# Temporal Validation Framework (TVI)  
## Public Method Note v1 — Conceptual + Testing Discipline

**Authors:** Carl Boon · BoonMind Labs  
**Status:** Live closed beta (invite) · Patent Pending  
**Date:** 2026-08-17  
**Site:** https://www.tvi-framework.com  
**Contact:** Info@BoonMind.io  

> **Not a full paper dump of the production engine.**  
> Coefficients, era tables, component weights, and server operators remain **proprietary**.  
> This note describes *what is measured*, *how claims are tested*, and *what residual tests currently support* — without enabling reverse-engineering of the score engine.

---

## 1. Abstract

Short-term popularity metrics (views, downloads, adoption spikes) systematically over-credit loudness and under-credit **staying power**. The Temporal Validation Framework (TVI) is a proprietary, multi-domain index family for modelling whether something behaves more like **infrastructure** or **attention noise** after hype decays.

Public residual tests use an **Axiom-style court** (pre-declared claim freeze → fixed case table → baselines → sealed results → red lights allowed). Engine coefficients are **not** published.

---

## 2. Motivation

Across media, methods, AI data, and firm narratives, dashboards optimise velocity and scale. Time still filters for density, reuse, and survival under stress. TVI formalises that filter as a scored model for comparative ranking **within a domain**, not as investment advice or a universal “truth score.”

---

## 3. Conceptual structure (public)

At a conceptual level (operators withheld):

\[
\text{Score} \propto f(\text{Saturation},\ \text{Temporal validation},\ \text{Structural resistance})
\]

| Concept | Public meaning | Not disclosed |
|---------|----------------|---------------|
| **Saturation** | How unavoidable something was *in its era* (relative, not absolute size) | Exact formulas, inflation factors, domain coefficients |
| **Temporal validation** | Whether time has re-endorsed the thing (persistence / resurfacing / legacy signals) | Exact weights, caps, event tables |
| **Structural resistance** | Era-adjusted difficulty of achievement | Full era schedules and numeric SRC tables |

**Zero-locking (conceptual):** weak saturation *or* weak temporal validation collapses the interpretive claim — volume alone is not enough.

**Domains (product names):** Culture/Media (TVI) · AI/Data (TDIS) · Business methods (TVI-B) · Investments (ISPS). Scores are **not** assumed comparable across domains without separate normalisation (see §6).

---

## 4. Testing discipline (Axiom-style methodology)

Borrowing process discipline from BoonMind Labs’ sealed-evaluation practice (**Axiom Workbench** methodology), *not* forcing TVI data through industrial actuator metrics:

1. **Claim freeze** written before scores  
2. **Fixed case table** with preferred winner declared in advance  
3. **Baselines:** volume-only, age-only, recency-only  
4. **Ablations:** saturation-only (equalised temporal inputs); sensitivity (−20% temporal factors)  
5. **Sealed artefacts** (local ledgers / JSON results)  
6. **Fail closed / red lights allowed** — including “not yet” product verdicts  

This is **court process**, not a claim that TVI is an Axiom industrial pack.

---

## 5. Residual results (high level, non-IP)

Two local residual batteries were run on the frozen server engine (`nowYear = 2026`). Detailed case inputs and internal components are **not** published.

### Round 1 — easy retrospective pairs
- Primary rank claim: **PASS** (preferred above rival)  
- **Age confound:** age-only also matched preferred often → residual not hard enough  

### Round 2 — same-era hard residual (primary test)
- **Same-era subset** (entities within ≤5 years): full-input rank of preferred **9/9**  
- **Volume baseline** on same subset: **8/9** → TVI **beats volume**  
- **Age baseline** on same subset: **~5/9** → age **does not** explain most wins  
- Saturation-only and mild sensitivity: ranks **stable** on this set  
- **Cross-domain raw scale:** still **incoherent** (culture-scale scores not comparable to invest/dataset raw magnitudes)

### Product legitimacy label (R2)
**YES_CONDITIONAL** — legitimate **within-domain** closed-beta rank aid with hard limits.  
**Not** a proven predictive market product. **Not** a universal cross-domain number.

---

## 6. Limits (required)

- Retrospective, not prospective prediction  
- Temporal inputs may include structured human judgment — not pure automatic scrape  
- No investment advice; ISPS is structural lens only  
- No publication of production coefficients or full case tables here  
- Public marketing must not over-claim residual results  

---

## 7. Access

| Layer | Status |
|-------|--------|
| Public site + this note | Free |
| Calculator | **Live closed beta** — [request access](https://www.tvi-framework.com/request-access) |
| Commercial use | License required · Info@BoonMind.io |

Built by **BoonMind Labs**. Tested under **Axiom-style** freeze/baseline/seal discipline. Engine proprietary · Patent Pending.

---

## 8. Citation

Boon, C. (2026). *Temporal Validation Framework (TVI): Public Method Note v1*. BoonMind Labs.  
https://www.tvi-framework.com/evidence · https://www.tvi-framework.com/papers/TVI_Public_Method_Note_v1.md

---

*End of public method note. Production operators remain confidential.*