The Brexit filter bubble on twitter in 2016 vs 2018
I know these type of flock diagrams are boring, but bare with me for a second, because this one should interest you.
What can be seen in the June 2016 snapshot is that there is little interaction between the remain and the leave bubbles on twitter. What is also noticeable is that the leave one is a lot more engaged with external bots constantly feeding information into the bubble.
In the light of Cambridge Analytica revelations this makes perfect sense. The goal was to motivate leave voters to go out and vote while keeping remain voters at home. Inside the remain bubble everything was fine, nobody takes these clowns serious, and remain is going to win anyway.
And this is exactly what opinion polls had a problem with. The FT analysis the day before said that they know who is going to vote how, but it is hard to predict who goes out to vote. If turnout is low, leave would win, as they are more motivated to go to the polls. If turnout is normal, remain would win. If it is unusual or very high, leave is likely to edge it. On the night we have seen queues in front of voting stations of districts where turnout during general elections is usually very low.
Lets fast forward to January 2018 and look at the same twitter flock now. What we can see is a much more engaged dialog. Not only are there now a lot less bots on both sides, but they also seem a lot less influential, either due to twitter policies or because the campaign simply stopped. The remain argument clearly has the upper hand now, which is hard for the users in the remain bubble to grasp though. Remember that in the past it always looked like that. For leave voters things must looks very different though as remain arguments making it into their bubble on a broad front. One needs to be very deep in the leave bubble to not have seen any negative posts recently.