Quality Assurance · The Audit Flow

Eight stages, and readiness is the result.

Stages

Inspection results

Readiness

COMPUTED

How MARS uses this

The Weekly Scorecard converts branch hit probabilities into EV and tags each week by threshold. The rolling line is what MARS actually trusts: single-week readings are treated as noise until the rolling window confirms direction. RED weeks trigger doctrine - reduced aggression and review - rather than negotiation.

How it benefits you

You learn to distinguish a bad week from a broken system. Variance stops triggering rule changes, and genuine expectancy decay gets caught while it is still one line on a chart instead of a hole in the account. The GREEN/YELLOW/RED language also makes weekly review fast and unambiguous.

Weekly EV tagged GREEN / YELLOW / RED against expectancy thresholds, with the rolling-EV line separating persistent edge from one lucky week.

Six results, not two

The inspection vocabulary refuses a binary.

Pass and fail are only two of six available results. Warning covers behaviour that is not strictly wrong but is risky, ambiguous, or visually misleading. Needs review covers the case where the auditor genuinely cannot determine pass or fail yet, usually because source confirmation is missing or maturity is ambiguous. Blocked covers a check that cannot complete because a dependency is broken. Not tested is honest about coverage. Collapsing these into pass or fail would force every uncertain observation into whichever bucket the auditor preferred that day.

The register is evidence

The original failure record is never edited.

When a defect is repaired, the pass proof goes into the retest log, and only the lifecycle fields on the original error row are updated: the latest retest reference, the retest status, the error status, the blocking flag, and the fix applied. The expected result, actual result, severity, scenario, source tab, and failure class stay exactly as recorded. Overwriting them would erase the only account of what actually went wrong, which is the thing the register exists to preserve.

Same defect still broken: same error identifier, a new retest record.
A genuinely distinct defect found during a retest: a new error identifier.
A regression is flagged as such, and counts as a blocker if it blocks advancement.

False blockers

Placeholder rows must not block advancement.

Readiness counts open blocking errors only where an error identifier is actually populated. Severity values, blocking flags, and status text sitting in empty template rows are not defects, and a counter that includes them will report a workbook as permanently unready for reasons nobody can locate. This is a real defect class in quality systems: the audit tool developing its own false positives and gradually training its operator to override it.

Approval scope

Not every valid scenario can certify.

Scenario approval is graded rather than binary. A scenario can be approved for formal certification, for limited or smoke-test scope only, for regression use after a known defect, or for broad workbook-level testing. Edge cases exist to stress unusual conditions and cannot certify unless explicitly scoped as edge-case work. Drafts and retired scenarios cannot certify at all. This is what stops an interesting test from quietly becoming a certificate.

Reference

The quality spine

Stage	Function	Output
1. Eligibility	Decide whether a timeframe and scenario combination is valid at all	Validity, reason, expected state, interpretation depth, approval status
2. Scenario registry	Store the reusable scenario records used downstream	A stable scenario identifier and its family
3. Contract library	Define what should happen for each scenario and output block	Expected behaviour, expected value, tolerance, source of truth
4. Run logging	Record the formal pass against a specific workbook version	A run identifier, date, version, scenario, and overall status
5. Inspection	Compare expected behaviour to actual behaviour by output family	Pass, fail, warning, needs review, blocked, or not tested
6. Failure governance	Store the failures discovered during inspection	Error identifier, severity, blocking flag, fix status
7. Repair proof	Document the rerun after a fix	Retest identifier, actual result, pass or fail, regression flag
8. Command summary	Show current readiness and advancement state	Ready or not ready, open blockers, next action

Doctrine

Connected inside MARS

This module doesn't work alone.

Connected — 01

Certified Validation QA (full module)

Open system page →

Connected — 02

The Five Principles

Open system page →

Connected — 03

System Architecture

Open system page →

Go deeper

Operator briefs on this territory.

Deep dive — 01

Eligibility → scenarios → runs → verdict: the eight-stage QA spine.

Eight stages in one direction. Governance first, comparison second, readiness last — and never rearranged.

Read the full brief →

Deep dive — 02

Never overwrite a failure: the error register is evidence, not a task list.

Fix the defect, never the record of it. Only lifecycle fields may move; the evidence pair is permanent.

Read the full brief →

Deep dive — 03

Test 0: a broken QA cockpit cannot certify anything.

The instrument gets inspected before the thing it inspects. Everything downstream inherits its faults.

Read the full brief →

Every module ships in the complete MARS package.

One price. Eleven workbooks, three TradingView indicators, and the full manual library — $497.

Get MARS Pro — $497 Explore the Platform