TL;DR – If you read nothing else, read this:

DON'T ENGAGE. IGNORE THEM.

That's it. That's the FAQ. Everything below is supporting detail and historical color.

Trolls feed on attention. Deny them attention and they wander off to break something else. There has never been a successful counter-troll strategy that involves engaging with the troll. There never will be. The fancy version of this advice is "don't feed the trolls." The unfancy version is the same.

The medium has changed – Usenet to web forums to blog comments to Twitter to whatever-platform-died-last-week – but the underlying behavior and the only effective response have not. If somebody farts in a room, the smell will eventually go away. If you stir it with a fan, it will not.

The rest of this document is for people who can't quite believe the above and need to walk through the reasoning, plus for the historical record of where this all came from.

– Ken (2026 update)


2026 note: The old / original FAQ originally ran monthly on Usenet from the mid-1990s through 2005 to help people deal with the alt.syntax.tactical crew, the meowers, the kook gallery (long may Archimedes Plutonium reign), and assorted other career trolls of that era. The Usenet bits are now historical – Google Groups closed its Usenet archive to the public in February 2024 and the alt.* hierarchy is a ghost town – but trolling itself migrated to every other medium and got worse. The methodology of recognition and non-engagement is unchanged. The historical AST crew descriptions and the original parody FAQ are preserved verbatim at the end of this document.

– Ken Hollis (GandalfDDI)

Contact: BlueSky @GandalfDDI, Mastodon @GandalfDDI. The old gandalf@digital.net address has been dead for years.

Archive-name: net-abuse-faq/troll-faq
Last-modified: 2026
URL: https://gandalfddi.github.io/trollfaq.html

This FAQ is about crossposts, trolls, and flames, and what to do about them. (Spoiler: see the TL;DR above. The answer is "nothing.")

 
Table of contents
=================
 
   o   Introduction
          o   Introducing alt.syntax.tactical (and friends, old and new)
   o   Defining troll, flames, and crossposts
   o   What to do
          o   E-mailing (and other direct contact)
          o   Killfiles, blocks, mutes, filters
   o   A description of the alt.syntax.tactical crew (and lesser trolls) – historical, preserved as written
   o   The alt.syntax.tactical FAQ – Know your opponents – historical, preserved as written
 
 
Introduction
============

Please feel free to repost, e-mail, mirror, or otherwise distribute this FAQ. Just don't pop it up on the screen of somebody who didn't ask for it.

The TL;DR at the top covers 95% of what you need. The rest of this document is structured as follows:

   o   Item 2 introduces the historical alt.syntax.tactical crew and the modern descendants you're more likely to encounter now (4chan raid culture, groypers, state-sponsored troll factories, AI-generated bot swarms, etc.).
   o   Item 3 defines the vocabulary – troll, flame, crosspost, sealioning, brigading, dogpiling.
   o   Items 4–6 are the practical advice. Item 6 is killfiles / mutes / blocks – the single most underused tool on every modern platform.
   o   Items 7–8 are preserved historical material from the original 1990s FAQ. Don't update what isn't broken.

 
Introducing alt.syntax.tactical (and friends, old and new)
============================================

The original alt.syntax.tactical crew decided every now and then to start a colossal crosspost / flame war. One of their early attacks involved trolling between alt.smokers, can.talk.smoking, alt.support.non-smokers, asthma, and SCI groups: take an original post from alt.smokers, pass it through an anonymous remailer (with the original e-mail name intact), and repost to all of the above. The original innocent poster got blamed for crossposts they never made. It migrated to rec.pets.cats and beyond. It is rumored the AST crew coordinated through a listserver, planning their next attack through back channels. Wouldn't want the plans posted out in the open, now would we.

Jonathan tells us that while alt.tasteless claimed credit for the rec.pets.cats war, it was really the first unofficial AST invasion. "It was all in good fun and we never meant to do more than rile up a few people who took things a bit too seriously." Sure, Jonathan.

Michael adds that the AST crew did not originate in alt.shenanigans – that group played jokes, but never anything like AST. AST was its own special creature.

There was also the MEOW brigade – essentially the same animal as AST, just a different branch of the same family tree that doesn't branch. They originated from a thread in alt.fan.karl-malden.nose about Henrietta Pussycat (the Mr. Rogers cat puppet who speaks in "meow meow go to bed meow? meow very tired meow"). A MEOW post is a full quote of the post above it with no additional content except "meow." Some of the meowers actually did useful anti-spam work in news.admin.net-abuse, but they appeared to believe the rules were for everyone else.

Honorable mention, as always, must be given to Archimedes Plutonium – the dean of the kook gallery, prolific Usenet poster, proponent of a theory in which the universe is one giant plutonium atom (the details vary by decade), and an absolute fixture of net.kookery from the late 1980s onward. His site (or what's left of it) was at archimedesplutonium.com but his life details and choices live on. I will let Archimedes speak for himself, as he always has, at length, regardless of who asked.

You can take the Kook Test – the catnip.org URL is long dead, but the test itself has been archived various places; search for "Hipcrime Vocab kook test" or look in the Wayback Machine.


The modern troll landscape (2026):

The shape of trolling changed dramatically as the audience moved off Usenet. Recognize these:

   o   4chan raid culture – /b/, /pol/, and the boards inherit AST's invasion-wave model and industrialize it. Pick a target, coordinate off-platform (Discord, Telegram, IRC), descend on the victim's social media, blog comments, or company contact form. The 2014 Gamergate harassment campaign was a watershed example. Tactics: doxxing, mass complaint-bombing (false reports to employers / Twitter / hosting providers), fake pizza orders, fake suicide hotline calls.

   o   Groypers – the Nick Fuentes / America First crowd. Tactic: show up at campus Q&A sessions with rehearsed "just asking questions" bigotry designed to either elicit a damaging answer or generate viral clip-bait when ignored. The descendants of AST "invasion waves," just with worse politics and better camera framing. Online they brigade, mass-report, and dogpile.

   o   State-sponsored troll factories – Russia's Internet Research Agency, the various Chinese 50-cent army formations, Iranian PSYOP, etc. These are paid jobs at scale. They look like organic users, push narratives in groups, and inflame existing divisions. Indistinguishable in the moment from a sufficiently dedicated bigot, only with a paycheck. They got significantly more sophisticated post-2016 and they don't go away when ignored – but ignoring them at the individual level is still your job.

   o   AI-generated bot swarms – the 2024 election cycle proved out at scale what was theoretical in 2020: LLM-driven persona accounts that can sustain conversation, write on-brand harassment, and operate at a volume no human troll factory could match. By 2026 a credible-looking pile-on can be assembled by one person with a credit card and an API key. Eliza was a primitive ancestor; the 2026 versions are not.

   o   Targeted harassment platforms – Kiwi Farms and its successors (it's been deplatformed and rehosted many times) exist specifically to coordinate long-term campaigns against individual targets, often trans or otherwise marginalized people. This is past trolling and into stalking; it has a body count.

   o   Swatting – calling in fake hostage / shooting reports to send armed police to a target's home. Genuinely lethal – people have died. The FBI takes this seriously; if it happens to you, document everything and report.

   o   Sealioning / JAQing off – relentless, polite-on-its-face demand for "sources" and "debate" designed to consume the target's time and energy. Often paired with a chorus of co-trolls who cite the target's exhausted refusal to keep engaging as proof of bad faith. The David Malki cartoon that named this is at wondermark.com.

   o   The classic kook – never goes out of style. The Internet has always had its Archimedes Plutoniums and Robert McElwaines and Serdar Argics. The platforms changed; the genre is eternal. Be glad of them; they remind us what real obsession looks like.

The methods evolved. The advice didn't: don't engage. Block, mute, report when it crosses into actual threats or coordinated harassment, and otherwise refuse to be the entertainment.

For the philosophical reasons why this is so hard for humans to do, you may correspond with an automated bot for a while without realizing it. The classic was ELIZA; the 2026 ones are much better. Sometimes you're not arguing with anyone at all. This used to be a punchline; it's a reasonable working assumption now.

 
Defining troll, flames, and crossposts
============================================

From the Jargon File (catb.org/jargon/html/T/troll.html):

troll v., n.

1. [From the Usenet group alt.folklore.urban] To utter a posting on Usenet designed to attract predictable responses or flames; or, the post itself. Derives from the phrase "trolling for newbies" which in turn comes from mainstream "trolling," a style of fishing in which one trails bait through a likely spot hoping for a bite. The well-constructed troll is a post that induces lots of newbies and flamers to make themselves look even more clueless than they already do, while subtly conveying to the more savvy and experienced that it is in fact a deliberate troll. If you don't fall for the joke, you get to be in on it. See also YHBT.

2. An individual who chronically trolls in sense 1; regularly posts specious arguments, flames or personal attacks to a newsgroup, discussion list, or in e-mail for no other purpose than to annoy someone or disrupt a discussion. Trolls are recognizable by the fact that they have no real interest in learning about the topic at hand – they simply want to utter flame bait. Like the ugly creatures they are named after, they exhibit no redeeming characteristics, and as such, they are recognized as a lower form of life on the net, as in, "Oh, ignore him, he's just a troll." Compare kook.

3. [Berkeley] Computer lab monitor. A popular campus job for CS students. Duties include helping newbies and ensuring that lab policies are followed. Probably so-called because it involves lurking in dark cavelike corners.

The use of "troll" in sense 1 is a live metaphor that readily produces elaborations – the warning "Do not feed the troll" is the canonical example.

Some people argue troll (sense 1) is properly narrower than flame bait – that a troll is categorized by containing some assertion that is wrong but not overtly controversial. See also Troll-O-Meter.


Troll typology (an old British classification still useful):
   o   Playtime – one-off trolling.
   o   Tactical – long-term trolling by a single character.
   o   Strategic – long-term trolling using multiple characters and different waves (the AST model).
   o   Domination – trolls who run mailing lists / forums / Discord servers themselves and weaponize moderation.

From the original alt.flame FAQ by Christopher Henrik Lund:
   a)   Flaming – rude comments, insults, personal attacks.
   b)   Trolling – fishing for flames. Smarmy love chatter, useless pieces of boring information, etc.
   c)   Cascades – endless meaningless threads where posters repeat the same phrase over and over with little variation. Amusing to participants, boring to everyone else.

Modern additions to the vocabulary:
   o   Dogpiling / brigading – coordinated mass-reply from a target's enemies, usually after a link is shared in an unfriendly community.
   o   Sealioning – bad-faith demand for sources and debate, performed politely, designed to exhaust.
   o   Quote-tweet ratio – the Twitter / X-era metric: more quote-tweets than likes means everybody piling on hates the post.
   o   Concern trolling – pretending to be a sympathetic insider while subtly undermining whatever the group is trying to do.
   o   Sockpuppeting – running multiple fake accounts to fake consensus.
   o   Astroturfing – sockpuppeting at scale, often paid.
   o   Doxxing – publishing a target's real identity, address, family, employer.
   o   Swatting – calling in fake emergencies to send armed police to the target.

Crossposting is posting the same message to more than one group / community / hashtag. Sometimes legitimate (a relevant cross-disciplinary post); usually a red flag.

 
What to do
============================================

A practical sequence, in order. Walk through this before hitting reply.

A) Don't reply. The TL;DR at the top is not a suggestion.

B) Walk away from the keyboard. Ten minutes. Make tea. Pet the cat. Whatever.

C) Check the headers / profile. If you're on a platform that shows account creation date, post history, and follower graph, look. An account created last week with three followers and a stock photo profile picture is not your real interlocutor. Often it's a bot, a sockpuppet, or a brigade account.

D) Search for the troll's prior pattern. Most trolls have a public history. A 30-second search of the username on the platform plus its predecessor platforms will usually tell you if you're dealing with a one-off or a career operator.

E) If the post has a real header / metadata trail and you suspect impersonation, see the spam FAQ for header-tracing methodology. The same skills apply – somebody may be using a forged e-mail address or impersonating a real person to get that real person in trouble. Don't go after the obvious named target; they're often the victim.

F) Stop sitting on your hands, but don't reply to the post.

G) Read about the Gullibility Virus. A classic comic editorial by Robert Harris on how hoaxes spread, still funny: virtualsalt.com/the-gullibility-virus.

H) Sit and think. The Internet is not real life. The newsgroup / subreddit / forum / Discord / Mastodon instance you're so involved in would forget you exist if you stopped posting for six months. (Most likely quicker.) Real life goes on with or without it.

I) Remember Neal Stephenson's line: "Arguing with anonymous strangers on the Internet is a sucker's game because they almost always turn out to be – or to be indistinguishable from – self-righteous sixteen-year-olds possessing infinite amounts of free time." (Cryptonomicon.) Update for 2026: "or sufficiently capable language models running on someone else's GPU budget."

J) Use the platform's tools. Block. Mute. Report. Filter. See item 6 – this is the single biggest underused option on every modern platform.

K) Whatever you do, DO NOT reply to the post.

L) Go do something else. Clean that green stuff off the sides of your fish tank like you have been meaning to do. The post will still be there if you change your mind. It will also not be there if you decide it never deserved your attention.

M) Document, if it's escalating. Screenshots, archive links (Wayback Machine "Save Page Now," archive.today), header dumps. If this becomes harassment, stalking, swatting, or coordinated brigading against you personally, you'll want evidence. Don't post the evidence at the trolls; save it for law enforcement and / or your lawyer.

N) If it's crossed into actual threats, swatting, or stalking, contact law enforcement. FBI IC3 in the US, your national equivalent elsewhere. For online platforms, also report directly to the platform's Trust & Safety team. Real threats are felonies in most jurisdictions and platforms generally cooperate with law enforcement requests.

After working through that list your blood pressure has hopefully dropped a few points. The above applies to all posts that are trolls – one-off or sustained, single-platform or coordinated.

DO NOT POST a reply. It will only get more people upset at you than at the original troll. If you feel like you absolutely have to post a public response, trim the audience – reply only in the original context, never quote-tweet or amplify, and never crosspost the troll's message into communities that hadn't seen it yet. The troll's goal is exactly that amplification. Don't be the loudspeaker.

 
E-mailing (and other direct contact)
============================================

As above: do not post a reply in the target newsgroup / forum / thread to the troll. It will only clutter the place and get the regulars upset at you for replying.

Raffaele notes there are groups where some regulars encourage the opposite of the golden rule – they consider non-engagement cowardice. As always, read the FAQ for that particular community. Local culture varies. Mostly though, even in groups that say they want pushback, the pushback should come from the long-time regulars, not from a newcomer who's just had their buttons pushed for the first time.

Looking up the troll's other behavior: Google is no longer the friend it was for Usenet; the Google Groups archive went read-only and then closed. Try DuckDuckGo, the platform's own search, or specialty archives. If the troll is a known kook, somebody usually maintains a public dossier. The Wayback Machine often has snapshots even if the original page is gone.

If the message came through an anonymous remailer, the e-mail address is almost certainly forged. Remailers are an important free-speech tool and should not be destroyed because some users abuse them. You can contact the remailer operator if it's serious, but expect a polite refusal to identify the user – that's the entire point. Modern equivalents include Tor-routed services, Cwtch, SimpleX, and various end-to-end-encrypted relays.

If you decide to contact the apparent sender directly, send a nice message and ask why they did it. The e-mail address may be forged and the "sender" may be the actual victim. The person may also genuinely be a newbie who didn't realize they were starting a flame war. Sometimes the cause is hopeless; sometimes it isn't. Don't assume.

If you get a nasty reply (or no reply and the posts continue), the next option is to contact the responsible operator – the platform abuse address, the ISP, the company employing the troll (if they're posting from work, which is a category of stupidity that still happens). See the spam FAQ for how to figure out who that is.

Andrew's rules for writing an effective complaint:

   1.   Keep it simple, keep it short. Abuse desks are buried.
   2.   Be polite. It is not the abuse desk you are angry at.
   3.   Include the full headers / message ID / URL. Without unambiguous identification of the offending content, they cannot act.
   4.   Thank them at the end. Gratitude before the fact helps spur effort on your behalf.
   5.   Be patient. Resolution may take a few days. Don't repost the same complaint; new and different complaints, however, are permissible.
   6.   If it has escalated to harassment or threats, call instead of e-mailing – or contact law enforcement directly. The 1990s advice was that long-distance phone charges might be worth it. In 2026 just use the abuse contact form on the platform's transparency page or file with IC3.

Sample complaint:

   To: abuse@example.com
   Subject: Harassment from your user / IP / customer
  
   Hello,
  
   Please find attached the headers and content of a harassing message posted to <forum / group / platform> on <date>. The message originated from your <customer / user / IP / hosted instance> and I'd appreciate your help in addressing it per your acceptable use policy.
  
   Thank you for your time.

 
Killfiles, blocks, mutes, filters
============================================

The single most underused tool on every modern platform.

The classic Usenet killfile told your newsreader to ignore certain authors, subject lines, or quoted patterns. The newsreader filtered the spool before you ever saw it. rn, trn, nn, tin, Gnus – every newsreader had this. Quiet, effective, invisible to the troll.

The modern equivalents:

   o   Twitter / X: Mute (quiet), Block (the troll sees they're blocked, which they may make a fuss about), Mute Words (your single most powerful tool – mute the troll's catchphrases and the topic itself).
   o   BlueSky: Mute, Block, Moderation Lists (community-curated mute / block lists), labeler subscriptions. The labeler system is what killfiles always wanted to be.
   o   Mastodon: Mute, Block, server-wide blocks via your instance admin, Filter (mute keywords). Different instances have different cultures – pick yours accordingly.
   o   Reddit: Block User, mute subreddit, hide post, RES (Reddit Enhancement Suite) tagging if you want to remember who somebody is.
   o   Facebook / Instagram: Unfollow (silent), Snooze (30 days), Mute, Block, Restrict (they don't know).
   o   Discord: Block user, ignore server, leave server. The server admin has more tools – if you are the admin, learn them.
   o   YouTube comments: Disable comments on your videos. Genuinely. Comments-off is a feature, not a failure.
   o   E-mail: Filter / rule by sender, subject, or domain. Auto-archive or auto-delete. Most modern mail clients have learning filters that get this right.

A few principles:

   o   Mute beats block when the troll might escalate if they notice. Mute is silent on your side.
   o   Block beats mute when the troll is harassing you and you want them off your timeline / mentions entirely, including in other people's threads.
   o   Keyword muting beats account-level muting for ongoing topics – mute the topic, not just the people. New trolls will join the pile-on; muting the topic catches them too.
   o   Block lists / labeler subscriptions let you adopt somebody else's curation. On BlueSky and Mastodon this is built in; it's how community defense scales.
   o   You do not owe anybody your attention. Blocking is not censorship; it's housekeeping.

If you follow all of the above steps you will be happier, the Internet will be a slightly quieter place, and the trolls will have had to find somebody else to bother.

Welcome to the Troll FAQ.