Can Performance Crises in Football Really Be Predicted Mathematically? Formal Claim Auditing, Epistemic Agency, and Operationalized Self-Research in High-Performance Sport

Scientific publications may contain analytically distinct layers: what was measured or calculated, how constructs were operationalized, what relationships were examined, and what additional interpretive, predictive, explanatory, or practical claims were derived from these results. These layers do not necessarily have the same empirical basis or inferential reach. This creates a methodological problem whenever conclusions extend beyond what was directly measured, tested, derived, or validated within the specified design and scope. This methodological paper proposes Formal Claim Auditing (FCA) as a structured approach for reconstructing and examining individual scientific claims. FCA treats the claim, rather than the study as a whole, as the primary unit of analysis. The proposed architecture distinguishes source statements, claim reconstruction, measurement and operationalization, premises, provenance or derivation, inferential transitions, scope, level of analysis, measured and claimed endpoints, and explicit non-inferences. Propositional and predicate-logical reconstruction is used only where the relevant formal requirements are specified; other symbolic representations are distinguished as structured inferential, relational, or process notation. The approach is demonstrated through a bounded analysis of published mathematical indices for performance crises in professional football. The analysis separates the mathematical quantities produced by the reported procedures from additional claims concerning performance deterioration, crisis classification, prospective prediction, psychological interpretation, and intervention. Each transition is examined according to the information, premises, validation requirements, level of analysis, and scope required to support it. FCA is presented as a methodological proposal, not as a validated system for determining truth, study quality, decision correctness, or human judgment capability. Its immediate methodological target is representational explicitness: making the measurement basis, operationalization, premises, provenance or derivation, inferential transitions, and scope of a claim explicitly reconstructable. The same architecture may also serve as an instruction framework for AI-assisted claim auditing, but AI-generated reconstruction or audit output is not treated as inherently correct or validated. Whether FCA or AI-assisted FCA improves human inspection, judgment, learning, or practical decision making remains an empirical question. The broader normative position of the paper concerns epistemic agency. FCA addresses the examination of externally produced scientific claims, while Operationalized Self-Research addresses the development of person-specific longitudinal knowledge. Their proposed complementarity reflects a common objective: to strengthen the conditions under which individuals can participate more actively in examining what is claimed about them and what can be learned from their own history. The intended endpoint is greater epistemic participation, not the replacement of scientific or professional expertise.