Evidence is the interface.

A rolling 24-month window ending on the current analysis date. Prepared datasets are recalculated during publication, scheduled analysis, or a manual editorial refresh — never on each visitor request.

Signals are normalized into the existing topic taxonomy and related infrastructure dimensions: identity, compliance, PQC, PKI, secure hardware, root of trust, AI / automation, DLT, verifiable credentials, institutional adoption, machine identity, IoT, autonomous transactions, auditability, and tokenization infrastructure.

Direction is derived from published Signal frequency, recency, six-month acceleration, primary-evidence coverage, deployment maturity, cross-stack connectivity, persistence, and convergence. ACCELERATING requires an increase over time, not merely several supporting records.

Confidence describes the evidence supporting an editorial trajectory hypothesis. High confidence requires at least six independent developments, four primary sources, four active months, strong primary-source coverage, and more than one organization. A single announcement can never produce high confidence.

The model records contradicting Signals when they exist and states what evidence remains missing. The absence of contradictory evidence is not treated as confirmation. Hypotheses can weaken or remain insufficient as new evidence arrives.

Automated analysis creates review candidates only. Hypotheses never auto-publish, and material changes to a published conclusion create a separate update for editorial approval. Earlier snapshots remain preserved.

STACK TRAJECTORY measures documented infrastructure development, not commercial success. It does not use price, market capitalization, trading volume, liquidity, expected returns, investor sentiment, or price targets.