TR Member Perks!

Updated Editor’s Note 11/7/2017 – In an effort to further commit to our editorial vision of quality content about nothing but games or the industry, we are leaving this note here to let you know that this article does not meet the standards of that vision as it exists today. This article may be poorly written, or it may be well-written but with charged political content, which we have stepped away from. It’s not the ideas we have a problem with, as we do not discourage any viewpoint, we are just moving away from this sort of content. This article no longer represents TechRaptor’s editorial vision today and into the future. You can read more about why we are doing this here.


Full disclosure: Both @The_Nimbus_ and I self-identify with Pro-GamerGate leaning. However, the sample tweets used in this article are picked completely at random, and not cherry picked to agree with our bias. This is so that readers can evaluate the validity of the study for themselves. All twitter handles have been removed from tweets for privacy. For the purpose of this study people are categorised as giving a Pro-GamerGate or Anti-GamerGate stance. While grouping this way is problematic, it is helpful in this context.

Some of you may have seen the Newsweek piece which broke down tweets sent to various GamerGate figures using BrandWatch, broken down by harassing, supportive and neutral tweets. This study found that around 85-95% of tweets using the GamerGate hashtag were neutral.

However, as this survey was carried out by a computer, it is hard to truly analyse the tweets. For a closer analysis we need a human eye. Luckily, @The_Nimbus_ has been working for the last 18 days (as of writing), studying 100 random tweets a day using the #GamerGate hashtag and breaking them down into 10 categories. These were the findings: (Click to enlarge)

Analysis of #GamerGate on twitter

I wanted to know more about the ambitious project and so spoke with the twitter user, who self-identifies as leaning pro-GamerGate. When asked why he began to investigate tweets he stated he “Saw a need” and “Just wanted to do something useful.”

A pastebin with all the tweets used records the categories tweets fall under, how they are defined and a list of every tweet and their category. Unlike other unverifiable studies, this has helped immensely in ruling out potential bias. However, as this is merely one person’s project and only samples a small amount of data it cannot be said to represent all Twitter activity, but is a good cross-section of the phenomenon.

As part of this study, harassment is defined as “a tweet with the intent to intimidate, insult, or silence any individual or defined group.” where if “it could be expected the person/group discussed would not see it, I did not deem harassment” Tweets which contain harassment from Pro-GamerGaters account for less than 2% of tweets and include:

“Back to #BriannaWu for a sec. Gamers should treat this bully, and she is a bully, like all other bullies: Don’t play w/her. #GamerGate”

“such maturity, it’s awe inspiring. #gamergate #NotYourShield” 

“but at least you are a good little zealot. please do more of it, so the rest can see how Anti #GamerGate rs are like.”

Harassing tweets from Anti-GamerGaters make up over 7% of tweets and include:

“every day #GamerGate has less credibility.”  

“Maybe if I use #Gamergate I’ll be an enlightened crusader for fair journalism that spouts bane memes and constantly use the word falseflag”

“The anti-woman hate-group known as #gamergate posted Felicia Day’s personal details online less than 1hr after her blog post on the subject. “

Luckily, there were only two tweets inciting violence. Both came from Anti-GamerGaters.

“#StopGamerGate2014 all gamers should be put to death. every single one”

“If you support #GamerGate, die. You don’t belong in the new world we’re going to create regardless of your terrorist tactics. #GetOverIt”

Polite debate over the topic of GamerGate, is listed under the debate category. Over 15% of tweets fall under this category. Some tweets include:

“I never brought up Eron. Patricia’s relationships are not “hearsay,” she explicitly admitted to covering an ex’s games. #GamerGate”

“You didn’t really challenge #GamerGate”

“Anita isn’t really part of #Gamergate’s crosshair, but she is mostly a obstacle right now. Same with Mrs. Wu.”

Tweets which self-promote or were tweeted by known trolls were grouped in the same category. Trolls were classed as accounts which “posts covering semi-random topics and all of them derogatory (racism, sexism, gamergate, religions, etc)” Self-promotion and trolling makes up 5% of #GamerGate tweets, and include:

“#gamergate defies parody. …”

“#FIFA15 #gamers #GamerGate”

“actually it’s about ethics in south africa plier journalism #gamergate”

Tweets about ending harassment and games journalism ethics were grouped together as @The_Nimbus_ identified these 2 categories as the 2 key goals of the #GamerGate movement. Over 31% of tweets fell into this category and include:

“Because a corrupt press uses it’s power to promote lies & fear monger the public must stand up to resist their abuses. #GamerGate”

“You do know that targeted idiocy like this is why people believe #gamergate is about harassing specific women right?”

affects sales. Sales affect developers. Developers affect the games you play. #GamerGate”

There were several tweets used to boost the morale of those involved in the #GamerGate scandal. These make up less than 6% of tweets. Some include:

“Remember Gamers, stay positive! Games are good. Games bring people together. Everybody is a gamer! #GamerGate”  

“Day 1468, still no signs of #GamerGate slowing. If only these journos had adopted a code of ethics we would have been saved.”

Those who support the #GamerGate movement, boycott websites they feel are corrupt, and email the advertisers of these websites to alert them to this perceived corruption. Tweets discussing this make up almost 3% of tweets and include:

“If #GamerGate stops emailing and informing, and continue waiting for the next Happening, then GG is nothing more than a spectator sport”

“some drama between ?leaders? No problem, I got my emails weaponised and firing in righteous cause #gamergate”

Finally, other tweets are simply tweets which do not fit into these categories. This can be anything from a thanks, to asking for more information or stating opinion. This is the majority of tweets at almost 36%. A small sample includes:

“if you read @- feed they claim he said Blizzcon was the opposite of #gamergate is that what he said?”

“#GamerGate song of the day …”  

“#Gamergate really IS the digital world’s ISIS!”

“just weird to see #GamerGate on the defensive over identity pols. “We’re not all white men!” as though something wrong with white men”

“October 30, 2014. Fullmac becomes self aware.”

I have a confession to make, anti-#GamerGate, call me names now but….. I killed Hrist in Valkyrie Profile. And I’ll kill her again! >:3″

“Ok seriously anti-#GamerGate could you please just… would you kindly stop.”

“In truth, given how many games are decried as sexist in spite of more women becoming gamers, their analysis is wrong. #GamerGate”

This is a random selection of tweets. For the full list of the 18 days of tweets which I examined visit here. You can also follow @The_Nimbus_ for daily updates on his findings.

I do not wish to analyse this data further as I wish people to read the data and decide for themselves whether a.) it is trustworthy and b.) what it means. However, for me personally it shows what I long suspected from my own observations of Twitter, that the vast majority of harassment on Twitter comes from those who oppose the movement, rather than the advocates as the mainstream media has suggested. Finally, it tells me that, while jokes are often made, #GamerGate is at least 1/3 about ethics in games journalism.


Georgina Young


British girl, currently in Japan. Surviving on a diet of retro games. Worshiping the god that is the Sega Megadrive. I like Nintendo.

  • Patrick Howard

    great study and article 🙂

  • Kelly Maxwell

    i never really cared for a numbers based analysis on people’s feelings. How do u even decide the difference between sarcasm and jokes vs harassment and malice? I don’t think the algorithms are this precise. Not even People are always sure when someone is joking on twitter. Not poo poo ing the article though, just the role statistics have played in trying to pin down intent.

  • Zepid

    No shock that the Anti-GG crowd harasses more. Leave it to cold logical machines to shed light on the reality of the situation.

  • Liz

    Yes, it’s hard to be truly random unless you’re using the Twitter API. Most likely, these Tweets are taken from specific individuals who are either followed by the author, whose Tweets they saw because they were RT’d by a follower so they were in their timeline or they went out to see the Tweets of a specific person whose views they wanted to include.

    It is actually very difficult and or expensive to be a true random study unless you want to collect data via the API. There is always a human decision on what to include and what to exclude.

  • Kelly Maxwell

    one thing that hasn’t been explored much are the casual -isms that our opponents utilize daily. It’s tiring to hear the refrain ‘well, one or some ppl did this (usually w/o evidence) & if they did; that means thats what they are about.” If you said this of a country, a gender, a race, or a species of animal, most people would look at you as if you were a loon. Yet somehow it’s “ok” to indict entire websites or hashtags in this manner? Preposterous.

  • Guest

    Love being the digital world’s ISIS. Means we get this awesome song.

  • GDI

    Well I’ve seen pro-ISIS in my country’s Yahoo comments accusing Christians of kidnapping their girls and turning them into prostitutes. They’re projecting accusations they’re guilty of themselves just like anti-GG.

  • Kain Yusanagi

    Most likely went to the hashtag’s timeline and did a sampling of the tweets involved there. Assuming it was tweets & retweets of someone they follow assumes they are completely incompetant without any evidence or reasoning for your claim.

  • BlueLight

    Smile, lol, /s are the standard ways to stating what’s said is a joke.
    If i were to say “I hope you die laughing Kelly.” then according to Poe’s law, then there is no reason not consider this harassment or malice. It is up to me the author to make sure you understand this is a joke. For instances “I hope you die laughing Kelly :-)” it is clear this is meant as sarcasm or a joke.

    This does not make bad jokes acceptable. For instances if i were joke with “I hope all blacks die! LOL” i would hope people would call me out on it.

    This all being said, it’s really hard to do any number crunching on these concepts.

    Do note all forms of sarcasm, jokes, harassment, or malice featured in this post were used as examples for educational use. These statements do not reflect the authors views on anything besides what is consider a joke vs harassment.