Around the time that Twitter suspended @realDonaldTrump, they also suspended ~100,000 or so other accounts. We ( @uwcip PhD student Andrew Beers) overlaid those suspensions onto a network graph of “influencers” in voting/election Twitter discourse (black sections of this graph).

On the political “right”, there is a section of highly active, pro-Trump accounts that have high #s of both friends + followers (probably from Trump train / follow back activity in the past). In our graph (and others) Qanon accounts are embedded into that larger MAGA network.

Twitter suspensions took out a massive section of those influential MAGA/QAnon accounts. Now, these weren’t rank-and-file accounts (in our graph), they were influencers (highly RTed by others). This graph doesn’t provide insight into accounts that weren't as visible/influential.

The graph was published in a recent Nature article that describes some of the dynamics of participatory propaganda: looking at top-down and bottom-up dynamics within the disinformation campaign to falsely claim election fraud in Election 2020. https://www.nature.com/articles/d41586-021-00257-y

PhD student Andrew Beers, @beeeeeers, generated this network graph (and devised the network property that allows us to focus on the influencers but retweet/retweeters to define the structure).

Answering other questions. The largest black (suspended) account in the graph (in the red, close to what was once an orange cluster) is @realDonaldTrump. The largish blue account on the peninsula that reaches out to the red is @thehill. (There are other media in that area).

Accounts are connected with an edge when >=20 accounts retweeted both at least 7 times (over the course of several months in tweets that included voting or election-related terms).

And here’s a much more detailed explanation of the network graph, from it’s creator, Andrew Beers ( @beeeeeers): https://twitter.com/beeeeeers/status/1358141547057848320