{"closest_prior_art":[{"name":"PEAX interactive clinical-data exploration system","overlap":"Integrates interactive subgroup/model exploration with statistical testing, exposes the cumulative number of tests for multiplicity adjustment, and proposes separating hypothesis generation from testing with a holdout set.","remaining_difference":"It does not automatically preserve every attempted analysis, apply a predeclared FDR rule, freeze a decision-specific claim, assign staged evidence statuses, or gate athlete-facing workload decisions.","source_ids":["SRC4"]},{"name":"Training-load questionable-research-practice guidance","overlap":"Identifies the domain-specific risk from numerous metrics, windows, cutoffs, outcomes, and models; warns that selective reporting, HARKing, multiple testing, and overfitting can mislead training decisions.","remaining_difference":"It diagnoses the problem and recommends caution but does not implement an auditable weekly claim-family registry, multiplicity screen, untouched confirmation partition, or action gate.","source_ids":["SRC2"]},{"name":"AthleteMonitoring Team Dashboard","overlap":"A first-party athlete-monitoring product combines customizable wellness, health, workload, time-window, and risk metrics and issues workload-reduction guidance, demonstrating the relevant decision environment.","remaining_difference":"The documented dashboard does not disclose an attempted-look inventory, search-family error control, discovery/confirmation status, or prospective confirmation requirement.","source_ids":["SRC3"]}],"contrastive_claim_falsifier":"The contrastive claim would be falsified by evidence that an existing athlete-monitoring workflow already captures the complete decision-linked analytic family, controls discovery error, freezes a selected claim, tests it on subsequently untouched sessions, and prevents non-urgent action until confirmation—or if the shadow pilot shows that these controls do not materially change evidential status or cannot prevent unlogged and contaminated analysis paths.","contrastive_claim_remaining":"For athlete-load decisions, neither an isolated significant result, coaching plausibility, ordinary provenance, nor model reproducibility is sufficient: decision readiness must reflect the full decision-linked search family and require a frozen prospective test on evidence untouched by discovery.","experiment_id":"eoa_inverse_innovation_exp13_second_slot_policy60_20260806","gates":{"adequate_source_search":{"rationale":"The bounded search covered the exact title and intervention, p-hacking/HARKing and data-dredging terminology, athlete-monitoring products and practices, and combinations of interactive provenance, multiplicity control, and holdout testing. Four opened sources span four publishers and include primary research and a first-party product source.","source_ids":["SRC1","SRC2","SRC3","SRC4"],"status":"PASS"},"bounded_next_test":{"rationale":"A four-cycle, one-squad shadow pilot can measure log capture, family-definition stability, holdout contamination, status changes, and reviewer reconstruction without requiring athlete-facing intervention. It tests operational feasibility and mechanism integrity, not clinical benefit.","source_ids":["SRC2","SRC3","SRC4"],"status":"PASS"},"distinct_testable_claim":{"rationale":"The components are established individually and PEAX combines test counting with contemplated holdout validation, but the search found no direct implementation combining complete athlete-analysis capture, FDR-screened discovery status, frozen prospective confirmation, and a non-urgent workload action gate. The pilot supplies explicit falsifiers.","source_ids":["SRC2","SRC4"],"status":"PASS"},"no_obvious_safety_or_authority_stop":{"rationale":"The authorized test is shadow-only, preserves clinician control and established acute-safety responses, forbids athlete-facing changes, and includes privacy and logging-related halt conditions. This is consistent with the literature's caution against prematurely modifying training from uncertain load evidence.","source_ids":["SRC2"],"status":"PASS"},"supported_problem":{"rationale":"Training-load literature directly documents many choices of metric, cutoff, window, reference category, and outcome, with risks of selective reporting, multiple testing, overfitting, and false discovery. A commercial dashboard shows that customizable athlete metrics can feed immediate readiness/risk interpretations and workload advice. Direct evidence of the proposal's exact internal weekly memo behavior was not located, so support is partial rather than complete.","source_ids":["SRC1","SRC2","SRC3"],"status":"PASS"}},"prior_art_disposition":"ADJACENT_PRIOR_ART","problem_evidence":{"finding":"The underlying multiplicity and analytic-flexibility problem is visible in athlete-load research and is relevant to operational dashboards. Carey et al. found false-discovery rates of 16–21% for discretized models and at least one false discovery in 42 of 100 simulations when three methods were available. Impellizzeri et al. catalogued numerous selectable metrics, windows, cutoffs, and outcomes and warned of selective reporting, HARKing, multiple testing, and overfitting. However, the retained sources do not directly audit weekly internal team recommendations or prove that null dashboard looks are routinely omitted.","source_ids":["SRC1","SRC2","SRC3"],"status":"PARTLY_SUPPORTED"},"research_id":"eoa_inverse_innovation_exp13_light_screen_20260806","schema_version":1,"screen_id":"E13P107","screen_survival":true,"search_lanes":{"component_combination":{"no_result_note":"No opened source implemented the complete combination of automatic analytic-provenance capture, decision-family FDR control, frozen prospective confirmation, staged claim status, and an athlete-facing action gate. PEAX was the closest component combination.","queries":["interactive data exploration log every query multiple testing holdout system","visual analytics provenance multiple comparisons holdout exploratory analysis","sport science holdout multiple testing athlete monitoring"],"source_ids":["SRC4"]},"direct_problem_and_intervention":{"no_result_note":"The exact title produced no direct match; broader intervention phrases located the domain problem and adjacent systems but not the proposed end-to-end gate.","queries":["\"Athlete Load-Response Claim Gate\"","athlete load response \"claim registry\" false discovery rate holdout confirmation","athlete monitoring data dashboard multiple testing preregistration prospective validation training load recommendations"],"source_ids":["SRC2","SRC3","SRC4"]},"products_practices_and_standards":{"no_result_note":null,"queries":["athlete monitoring dashboard customizable workload wellness risk recommendations","athlete management system performance data alerts reports decision making","consensus athlete load monitoring individualized training recommendations"],"source_ids":["SRC3"]},"synonyms_and_historical_terms":{"no_result_note":null,"queries":["sports science HARKing p-hacking garden of forking paths training load","training load data dredging selective reporting multiple comparisons","false discovery rate training load athlete"],"source_ids":["SRC1","SRC2"]}},"sources":[{"claims_supported":["Training-load model specification and discretization choices can inflate false discoveries.","Using three alternative discretized methods produced at least one false discovery in 42 of 100 null simulations.","Out-of-sample testing is needed to avoid positively biased model evaluation."],"publisher":"American College of Sports Medicine","source_id":"SRC1","source_type":"PRIMARY_RESEARCH","title":"Modeling Training Loads and Injuries: The Dangers of Discretization","url":"https://david-carey.github.io/files/%5BMSSE%5D%20Carey%20-%20Modelling%20Training%20Loads%20and%20Injuries%20The%20Dangers%20of%20Discretization.pdf"},{"claims_supported":["Training-load research exposes analysts to many selectable measures, windows, cutoffs, computations, and outcomes.","Trying several windows is prone to multiple-testing bias and overfitting.","Selective reporting, p-hacking, and HARKing can mislead clinicians, making training modification from the available evidence premature."],"publisher":"Journal of Orthopaedic & Sports Physical Therapy","source_id":"SRC2","source_type":"SECONDARY_RESEARCH","title":"Training Load and Injury Part 2: Questionable Research Practices Hijack the Truth and Mislead Well-Intentioned Clinicians","url":"https://www.lukebornn.com/papers/impellizzeri_jospt_2020b.pdf"},{"claims_supported":["A commercial team dashboard combines athlete wellness, health, internal-load, ACWR, and week-to-week workload metrics.","Dashboard fields and calculation periods are customizable.","The product presents immediate risk/readiness interpretations and recommends reducing workload for high displayed risk."],"publisher":"AthleteMonitoring","source_id":"SRC3","source_type":"FIRST_PARTY_PRODUCT","title":"The Team Dashboard","url":"https://support.en.athletemonitoring.com/support/solutions/articles/13000039236-the-team-dashboard"},{"claims_supported":["Interactive high-dimensional exploration creates data-dredging and false-discovery concerns.","PEAX displays the cumulative number of statistical tests to support multiplicity adjustment.","The authors identify a training/test holdout split as a feasible way to generate hypotheses on one partition and test them on another."],"publisher":"Pacific Symposium on Biocomputing","source_id":"SRC4","source_type":"PRIMARY_RESEARCH","title":"PEAX: Interactive Visual Analysis and Exploration of Complex Clinical Phenotype and Gene Expression Association","url":"https://psb.stanford.edu/psb-online/proceedings/psb15/hinterberg.pdf"}],"world_novelty_boundary":"This coarse, bounded public-web search supports only an adjacent-prior-art disposition. It cannot establish world novelty, patentability, market size, expert acceptance, implementation feasibility beyond the proposed pilot, or realized athlete or organizational value."}