How is GOP congressional candidate @Barge4Congress (GA-14) getting hundreds of retweets with only 152 followers, and why do his tweets garner more retweets than likes? #TuesdayThoughts #Lobsterfest
cc: @ZellaQuixote
cc: @ZellaQuixote
Also, why is 96% of the amplification of a Georgia congressional candidate's tweets coming from accounts claiming a location other than Georgia? (Accounts with empty or fictional locations excluded.)
One possibility, and we think it's likely: #MAGA retweet DM rooms. In an interesting coincidence, 269 of 421 accounts (63.9%) that retweeted @Barge4Congress are mutual followers with one or both of the accounts recruiting members for #Mighty200+ retweet rooms.
We looked at these accounts to see if any of the other tweets they retweeted received far more retweets than likes, and found 13699 tweets from the last two months with at least 150% as many retweets as likes. (Tweets with fewer than 10 retweets were excluded from this analysis.)
For the sake of comparison, most tweets get roughly 37% as many retweets as likes, based on a quick scan of 45 million tweets from various datasets we've accumulated over the years, so the 150% ratio seen here is. . . disproportionate.
The accounts doing the retweeting-without-liking are mostly random #MAGA accounts with 5K-50K followers, as are the accounts most frequently retweeted. Things get more interesting when we look at the verified accounts being amplified; almost all are GOP congressional candidates.
Is this more-retweets-than-likes phenomenon a fingerprint of the use of coordinated retweet rooms? To test this hypothesis, we took a peek behind the curtain in a #MAGA retweet room and found that most tweets being amplified indeed received substantially more retweets than likes.
We likewise found that most of the verified accounts whose tweets were being retweeted by the room were in fact congressional candidates. Here is a sample, along with the ground rules of the retweet room, which are consistent with the retweeting-without-liking behavior.
Some examples of tweets from datasets we've recently studied that have substantially more retweets than likes, and may have been amplified via these DM rooms.
