A declaration nobody checks is not a conflict policy.
Most double-blind processes ask the reviewer to step aside if they recognise the work. That asks the one person who must not know the authors to first decide whether they know the authors. We do the checking before anyone reads anything.
To do this at all, one system has to hold three things at once: who
registered, who wrote each paper, and what each reviewer declared. That is
why it cannot be bolted onto a ticketing tool.
First
The reviewer declares, before seeing anything
Supervision, shared group, recent co-authorship, a competing grant, a personal tie. Named people and named institutions.
Second
The author declares every co-author
Name and email, at submission. Used to exclude reviewers and to credit authorship, and for nothing else, ever.
Third
The two lists are crossed automatically
Conflicted pairings are removed from the pool before the random draw. Neither side is told the pairing was ever considered.
Fourth
And still, a way out
If a reviewer infers the authorship anyway, one click returns the paper. No reason required, no name to confirm.
A blind interface over a signed PDF is theatre
So the file is treated as part of the blind, not an attachment to it.
i
Authorship metadata is stripped from the upload.
ii
A blinded version is required — no named title page, no identifying acknowledgements.
iii
Before the submission is accepted, the text is searched for the co-author names the author just declared, and the author is shown exactly where they appear.
What we will not do
Assign past an exclusion to clear a queue
If a session runs out of eligible reviewers, you are told, and offered two honest ways out: borrow from another session, or bring someone new in.
Show the reviewer why they were skipped
An exclusion reveals authorship by implication. The trail is auditable by you, invisible to them.
Change the rules halfway through review
You choose whether reviewers score, recommend or decide, and whether a session reads every paper together or shares them out. Once the first paper is handed out, that choice is fixed, so no paper is judged by a different standard from its neighbours.