Methodology
TACIS™ Scoring Methodology
Four-axis scoring architecture designed for volatility suppression, arbitration defensibility, and institutional-grade transaction intelligence. Every design decision serves transparency and decomposability.
Design Rationale
Why Linear Scoring Fails
Conventional vehicle inspection scoring uses linear aggregation — summing individual checkpoint results into a single number. This introduces systemic flaws that compromise transactional reliability.
Axis Masking
A vehicle with excellent cosmetic condition can mask critical structural deficiencies in a linear score. The composite number obscures the distribution of condition across axes.
Outlier Sensitivity
Single catastrophic findings disproportionately distort linear composites, producing scores that do not reflect the vehicle's actual condition distribution.
Arbitration Vulnerability
Linear scores are difficult to defend in dispute resolution because they cannot demonstrate how individual axes contribute to the composite assessment.
Context Blindness
Linear scoring treats all checkpoints equally regardless of transactional significance. A paint scratch and a frame misalignment carry identical weight.
Core Architecture
Multi-Axis Assessment Architecture
TACIS decomposes condition assessment into four weighted axes, each producing an independent score before composite aggregation. This preserves axis-level transparency while enabling structured composite scoring.
Structural Integrity
35%Frame, welds, subframe, crumple zones, corrosion mapping, crash repair detection, paint depth measurements. Highest weighting reflects the irreversibility of structural damage and its safety implications.
Mechanical Systems
30%Powertrain, brakes, suspension, steering, electrical, HVAC, exhaust, cooling, and fuel systems. Second-highest weighting reflects safety significance and functional impact on vehicle operation.
Cosmetic Condition
20%Paint quality, panel condition, glass, interior surfaces, wheels, trim, headliner, cargo area. Lower weighting reflects reversibility of cosmetic findings and their lower transactional risk profile.
Documentation & Provenance
15%Title status, liens, service history completeness, maintenance patterns, warranty claims, accident records, ownership chain integrity. Weighted for institutional audit and provenance verification requirements.
Composite Formula
TACIS = (SI × 0.35 + MS × 0.30 + CC × 0.20 + DP × 0.15) − CVM × ANC
Subject to Red Flag Cap enforcement
Volatility Architecture
Volatility Clustering & Suppression
When multiple findings cluster within a single axis, TACIS applies suppression logic to prevent cascading score degradation. This operates on three principles.
Cluster Detection
When findings within a single axis exceed the calibrated density threshold, the scoring engine identifies the cluster and flags it for suppression analysis. This prevents related findings from compounding beyond their actual condition impact.
Dampening Coefficient
Clustered findings receive a dampening coefficient that reduces their marginal impact on the axis score. This prevents double-counting of related condition indicators — a cracked bumper and a bumper misalignment are related findings, not independent events.
Cross-Axis Stabilization
The composite score incorporates cross-axis stabilization to ensure that suppressed findings on one axis do not create artificial inflation on others. The total condition picture remains balanced and representative.
Volatility Measure
Condition Volatility Measure (CVM)
CVM quantifies the degree of condition asymmetry across axes. A vehicle with uniform condition across SI, MS, CC, and DP receives minimal penalty. A vehicle with severe axis divergence receives proportional penalty.
CVM Formula
CVM = avg(|SI-MS|, |SI-CC|, |SI-DP|, |MS-CC|, |MS-DP|, |CC-DP|)
Penalty = CVM × 0.08. The penalty is subtracted from the weighted base score before age normalization.
Low CVM — Consistent Condition
SI: 84 · MS: 82 · CC: 80 · DP: 79
CVM: 2.3 → Penalty: 0.2 pts
Axes are balanced. The score accurately represents overall condition.
High CVM — Severe Asymmetry
SI: 55 · MS: 92 · CC: 88 · DP: 70
CVM: 19.5 → Penalty: 1.6 pts
Structural compromise masked by strong mechanicals. CVM reveals the true risk.
Age Calibration
Age Normalization Coefficient (ANC)
ANC calibrates scoring expectations to vehicle age. A 2-year-old vehicle with minor wear patterns is assessed differently than a 15-year-old vehicle with the same findings. Without age normalization, newer vehicles would be penalized for expected aging patterns.
Vehicle Age > 10 years
Modest tolerance for age-appropriate wear
Vehicle Age 5–10 years
Slight adjustment for mid-life condition patterns
Vehicle Age < 5 years
Full scoring standard — recent vehicles held to baseline
Safety Override
Red Flag Cap
The Red Flag Cap is a non-negotiable safety enforcement mechanism. When triggered, it overrides the computed score to prevent mathematically favorable composites from concealing critical safety findings.
IF red_flags ≥ 3 AND computed_score > 69 THEN final_score = 69
A “red flag” is defined as any finding classified at severity level 5 — the highest severity tier, reserved for conditions that represent immediate safety concern, structural failure risk, or material misrepresentation.
When three or more severity-5 findings are present, the vehicle cannot score above 69 regardless of how strong the remaining axes are. This ensures a vehicle with critical frame damage, active safety system failure, and undisclosed flood damage cannot achieve a VERIFIED WATCH or higher band.
The cap is applied after all other computations (base score, CVM penalty, ANC adjustment) and cannot be overridden by any parameter or configuration. It is a constitutional rule of the scoring engine.
Institutional Framework
Arbitration Defensibility Framework
TACIS scoring is designed from the ground up for defensibility in dispute resolution. Every score can be decomposed into its constituent elements, traced to specific evidence, and explained without proprietary black-box logic.
Score Decomposition
Any composite score can be broken down to show individual axis contributions (SI, MS, CC, DP), the CVM penalty, ANC adjustment, and the specific checkpoint-level findings supporting each axis.
Evidence Traceability
Every finding links to timestamped, SHA-256 hashed photographic evidence. The chain of custody from evidence capture to score output is fully documented and independently auditable.
Methodology Transparency
Axis weightings, suppression parameters, CVM formula, ANC factors, and Red Flag Cap rules are published. No proprietary black-box algorithms. Defensibility requires transparency.
Version Control
Every assessment references a specific methodology version. Changes to scoring parameters are documented and versioned. Prior assessments are never retroactively modified.
Inspection Protocol
147-Point Condition Verification Protocol
The CVP provides the evidence foundation for TACIS scoring. Each of the 147 checkpoints maps to one of the four scoring axes and is graded on a 0–5 severity scale.
Structural Integrity
~45 checkpoints
Frame integrity, unibody, subframe, structural welds, crash repair detection, panel alignment, corrosion mapping, structural paint depth.
Mechanical Systems
~42 checkpoints
Engine, transmission, drivetrain, exhaust, cooling, fuel, electrical, HVAC, brakes, steering, suspension — component-level severity classification.
Cosmetic Condition
~35 checkpoints
Exterior paint, panel condition, glass, trim, wheels/tires, interior surfaces, seats, headliner, cargo area, odor assessment.
Documentation & Provenance
~25 checkpoints
VIN verification, title status, lien records, registration consistency, service history, maintenance patterns, ownership chain.
Total: 147 inspection checkpoints across 4 scoring axes. Severity scale: 0 (no issue) through 5 (critical / safety).
Confidence Model
Confidence Grade™ Methodology
The Confidence Grade measures how much of the 147-point protocol was fully executed and evidenced. It does not measure score quality — it measures evidence depth. A perfect TACIS score with LIMITED confidence means the score is based on incomplete evidence.
HIGH Confidence
≥85% evidence coverageFull protocol completion with comprehensive photographic evidence across all four axes. Minimal access constraints. Score carries institutional-grade reliability for all transaction types including lending, insurance, and arbitration.
MODERATE Confidence
60–84% evidence coveragePartial protocol completion with documented access constraints. Score is directionally reliable but may not capture the full condition spectrum. Inaccessible areas are noted in the Known Unknowns section. Suitable for preliminary transaction decisioning.
LIMITED Confidence
<60% evidence coverageSignificant access limitations prevented comprehensive verification. Score should be treated as preliminary and directional only. Additional assessment recommended before institutional decisions. Known Unknowns section expanded to document all evidence gaps.
Band System
Comparable Condition Bands
The five-tier classification system maps TACIS scores to standardized bands for at-a-glance transaction decisioning. Bands are fixed — no manual overrides.
90–100
VERIFIED PRIME
Exceptional condition. Minimal findings. Ready for premium transaction terms.
80–89
VERIFIED SOLID
Strong condition with minor findings. Standard transaction terms appropriate.
70–79
VERIFIED WATCH
Moderate findings present. Review specific flags before proceeding.
60–69
VERIFIED RISK
Significant findings. Discounted terms or additional verification recommended.
<60
VERIFIED CRITICAL
Critical findings. Exercise extreme caution. Full review required.
Governance
Methodology Version Control
Every VINDEX assessment references a specific methodology version. When scoring parameters, axis weights, protocol requirements, or enforcement rules change, the methodology version is incremented. Prior assessments are never retroactively modified.
Every published report includes a methodology version identifier in its metadata.
Version changes are documented with specific parameter deltas (e.g., “SI weight changed from 35% to 36%”).
Scores from different methodology versions are not directly compared without version context.
Lifecycle intelligence accounts for methodology version transitions in longitudinal tracking.
Transparent methodology. Defensible scores.
Every scoring decision is documented, decomposable, and version-controlled.