Docs / Appeals
Appeals
An appeal asks for an editor's decision to be looked at again. The party the decision went against signs it from an account with a linked ORCID iD, with reasons, within 30 days. No AI reads it: it goes straight to the editors of the field, and an editor who decided nothing in the case (removing or withholding anything in it counts) grants or denies it. Each decision can be appealed once (for a contest, its triage and its outcome once each), and a fresh decision of the same kind made after a granted appeal is final, so a case always ends. A later decision of another kind is a new decision: after a granted appeal of a rejection at screening, the editor's decision on the merits can still be appealed once.
The decision's page shows the way to appeal while you may, and the appeal's page shows where it stands. PubPhys notifies in the site: the decision's signer and the other parties to it, and the editors who may decide it, when it is filed; the appellant, the parties and, once the item was published, its followers when it is granted, denied or made moot; the signers of decisions a grant supersedes beyond the appealed one (who is told what). An appeal on an item never published stays private, and so do its notifications.
What can be appealed, and by whom
| Decision | On | Who may appeal | If the appeal is granted |
|---|---|---|---|
| Not accepted at screening | A result | Its author | Back to held: an editor releases it or rejects it at screening again. |
| Reject | A result | Its author | Back to published and waiting for an editor's decision; a result counts again meanwhile. |
| Not accepted at screening | A contest | The contest's filer | Back to held: an editor releases it or rejects it at screening again. |
| Not substantive | A contest | The contest's filer | Back to triage, due 14 days after the grant. |
| Dismissed | A contest | The contest's filer | Back to open, waiting for a fresh outcome. |
| Upheld | A contest | The contested result's author (for a contest of a decision, the author of the result it decided on) | The uphold and the credits and reversals signed with it stop counting, as do decisions made later in the places those reversals had freed; what they reversed counts again. The contest is back to open. |
| Not accepted at screening | A problem or a revision | Its author | Back to held: an editor releases it or rejects it at screening again. |
| Reject | A problem or a revision | Its author | Back to published and waiting for an editor's decision; a result counts again meanwhile. |
| Reversed | A problem or a revision | Its author | The reverted edit is displayed again with its acceptance, if the problem has not changed since. |
| Not accepted at screening | A topic or a revision of it | Its author | Back to held: an editor releases it or rejects it at screening again. |
| Reject | A topic or a revision of it | Its author | Back to published and waiting for an editor's decision; a result counts again meanwhile. |
| Appeals of decisions on | Open while |
|---|---|
| New problems (revision 1 of a problem, phase 3's screening rejections included) | new problems are accepted |
| Later revisions of problems, and reverts | revisions are accepted |
| Topics and their revisions | topics are accepted |
| Results and contests | results are accepted |
While appeals of a kind are closed, no new appeal of that kind is signed; appeals already filed are still decided.
Not appealable: an acceptance, a recommendation or a recommendation against (a contest is the way to challenge those), triaging a contest as substantive, a credit or a reversal on its own (they go with the uphold), an appeal decision, a fresh decision of the same kind made after a granted appeal, and a removal (removals are administrative: write to the administrator). Phase 3's screening rejections of problems are appealable like an editor's. On a later revision of a problem, an appeal becomes moot once the problem changes again (revisions); so does one on a topic once the topic changes again (topics). The platform's own refusal of a revision or a topic whose base moved meanwhile is not appealable on its own.
The 30 days
The window runs from the decision's signature, or, if appeals of its kind are closed then, from the moment they next open (above). If they close during the 30 days, the 30 days start again in full when they reopen. A window that had already ended before the closure stays ended. For an uphold whose contest was removed, the 30 days run from the removal, and a closure restarts them in the same way. Decisions on new problems made before 14 October 2026, when their appeals stopped waiting for results to open, may be appealed for 30 days from that date.
To the party who may appeal, the decision's page shows the last day to appeal, or says that appeals of its kind are closed at the moment and that the 30 days will start when they reopen.
If a contest is removed during the 30 days after its uphold, the result's author has 30 days from the removal instead: the uphold stays in force after a removal, and the appeal is how to challenge it. The parties and the editors who may decide the appeal can read the removed text on the contest's page until then, or until the appeal is decided. A decision no longer in force (reversed, or superseded by another appeal) cannot be appealed; nor can one on a removed result or problem, nor a decision on a removed contest other than its uphold (below).
What you sign
Your browser builds an appeal record naming the decision it appeals, with your reasons (at most
4,000 characters), and one ORCID sign-in signs it. It has no files and declares no
AI. Only the party the table names may sign it.
Who decides
An editor of the field, or the chief editor, who:
- is not the appellant, and not the author of the result or problem concerned, nor of another live result on the same problem;
- did not file the contest the appeal concerns;
- decided nothing in the case before: no decision on the result, its contests, the decisions on them, or its other appeals (for a problem, on the problem and its appeals), and no removal or withholding in it (PubPhys's records written on an editor's instruction count for that editor).
When the editor who instructed a screening rejection made before signed decisions is not on record, the chief editor decides the appeal. If every chief editor is excluded from it by the rules above or suspended, or none is appointed, an editor of the field who decided nothing in the case decides it instead. If the chief editor cannot decide because of an interest of their own that is not yet public, the appeal waits; it resolves when the chief editor's interest ends.
An appeal that nobody may decide is flagged no eligible editor, like a result or a contest, and the chief editor lends an editor of another field (lending).
With one editor
PubPhys has one editor for now; what that editor may not decide waits for another.
An appeal goes to someone other than the editor whose decision it concerns. While PubPhys has one editor, an appeal of that editor's decision is published as usual and waits for a second editor: it is not dismissed and has no deadline. Its page says "No editor may decide this yet." until an editor who may decide it is appointed or lent.
During PubPhys's launch, results, contests and comments open in a field once it has one editor (the chief editor counts). With one editor, a decision cannot yet be contested or appealed to anyone else: contests and appeals are published and wait for a second editor, with no deadline. Referees are appointed as the site grows; a solution is verified only after two referees recommend it.
The fresh decision after a granted appeal is made by someone else again: not the editor whose decision was overturned, and not the editor who granted the appeal.
What a grant does
The appealed decision stays in the public log, marked superseded, and stops counting. For an uphold, so do the credits and reversals signed with it, and any decision made later in a place one of those reversals had freed (for example a new acceptance after an acceptance was reversed); the reversed decisions count again. The editor sees this list before signing; if it changes before the signature, the signature is refused and the grant is staged again. An appeal pending against a decision the grant supersedes becomes moot: nothing is left to decide. A held item carries the reason "appeal granted".
If a contest is removed while an appeal of its screening rejection, triage or dismissal waits, that appeal becomes moot at the removal: nothing on a removed contest can be restored, so no grant can move it back to held or to triage. An appeal of an uphold stays open (above). While such an appeal waits, the contest is removed only by an editor who could decide it, and never by the signer of the appealed decision.
Who sees an appeal
An appeal is public, with the decision on it, once the item it concerns was ever public. An appeal of a screening rejection of an item that was never published stays visible only to the appellant and the field's editors. An editor may withhold an appeal's text, or a decision's reasons, for copyright, personal data, abuse or spam; the record and its effect stay.