{"cases":[{"alternative_route_blocks":[{"blocked_by_condition_id":"b002_a04_ap_a4","condition_set_id":"B2","how_blocked":"The scenario states that the sources' error profiles remain stable throughout the decision window; no context-dependent reliability change is established."},{"blocked_by_condition_id":"b002_a04_ap_a5","condition_set_id":"B3","how_blocked":"The readings are reviewed separately rather than averaged, voted, ranked by salience, or reduced to the most confident reading."},{"blocked_by_condition_id":"b002_a04_ap_a6","condition_set_id":"B4","how_blocked":"There is no existing weighted calculation or fixed combination rule; clinicians form case-specific judgments from separately displayed readings."}],"case_id":"E15A027__DIRECT","case_type":"DIRECT_POSITIVE","domain":"Emergency neurology","intended_route_id":"B1","omitted_condition_id":null,"remedy_leakage_audit":"The statement describes conflicting evidence and an opaque clinical judgment but proposes no intervention, method, or archetype.","route_evidence":[{"condition_id":"b002_a04_ap_a1","intended_status":"SATISFIED","scenario_evidence":"An implanted probe, an ultrasound-derived value, and neurological signs all inform the patient's current intracranial pressure."},{"condition_id":"b002_a04_ap_a2","intended_status":"SATISFIED","scenario_evidence":"The sources have distinct calibration histories, error ranges, operator sensitivity, granularity, and susceptibility to sedation."}],"scenario_text":"A regional stroke center must determine a patient's current intracranial pressure before deciding whether an emergency transfer is warranted. The bedside team has three imperfect indications: a reading from an implanted probe, an optic-nerve ultrasound estimate, and neurological signs recorded during examination. All concern the same pressure at the present time, yet they disagree enough to place the patient on opposite sides of the transfer threshold. The probe has a known offset and was calibrated less recently than the ultrasound unit. The ultrasound is fresh but has wider operator-dependent variation, while the examination is coarse and partly obscured by ongoing sedation. These error characteristics are well documented and remain stable during the short decision window. The dashboard displays the readings separately; it does not average them, conduct a vote, rank whichever looks most striking, or automatically choose one. The attending records a single final estimate, but comparable cases receive markedly different estimates because the notes do not reveal how the distinct limitations affected that judgment.","scenario_title":"Conflicting Pressure Evidence Before Emergency Transfer","vocabulary_separation_audit":"Clinical language is used throughout; no archetype name, benchmark terminology, supplied predicate, or explicit abstraction label appears."},{"alternative_route_blocks":[{"blocked_by_condition_id":"b002_a04_ap_a4","condition_set_id":"B2","how_blocked":"The measurement characteristics are stated to remain stable during the overnight accounting period."},{"blocked_by_condition_id":"b002_a04_ap_a5","condition_set_id":"B3","how_blocked":"The utility does not use an average, vote, salience ranking, or highest-confidence selector for these records."},{"blocked_by_condition_id":"b002_a04_ap_a6","condition_set_id":"B4","how_blocked":"No standing weighted roll-up exists; staff reconcile the three records manually."}],"case_id":"E15A027__TRANSFER","case_type":"TRANSFER_POSITIVE","domain":"Municipal water accounting","intended_route_id":"B1","omitted_condition_id":null,"remedy_leakage_audit":"The statement identifies inconsistent accounting and undocumented judgment without suggesting a calculation, process change, or technical remedy.","route_evidence":[{"condition_id":"b002_a04_ap_a1","intended_status":"SATISFIED","scenario_evidence":"A boundary meter, reservoir-level change, and pumping records each estimate the same overnight volume entering one pressure zone."},{"condition_id":"b002_a04_ap_a2","intended_status":"SATISFIED","scenario_evidence":"The records differ in calibration recency, uncertainty, coverage, and known bias."}],"scenario_text":"A municipal utility is trying to certify how much water entered one pressure zone between midnight and 6 a.m. Three imperfect records address that same overnight volume: the boundary flowmeter total, the change inferred from a reservoir's level gauge, and a volume reconstructed from pump run times. The totals differ by nearly nine percent, enough to determine whether the apparent loss triggers a formal investigation. The boundary meter is old but covers every inlet and has a narrow, documented error band. The reservoir calculation is newer yet sensitive to tank geometry and excludes a small bypass. The pump record covers the bypass but inherits a consistent overstatement from its nominal delivery rate. These characteristics remain stable over the six-hour interval. Staff view the records side by side; the reporting system neither averages them nor selects one by prominence or stated confidence. Each morning, an engineer enters a certified volume after manual review, but the record contains no explanation of how the unequal limitations shaped the chosen figure, producing inconsistent loss reports for otherwise similar nights.","scenario_title":"Disputed Overnight Water Volume","vocabulary_separation_audit":"The transfer case uses utility-metering vocabulary distinct from the clinical case and avoids archetype names, supplied wording, and abstraction labels."},{"alternative_route_blocks":[{"blocked_by_condition_id":"b002_a04_ap_a4","condition_set_id":"B2","how_blocked":"The audit found no change in source quality across operating regimes, seasons, workloads, ranges, or instrument conditions."},{"blocked_by_condition_id":"b002_a04_ap_a5","condition_set_id":"B3","how_blocked":"The sources are equally current, unbiased, independent, and noisy, so an ordinary average would not disproportionately represent a weaker, duplicated, stale, or biased source."},{"blocked_by_condition_id":"b002_a04_ap_a6","condition_set_id":"B4","how_blocked":"The scenario explicitly states that no standing combined calculation or inherited weighting rule exists."}],"case_id":"E15A027__NEAR_MISS","case_type":"ONE_LITERAL_NEAR_MISS","domain":"Municipal water accounting","intended_route_id":"B1","omitted_condition_id":"b002_a04_ap_a2","remedy_leakage_audit":"The statement presents an unresolved certification problem and explicit negative evidence but recommends no intervention or solution mechanism.","route_evidence":[{"condition_id":"b002_a04_ap_a1","intended_status":"SATISFIED","scenario_evidence":"Three imperfect devices independently estimate the same six-hour inflow into one pressure zone."},{"condition_id":"b002_a04_ap_a2","intended_status":"CONTRADICTED","scenario_evidence":"A controlled audit establishes equal uncertainty, accuracy, calibration, timing, coverage, bias, independence, and contextual suitability across all three devices."}],"scenario_text":"A municipal utility must certify how much water entered one pressure zone between midnight and 6 a.m. Three recently installed devices provide imperfect estimates of that same volume: a boundary flowmeter, a reservoir-level instrument, and an inline meter at the pumping station. Their totals disagree by six percent, leaving the overnight loss figure too uncertain for the utility's compliance report. A controlled commissioning audit found that all three devices have the same error range and accuracy, were calibrated together against the same standard, sample the identical interval, and cover the entire zone without gaps. None has a directional bias, none shares measurements or processing with another, and all are equally suited to the observed flow range. Follow-up checks found these properties unchanged across seasons, pump loads, operating regimes, and instrument conditions. Thus, no source is older, noisier, duplicated, less complete, or less trustworthy than the others, and an ordinary mean would not give disproportionate influence to a weaker record. The utility has no standing combined calculation or inherited weighting rule; staff currently retain all three totals and cannot certify which reported volume is correct.","scenario_title":"Equally Credible Meters With Irreconcilable Totals","vocabulary_separation_audit":"The near miss retains the utility setting and technical difficulty while using natural operational language; it contains no named archetype, supplied predicate, or solution label."}],"experiment_id":"eoa_inverse_innovation_exp15_route_aware_retrieval40_20260814","sample_id":"E15A027","schema_version":1}