Append-only · never edited
Every published thesis is logged the moment it goes out — before the outcome is known — and cannot be deleted or rewritten. Right, wrong, or still open. It is not a performance claim. It is a constraint: the reason to trust the next paper is that you can read every previous one, including the ones that did not work.
Deliberately no win rate. A percentage invites you to read this as a performance claim, which is the one thing it is not.
| Logged | Thesis | Resolved | Status |
|---|---|---|---|
| — | Loading the record… | — | — |
Logged at publication, not after. A thesis enters the record the day the paper goes out, while the outcome is genuinely unknown. Nothing is added retroactively.
The wrong-if is written first. Each entry carries the condition that would falsify it, decided in advance. That is what makes "Wrong" a fact rather than an opinion about a fact.
Nothing is deleted or edited. Corrections append with their own timestamp. The original line stays where it was, wrong and legible.
Open stays open. A thesis that has not resolved is not quietly dropped because it stopped looking good. It sits there until it resolves.
No win rate is published. A percentage turns this into a marketing number, and a marketing number is exactly what a record like this exists to replace.