Trying to predict the outcome of today’s UK election? Spent a last frantic few moments comparing the proposals from the various candidates or parties? Interested in understanding the demographics of the voters? Wondering whether there’s a correlation between a constituency’s average income and number of Tory votes, or its diversity and UKIP support? Certainly maths is everywhere and we could spend hours and hours analysing, modelling and predicting – as the newspapers and numerous other blogs have been doing for the past months.
Here at chalkdust, however, our curiosity was piqued by something (only slightly!) less controversial and certainly a lot more fun: determining the similarity between the Twitter accounts of two of the main protagonists of the 2015 election, Ed Miliband and Nigel Farage, who live on opposite ends of the political spectrum.
Adding up the number of accounts being followed by Ed and Nigel, we get a total of $1,899$ Twitter accounts, with $62$ of these being followed by both, including, perhaps bizarrely, the liberal-leaning newspaper The Guardian. This gives us a following similarity of only $62 \div 1,899 = 0.032$, which isn’t very similar at all. What does that mean? It means that every time they look at their Twitter accounts their newsfeed would look completely different: they have a different perception on what is being said in the big wide world of Social Media.
If we focus on who is following them, we get a very similar result. Although a far more active Twitter poster than Ed, Nigel has less than half the number of Ed’s followers. We will leave it as an exercise to the reader to draw conclusions from the two facts in the preceding sentence; but we will have a look at their popular similarity: that is, how many people follow both politicians. Between the two of them, they have almost $680,000$ followers and from this multitude of accounts, only $402$ follow both. Hence their popular similarity is $402 \div 679,401 = 0.0006$. How shall we interpret this? Well, to put it bluntly, when one candidate tweets, the followers of the other couldn’t care less. There’s only a very small intersection between the followers of both, and they are most likely to be news reporters.
Let’s take the leader of the Scottish National Party, Nicola Sturgeon, as perhaps a more interesting figure of comparison given the potential alliance between Labour and the SNP: she is very active on Twitter and has following similarities of $0.035$ and $0.02$ with Ed Miliband and Nigel Farage respectively; and popular similarities of $0.0002$ with Nigel and $0.0004$ with Ed, both less than that between Nigel and Ed. So a greater proportion of people follow both Ed and Nigel, than a combination of Ed or Nigel and Nicola. We are unsure as to the implications of this for the future of the United Kingdom.
The Following Similarity between the candidates is
|Following Similarity $\times 100$||Nigel Farage||Nicola Sturgeon|
and their Popular Similarity is
|Popular Similarity $\times 1000$||Nigel Farage||Nicola Sturgeon|
There are numerous websites you could use to further explore the UK General Election from this slightly different angle and shed more heavily refracted light into the murky world of politics. We made use of twiangulate.com, and you can read more about how to measure the similarities between Twitter accounts, in Music Playlists, Facebook and Twitter from our first edition.
chalkdust is a maths magazine and takes a completely statistical and unbiased approach to politics, which is a bit ironic really. Any views expressed in this article are therefore entirely factual and we would like to make it clear that we do not endorse Twitter popularity as a criterion for deciding how to cast your vote.
You might also like…
- We chat to Andrea about crime, maths and her fluid career
- How the queue size tells you when the next bus is coming
- Maths and Decision Theory to be the best at playing Guess Who?
- Here is the reason why the golden era of Twitter is gone
- How can we differentiate a function 9¾ times?
- Three ways to obtain and generalise a beautiful fractal