Abuse Detection Mechanism

open reporting is easy to weaponise a user who submits a stream of unfounded reports against another account costs you moderator time and, if the reports are targeted, becomes a harassment vector in its own right abuse detection watches how a reporter's submissions actually resolve when someone accumulates a set number of reports that moderators dismissed as no policy violation , nima applies a moderation action to that reporter automatically configure it in configure > trust & safety > abuse detection how a rule works each rule is a threshold and a consequence select add new rule and set three fields number of reports with no policy violation — how many of this account's reports must have been dismissed before the rule fires action triggered — the moderation action nima applies to the reporting account any action you have configured is available here; see actions docid\ lsfkyoqsi365x alowit1 duration — how long the action stays in force, in hours save the rule with the tick, or remove it with the bin icon you can run several rules at once to build a graduated response — for example, a warning at five dismissed reports and a 72 hour suspension of reporting rights at twenty choosing thresholds set the first threshold well above your normal false positive rate reporting in good faith and being wrong is ordinary user behaviour — a moderator dismissing a report is not evidence of abuse, and a threshold set too low punishes conscientious users and suppresses genuine reports a practical starting point is to look at your dismissal rate per reporter over a month and set the first rule above the great majority of accounts, so only clear outliers are caught tighten it once you can see how many accounts the rule actually affects abuse detection acts on the reporting account, not on reported content the action you select should be one that makes sense applied to a reporter — restricting their ability to submit reports, for instance — rather than a content level enforcement