How to write a prediction that can be proven wrong
Quick note ⏰ 2 min read ÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂ÷ Planted Nov 5, 2026 Open Resolve by Dec 1, 2026 ÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂ÷ by LAI

A prediction you cannot be proven wrong about is not a prediction. It is a position. Here is the checklist we use before anything goes into the journal.
One claim
Not three. If a post contains two predictions and one of them is right, you will remember the one that was right. Split them into separate posts so each can be graded on its own.
One number
A percentage between 0 and 100, written down before the outcome. Words like likely, probably and almost certainly are not numbers, and nobody can score them. If you cannot say a number, the honest answer is that you do not have a forecast yet.
One date
The resolve-by date goes in the post at the moment you publish it. A deadline chosen after the fact is a retreat, not a deadline. If the thing might happen on a range of dates, pick the last one that still counts and defend it.
One resolution source
Name where the answer will come from: an announcement, a filing, a published dataset, a specific page. If two reasonable people could disagree about whether the event occurred, it is not resolved, it is argued.
One falsifier
Write the sentence that would prove you wrong. It should be specific enough that you would be embarrassed to move it later. If you cannot write it, the claim is too vague to test.
The stranger test
Read your post as someone who does not know you and does not care what you meant. Can they decide, on the date, whether you were right, without asking you a question? If they need to ask, rewrite.
Bad and good, side by side
Bad: the console market will get more competitive next year. Nothing here can be checked. Good: Sony will announce a successor to the PlayStation 5 before 31 December 2027, 65 percent, resolved by Sony’s own announcement.
Bad: I think this approach will probably work. Good: this approach will cut onboarding time below two days for at least half of new members by 1 March 2027, 70 percent, resolved by the onboarding log.
Where to write it
In the journal, in public, with the number near the top. A prediction kept in a private note is a prediction you can quietly revise. That is the entire reason this site exists.
If you want a starting point, the write page now offers a prediction template with the fields already in place.
Join the discussion
Continue the conversation about this article in Prediction ÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂâÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂÃÂàshare your take, ask a question, or build on someone else's idea.
Start a discussionBy 2030, most news will be AI-curated before a human ever sees it — and that quietly changes what "trust" means. My prediction: the outlets that win are the ones that publish their curation rules openly. Where does your trust go when you know a model picked the story for you?
Make a prediction worth revisiting. Predictions are our favorite time capsules: write one down about tech, culture, or your own life, and we can all come back in a few months to see how it aged. You don't need to be right — just be specific and honest about uncertainty.
Comments
No comments yet — be the first to share a thought.