{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp06_four_proposal_generalization60_20260803","source_assessment_id":"bounded_rivalry_governance__accounting_auditing:P3:v0","cell_id":"bounded_rivalry_governance__accounting_auditing","proposal_index":3,"qualification":"EMPIRICAL_PARTNER_CANDIDATE","criteria":{"specific_differentiated_claim":{"status":"YES","reason":"The falsifiable claim is that a preregistered, identity-blinded holdout tournament can outperform incumbent controller selection and a non-blinded expert panel on stability, usage linkage, reproducibility, and sponsor-benefit association."},"credible_problem_signal":{"status":"YES","reason":"Independent consulting evidence, primary research, and managerial-costing guidance support recurrent allocation disputes, opacity, distorted incentives, and consequential decision effects, while appropriately leaving the proposal's narrower strategic-rivalry premise for empirical testing."},"identifiable_partner_or_adopter":{"status":"YES","reason":"A large organization with a corporate controller, governed shared-service allocation data, and an independent audit or model-validation function is a concrete partner class; HHS demonstrates an identifiable institutional structure with cost-modeling authority, analysts, data repositories, and review cycles."},"partner_access_is_necessary":{"status":"YES","reason":"The decisive premises require proprietary candidate models, transaction and usage data, sponsor identities and exposure, selection records, staff effort, and organization-specific authority boundaries that public research cannot supply."},"safe_authorized_first_step":{"status":"YES","reason":"A controller-authorized, closed-year, nonposting shadow analysis using existing governed data is reversible and expressly excludes ledger, budget, compensation, tax, transfer-pricing, personnel, and external-reporting use, with clear halt and rollback conditions."},"bounded_decisive_empirical_design":{"status":"YES","reason":"The study is limited to one reconciled pool, one closed year, a frozen withheld quarter, and at most three alternatives, and directly compares the incumbent basis, a non-blinded independent-panel choice, and the blinded tournament using preregistered measures and stopping rules."},"no_material_negative_gate":{"status":"YES","reason":"All substantive pipeline gates are YES. Cost scope is uncertain, but this is not the sole uncertainty and actual partner-specific labor, tooling, validation, and governance costs can be measured within the bounded test."},"not_merely_more_research":{"status":"YES","reason":"The next step specifies the partner roles, governed records, three comparators, frozen protocol, outcome measures, minimum advancement rule, prohibited uses, and explicit falsification and stopping conditions."}},"uncertainty_types":["PROBLEM_PREVALENCE","ADOPTER_PULL","INCREMENTAL_EFFECT","WORKFLOW_FIT","DATA_ACCESS","COST_SCOPE"],"partner_profile":"A large enterprise or public institution with one material shared-service cost pool, competing historical allocation proposals, a controller authorized to approve a nonposting retrospective study, governed ledger and operational-driver data, and an independent internal-audit or model-validation team.","required_access":"Controller authorization; one closed year's reconciled pool, ledger, usage-driver, lineage, exclusion, and model records; incumbent and up to three alternative models; sponsor and beneficiary mappings available to independent validators; selection-process records; authorized validation staff; and tracked accounting, data, tooling, and governance labor.","bounded_empirical_test":"For one closed year and one shared-service pool, preregister eligibility, holdout quarter, scoring weights, stress scenarios, minimum meaningful differences, and stopping rules. Using the same candidate models, compare the incumbent controller-selected basis, a conventional non-blinded independent accounting-panel selection, and the identity-blinded holdout tournament. Measure reconciliation, independent reproducibility, out-of-period stability, causal usage linkage, discretionary-assumption sensitivity, sponsor-savings correlation, unexplained transfers, data and labor burden, and direct-tracing feasibility without posting or operationally using any result.","success_condition":"The tournament is independently reproducible, preserves reconciliation and auditability, and materially improves at least one preregistered decision-quality measure over both comparators without materially worsening the others, increasing sponsor-benefit association, or exceeding the capped resource envelope; direct tracing is not feasible at lower authorized burden.","falsification_condition":"Do not advance if common-pool reconciliation, authorized access, validator independence, sponsor blinding, or prospective-target validity fails; rankings are unstable across reasonable holdouts or stress tests; sponsor benefit remains the dominant rank predictor; no model clears auditability gates; the tournament fails to improve on both comparators; or direct tracing or pool decomposition is feasible at acceptable burden.","rationale":"The proposal occupies the narrow empirical-partner lane: broad problem evidence and adjacent technical practice are credible, but the differentiated mechanism's key premises and comparative effect are organization-specific and inaccessible through ordinary web research. A controller-authorized shadow tournament can resolve those uncertainties safely and decisively before any live accounting-policy change."}