,
25 tweets,
14 min read
Read on Twitter
[Thread] Evening to those following #BrexitShambles. In part 2 of my @BorisJohnson Tweet Analysis (I absolutely refuse to use the term 'deep dive'), I have analysed approximately 900,000 of Bojo's Twitter followers. Why you ask? Well who knows what you may find... #HouseOfCommons
@BorisJohnson Of the 900,000 followers analysed (he has around a million), around 53,000 joined Twitter in July and August 2019. That's around 6%. You can see a clear spike at that time on the graph, well above the average no. created per month, and one standard deviation. What does this mean? 
@BorisJohnson Remember, the graph shows when people JOINED Twitter, not FOLLOWED Boris Johnson. If this was a sample of Tweets in a hashtag, it would be highly likely that most of these accounts are fake followers or bot accounts, designed in this case to boost Bojo's follower base, and ego
@BorisJohnson Of course it may also be that people, enraptured by Boris Johnson's charisma, and the state of British politics in general, decided to open Twitter accounts in droves in order to popcorn the hell out of Britain's inevitable apocalypse. This is possible, but let's look closer...
@BorisJohnson Below we can how many Twitter accounts were created per day between March 2019 and August 2019. This is the period preceding and following Boris's campaign to be PM. There is a gradual increase of new users joining after Johnson announces he will run in May. However, there is a
@BorisJohnson a large jump in new users joining Twitter on the day/the day before Johnson announces that Britain will leave the EU by the 31st with or without a deal. You can see on 22 July, 728 new users joined Twitter, while on 23rd this jumped to 1119. That is a substantial increase - Why?
@BorisJohnson It could be that Johnson's escalating rhetoric about a no-deal Brexit put the PROJECT FEAR POPCORN voyeurs into a Twitter-joining frenzy...After all, who doesn't want to see the Union break apart? #BrexitShambles
@BorisJohnson Alternatively, we could hypothesize that there may be something nefarious at play. Could it be that once Johnson committed to a no deal Brexit supporters of a hard-Brexit mobilized their disinformation campaign to increase Johnson's following and promote his #LeaveOct31 tweets?
@BorisJohnson Conspiracy you cry? Well, remember in my last Bojo thread-a-thon, we found that the most cohesive community promoting his #LeaveOct31 tweets were accounts that had either #MAGA or Trump in their name. In fact, around 8% of his retweets were from MAGA accounts.
@BorisJohnson Indeed, MAGA and Trump accounts were the most common biographical descriptors of accounts retweeting Boris's tweets. They even outranked accounts with the word Brexit. Again, remember this is a British politician .... So I scaled this idea up to look at all Boris's followers...
@BorisJohnson Ignoring for a second the miscoded emojis below, we can see that MAGA and Trump still outrank Brexit in follower biographies. UK is up there - understandably (Weirdly, music, lover and student are some of the most frequent words)....But of the approx 900,000 followers analysed,
@BorisJohnson how many can be considered hardcore pro-Trump accounts? Well let's examine only those accounts that have any biographical data (i.e remove all the accounts that just have blanks for their bio). That comes to around 503,000. Of these, around 14,000 accounts are pro-Trump
@BorisJohnson By pro-Trump I mean any account containing either #MAGA #QANON #TRUMP and a few other specifically pro-Trump phrases. Now, 14,000 may seem like a small amount, after all, it's only around 3% of 503,000 - but this is why it's actually interesting. I shall explain why...
@BorisJohnson First we need to make a distinction between those who follow with little interaction and those who actively retweet those that they follow. It would be fair to say that those who actively retweet are more interested in promoting and sharing specific messages...
@BorisJohnson Now let's assume that those who retweeted a tweet by Bojo were a representative cross section of Bojo's followers. This would mean that of any one of his tweets, approximately 3% of retweets might be expected to be pro-Trump accounts. HOWEVER...
@BorisJohnson We also know that people retweet based on interest, agenda, or otherwise. What is striking though, is that I've done some updated analysis on one of Boris Johnson's recent set of tweets. To be specific, it's this tweet.
@BorisJohnson Because Boris Johnson can't be trusted to run his own Twitter account, he has hired someone to say the exact same tweet four times, on Aug 14, 20, 12 and today. The sample I ran returned around 9000 retweets of this text. Now remember, we might expect 3% of those retweets to be
@BorisJohnson from pro-Trump accounts on the basis that retweets are representative of the follower demographics. But this is not the case.... #BrexitShambles
@BorisJohnson If we remove the null values from the 8000 or so retweets, we are left with around 6500 accounts that have biographical data. Of these 6500, around 1500 can be considered seriously pro-Trump accounts. That's a whopping 23%. Unsurprising, if we look at the corpus data, we see that
@BorisJohnson the most common terms in the biographical data of accounts that retweet Bojo's #leaveOct31 tweets are 1) MAGA, 2) Trump, 3) Love 4) Conservative 5)Brexit . Indeed, MAGA appears twice as many times as Brexit - (997 compared to 467). That's quite astonishing really.
@BorisJohnson Essentially the takeaways can be summed up as: While only 3% of all Boris Johnson's followers might be pro-Trump accounts, a disproportionately high number are invested in promoting his no-deal message on Twitter. In his most recent tweet, 25% of retweets were pro-Trump accounts!
@BorisJohnson Boris Johnson's no-deal messages and #LeaveOct31 tweets are essentially a honeypot for suspiciously right-wing pro-Trump accounts. By suspicious, I mean accounts with little biographical diversity tweeting largely the same message.
@BorisJohnson And to be clear, these accounts are not the unusual spikes we saw in July and August. Those spikes might be attributable to fake followers designed to boost Boris Johnson's followers during his leadership and no-deal bid. The pro-Trump accounts are a separate beast.
@BorisJohnson So yes, this adds to my previous suspicion. There is inorganic Twitter activity from pro-Trump accounts pushing for a Brexit by 31st at any cost. Data Fanciers can download Bojo's followers here: drive.google.com/open?id=1y46ht… Retweet data here: drive.google.com/open?id=1C0Rn8…
Thanks for listening folks! I'm gonna catch some 😴😴
