Anyone can claim they called it. The hard part is proving you called it before — in a way a stranger can check without trusting you, the server, or the platform. This is the plain-English method behind every sealed call on this site: SHA-256, a public manifest, and a Bitcoin anchor.
Predicting an event is the easy part. Proving you predicted it — before it happened, in the exact words you now claim, without quietly editing the record afterward — is the part almost nobody does, because it is genuinely hard. After any major event the internet fills with people who 'called it.' Most are not lying so much as misremembering, rounding a vague hunch up into a precise hit. The only defense against that, for me and against me, is a timestamp a stranger can verify without taking my word for anything. This essay is the plain-English version of how that works, the same method that sits under every graded call in this corpus.
Start with why the obvious answers fail. A dated blog post can be edited and the date changed. A screenshot can be fabricated in a minute. 'I said it on a call' is unverifiable by anyone who wasn't on it. Even a public post on X — better, because it's harder to backdate — still asks you to trust the platform's clock and its promise not to alter history, and it can be quietly deleted. Every one of these requires trusting some party. The goal is a proof that requires trusting no one.
The first move is to turn the claim into a fingerprint. A cryptographic hash function — I use SHA-256 — takes any text and returns a fixed 64-character string. Three properties make it useful: the same input always yields the same output; changing a single character anywhere changes the output completely; and you cannot run it backwards to recover the text. So the hash of a forecast's exact identifying fields — on this site, `objectId | dateIssued | title | claim` — is a fingerprint that only that exact, unaltered forecast can produce. Alter one word later and the fingerprint no longer matches.