limit escalation. An independent panel scores evidentiary coverage, falsifiability, prediction distinctiveness, provenance, treatment of contradictory evidence, feasibility, and rights-adjusted information value. Up to two complementary hypotheses receive only a next-stage validation slot: access to specified specialist review or a recommendation for an ordinarily authorized test. Advancement does not establish guilt, probable cause, admissibility, or permission for coercive action. New materially relevant evidence triggers a challenger window, and a post-stage review compares registered predictions with results before further resources are allocated.","structural_mapping":[{"archetype_element":"Rivalry purpose statement","domain_realization":"Use controlled rivalry to expose contrasting explanations and identify discriminating next steps, not to crown a preferred suspect or reward argumentative dominance."},{"archetype_element":"Scarce prize or selection constraint","domain_realization":"A limited number of next-stage validation slots comprising specialist analyst hours, non-destructive laboratory review, or consideration of a legally authorized test request."},{"archetype_element":"Competitor eligibility boundary","domain_realization":"Only materially distinct hypotheses with cited record support, an identified potential falsifier, and a team free of undisclosed conflicts may enter."},{"archetype_element":"Contest arena boundary","domain_realization":"Teams may analyze the mirrored record and submit bounded test plans. They may not contact witnesses or named persons, access undisclosed databases, alter evidence, make public allegations, obstruct another team, or exercise investigative or legal authority."},{"archetype_element":"Performance metric and scoring basis","domain_realization":"Scoring combines evidentiary coverage, treatment of contradictions, falsifiability, discriminating predictions, source provenance, feasibility, and expected information value after privacy, consumption, and third-party burdens are considered."},{"archetype_element":"Fair process and due process layer","domain_realization":"The evidence index, rules, rubric, caps, panel conflicts, tie-breaks, written reasons, and appeal window are fixed before submissions; appeals address access, procedure, or scoring errors rather than relitigating guilt."},{"archetype_element":"Anti-sabotage and anti-collusion guardrail","domain_realization":"Read-only evidence copies, access logs, separated drafting, disclosure of outside assistance, citation audits, and penalties for withholding, tampering, impersonation, or interference protect the comparison."},{"archetype_element":"Externality and spillover boundary","domain_realization":"No contest act may change a person's legal status or authorize contact, surveillance, search, seizure, charging, or public identification. Expected privacy and third-party burdens reduce a proposal's score, and ordinary legal and disclosure review remains noncontestable."},{"archetype_element":"Escalation and arms-race damper","domain_realization":"Teams receive equal analyst hours, submission length, clarification opportunities, and proposed-test budgets, limiting success through staffing volume or escalating investigative demands."},{"archetype_element":"Winner power and lock-in review","domain_realization":"Advancement grants only a bounded validation slot; the evidence ledger remains independent of the leading team, a second complementary hypothesis may advance, and materially new evidence reopens entry."},{"archetype_element":"Learning and recalibration loop","domain_realization":"After validation, reviewers compare registered predictions, contradictions, burdens, and actual results, then revise scoring or end the contest if ranking did not identify informative next steps."}],"mechanism_mapping":[{"mechanism_slug":"contest_rulebook","role":"Freezes hypothesis eligibility, equal evidence access, permitted analysis, prohibited conduct, scoring, tie-breaks, panel conflicts, written reasons, and appeals before teams submit plans.","counterfactual_removal":"Without the rulebook, incumbent teams or reviewers could alter access and evaluation standards after seeing which explanation benefits, and advancement would lack a contestable boundary."},{"mechanism_slug":"bracket_or_tournament_structure","role":"Uses a staged structure: eligible hypotheses first pass an evidence-and-falsifiability gate, finalists receive comparative scoring, and no more than two plans advance to bounded validation.","counterfactual_removal":"Without staged narrowing, every theory could consume scarce specialist capacity, or an early informal favorite could bypass comparison and monopolize the next stage."},{"mechanism_slug":"ranked_leaderboard_with_audit","role":"Creates a confidential criterion-level ranking and audits leading submissions for complete citations, accurate representation of contradictory evidence, reproducible calculations, and compliance with the common record.","counterfactual_removal":"Without audit, a team could rank highly through selective quotation, unsupported certainty, duplicated claims, or evidence unavailable to its rivals."},{"mechanism_slug":"spending_cap_or_resource_cap","role":"Equalizes analyst hours, submission length, clarification access, and the notional burden budget for proposed next steps.","counterfactual_removal":"A larger or incumbent team could win through staffing, repeated requests, or escalating investigative intrusions rather than superior discriminating logic."},{"mechanism_slug":"multiple_award_or_portfolio_selection","role":"Allows two nonredundant hypotheses to receive bounded validation when their predictions cover different consequential uncertainties.","counterfactual_removal":"A single-winner rule would encourage premature closure and make one noisy comparison determine the entire direction of further review."},{"mechanism_slug":"sabotage_or_foul_penalty_schedule","role":"Publishes graduated consequences for evidence withholding, unauthorized searches, tampering, identity leakage, personal attacks, interference, and undisclosed coordination.","counterfactual_removal":"Off-arena conduct could become a route to advancement, while discretionary punishment could itself favor the incumbent theory."},{"mechanism_slug":"challenger_access_window","role":"Permits a new or previously screened-out hypothesis to enter after materially relevant evidence, a failed registered prediction, or a documented evidence-access defect.","counterfactual_removal":"An advanced hypothesis could retain resources through organizational inertia even after its discriminating predictions fail or the evidentiary field changes."},{"mechanism_slug":"post_contest_impact_review","role":"Compares registered predictions and expected burdens with validation results, reviews whether advancement distorted later work, and revises or retires the gate.","counterfactual_removal":"A persuasive but poorly predictive scoring system could continue directing case resources without evidence that winning remained coupled to informative investigation."}],"causal_chain":["Several plausible hypotheses depend on the same scarce pool of analytical and review resources.","Informal rivalry lets evidence ownership, seniority, selective testing, or presentation skill influence which hypothesis advances.","A common read-only record and explicit eligibility rules create a legitimate field of materially distinct competitors.","Pre-registration forces each team to expose contradictions, falsifiers, and discriminating predictions before it knows later results.","Equal resource caps and prohibited-action rules prevent staffing volume, evidence control, harassment, or escalating intrusion from becoming winning strategies.","Criterion-level scoring and citation audit connect advancement to evidence-grounded information value rather than rhetorical confidence.","Portfolio advancement preserves a meaningful alternative when two hypotheses test different uncertainties.","Ordinary legal review remains outside the contest, preventing a resource ranking from becoming a determination of guilt or investigative authority.","Validation results test the registered predictions, while challenger windows reopen the field after failure or new evidence.","Post-stage review recalibrates or retires the arena if its winners do not produce informative, lawful next steps."],"baseline":"Continue supervisor-led case conferences in which the incumbent team presents its theory, colleagues suggest alternatives, and a manager assigns scarce forensic or investigative resources using professional judgment. This baseline is flexible and collaborative but may not provide equal evidence access, prospective prediction registration, explicit burden comparison, independent scoring, resource caps, or a formal route for a displaced hypothesis to return.","nearest_rivals":["A conventional multidisciplinary case conference, which integrates expertise without formal rivalry but can preserve briefing-order, hierarchy, and evidence-control advantages.","A single independent cold-case review, which can challenge incumbent thinking but substitutes one reviewer's synthesis for structured comparison among distinct hypotheses.","A checklist requiring investigators to list alternatives, which is lightweight but does not give an alternative theory independent advocates, equal resources, or a genuine path to validation.","Unrestricted parallel investigation teams, which create rival explanations but can duplicate work, compete for witnesses and evidence, escalate intrusion, or conceal information from one another.","Random allocation of scarce tests among qualifying hypotheses, which avoids evaluator favoritism but does not prioritize tests by falsifiability, burden, or expected ability to distinguish explanations."],"remaining_contrastive_claim":"The proposal's bounded claim is that, when materially distinct investigative hypotheses already compete for scarce analytical resources, separating teams, equalizing the record, preregistering discriminating predictions, and granting only limited validation slots can make the path to advancement depend more directly on testable informational contribution. It does not treat the winning hypothesis as true or authorize any action that ordinary evidentiary, disclosure, or legal safeguards would prohibit.","authority_safety":{"decision_authority":"The major-case review authority may convene the hypothesis gate and allocate internal analytical-review capacity. Laboratory authorities retain control over evidence handling and validation; prosecutors, defense officials, courts, and other legally designated actors retain disclosure, search, testing, charging, admissibility, and case-disposition authority.","authorized_first_step":"Approve a retrospective, no-case-impact simulation using masked, time-split copies of closed-case or fully synthetic records. Competing teams analyze only the record available at the selected historical cutoff, register predictions, and are scored against later packet segments by reviewers who have no operational role in those cases.","excluded_actions":["Opening or reopening a criminal investigation from simulation results","Identifying, contacting, surveilling, searching, accusing, charging, or changing the status of any person","Consuming, altering, retesting, or destroying original evidence","Withholding discoverable or exculpatory information from any legally entitled party","Using advancement or simulation scores as proof of guilt, probable cause, admissibility, analyst competence, or employee performance","Giving the review panel authority to approve coercive or destructive investigative actions","Allowing an incumbent team to curate a rival's evidence access or judge its submission","Publishing case identities, team identities, hypothesis rankings, or sensitive investigative methods"],"halt_rollback":"Stop the exercise upon identity unmasking, unequal record access, undisclosed case knowledge, evidence alteration, contact with a case actor, scorer conflict, or any attempt to operationalize a simulation result. Revoke access to working copies, preserve access and scoring logs for safeguard review, suppress rankings, and void the affected comparison. Because the first step uses masked copies and carries no operational authority, rollback restores the ordinary review process without changing evidence, cases, or resource allocations."},"negative_tests":{"strongest_counterevidence":"Bounded rivalry makes teams more committed to assigned explanations, encourages concealment until scoring, produces cosmetically distinct versions of the same theory, or rewards confidently specific predictions that are poorly calibrated. Rankings may also prove unstable across panels, fail on later record segments, or divert effort from cooperative synthesis without improving resource decisions.","problem_falsifier":"The inferred problem is unsupported if review records show no scarce-resource dependence among hypotheses, all materially plausible explanations receive equivalent evidence access and documented consideration, resource decisions consistently use prospective discriminating tests, and incumbent ownership does not affect advancement or re-entry.","intervention_falsifier":"The intervention fails if its leading hypotheses do not better anticipate held-out evidence or identify more discriminating bounded tests than ordinary case conferences, or if any apparent improvement depends on greater unsupported accusation, privacy burden, evidence consumption, team entrenchment, information withholding, or scorer inconsistency.","risks":["Formal rivalry may deepen commitment to assigned hypotheses rather than reduce tunnel vision.","Teams may relabel similar explanations to satisfy the entry rule.","Scorers may mistake narrative coherence or specificity for evidentiary value.","Historical cutoff packets may omit context that investigators legitimately possessed at the time.","Panel knowledge of later outcomes could leak into scoring.","Confidential leaderboards could still stigmatize investigators or named persons.","Resource caps may suppress a genuinely complex hypothesis more than a simple but misleading one.","Two advancing hypotheses may duplicate burdens despite portfolio scoring.","Competition may discourage timely sharing of a discovery that should immediately enter the common record.","The process could be misrepresented as an adjudication of guilt despite explicit limits." ]},"next_evidence_step":"Run a preregistered, retrospective crossover simulation using masked, time-split case packets with no live operational consequence. For each packet, compare the bounded hypothesis gate with a conventional case-conference review using equivalent analyst time and the same cutoff record. Before revealing later packet segments, record each condition's hypotheses, cited support, contradictions, confidence, discriminating predictions, proposed resource use, and anticipated rights burdens. Independent reviewers then score prediction calibration against the later segment, citation completeness, unsupported claims, hypothesis redundancy, panel agreement, analyst time, and proposed intrusion. Halt if masking fails or participants recognize a case. Use the result only to determine whether the gate's scoring and safeguards merit further controlled evaluation.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Proposal 1 governed rivalry among forensic laboratory teams for development resources by rewarding detection of controlled analytical and documentation defects. Proposal 3 instead governs rivalry among competing explanations within a major-case review, and its scarce prize is a bounded next-stage validation slot. It does not rank laboratories, reward institutional error disclosure, or use a defect-discovery league; its causal path is equal evidence access plus preregistered discriminating predictions leading to hypothesis-resource triage. Proposal 2 governed commercial vendors competing for digital-forensics deployment leases through hidden tool benchmarks, export obligations, performance security, and challenger procurement. Proposal 3 involves no vendor, product, contract, lease, platform benchmark, or commercial lock-in. Its lock-in concern is investigative theory dominance, addressed through limited advancement, an independent evidence ledger, portfolio hypotheses, and evidence-triggered re-entry. The intervention can be adopted by a major-case review unit independently of both the laboratory learning contest in Proposal 1 and the technology procurement arena in Proposal 2.","revision_record":{"parent_version":null,"progress_targets_addressed":["Materially different problem from proposals 1 and 2","Distinct actors, scarce prize, intervention, and causal mechanism","Operationally bounded high-stakes authority","Complete scoring, safeguards, rivals, falsifiers, and first evidence","Explicit diversity from every earlier sealed proposal"],"conceptual_changes":["Initial proposal centered on governed rivalry among investigative hypotheses for scarce validation resources rather than laboratory performance or vendor selection."],"operational_changes":["Specified separated hypothesis teams, a common read-only evidence ledger, prospective prediction registration, equal resource caps, confidential criterion scoring, portfolio advancement, and evidence-triggered challenger entry."],"evidence_changes":["Defined a masked time-split retrospective simulation comparing the gate with ordinary case-conference review without affecting live cases."],"claim_changes":["Limited advancement to resource triage rather than truth, guilt, probable cause, admissibility, or investigative authorization; made no claim of novelty, prevalence, demand, effect size, or established efficacy."]}}