The WROITER method
A number alone tells you nothing. The method behind the diagnostic exists so you can see what the tool reacted to, decide whether that reaction is meaningful, and avoid the mistake of treating a score like a verdict.
The sequence
- Paste a sample into the diagnostic. At least 50 words; 150+ is better.
- Read the flagged patterns and rhythm metrics before you look at the number.
- Check the limitation notes for anything that might explain the result without invoking AI.
- Then decide: revise, investigate, or move on.
What the score tells you — and what it does not
A higher score means stronger overlap with AI-typical surface patterns. It is a prompt for closer review. It is not proof of authorship, cheating, or intent. For the full interpretation boundary, see How It Works.
Read by question
- How It Works — score construction, signal families, and the interpretation boundary.
- Pattern Library — what each flagged pattern looks like, with examples and rewrite guidance.
- Limitations — false positives, false negatives, genre traps, and safe review policy.
- Changelog — public record of method and interpretation changes.
Evidence before enforcement
Do not turn a score into a decision without reading the evidence first. Do AI Detectors Work? covers the broader reliability case. The False Positive Hall of Fame shows that false positives are not a theoretical edge case — they happen to real text written by real people.
Why diagnostic-first
WROITER does not publish bypass tactics or "undetectable writing" positioning. The philosophy is blunter than that: show what the system is reacting to, help writers revise toward clarity, and keep reviewers from confusing probability with proof. If you want the practical workflow for checking a specific piece of text, start with How to Check AI Writing.