Docs / Contests
Contests
A contest is a signed objection on stated grounds to a result, or to an editor's or referee's decision on one. Editors of the field with no stake in it triage it and then uphold or dismiss it. An upheld contest is what changes a result's standing: how the problem's status follows. Some of its decisions can be appealed once, within 30 days (how to challenge each stage).
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 contest's page shows its stage and what happens next, and the editorial queue shows contests waiting for triage, overdue first. PubPhys notifies in the site: the filer, the contested result's author and the contested decision's signer at each stage; the filer and the editors who may release it when the screening holds it; editors with no stake in it when it waits for triage or for its outcome; the field's editors on day 7 of a contest still waiting for triage, and the chief editor on day 14 (who is told what).
Who may contest
- Anyone signed in with an ORCID iD, in a field that takes results (opening a field). A result accepted while launch staffing was on can still be contested after it ends, even if its field has closed.
- Not your own result, and not a decision you signed. The author of a result may contest a decision on it.
- At most 3 contests per ORCID iD in any 7 days, counted when the contest is signed.
What can be contested
- A result that counts for its problem: published, not rejected, withdrawn, refuted or removed, answering the problem's current core question.
- An effective decision on a public result: an editor's acceptance, a referee's recommendation or recommendation against, or a credit. Its grounds can only be error, prior art or missing credit.
- Not: a contest, a decision that was reversed, a rejection or a screening decision (those are appealed), triage and outcome decisions, removals.
Grounds
| Grounds | Meaning | If upheld against a result |
|---|---|---|
| Error | The result, or the reasoning of the contested decision, contains a mistake that invalidates it. | The result is refuted: it no longer counts. |
| Gap | A step is missing or unjustified, so the result does not establish what it claims. | The result is incomplete: a solution or a disproof then counts as progress only. |
| Prior art | The result was published before, by someone else. | The earlier work is credited on the result (a signed credit decision); its standing is unchanged. |
| Missing credit | The result uses earlier work without crediting it. | As for prior art. |
| Irreproducible | A computation, simulation or experiment the result relies on cannot be reproduced from what it gives. | As for an error. |
Against a decision, an upheld contest reverses it (and, for prior art or missing credit, also credits the earlier work on the result): a reversed acceptance leaves the result waiting for an editor again, and its editor cannot decide it again; a reversed recommendation no longer counts, and its referee cannot referee the result again.
Quoting the disputed passage
A contest quotes the passage it disputes exactly, from the signed record it contests: the text or references of a result, the
reasons of a decision. The record names the field and the place (body:120-184: characters 120 to 183, counted in
Unicode code points from 0 over the text exactly as signed, in Unicode NFC). A browser counts a selection in UTF-16 units, in
which a character outside the Basic Multilingual Plane, such as a mathematical letter like 𝐀, takes two; the page
converts, so it counts once. The page fills this in from the range you select (not merely the first place the same words occur); PubPhys checks that the quote is exactly
there. At most 2,000 characters; the objection itself at most
20,000, references at most 4,000.
Source files as for a result (source files). If the text it quotes is later removed or withheld (a removed result, a decision's withheld reasons),
the contest's quote is withheld with it; the place stays. The contest's own signed record holds the quote, so its content is
withheld too (the bundle is then served without it, for that reason); its objection and references stay on its page. If this
happens before the screening reads the contest, nothing is sent and it is held for an editor.
What you sign, and the AI screening
Your browser builds a contest record and computes its fingerprint; one ORCID sign-in signs it, as for a result
(posting a result). You choose whether an AI may screen
it before publication: whether it is a serious, intelligible objection in physics on its grounds (not spam, abuse or a
crank claim); only its grounds, the quoted passage, its text and references are sent, never what it
contests (details). The screening never judges whether
the contest is right. Without consent, or if it is unsure, an editor looks at it first
(hold reasons).
The stages of a contest
Who may appeal (column "How to challenge") is the filer, or for an uphold the author of the contested result (for a contest of a decision, of the result it decided on), within 30 days, once per contest and decision: a fresh decision of the same kind, made after a granted appeal, is final.
| Stage | Starts when | Who acts, and what they can do | Recorded | What they see | How to challenge | Next |
|---|---|---|---|---|---|---|
| Being screened | The filer signs it. | Nobody; the screening runs by itself. | The signed contest in the public log; the screening report. | The filer and the field's editors: the contest. The screening report: the field's editors only. | Nothing is decided yet. | Published if the screening is confident; otherwise held. Still in screening after 7 hours, it is held (AI unavailable). |
| Held | No consent, doubt, a suspected attempt to steer the screening, AI unavailable, a source file that cannot be sent, or a quote withheld before the screening. | An editor of the field who may decide the contest releases it or signs that it is not accepted at screening. | Release: the audit log (internal: only the people who operate PubPhys see it). Not accepted: a signed review record. | The filer and the field's editors: the contest and why it is held. The screening report: the field's editors only. | Nothing is decided yet. | Waiting for triage, or not accepted. After 14 days held (held again after a granted appeal, it counts from the grant), it is listed in the chief editor's section of the queue and the administrator is alerted. |
| Not accepted at screening | An editor signed it on a held contest. | Nobody. | The signed decision. | The filer and the field's editors: the contest and the decision; it was never published. | The filer may appeal; the appeal stays private. A granted appeal sends it back to held. | Final unless appealed. |
| Waiting for triage | Published by the screening or released. | An editor of the field signs substantive (a specific, checkable objection that would change the result's standing or credit, or the contested decision, if right) or not substantive, within 14 days of publication. If no editor may triage it when it is published, its page says "awaiting an editor who may decide it", and the 14 days start only once one may (who decides). Triage says nothing about who is right. | A signed review record on the contest. | Everyone: the contest, its quote and its text. | Nothing is decided yet. | Open, or closed. Past the date it is listed first in the queue, also in the chief editor's section, and the administrator is alerted. |
| Closed: not substantive | An editor signed not substantive. | Nobody. | The signed decision. | Everyone: the contest and the decision with its reasons. | The filer may appeal; a granted appeal sends it back to triage with a new 14-day date. | Final unless appealed. |
| Open | An editor signed substantive. | An editor of the field who may decide it and did not triage it signs uphold or dismiss, with reasons. There is no deadline. | The signed decision. | Everyone. While it concerns a solution or disproof that still counts, or a decision on one, the problem shows contested; if the result stopped counting, the contest stays open but no longer marks the problem. | Triaging it substantive is not appealable. | Upheld or dismissed. |
| Upheld | An editor signed uphold, with its companions in the same signature. | Nobody. | The uphold, and the credit or reversal decisions signed with it, each a review record. | Everyone. | The contested result's author may appeal the uphold with its companions; a credit or a reversal is not appealed on its own. A granted appeal supersedes them all and the contest waits for a fresh outcome. | The result's standing, credit or the contested decision change at once (grounds). |
| Dismissed | An editor signed dismiss. | Nobody. | The signed decision. | Everyone. | The filer may appeal; a granted appeal sends it back to open. | Changes nothing. Final unless appealed. |
| Removed | An editor removes it for copyright, personal data, abuse or spam, at any stage. | Nobody; an open contest closes. For an undecided contest, or one with an appeal of a decision on it waiting, nobody who could not decide it may remove it: not its filer, the signer of the decision it contests or of the appealed decision, anyone excluded from its outcome (its triage editor, while it is open) or from the result; the waiting appeal becomes moot, unless it appeals an uphold. Otherwise, not its filer, nor the contested result's author or the author of another live result on the problem. | A review record written by PubPhys on the editor's instruction. | Its text and quote are withheld for good; a contest that was public keeps its page, which says why. Until no appeal can still need it (30 days after the removal, or once an appeal of its outcome is decided), the page shows the removed text to the parties only: the filer, the contested result's author, the signers of the contested decision and of the contest's outcome, and, while an appeal on the contest is pending, the editors who may decide it. Then that copy is deleted. | Not on the site: by email to the administrator. A removal never undoes a decision already signed on the contest. | Final. |
Decisions on a contest
| Decision | On what | Signed with |
|---|---|---|
| Substantive | A published contest not yet triaged. | On its own. |
| Not substantive | As above. | On its own. |
| Uphold | An open contest. | Its companions: a reversal for a contest of a decision; at least one credit for prior art or missing credit. One upheld contest per signature. |
| Dismiss | An open contest. | On its own. |
| Credit | The contested result (for a contest of a decision, the result it decided on). It names the earlier work, at most 1,000 characters. A result credits each work once: a credit naming a work it already credits is refused, in any signature. | Only with the uphold of an open contest on prior art or missing credit. |
| Reverse | The contested decision, still in force. | Only with the uphold of the open contest against it. |
An uphold missing its companion, or a credit or reversal without its uphold, is refused when the editor opens the signature and again inside it: nothing is signed. If the contested decision was reversed meanwhile, or the result removed, the contest stays decidable and its uphold needs no reversal or credit. If the contested result stops counting (withdrawn, removed, or the problem's core question changed), the contest stays decidable but no longer affects the status.
Who decides
Editors of the result's field, and the chief editor, except:
- the contested result's author, and the author of another live result on the problem;
- the contest's filer;
- whoever signed the decision the contest is against, and whoever signed any editor's decision on the result, even one since reversed;
- for a contest of the result itself, whoever recommended on it, for or against;
- whoever signed an outcome of any contest of the result before: an uphold, a dismissal, a credit or a reversal;
- whoever decided an appeal of a decision on the contest;
- the editor who triaged the contest, from its outcome (triaging another contest of the result does not exclude).
So the editor who accepted the result does not triage a contest of it, and an editor who decided one contest of a result does not decide another; an editor who only triaged one may decide the others. Who is excluded from what.
A contest nobody may decide is not dismissed: it is published, shown on the result and waits, with no deadline, until an editor who may decide it is appointed or lent. With one editor, every contest of that editor's decisions waits this way.