Investigation Into Euro 2020 Racist Social Media Posts

It's taken a while but we have finally completed our full analysis on the post-Euro2020 Social Media Hate aimed at Black England players.

We previously reported on our interim progress here, along with a subset of example data. 

In the aftermath of this event our scraping scripts and scans (the process was explained at the time) scraped several thousand racist posts on various social media platforms.

In addition, as per our request, several hundred were also contributed to our email address.

We collated a total of 8000 such posts. In addition, we collated another 2000 posts aimed at F1 driver Lewis Hamilton the next week, for a working sample of 10,000 posts.

In the light of some media stories about 'the majority of posts coming from outside the UK', we have primarily focused on identifying locations of those accounts which could be sufficiently localised.

Regarding these media stories : we find the opposite. There are several possible or indeed probable reasons for this:

  1. We focused on Twitter and Facebook, primarily, as the biggest platforms whereas most media stories also encompassed Instagram. It is indeed possible that Instagram posts were more International, we cannot say either way. But from the two biggest platforms, Twitter and Facebook, the posts were mainly from the UK, of those we could localise.

  2. The media stories about 'the majority of posts coming from outside the UK' came out quite rapidly after the events, leading us to suspect that not a great deal of actual analysis work had been done on accounts which were more difficult to localise. For Twitter and Facebook, specifically, we have done this work - and come to an opposite conclusion.

  3. Part of the source of 'the majority of posts coming from outside the UK' came from the Social Media companies themselves. They have not shown themselves in the past to have a history of being entirely truthful when it comes to such hatespeech and abuse on their own platforms.

  4. A great number of accounts are not identifiable as to location with sufficient data. It may well be that media stories have made assumptions on the locality of these accounts based on insufficient data, and tended, for whatever reason, to prefer a judgement of 'Non-UK' source. As to why, we can only theorise.

We have collated and analysed all the 10,000 posts, and can report the following totals:

Posts Collated: 10000
Posts Analysed: 10000
Posts Deemed Actionable: 5019
Posts Origins Identified: 1667 (
33% of those deemed actionable) of which:
  • UK Based: 87% (rounded)
  • Europe-Based: 8% (rounded)
  • ROTW: 5% (rounded)
Action Taken: 1064, of which:
  • Action Taken via Police: 175
    (Reported to regional Police Force with original posts and evidence of identity)
  • Action Taken via Employers: 889.
    (Reported to employer with original posts and evidence of identity)

Of the 175 cases referred to police, we have heard officially back from some of them. In many cases we were advised that police action was already in place due to police being aware of the issues via other means, before we informed them.

Our apologies for the time taken to complete this work, but we are all volunteers in full-time employment and we strov to identify as many accounts as possible. At the conclusion, we were forced to write off many accounts as untraceable with sufficient evidence. We continue to monitor such accounts for any future clues as to identify and location.

Our sincere thanks to those who helped to make this happen.

If you wish to contribute, our Paypal link is here :

 
Thank you.

[We are more than happy to discuss our methodology and conclusions with academics or persons involved in social media analysis at any time. Please feel free to contact us on this matter]