How I score my essays

Every essay on this site carries a percentage score in its header. That score is the result of running my factual claims through an AI evaluation prompt designed to catch me when I am wrong, sloppy, or being unfair to my sources.

Why? Because I write about whatever catches my curiosity, and every subject has facts that can be checked. It is dangerously easy to sound authoritative while being completely wrong. I am not an academic. I am a curious person with strong opinions and access to a keyboard. That combination should come with a warning label.

Instead of a warning label, I built a scoring system.

What the score means

The score is a factual accuracy assessment, not a quality rating. A beautifully written essay with a wrong date still gets marked down. A dry, boring paragraph with perfect citations still scores well.

85-100% Claims are well-sourced and historically accurate. Minor interpretive disagreements may exist but facts hold up.

60-84% Mostly accurate but contains simplifications, contested claims, or areas where stronger sourcing is needed.

Below 60% Significant factual issues. I have probably let enthusiasm outrun evidence. Read with healthy scepticism.

The prompt

Transparency matters more than the score itself. Here is the exact prompt I use to evaluate each essay:

Evaluation Prompt

You are a rigorous, impartial fact-checker. You have been given an essay to evaluate for factual accuracy.

Your task:
1. Identify every factual claim in the essay (historical dates, attributed quotes, named events, cited sources, statistical claims, and causal assertions presented as fact).
2. For each claim, assess whether it is:
   - VERIFIED: Well-established and accurately stated
   - PARTIALLY ACCURATE: Contains a kernel of truth but is misleading, oversimplified, or missing critical context
   - UNVERIFIED: Cannot be confirmed or is a matter of significant scholarly debate presented as settled fact
   - INACCURATE: Demonstrably wrong
3. Ignore rhetorical style, literary quality, and the author's personal opinions or interpretations. Only evaluate statements presented as factual.
4. Provide a final accuracy score from 0-100 based on the ratio of verified claims to total claims, weighted by significance.

Output format:
- List each claim with its assessment and a brief explanation
- Provide the final numerical score with a one-paragraph justification

This prompt is currently run using an AI model from Anthropic. Scores may vary across models and are a snapshot, not a permanent verdict.

Why this matters

Whatever I am writing about, getting the facts wrong has consequences. If I misrepresent a thinker, someone who knows better catches it and dismisses everything else I have to say. If I get a technical detail wrong, the argument built on top of it crumbles. If I quote someone out of context, I undermine the very integrity I am trying to build.

This is my commitment to you: I will try to earn the attention you give me. When I fail, and I will, the score will tell you.

- Mayo