What moves a score, and what doesn’t.
A correction isn’t an appeal to anyone’s judgment — it’s new evidence entering the same evidence-and-independence engine that produced the original number. If the evidence changes, the arithmetic changes with it.
A correction never means deleting or hiding anything. The original claim and its full evidence trail stay visible after a correction lands — the record shows its own history, it doesn’t erase what came before.
Two kinds of evidence, two different rules.
Every correction ASURIQ receives is one of two things, and the rule for which one is fixed — it’s what keeps a public correction record from turning into ASURIQ making claims of its own.
Facts about public record: what a filing says, what a source’s ownership structure is, what a database contains, what a public video shows at a given timestamp. These are observations about a public artifact, not anyone’s opinion of it — which is why ASURIQ can ingest this kind of evidence automatically, the same way it ingests evidence for the original score.
A hypothesis, a synthesis, a connection between two facts that isn’t itself a documented record yet. This kind is always user-filed, always per item, always opt-in — it is the contributor’s speech, credited to them, never filed by ASURIQ on its own behalf and never presented as ASURIQ’s own finding.
Most corrections submitted here are documented facts — a documented public fact the original score didn’t have. If what you’re submitting is your own reading of the evidence rather than a fact you can point to, say so — it’s logged as interpretation, credited to you by name, and it never gets treated as ASURIQ’s own conclusion.
The Northfield Robotics certificate counts two of its four sources as one voice because both trace to the same parent company. Public proof that a third source shares that same ownership is exactly the kind of documented evidence that would move its independent-group count — the correction mechanism runs on the identical math the original score did.
Send the evidence, not the argument.
Include what you’re correcting, the public source behind the correction, and whether it’s a documented fact or your own interpretation of the evidence.
A dedicated submission form is on the way; for now this reaches a person directly.
Not a support ticket. A record.
Every correction ASURIQ receives, and what happens to it, is part of the public record — not a private queue between you and a support inbox. The same transparency the certificates carry applies to the process that can change them.
This is deliberate, not incidental. A private appeals channel is exactly the kind of editorial discretion this entire model is built to avoid — publishing the process is what keeps “the math re-runs” a checkable claim instead of an assurance you have to take on faith.
Currently in friends-and-family beta. Built on peer-reviewed cognitive architecture.