I got algobanned from Twitter the other day. At least, this seemed like the most natural thing to call it. To me, it seems like an obvious term to coin. In the slang of Internet forums we have many types of bans: "temp-bans", "perma-bans", "shadow-bans" and one can "be permabanned" (permanently banned) or "hell-banned" (another term for shadowbanned) and so on and so forth.
But I realized that I'm not actually sure I've heard this expression anywhere, so I did some casual searching. Certainly it's only becoming more relevant with time, as algobanning becomes the common practice, and users are getting increasingly frustrated with it.
So how far back does the phrase go? Is it continually coined every so often? (spoiler: yes) Here are some stats for where it was used in the same sense I did: where an "algoban" is the type of ban, but not "the algo banned me" where "[the] algo" is the subject and "ban" is an action it took.
For "algobanned", all one word:
- 0 results on Google
- 1 result on Twitter's own search:
August 15, 2020
"I’m pretty sure I’m either shadowbanned or algobanned or [...]"
- 0 results on Facebook and Reddit search
For "algo banned", an exact two-word phrase:
- Google: 3 (of 52 raw results)
- Twitter search: 8 (of 24 raw)
And how about the full length equivalent, "algorithmically banned"
- About 20 results on Google (of 54 raw results)
- 8 results on Twitter search (of 24 raw)
An archived 4chan thread from April 2016
"[...] just so i dont get algo banned from omegle"
An unexpected use from
February 21, 2017in this forum post that seems to be about the Pokemon competitive metagame
as for now, tapus are algo banned.. but who knows, things might change completely next week :hor:
Jan 4, 2021
I totally believe their story. I got algo banned before running a single ad for an app that Apple approved. (Higher bar than FB.) Stonewalled on appeal. Makes no sense.
Here's a curious example from a spam domain. Google reports first indexing the page in
Her current projects try combining machine learning and crowdsourcing manual trail of algorithmically banned brain volumes through gamification and expository reality, identifying biomarkers of epileptogenesis after operating brain injury, and using the effects of high enough exercise on the ring.
It's clearly generated in some way. If it was generated by a transformer model like GPT2/3, it would be interesting that the model picked up on the idea something can be "algorithmically banned". If it generated that spontaneously without seeing it in the training data, that would be even more interesting.
The oldest I can find is a Webmaster World forum post
July 30, 2005and it talks about Google algorithmically banning sites from its search index. This is a common theme with the posts older than a couple of years.
- Algorithmically Banned on 7/28 (Unless Google has army of eval.google trainees that weren't properly trained)