Shadow banning was not built to deal with bots. The practice was popularized in various communities long before bots were a serious problem.

It was built to deal with jerks who, when moderated, would sign up under a new account, or would repost the same content over and over again, or who would invade discussions, distract them with vitriol or out of context nonsense – even if each post as original.

The whole point of the concept was that the jerks would FEEL like they were contributing to the discussion, but maybe being ignored, while maintain a static history or login that made it easier to moderate them.

Platforms say it's for spam [1], which most people interpret to mean bots.

I've never seen a platform provide an explanation like the one you give here.

[1] https://old.reddit.com/r/TheoryOfReddit/comments/8k5qh3/if_y...

You can look up the Wikipedia article on the practice: https://en.wikipedia.org/wiki/Shadow_banning

It has a fairly rich history, and evolution, as well as counter-measures, and counter-counter-measures and counter-counter-counter-measures.

That article barely scratches the surface of what's going on. Reddit alone has tons of secretive ways to remove or hide content, some of which I list here [1]. Others I mention in my talk [2].

[1] https://www.reddit.com/r/reveddit/comments/sxpk15/fyi_my_tho...

[2] https://cantsayanything.win/2022-10-transparent-moderation/

I'm talking about the origination and evolution of a specific practice, called, specifically, "shadow banning".

Yes, other techniques exist. Sometimes people, being imprecise, ignorant and/or lazy, call those techniques "shadow banning", but that isn't what they are.

> Sometimes people, being imprecise, ignorant and/or lazy, call those techniques "shadow banning", but that isn't what they are.

What umbrella term would you use for moderator actions that are kept secret from the author of the content? "Moderation" on its own can be secret or transparent, so by itself isn't an apt description.

I think that might have been its original purpose, but I can't help but believe that this is evil.

People are quick to jump to the "It's their website, they don't have to do anything yada yada" but there is an implied social contract between the users and the owners that for whatever tier of service (free, paid, ad-supported, whatever) you are using you will receive an equal amount and quality of service as every other user on the platform.

I'm not against moderation, I'm against secret moderation that wastes the time of other people's lives. Just be forthright and tell people they are muted but can still participate.

You are vastly stretching the definition of shadowbanning if you're posting a public link that everyone (even people who aren't logged in to Twitter) can see, and then claiming that it's suppressed because it doesn't show up in the replies of a totally different account's Tweet.

It's not a stretch, and it fits the definition Twitter provided. You wouldn't be able to find the tweet unless I'd linked it. It's shadow detached from the parent.

Personally I prefer the term shadow moderation.

FYI, I think your tweet must be hidden if the user is logged out.

I'm logged in, and can see your tweet in the replies just fine. Still not shadowingbanning.

You're mistaken. I tweeted twice, once with the .win link, once without. The one with the .win link does not appear even when logged in as a different username. It does appear for me when I'm logged in as the same username.

So there is shadow moderation going on. I'm not so bothered about whether it's called a shadow ban or shadow moderation. Also, I can see why some former Twitter employees want this distinction to be understood given the post they previously put out. I just think that's a hard sell for the public who is just coming to understand this issue. Even I didn't know about Twitter's 2018 post.

Finally, there was already substantial dispute over whether or not Twitter's initial claims were accurate. See "Shadowbanning is not a thing": black box gaslighting and the power to independently know and credibly critique algorithms" by Kelley Cotter:

https://kelleycotter.com/wp-content/uploads/2021/11/Shadowba...

I just want to thank you for creating reveddit.

They explicitly say they don't shadowban people. Pattern matching for moderation is a completely different thing.

The public is still going to interpret that as shadow moderation of people's content.

Shadowbanning people means banning people, it's obviously separate from moderation of individual tweets. AFAIK the only subreddit that has banned me was the_donald, but I can obviously have my content removed if I post to a news subreddit from domains they consider unacceptable.

It is honestly silly to be surprised. Any community with just a few thousand users will ban entire domains, it's not surprising that they resort to banning entire TLDs at the billion user level.