No, Secret Russian Agents Probably Aren't Behind 'The Last Jedi' Hate

The Last Jedi
October 2, 2018

Some people don't like Star Wars: The Last Jedi. This won't come as surprise to many of you: the film currently has a "Rotten" audience score on RottenTomatoes despite good critical reaction, indicating that at least 100,000 people went to site specifically to complain about a Star Wars movie. The Free Beacon's in-house critic didn't like the movie. But evidently, the loud and public backlash to the movie was actually a Russian plot!

The Hollywood Reporter has a shocking piece out today: "'Star Wars: The Last Jedi' Negative Buzz Amplified by Russian Trolls, Study Finds." The findings spread quickly after being picked up by the Drudge Report: "A Study Says About Half Of ‘The Last Jedi’ Haters Online Were Actually Russian Trolls," reports UpRoxx, Comic Book Resources writes that "Russian Bots May Have Derailed The Last Jedi," Business Insider writes that "A lot of the criticism of 'Star Wars: The Last Jedi' actually came from Russian trolls and bots."

This surprised me, given that I dislike The Last Jedi AND the Russians. Was it true? Had I been taken in by the Ruskies? I decided to go to the actual study from USC PhD student Morten Bay, titled "Weaponizing the haters: The Last Jedi and the strategic politicization of pop culture through social media manipulation."

First, it's important to note that study didn't actually study general tweets about Star Wars, given how many thousands of those there are in a given day. Instead Bay chose to study "tweets addressing The Last Jedi director Rian Johnson directly. Using Twitter’s Advanced Search functionality, I retrieved 1273 tweets tweeted directly at the director’s Twitter handle, '@rianjohnson' between December 13, 2017 and July 20, 2018."

This is probably not the best metric to draw broad conclusions about Star Wars Twitter or fandom in general (and Bay admits as much). Just from using the platform casually, in my experience tagging people in Twitter is pretty uncommon and "@-ing" someone is a method of communication disproportionately used by trolls who get their kicks knowing that you get a notification of their message. The average Twitter user tweets something like "I don't like Star Wars," the average troll tweets "@rianjohnson you ruined my childhood you overrated hack."

From there, Bay went through each tweet individually and determined whether the tweet is positive, neutral, or negative. He then went through the accounts that sent the negative tweets and categorized them as "Political agenda, Troll/Sock Puppets/bots and Real fantagonists" based on their other Twitter activity.

First of all, that strikes me as unnecessarily binary (or trinary, as it may be). It's perfectly possible that someone is a legitimate Star Wars fan who disliked the film AND a right-winger. For my part, I stated my dislike of The Last Jedi after a midnight showing and before I could even begin to imagine it would become a right-left political issue. But Bay's analysis would assume I had a political agenda for hating the film.

My second gripe is that individual coding of tweets requires a lot of subjectivity. Take for example what Bay forwards as an example of an "anti-homosexuality message" that would have a user coded as having a "political agenda": "a user tweeted to comedic actor Seth Rogen after he reacted negatively to the ‘Remake The Last Jedi’ project: 'You know how I know you're gay? You don't want to remake last Jedi.'"

Uh, that's a reference to a running joke from The 40 Year Old Virgin, in which Rogen and Paul Rudd's characters jokingly trade insults about how they know the other is gay for increasingly absurd reasons, including liking Coldplay and making spinach dip. It's hard to believe that in context the random Twitter user was seriously accusing Rogen of being a secret homosexual or that he actually had any sort of political motivation.

Moving on, Bay takes the "Troll/Sock Puppets/bots" subset and singles out Russia accounts ... kind of. Contrary to the certainty of the news reports based on his report, Bay notes you can't prove the accounts are Russian, he suspects they're Russian based on certain clues: they use "auto-generated handles consisting of a very common, English name followed by a series of seemingly random digits," they had "not uploaded a profile image," they "have an extremely high retweet rate," they go dormant and delete tweets, they tweet during work hours in Russia, etc.

That isn't air-tight of course. An English name followed by random digits is also the handle you get automatically assigned upon signing up for Twitter. That and a lack of a profile picture are "a combination which ... [are] a typical characteristic of Russian troll accounts," Bay writes, but it's also typical of someone who just signed up for the platform. (Indeed, as I was sharing the study on Twitter last night, a follower "@Tiffany44541244" complained "The number of ppl who think numbers in your twitter handle (assigned by twitter) = Russian bot is insane." If she's a Russian, she hides it well.)

But set all that aside. Lets assume tweets directed at Rian Johnson are representative of broader Star Wars fandom, that Bay's judgment is unimpeachable, and the the accounts he fingers as Russians are just that. How many Russian trolls did he find?

Using the Botometer mentioned in the Method section, 11 out of the of 206 accounts expressing negative sentiments were identified as bots. I identified 33 of the 206 negative accounts as trolls and/or sock puppets... 16 of these 33 troll/sock puppet accounts appear to be Russian trolls, or at least possess several of the Russian troll characteristics presented above.

Sixteen. Of over a thousand tweets directed at one individual over an eight-month period, of 206 accounts who didn't like The Last Jedi, of the 33 accounts who appear to be trolls, sixteen MIGHT be Russians.

Just while writing this piece, I searched for "Rian Johnson" in Twitter's search feed to see how many negative tweets about The Last Jedi popped up. There are about sixteen in a twenty-minute span. In the grand scheme of things, we're talking an absolutely minuscule amount of semi-verifiable Russian activity. The articles suggesting there's evidence Russians manage to "amplify" hatred of movie are drastically overstating things, the articles claiming "half" of Last Jedi hate is from Russians are simply flat-out wrong.

Were there Russian accounts spreading hate of The Last Jedi? That's believable and consistent with what we know. Is there concrete evidence they had an appreciable impact on the larger debate over the film? No, not all. Sadly, those guys you see on social media saying they didn't like your favorite Star Wars movie are probably just people didn't like your favorite Star Wars movie, not Russian operatives.

Published under: Russia , Star Wars