Are You Shadowbanned ? Or Jumping to Conclusions ?

By | March 25, 2018

The issue of shadowbanning on Twitter seems to have hit a level of consciousness that at times reminds of the days when fat fingers would do something unexpected on a computer and the immediate conclusion people would jump to was that they had been infected with a mysterious virus.

ShadowBanI’ve recently pointed a number of people at a great article describing why the Twitter algorithms will flag you to be shadowbanned and what to do about it.  It can be found here: Shadow Banning On Twitter

It can be summarised very succinctly as: don’t look like a bot.

[ETA: Pay particular attention to bullet #4 in that article: “Be careful who you retweet. If you are RTing banned people or known spam accounts you could in turn get banned yourself.

You can make yourself squeaky clean according to the Twitter rules and still find yourself banned due to RT-ing other people who are banned.  Usually, those other people are unaware of being banned and this becomes a huge rats nest to navigate.]

The article only mentions that shadowban checkers are unreliable without saying why.  Here’s some notes on a seemingly reliable method to properly check for yourself, or have someone else do it for you.

All that the shadow ban checkers do under the hood is call the Twitter Search API, they then scan the returned results to see if a tweet turns up.

One way of trying to not look like a bot is to set your account to protected.  Limiting the reach of a tweet is about the last thing a bot will do

This renders shadowban checkers useless.  The problem is that they access the Twitter Search API under their own account, which means that they can’t see a protected account because they don’t follow it.  Hence, when they’re now used on your Twitter handle they see no results and proclaim that you might be shadow banned.

The definitive check is to open a new browser window on a machine on which you are already logged in to Twitter and go to the Twitter Search page https://twitter.com/search and enter the handle to search for.  You must be a follower of the handle you are searching for.  And the search term must be of the form “from:@user”.

Performing this search will let you see search results for protected tweets.  If you are shadow banned it will show no results, even though you can happily see the TL for this handle in a different window.  It seems to be a simple and reliable test of shadow ban status.  If you see search results, the person is not shadowbanned.  If you don’t see search results, the likely are.

By keeping a TL window open and a Twitter Search window it’s possible to have an easy visual check just by looking at the tabs.  Provided the Search window keeps ticking over with new tweets in synchronisation with the TL window, then all is ok.  Some context is required here, the search results will not list retweets, only actual tweets and replies.  But it’s easy enough to correlate that with a TL window.

[ETA: I’ve also noticed on the Twitter Search page that when typing in a handle to search for tweets, a shadowbanned handle no longer appears in the autocomplete listing.]

[ETA II: It does look as if one can get a good indicator as to whether a protected account is banned or not.  When entering the search term for the user in the form “from:@username” if the dropdown list shows the user name it is an indication that they are not banned.  Note: clicking on a name in the drop down list negates the search term you’ve just been typing in the box.  It will bring up the user’s home page and if you’re not paying attention this can look as if the user is not banned when they actually are.]

[ETA III:

  1. Why do some accounts appear to be immune ?

A working theory for this goes as follows:

It may be something to do with the handle name.  Accounts with names to include the likes of ‘Miss’ or ‘Mistress’ or ‘Goddess’ or “Lady” or “Domme” or “Domina” are known to have suffered being banned.  Some accounts with no suggestive SW connotations in the handle are amongst those observed to be seemingly immune from a ban, no matter what rules they flout.  There’s also a question as to whether this applies just to the main handle or also to the screen name ?

I’m sure we will encounter anomalies to this theory and it doesn’t help those who’s handle Twitter filters might look at more aggressively.  Finding the rule of immunity is important.]

Update

A new BuzzFeed article Twitter Is Going To Limit The Visibility Of Tweets From People Behaving Badly is a report from Twitter shedding a bit of light on how Twitter will detect [some] behaviour that lead to a shadow ban.

On Tuesday, Twitter announced a massive change to the way its conversations will work, evaluating not just the content of individual tweets, but the way users behave more broadly on the service. Twitter will now use thousands of behavioral signals when filtering search, replies, and algorithmic recommendations. If it believes you are trying to game its system, or simply acting like a jerk, it will push your tweets lower down. It’s the biggest update so far in the company’s push to create healthier conversations, an initiative announced by its CEO Jack Dorsey in March

  • whether you tweet at large numbers of accounts you don’t follow
  • how often you’re blocked by people you interact with
  • whether you created many accounts from a single IP address
  • whether your account is closely related to others that have violated its terms of service

This comment from Jack Dorsey makes sense in light of some changes I thought I’d sensed in recent weeks:

A lot of our past action has been content-based, and we have been shifting more and more toward conduct and behaviors on the system

It had seemed that quite a few SW accounts were falling foul of some arbitrary content rules, especially with images.  I’ve not seen some of those accounts jumped on recently has had previously been the case.

Obviously Twitter are not disclosing anywhere near enough information to make really informed decisions, but anything is better than nothing.

There’s more information in a Slate article Twitter Will Start Hiding Tweets That “Detract From the Conversation” about how they will mute tweets in a conversation.

Who decides and how that a tweet should not be part of a conversation is potentially worrying.

The annoying thing is that twitter haven’t yet worked out how or if they are going to notify users that they are on the naughty step.  Without that, people simply won’t know and it will complicate checking shadow ban status.  I think that all most people have wanted is to know they are banned and to be pointed at what caused the violation and the steps required to lift the ban.

Secrecy has led to conspiracy.

One thought on “Are You Shadowbanned ? Or Jumping to Conclusions ?

  1. Pingback: Twitter – Analysing Shadow Bans – Insights and Ramblings of melody

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.