How the work earns confidence

Evidence Labels Should Describe the Work—Not Decorate It.

Off Leash Marketer is a solo publication by R. Garcia. This methodology explains what each evidence label means, what it does not mean, and what must change before a claim earns a stronger label.

Evidence vocabulary

Four labels. Four different boundaries.

A stronger-sounding sentence does not create stronger evidence. Each upgrade requires new work that another person could inspect.

field tested

Observed in a controlled or real implementation and described within that boundary.

docs reviewed

Supported by current primary documentation but not presented as first-hand testing.

analysis

A reasoned interpretation of available evidence, with assumptions and limitations named.

opinion

A disclosed point of view, not presented as an established fact.

Upgrade path

The claim moves only when the evidence moves.

Opinion → Analysis

Add the reasoning chain, plausible alternatives, important assumptions, and a check that could overturn the conclusion.

Analysis → Docs-reviewed

Check current primary documentation, cite it near the supported claim, and separate documented behavior from interpretation.

Analysis or docs-reviewed → Field-tested

Run a controlled test and record the configuration, procedure, expected result, observed result, boundary, and limitations.

Any label → Lower label

Downgrade when documentation changes, a result cannot be reproduced, or the article claims more than its evidence supports.

Solo publication

One accountable human. No invented review board.

R. Garcia is the accountable author and publisher. The visible review date means the author deliberately revisited the article; it does not imply an independent peer review.

Theo is an editorial character and honest second opinion—not a human expert. Any future AI-assisted persona will be disclosed as a method or tool, never given fabricated experience or represented as an independent reviewer.

About R. Garcia
Required article record

What every article must leave visible.

  • Accountable author, publication date, author-review date, and evidence label
  • Symptom, likely break, and first check
  • Three article-specific checks before acting
  • A direct challenge to the primary conclusion
  • Important limitations, uncertainty, and evidence boundaries
  • Primary documentation or a reproducible test when the label requires it