This site combines measured market data, research summaries and AI interpretation. The research program has explored many Bitcoin strategies; most historical candidates were not promoted. September 12, 2026 methodology update: later audits found implementation defects in timing, selection, accounting and evaluation. Registration, test counts and historical verdicts are not blanket certifications. See the dated correction to the ML article.
The intended protocol records hypotheses, thresholds, costs and success criteria before evaluation. Defects and post-lock changes must be disclosed; previous attempts remain part of the research history. A written protocol is not proof that every historical experiment followed it correctly or that its implementation matched the stated method.
Relevant checks include inversion tests, random baselines, parameter perturbation and stability checks. They must themselves be correctly implemented. The old pipeline recorded 49 NO-GO outcomes; that is a historical result, not proof that the tested feature universe contains no tradable information. A completed revalidation of the full pipeline is not claimed here.
Research ledgers record attempted hypotheses so failed ideas and selection history are not silently forgotten. Multiplicity adjustments depend on the actual test family and valid input statistics. A large cumulative count alone does not establish program-wide false-discovery control.
Prospective observations and eligible published calls are timestamped and evaluated under stated rules. Missing observations, methodology changes and limited samples must remain visible. The BTC call scoreboard scores published levels at T+1; it is not an account execution or realized-profit record.
The daily brief is generated by Claude (Anthropic) reading the pipeline's metrics, regime models, and derivatives data each day. It is labeled as AI analysis because that's what it is. The AI can be wrong, and its commentary is not itself a validated trading signal. Correct source data does not make historical research claims valid; unresolved results must be withheld from live evidence.
Public: the daily brief, regime states, metric dashboards, the Alt-Euphoria Gauge, published-call track records, and research write-ups including nulls. Not public: proprietary model internals and execution configurations — including whether and how any published call was traded. The boundary is deliberate and disclosed.
