Live · closed beta

Evidence & testing

How TVI is tested without publishing the engine. Built by BoonMind Labs. Residual packs use an Axiom-style court: freeze the claim, fix the cases, score baselines, seal results, allow red lights.

Patent Pending · Experimental · Not investment advice · Coefficients proprietary

What we publish

Conceptual method (saturation · temporal validation · structural resistance)
Testing process (freeze → cases → baselines → seal)
High-level residual outcomes (pass / fail / conditional)

What we do not publish

Production coefficients, era tables, or full operators
Full private case tables with internal components
Anything that reverse-engineers the server engine

Public method note

Cite this page and the method note. The engine repository is private and is not a public receipt.

Read method note (v1)

Residual summary (non-IP)

Round	Focus	Outcome
R1	Easy retrospective pairs	Rank PASS; age confound (not hard enough)
R2	Same-era hard pairs (primary)	PASS — preferred rank 9/9; beats volume 8/9; age ~56% (confound broken)
Product label	Within-domain beta	YES_CONDITIONAL — not universal score, not investment oracle

Cross-domain raw score scale remains incoherent; public claims stay within-domain and experimental.

Limits

Retrospective rank tests, not prospective prediction packs
Structured judgment may enter temporal inputs — not pure auto-scrape
No SoftBank / trading “proof” narrative as sealed evidence
PROMOTION of strong marketing claims requires human review beyond residual smoke

Access

Calculator is live closed beta by invitation. No public self-serve checkout required for beta.

Request access Info@BoonMind.io Calculator (invite)