Excited to announce new working paper w/ @DarrenLinvill and @_aemoore -- "Touching Trolls -- How Real People Respond to a Coordinated Information Operation and Why They’re So Nice" Short Thread (1/?) Paper here: http://pwarren.people.clemson.edu/Touching_Trolls_080520.pdf

Comments very welcome.

Blurb: "Info Ops get more engagement by making friends than starting fights, and they seem to know it."

Darren and I have been trying to make this point for general audiences, from our intuitive feel of what was going on. The data seem to bear it out. https://www.rollingstone.com/politics/politics-features/russia-troll-2020-election-interference-twitter-916482/

People have done a lot of work about what coordinated info ops say and do. We want to look at how real people respond to those campaigns.

We look at the population of outsider mentions of Internet Research Agency accounts in the month before the 2016 election .

These mentions fall quite naturally into 4-5 categories
1) Support -- "Right on, Dude!"
2) Attack -- "You're a idiot/troll!"
3) Comment -- "I like bananas"
4) Troll whistle -- " @Troll1, @katestarbird, @yoyoel"
5) ??? -- ???

Supportive replies are the overwhelming majority, especially for the identity-group trolls. People liked the IRA trolls, at least the people who engaged with them. That's true on both the left and the right, at similar rates.

Why make friends instead of fight? Engagement! Turns out that you catch more flies with honey. No matter how you measure engagement (RTs, Likes, Followers, Mentions), on days in which Trolls get a high (but not 100%) share of supportive mentions, they get more engagement.