We all predict
with hindsight.
The costliest mistakes aren’t made in the dark. They’re made in broad daylight — by all of us, acting on yesterday, certain it’s today.
A great many of the numbers the world steers by are trailing averages — the average of the last so-many readings of something: the last year of inflation, the last 200 days of a stock price, the last several quarters of growth. A trailing average is built entirely from the past, so it can only ever follow what it measures. It cannot lead. Usually that lag is harmless. But in certain high-stakes systems — retirement income, bank-safety rules, public-health alarms, interest-rate policy — people read the lagging number as if it described the present, act on it strongly, and are harmed in ways that surface so much later, and so far from the cause, that no one ever traces the harm back.
A trailing average — the smoothed, backward-looking number central banks, regulators, and pension funds steer by — is a rear-view mirror. Nothing wrong with a rear-view mirror; the trouble is what people do with it. They mistake it for the windshield.
When the road ahead and the mirror picture drift back toward each other, almost everyone assumes the road is returning to normal — reverting, settling. But it’s backwards. The road isn’t coming back to you. Your mirror is slowly crawling toward where the road used to be.
And if you steer hard by that mirror, you oversteer. You yank the wheel for a curve you’ve already passed. The correction becomes the crash.
The Lagging Truth exists to make that lag visible — before it costs you.
Ten papers mapping where the lag is harmless and where it bites — each with a plain-English companion anyone can understand.
Dated, public, falsifiable calls registered before the outcomes are known — with the rules for judging them wrong written down in advance.
Sealed data, machine-verified numbers, adversarial review, and a public corrections log — so nothing here asks for your trust.
Upload your own sales history and see your trouble months, your limit, and your playbook — free, in about a minute, entirely in your browser.
Stay in the loop
One essay a week translating the research into plain English — plus the launch announcements for new papers, prediction gradings, and the diagnostic. Free, unsubscribe anytime. Read past essays on the Essays page. If the confirmation email lands in spam, moving it to your inbox tells your provider to trust the rest.
The research leaned heavily on AI, and its author is an independent researcher, not a credentialed economist. Every claim is published with the code and data to check it — and experts are invited to try. More about that here.