Wikipedia:Zero tolerance

The following is a short reflection on no-warning blocks. It is intended primarily for new admins, editors who frequently interact with administrators and/or perform anti-vandalism patrols, and anyone considering RfA.

What is zero-tolerance behavior?
In practical terms, zero-tolerance behavior simply refers to certain types of editing that are viewed by the community as beyond the pale, or so gravely disruptive that they may justify an immediate block without warning. Often this has not been spelled out in policy or guidelines but has nonetheless become de facto accepted practice supported by years of generally unchallenged precedent.

When considering a no-warning block
In all cases administrators should exercise restraint where practicable, and carefully consider the circumstances including possible mitigating or extenuating facts before imposing any block, much less a zero-tolerance one. Even with naked vandalism, in all but the rarest of cases some form of warning is desirable. All of which said, WP:BEFOREBLOCK clearly states that in cases of obviously bad-faith behavior, warnings are not required and an immediate block is within the discretion of administrators.

Common zero-tolerance activities
The following is a list of some of the more common types of behavior that can get an editor blocked without warning.
 * Threats of harm to self or others: This behaviour should be dealt with by following the procedures at Responding to threats of harm. Depending on context, telling other editors to die or wishing harm on them or their families, might also fall under this heading.
 * Legal threats: Any posts that clearly articulate a legal threat against any editor, the community or the Wikimedia Foundation are likely to end with a fast block. In cases where there may be mitigating or extenuating circumstances the blocks are often lifted upon retraction of the threat.
 * Illegal activity: Engaging in, or soliciting/promoting on behalf of, activities that are obviously and gravely illegal (especially in the United States) can result in a zero-tolerance block. To be clear, we are not talking about running a stop sign or drag racing. Think promotion of terrorism or child pornography. Additionally, anything which may be reasonably construed as endangering minors, or otherwise violating WP:CHILDPROTECT.
 * Porn-image vandalism: This is almost always going to get the perpetrating editor an instant block.
 * Use of slurs and epithets: We are not going to write out a list of no-no words. Anyone who is an administrator presumably has the requisite judgment to look at a word and know whether or not it is one that can have no other purpose than to insult or degrade a person or group on the basis of their nationality, race/ethnicity, religious beliefs, disability, sex, gender identity, or sexual orientation. Likewise there are certain symbols that racists and trolls sometimes use that can also be understood as a de-facto slur.
 * Socking and/or IP block evasion: Probably the most common form of zero-tolerance behavior.
 * Red hot and now: When an editor is on an "in progress" campaign of obvious vandalism, warnings are not required. Rampaging vandals should usually be blocked immediately.
 * Defamation of living persons: An editor who adds extremely defamatory material to a biography of a living person (e.g. "(subject) is a registered sex offender") may warrant an immediate block. Such content should always be revision-deleted.
 * Other obvious NOTHERE behavior: This can be something of a catchall for trolls, vandalism-only accounts, and their near-equivalent IPs. If someone just created an account thirty minutes ago and in the last ten minutes have blanked a half dozen articles, or had obviously malicious edits repeatedly obstructed by the edit filter, warnings are probably unnecessary.

Reporting zero-tolerance behavior
Where to report will depend somewhat on the specifics of the situation. With a few exceptions most of what is listed above can be handled at WP:AIV. Threats of harm should generally be handled discreetly where possible. A threat of harm needs to  escalated to the WMF emergency email, Special:EmailUser/Emergency. If you are the reporting editor try to find an admin who is online and active. This tool may be helpful in locating admins who have recently edited. Socking is typically handled at WP:SPI unless it is really obvious, in which case either AIV or WP:ANI would work. Legal threats are usually dealt with at ANI. And of course if you know that a given admin is online, you can always drop them a line on their talk page or ping them to the relevant discussion.

Duration of blocks
Generally this falls under the heading of administrator discretion. Due to the serious nature of zero-tolerance disruptive editing, indefinite blocks are not uncommon, even in the case of a first block. With IPs, it is not unusual to impose a block of longer than usual duration in the case of first offenders. As an example: when dealing with porn-image vandalism, IP blocks are rarely less than a month.

Summary
There are a handful of behaviors that the community regards as so disruptive that an immediate no-warning block is accepted, and in some cases may even be expected. With the exception of socking and block evasion, these activities are, happily, less than common. Admins should exercise restraint where such is possible, and carefully consider the background for a full understanding of the situation before imposing any block, much less one without warning.