Case Study 1 — The Envelope: Theo's Belief Inventory, Graded

Theo Vasquez is an illustrative composite carried through this book. The inventory format, the confidence–accuracy pattern, and the sources he traces are drawn from a clinical exercise I have run many times; the person is constructed.

Setup

Seventeen months earlier, in Chapter 1, Theo wrote ten beliefs about food on a sheet of A4, added where each came from and how confident he was, folded it, and put it in a kitchen drawer.

He was not told what the exercise was for. He was told only to be honest, to write beliefs he acted on rather than beliefs he could defend, and not to look anything up.

That last instruction is the whole design. An inventory of researched positions measures your library. An inventory of unexamined beliefs measures your intake — and it's the second one that governs what you actually eat.

The sheet

Nine of ten never checked. The one exception — the detox — he "checked" by doing it, which is Chapter 3's bundled-intervention problem in the form of personal experience.

The grading

He graded them himself, in the room, with the book's verdict scale in front of him.

Result: 8 wrong or substantially overstated · 1 genuinely "it depends" · 1 correct doubt.

The correlation

This is the finding.

The only belief he was right about was the only one he doubted.

💡 Aha moment. ⚠️ This is not a fact about nutrition. It is a fact about how confidence is generated.

Theo's confidence didn't come from evidence — nine of the ten were never checked. It came from repetition and source-authority. The 5s were the beliefs he'd heard most often, from the most official-sounding places: school, his mother, the news, his GP, "everyone."

And repetition is exactly what a myth is good at. A claim survives in the culture because it gets repeated, so frequency of exposure is a measure of stickiness, not accuracy — and Chapter 1's pipeline explains why the two can be actively inversely related, since the claims that travel furthest are the ones that were simplified and over-extended at each handoff.

So "I've heard this everywhere" is not weak evidence. It is evidence about the wrong thing.

The source trace

Step 4 of the checkpoint asks for a pattern in the "where I got it" column. Theo's:

Two observations he made himself:

"The ambient ones are the worst, and they're the ones I'd never think to question, because there's nobody to question." ⚠️ A belief with no attributable source cannot be evaluated by evaluating its source — which removes the single most practical tool in Chapter 1 §1.5.

"And number 10 came from a doctor." The eggs-and-cholesterol advice was mainstream clinical guidance when he received it. It was later revised. ⚠️ He was not misinformed; he was informed, and then not updated.

That second category deserves its own name, because it isn't a myth at all. It's stale guidance — correct when issued, superseded since, and never retracted to the person holding it.

There is no mechanism in the entire health system for telling you that something you were told in 2009 is no longer thought to be true. Guidelines are updated. Patients are not.

What he did next

He didn't change his diet. That had already happened over seventeen months, and none of it happened because of this exercise.

What changed was smaller and more useful:

1. He unfollowed two accounts — not because they were wrong about everything, but because tracing his errors put both of them behind more than one.

2. He added a rule for himself: "If I've never heard the person say 'I don't know,' I stop reading them." ⚠️ Chapter 1 §1.5's test, adopted as a filter rather than an assessment.

3. He wrote a new sheet. Ten beliefs he holds now — several taken from this book, which he noted with visible irritation. Dated. Back in the drawer. Three-year reminder set.

4. And he told his daughter about number 10. Not the egg advice. The fact that guidance changes and nobody comes to tell you.

Analysis

1. Confidence and accuracy were inversely related across ten beliefs, written by one person, without anyone trying to trick him. ⚠️ This is the single most transferable finding in the chapter, and it generalizes far beyond food.

2. Confidence tracked repetition, not evidence. Nine of ten were never checked. The 5s were the most frequently heard.

3. Ambient beliefs are the most dangerous because they have no attributable source — which disables source-evaluation entirely. You cannot ask "what happens to their income if you decline?" of "everyone."

4. One belief wasn't a myth at all. ⚠️ Stale guidance is a distinct category, and the system has no retraction mechanism for individuals. Expect several of your own.

5. Most verdicts were 🟠, not ❌. Six of the eight errors were over-extensions of something real — consistent with §17.1's interlude, and the reason a "myth-busting" frame is slightly wrong for most of what's actually in people's heads.

6. And the exercise's value was diagnostic, not corrective. It didn't change what he eats. It changed how he evaluates a source — which is the only thing that will still be working in ten years.

Discussion Questions

1. Theo's confidence tracked repetition rather than evidence. Is there any environment in which repetition IS a reliable signal of accuracy? What has to be true of the information ecosystem for that to hold — and does any part of the nutrition landscape meet it?

2. "Ambient" beliefs — no attributable source — were 4/4 wrong. Why might source-less beliefs be systematically less reliable than sourced ones, even bad sources? Does this argue for or against Chapter 1's emphasis on evaluating sources?

3. ⚠️ Belief 10 came from a doctor and was mainstream guidance at the time. Whose responsibility is it to update a patient? Design a realistic mechanism. What does it cost, and why doesn't it exist?

4. Six of eight errors were 🟠 over-extensions rather than ❌ falsehoods. Does calling this exercise a "myth audit" misframe what it finds? Propose a better name and defend it.

5. Theo's response was to unfollow two accounts and adopt a filter — not to change his diet. Is that the right lesson to draw, or is it a way of feeling improved without changing anything?

6. He set a three-year reminder because some of this book will be wrong. ⚠️ Which three claims in Chapters 1–16 would you bet on being revised first? State your reasoning, and what evidence would do it.

Your Turn

If you did the Chapter 1 inventory, you've already done the Project Checkpoint. This adds one layer.

Build the correlation table for your own inventory:

Then trace the sources, and answer three questions:

1. What's your ambient category? Beliefs you can't attribute to anyone. How many, and how did they score?
2. What's your stale-guidance category? Things you were told by a legitimate source that have since been revised. ⚠️ Look particularly at anything you learned before about 2010.
3. Which single source is behind the most errors? Not because you should stop reading them — because you should know.

And the harder one, which most people skip:

Name one belief on your sheet that you graded generously. Everybody has one — the belief you marked 🟡 when the honest grade was 🟠, because 🟡 lets you keep it.

You don't have to change it. Just write down which one it was.