Rape Culture and Twitter Abuse: Visualising Twitter’s reaction to the Ched Evans Rape Case

Rape Culture and Twitter Abuse: Visualising Twitter’s reaction to the Ched Evans Rape Case

Introduction and Discussion

On Friday 20th April 2012, Sheffield United striker Ched Evans was jailed for five years on Friday for raping a 19-year-old woman at a hotel near Rhyl, Wales.

An extract from an article by The Guardian’s Amanda Bancroft summarises the controversy and general sentiment surrounding the events:

On Friday afternoon, the Sheffield United footballer Ched Evans was sentenced to five years in jail for raping a 19-year-old woman. His co-accused, Port Vale footballer Clayton McDonald, was found not guilty.

Almost immediately, the #ChedEvans hashtag appeared on Twitter, and later, #JusticeForChed. Some tweets questioned why one defendant was found guilty and the other not. Others blamed the victim, particularly focusing on her being drunk; some could only be described as vile. Tweeters included fellow Sheffield United footballer Connor Brown who has since deleted his attack on the victim, which included calling her a slag, and which intimated she made her complaint for financial reasons.

Rape culture, which includes victim blaming, sexual objectification and trivialising rape, was demonstrably alive and well on both hashtags, and continued all weekend.

Although the Twitter community had, since the hashtags started, attempted to correct the views of supporters of Ched Evans by reminding them that a jury had heard the full facts; that victim blaming was never right; that alcohol consumption can and does render a person incapable of consent, the community rounded even more so once the tweets started including the alleged name of the victim.

This is despite the victim being entitled to lifelong anonymity under Sec 1 of the Sexual Offences (Amendment) Act 1992.  While police have warned/made arrests following comments on Twitter which identified his victim, this also brought about an outcry of dismay and disgust at what is percieved as a misogynist purveyance of rape culture and abusive behaviour towards rape victims on Twitter.  The Huffington Post reported:

…rape charities spoke of their disgust at the online abuse.

Fiona Elvines, from the Rape Crisis Centre South London, condemned the individuals who had revealed the victim’s name, saying that it was “really, really worrying that people aren’t taking this seriously.”

“This happened while the woman in question was recovering not just from the rape, but from the difficulty of the trial, which survivors of rape often call a second form of abuse.

“It’s not just in the individual’s interest that anonymity is maintained but in the public interest. Anonymity helps put rapists behind bars.

“It’s because of rape myths and the social stigma that survivors of rape are given anonymity.

“[The comments on Twitter] are creating a cycle where survivors of rape are being blamed. The CPS need to send a really strong message out, that you can’t just break the law because you think you are supporting your football club.”

Her words were echoed by Vivienne Hayes, Chief Executive of the Women’s Resource Centre who told The Huffington Post UK she was “appalled” that the woman “who suffered such a horrific crime, has had her name published on social media sites and smeared in such a repulsive manner.

“Rape is a serious crime with severe consequences for the victim, for this to happen is akin to being attacked all over again.

“Rape convictions in this country are far too low, and what women will want to come forward if this is the fate they will suffer?”



In a reflection of these concerns at the time, I also came across a thought-provoking article called ‘Misogynistic Twitter Bile About Ched Evans Case Shows Law Needs to Change‘. Katie Russel of the Huffington Post discussed the issues above,and how in a similar way to the Liam Stacey and Hunger Games Twitter Controversies, concerned citizens started to take it upon themselves to document the abusive tweets that were being made against the victim, so that should a prosecution occur and those tweets were in the meanwhile destroyed by their original posters, there would be evidence of the abusive messages.

The Tumblr blog discussed “Little Tweets of Misogyny” – Archiving the ugly side of Twitter” has now decided that it has fulfilled its purpose of standing testimony t the negative abuse aimed at the teenage rape victim in this case, and has closed down the site with the message below:

Ched Evans Rape Controversy: Little Tweets of Misogyny" - Archiving the ugly side of Twitter

*** Update***

On 5th November 2012, Nine people have each been told by a Magistrate’s court in Prestatyn, Wales, to pay £624 to the woman raped by footballer Ched Evans after they admitted naming her on Twitter and Facebook. As reported by the BBC, the former Sheffield United and Wales striker was jailed for five years in April for raping the 19-year-old. The law grants victims and alleged victims of rape lifelong anonymity. Men and women, aged between 18 and 27 from north Wales and Sheffield, have been charged and accused of revealing the victim’s identity. They were all charged with publishing material likely lead members of the public to identify the complainant in a rape case, contrary to the Sexual Offences (Amendment) Act 1992. Magistrates heard the victim was accused of “crying rape” and “money-grabbing” on the social networks. The nine who pleaded guilty claimed they were not aware naming her was a criminal offence.



Having read the initial articles and more tweets referring to #chedevans, I decided to investigate this case a little further and scope out the sentiment around it. Below is a Twendz mine of sentiment around the controversy on 24th April 2012.  It shows that sentiment on Twitter was pretty divided, with 31% negative sentiment around the hashtag #chedevans, 44% neutral and 24% positive.  This is broken down even more interestingly, with an 88% neutrality over the arrests made of twitter users who outed the teenage rape victim.  100% negative sentiment over the term “damage” – a frequent term used in articles discussing the damage being caused over the perception of rape victims over the outing of the teenager in this issue, and to any confidence rape victims will have in the English justice system if this is not handled correctly by the powers that be.

In relation to the naming of the rape victim in this case, Twendz gives a somewhat suprising result of 65% neutrality and 24% positivity. Objectively, it would seem that the torrent of abusive tweets towards the victim and support by some twitter users of the actions of Ched Evans and the group with him accounts for the 24% positive sentiment – versus the increased media reporting, which may be translated by the Twendz sentiment algorithms as neutral, would explain the 65% neutrality result. However, my subjective and outraged reaction to just how abhorrent this example of misogynistic rape culture and apologism has reared its ugly head in this case, leads me to be a little disappointed at these and the other forthcoming results.

The term Rape, has a 20% negative sentiment versus a 40% sentiment for both positive and neutral responses.  Whereas again, rather disappointingly, the term Victim has a 68% neutrality, and a 16% negative and positive sentiment.  On reflection, upon reading the broad spectrum of tweets, blogs and news articles, I really shouldn’t be surprised at the outcome of the Twendz sentiment analysis, as the debate at the time of writing is very strong, with incensed actors at each end of the pro and against spectrum, something that you can see just by reading down the screen-grab of tweets that were running along side the sentiment analysis results scraped from Twendz:

Chad Evans Rape Controversy: Twendz Overview

Mapping the Twitter conversation network

Stepping back into an objective frame of mind and wanting to do my own visual analysis of this evolvin controversy, I decided to use Hawksey’s Google TAGS spreadsheet to get a reatime scrape of the Twitter conversations that were occurring around the #chedevans hashtag.  My initial scrape gave me evidence of just how topical this issue was when immediately, Twitter outputted almost the maximum feed of 1500 tweets, most occurring within the 2-day period from Monday 23rd to Tuesday 24th May 2012.  A summary of the scrape below shows that there were quite a number of top conversationalists, each tweeting between 20 and 6 scrapes at the higher end of the spectrum, all the way down to between 3 to 1 scrapes over the course of 48 hours:

TAGS Scrape:

Ched Evans: Twitter scrape summary

Pie chart of Top Tweeters:

I used the above data to create a Google Analytics pie chart of the activities of the top conversationalists in this controversy below.  What was most interesting was prominent the cluster of top conversationalists are.  You can see in the different shades of red and blue, the small number of Twitter users who have tweeted up to 20 times or more around the same topic, in comparison to the much larger group of Twitter users, some 93% who had a much lower tweet frequency rate, having tweeted just once or twice about the issue:

Top Tweeters in the Ched Evans Rape Victim Controversy

However, as visual as the pie chart above is, I was still bugged by not being able to more interactively show the how the key actors in this evolving debate were tweeting. Then I remembered an animation I had seen a link to on a “posts of interest” update on Martin Hawksey’s blog back in February 2012.  Hawksey noted Jim Vallandingham’s innovative use of the D3.js JavaScript library to create bubble charts that visualised data showing the education-based donations the Gates Foundation has made. Vallandingham has written a very comprehensive post on how he re-wrote the original New York Times model of a bubble chart of the proposed budget for 2013 in the CoffeeScript language, and also very helpfully posted his visualization code is on github. I decided to see if I could test my JavaScript hacking skills and utilise this CoffeeScript version to visualise the Twitter conversation going on around the Ched Evans rape controversy.

Creating an Animated Bubble Chart in D3

Below is a screengrab of Jim Vallandingham’s orignal bubble chart:

Original Vallandingham D3.js Bubble Chart


My first task would be to hack the code to pull the data from my own Twitter scrape, and programme the animation so that it visualised clearly the different Twitter users, their different tweet frequency rates, and the number of tweets they had made.

Programming in CoffeeScript, an interesting exercise in force and gravity

The first thing to do was to match the node objects to the names of the header columns in the .csv sheet.  You can see this from lines 52 to 56 below:

Chad Evans Rape Controversy: Coffee Node Objects

Then I had to work out the gravity and force layout and directions that the animation would pull the nodes in.  See lines 118 to 147 below:

Chad Evans Rape Controversy: Coffee Node Force and Gravity

After that, I worked on the colours and of the different node categories (see bubble chart and lines 31 – 34), and the rate at which each mode would change and darken when a user hovered their mouse over it (see line 86, and how I ‘brighten up’ each node, but then command that when a users’ mouse hovers over a node, each one darkens down at a rate of 5 to the initial brightness):

Chad Evans Rape Controversy: Coffee Node Colours and Hover Styling

Then I programmed the wording that would appear in each tool tip as the users hovers thier mouse over each node (see lines 182 – 185):

Chad Evans Rape Controversy: Coffee Node Tooltips

One one of the final things I did was programme the positioning of the nodes when seperated into their “High”, “Medium” and “Low” tweet rate frequencies:

Chad Evans Rape Controversy: Coffee Node Display Frequency Positioning


Results and Conclusion

I have embedded the resulting animation in an iframe below.  What is good about it, is that as well as showing the large conversation network surrounding the Ched Evans rape controversy, it allows you to select individual nodes to see the main actors, and, upon click on the “Tweet Frequency” tab, allows you to further interact with the data and separate out the small group of top conversationalists from the larger cluster those who tweeted about the issue just once or twice.

Click on the “Top Tweeter” and “Tweet Frequency” tabs to explore the conversation network:

Click on the image below to launch the animation so you can see the force and gravity effects from start to finish:

Ched Evans Rape Controversy: Bubbles Chart Animation

At the time of publishing this post, the web sphere is alive with commentary on the ongoing Twitter storm, arrests of the Twitter users who illegally identified the rape victim, the controversial actions of Evans’ club, the emotional distress of the victim and her family, and the move by Evans and his legal team to appeal against his conviction:
Topsy News Feed: Ched Evans Rape Controversy (April 24th 2012)

In all of this, as objectively interesting as the technical and visual analysis of this controversy has been, my subjective thoughts tend to agree with The Guardian’s Amanda Bancroft:

While it may be without doubt that those who used Twitter in an unlawful way over this issue should be punished, and it is fair to say that the law is constantly being tested in its application in our new media age, what this weekend has demonstrated is how alarmingly alive and pervasive rape culture is. Isn’t the biggest question what we do about that?

It also brings me back to the on-going issue of the #ididnotreport hashtag and the inherent culture of “We do not Believe You” behind it.  In light of these and other related events, and in agreement with Julian Norman, it would seem that #ididnotreport is:

A very necessary campaign, because one of the ongoing problems with reports in the criminal justice system is that they are not believed.

Where it’s easy for commenters to dismiss one woman as a liar, it’s less easy to dismiss thousands of accounts.

While it seems that digital social networks can and have been used to abuse others, one of the more positive things to come out of this study is just how much other users are determined to pro-actively document this abuse so that it can eventually be used to prosecute abusers. Tom Phelan argues in his paper on How Social Media is Shaping Crisis Communication:

Social media has revolutionized how people and organizations receive and transmit news
during a crisis situation. It benefits crisis communication in that it, “… allow[s] for
wide-scale interaction between members of the public that has qualities of being collectively
resourceful, self-policing and generative of information that cannot otherwise be easily obtained
Sutton, Palan & Shklovski, 2008) [this] sets the precedence for today’s need for instant information. It
inherently provokes transparency in communication in near real time to be disseminated to
stakeholders. It prompts discussion, debate and feedback from those that care about the crisis”
(Prentice & Huffman, 2008).

This can also arguably be applied to users’ self-policing and self-regulation of abusive and potentially illegal behaviour on social networking websites. While some may behave in a negative manner, it has been heartening to see that others have been resourceful enough to gather and supply information to the authorities.



  1. Pingback: Visual Analytics: Comparison of @SCOREProject and @UKOER (and template for making your own) JISC CETIS MASHe

  2. Pingback: Her Master’s Voice? Misogyny in (anti) social media | Women's Views on News

Comments are closed.