The method

How we check, and what the tags mean

This page is the contract. It says what happens to a claim before it reaches you, what each tag is allowed to mean, where sources come from, and what we do when we get something wrong.

The four steps

Every question goes through the same four steps in the same order. The order matters. Checking comes last on purpose, because you cannot check a claim properly until you know what the argument is actually about.

  • Pick one honest question A question, not a topic. It has to be something a real person says out loud, and it has to be answerable wrongly. Is your house an asset? is a question. Real estate is a topic. If both a yes and a no would change what somebody does on Monday, it qualifies.
  • Put the famous books in the same room We pick the books people have actually read, not the books that agree with us. Fame is the selection criterion, because those are the books doing the work in the real world. We read the book itself rather than a summary of it, and we try to state each position in the form its own author would recognise.
  • Separate what they agree on from what they argue about This is where most of the value is. Books that appear to be shouting at each other usually agree on far more than they admit, then split on one buried assumption, often about human behaviour rather than arithmetic. We name that assumption in plain words. Once you can see it, you can usually work out which book is talking to someone like you.
  • Check the claims, then tag them Every factual claim is looked up and labelled Established, Contested, or Author’s Claim. The sources are listed with the date they were checked. Where a claim is country specific, we say which country the check applies to, because rules on tax, accounts and investing differ by country and change over time.

What we do not do

We do not rank the books, we do not declare an overall winner, and we do not tell you which one to buy. A verdict here is about a claim, never about a person. Books can be badly evidenced and still be worth reading, and they can be well evidenced and still be wrong for your life.

The three tags

Established, Contested, Author’s Claim

Three labels, applied the same way every time. The tag is a statement about the evidence, not about the author.

Established

Checked against the public record

Several independent, reputable sources agree, the claim is stable across them, and we can point you at each one. Definitions and arithmetic also land here, because they can be reproduced rather than trusted.

Established does not mean permanent. It means that as of the date we checked, this is what the record says.

Contested

Publicly disputed, both sides shown

Serious, informed people disagree in public, and the evidence genuinely points more than one way. We show the disagreement and name what it turns on, rather than picking a side and hoping you do not check.

Contested is not a polite no. It is a live argument you should know you are standing inside.

Author’s Claim

Asserted by the book, not independently verified

Personal results, private figures, unnamed mentors, and stories with no paper trail. We can confirm the book says it. We cannot confirm it is so, and we will not pretend otherwise.

This is not an accusation. Plenty of true things are unverifiable. It is a note about what evidence exists.

Worked examples

Nine examples of the tags in use. Use the filter to show one kind at a time. The examples below are written in general terms on purpose. We do not attribute a quotation to a book unless we have that book open in front of us and can give you the page.

Showing all nine worked examples.

Established

“Compound interest means you earn returns on your previous returns, not only on the money you first put in.”

Why this tag. This is a definition plus arithmetic. You do not have to trust anybody. You can reproduce it on paper in about a minute, and every reference source describes it the same way.

Established

“Fees are deducted whether or not the investment goes up.”

Why this tag. It is stated in the published documents of the products themselves, it is consistent across providers and countries, and any reader can verify it in the fee disclosure of a fund they already hold.

Established

“Carrying a balance on a credit card is one of the most expensive ways an ordinary household can borrow.”

Why this tag. Lenders are required to publish their rates, and the comparison against other consumer borrowing is public and checkable in every market we looked at. We cite the rate disclosures and the date, because rates move.

Contested

“Pay off your smallest debt first, not the most expensive one.”

Why this tag. The arithmetic clearly favours clearing the highest rate first. A serious body of behavioural work and a lot of practitioner experience argues that finishing something early keeps people going. Both sides are published, and the disagreement is about people, not about interest.

Contested

“Renting is throwing money away.”

Why this tag. Whether buying beats renting depends on price levels, borrowing costs, how long you stay, transaction costs, maintenance and what a renter does with the difference. Reputable analyses reach opposite conclusions in different markets and different decades, which is exactly what contested means.

Contested

“Everyone should hold six months of expenses in cash.”

Why this tag. The recommended figure moves depending on who is writing, how stable the income is, whether there is a public safety net, and what the alternative uses of the money are. There is broad agreement that some buffer helps. There is no settled number, so we will not report one as if there were.

Author’s Claim

“I turned a small sum into a fortune using exactly the method in this book.”

Why this tag. Private finances, no filings, no independent record, and no way for a reader to check the starting figure, the timeline, or what else was going on. It may well be true. It is not evidence that the method causes the result.

Author’s Claim

“A wealthy mentor taught me these rules when I was young.”

Why this tag. An unnamed person in an undated story. Often the mentor is described as a composite or a teaching device by the author elsewhere. We can verify that the book says it. We cannot verify the person, so the tag says so.

Author’s Claim

“Thinking about wealth in the right way causes wealth to arrive.”

Why this tag. It is presented as a law, and the support offered is testimony from people it worked for. Testimony from people it did not work for is not collected, which is what makes the claim unfalsifiable as written. That is an Author’s Claim, not an Established one.

Two rules about tags

One, the tag belongs to the claim, not the book. A single book can hold all three tags on the same page, and most good ones do.

Two, we take the boring tag. If a claim is well evidenced but dull, it is still Established. If a claim we like has no support, it is still an Author’s Claim. The moment the tag starts following our preferences it is worthless.

Sources policy

Where the checking comes from

A source has to be public, independent of the claim, and reachable by you. If we cannot show it to you, it did not count for us either.

What we treat as strong

  • Official statistical agencies, central banks, tax authorities, and financial regulators, for anything about rules, rates, or measured aggregates.
  • Peer-reviewed research, read in full rather than through a press release, with attention to what the study actually measured and who it measured.
  • Primary documents: company filings, fund prospectuses and fee disclosures, court records, published legislation.
  • Long-running, published comparisons produced by organisations that show their methodology.
  • The book itself, quoted briefly and accurately, for what the book argues.

What we treat with care

  • Reporting in reputable outlets, which is useful for finding the primary source and weak as a final source on its own.
  • Industry research, which is often good work with an interest in the answer. We say who paid for it.
  • Anything behind a paywall. We still use it, and we say plainly that it is paywalled so you know what you are being asked to take on trust.

What does not count as a source

  • A different book repeating the same claim. Repetition is not verification.
  • A statistic with no origin, including one we would like to be true.
  • The author’s credentials, seniority, or track record.
  • An anecdote, a screenshot, a testimonial, or an unnamed mentor.
  • A number generated or recalled by an AI assistant. Useful for finding candidate sources, never a source.

Four rules we hold ourselves to

  1. Date everything. A check is only true as of the day it happened, so the date is published with the claim.
  2. Say which country. Rules on tax, accounts and investing differ by country and change over time. A check done in one place does not travel.
  3. Quote briefly and fairly. Direct quotations are short, attributed, and used for commentary and criticism. We analyse the ideas, which copyright does not protect, rather than reproducing the expression, which it does.
  4. Show the disagreement. If two strong sources conflict, that conflict is the finding, and it gets tagged Contested rather than resolved for the sake of a clean sentence.

Where the method is weak

Three honest limits, so you can discount us appropriately.

  • Fame is a biased filter. Choosing the most famous books means the shelf we examine is the shelf that sold, not the shelf that was most correct. Good, quiet books get missed.
  • Established can go stale. Rates, rules and thresholds move. An old check with a clear date is honest. It is still old.
  • Contested can be misread as neutral. Showing both sides does not mean the sides are equally weighted. Where the weight is lopsided, we say so in the note, and the note is the part worth reading.

Corrections policy

What happens when we get it wrong

A project that tags other people’s claims has to be the easiest project in the world to correct. Otherwise the tags are decoration. Corrections get published, not deleted.

How to report an error

Use the contact page, or reply on any of the channels in the footer. The more of this you can include, the faster it moves.

  • Where the error is: the page and, if you can, the sentence.
  • What you believe is wrong about it.
  • What you think is correct, and how you know.
  • A source we can reach, if one exists. If it does not, say so and send it anyway.

You do not need a source to report an error, and you do not need to be polite about it. A blunt message with a link is worth more than a careful message without one.

What we do with it

  1. We acknowledge it. Even the ones we end up disagreeing with.
  2. We check it ourselves, against the sources policy above, not against how much we would prefer to be right.
  3. We classify it. Three levels, below.
  4. We publish the outcome on the page where the error appeared, dated, and we leave the record of it in place.
How a correction is handled depends on what the error changes.
LevelWhat it isWhat we do
Typo Spelling, a broken link, a wrong page number, a formatting slip. Nothing about the meaning changes. Fixed quietly and quickly. No note needed.
Clarification What we wrote was defensible but easy to misread, or it was missing a condition such as a country or a date. The text is amended and a short dated note is added at the foot of the page saying what changed.
Correction A factual claim was wrong, or a tag was wrong, or we described a book’s position unfairly. The error is struck through or clearly marked, the correct version is published next to it, the tag is changed if needed, and the note is dated and signed. The original wording stays visible.

Retagging

Tags can move both ways. Something tagged Established can become Contested when good contrary evidence appears. Something tagged Author’s Claim becomes Established the moment somebody shows us a public record that supports it. When a tag changes, the old tag and the reason for the change stay on the page.

If you are a rights holder

This project produces original commentary, criticism, and educational synthesis, and any direct quotation is brief and attributed. If you hold rights in something referenced here and believe the use exceeds fair use or fair dealing, please get in touch. It will be reviewed promptly and in good faith. Details on the disclaimers page.

The standing invitation

Nothing here should be treated as the final word. It is an invitation to read further, and to come back and tell us where we went wrong.

Report an error

Next

See the method applied

The free resources turn this page into things you can actually use: a claim-checking checklist, the questions to ask any advice book, and a guide to judging a source in under a minute.

Free resources Questions and answers

Reminder: this is what the books argue, not advice for your money. Check it, and talk to someone qualified before you act.