Looking for statistics for deaths in videos games broken down by sex.

August 27, 2014 6:40 AM

Everyone knows about the controversy around the exploration of sexism in gaming. I've (foolishly?) waded into the fray, and I am having a really hard time finding any stats regarding what percentage of female characters die violent deaths vs. male characters. I do not want to argue without solid facts. Anyone know if this data exists? Thanks!
posted by Fenriss to Society & Culture (9 answers total)

The only data I've found on female characters has focused on the issue of the gender of protagonists, not acts of violence. The best gender-of-protagonist study I've found is here; it only addresses FPSes (since it is the simplest; what 'gender' are you when you play From Dust or Civ V or Thomas was Alone?).

Basically, it might exist, and Game Studies might be the place to find it, but I can't recall any such studies.
posted by flibbertigibbet at 6:53 AM on August 27, 2014

Are you talking about deaths at the hands of the gamer? It seems that would be impossible to study, given that each game would have different outcomes--e.g., take Grand Theft Auto, with thousands of NPCs walking around. But I'd wager the female deaths are far outstripped by the male deaths, given that male NPCs are much more common in most FPSs like CoD. It's a lot of male soldiers.
posted by Admiral Haddock at 7:06 AM on August 27, 2014 [4 favorites]

It seems as if the only way this question is answerable, as has already been implied, is if you're asking about the gender ratio of "name", or storyline-significant NPC's, being killed as part of the narrative, or alternately (perhaps, but this gets tricky and is also fairly rare) the ability of the player to kill those NPCs outside of scripted events.

Otherwise, Admiral Haddock has the right answer.
posted by Poppa Bear at 7:27 AM on August 27, 2014

the argument is also not about pure numbers, but also the pinup/sexualized nature of how the women npcs are treated in violence and/or death. not a lot of dudes getting murdered in their underwear or raped to further the development of the main character.
posted by nadawi at 7:30 AM on August 27, 2014 [5 favorites]

Best answer: I think it would be possible to get a meaningful statistic, if you set up some definitions before compiling it. You could, for example, say you want to measure the number of characters with a potentially violent death (so that the individual nature of play iterations isn't an issue; the character is counted if it could die that way).

You could also decide to measure only NPCs, or include player characters. I think the NPC number might be particularly interesting, since you could argue that NPCs are more object-like in general.

Of course the basic # male characters would probably be much higher than female characters, so it sounds like the numbers you want is _percent_ of males that die violently (# males / # violent potential male deaths), vs. the percent of females that die violently (# females / # violent potential female deaths).

Finally, you could perhaps take some kind of random sample of games and come up with a good number yourself. However, you'd definitely want to create the random sample using some kind of credible and transparent method, for example a computer script (so that the people who argue with you could look at the code and agree it was random). Make sure the pool of games from which you take the random sample is a relevant one, and make sure you sample enough games to establish some credibility -- all this you probably know, but I thought I'd mention it just in case you don't.

Re: nadawi's comment - yes, it would probably make sense to measure that kind of thing, too. # male characters killed in underwear: 1; # female characters killed in underwear: 48 -- that would be a pretty good point, I'd think.
posted by amtho at 7:55 AM on August 27, 2014 [1 favorite]

I'm not much of a gamer and I haven't touched controversial stuff like GTA, but I had a mad obsession for the Half Life series and it's just blindingly obvious that male deaths vastly outnumber female ones in that, simply because most - actually maybe all - of the enemies are male (or alien). I think most of the "good guys" are male too but in any case the way they die is the same whether male or female. To the best of my recollection this also applies to old games like Unreal, Doom and Quake. I've also played Bioshock and my impression was that (Little Sisters aside) the male/female characters were fairly evenly distributed and, again, died in the same sorts of ways.

I am aware that I am both showing my gaming age and not providing hard stats, but I think - apart, possibly, from Bioshock - the male deaths so significantly outweigh the female ones that such stats might be overkill for the particular games mentioned.
posted by Decani at 8:56 AM on August 27, 2014

To make a more manageable dataset, you could also define the subset of games you look at-- so, for example, you might look at "all PCGamer Game of The Year titles, 2000-2013" or "all games that sold more than X copes and had a Metacritic score of 80 or higher, 2010-present" or similar.

Also, as mentioned above, you will see a difference between the deaths of male/female "enemies," which are generally mostly male, and the deaths of male/female NPCs who are presented as part of the plot and/or game world. You may want to focus your attention on named/significant NPCs and look at percentages rather than gross numbers.

Some other factors you may want to consider:
-sexualized versus nonsexualized violent death (killed in underwear, body posed in a "sexy" position, lots of moaning during scripted events, killer using sexual terms during scripted events, etc.)
-female NPCs presented as sex workers in some way (either explicitly written/presented as prostitutes/courtesans/exotic dancers or referred to as such by other characters)
-already-dead bodies of women used as "edgy" set dressing in game levels
-presence of a prostitution mechanic and/or brothel level and/or red-light district in-game, especially when presented to titillate without any real connection to the plot (i.e., the character you have to go see next just happens to be hanging out in a strip club instead of, say, a diner)
-whether a character exists primarily to die and/or be an origin story plot point or character motivation (think, for instance, of the many, many, many male game protagonists who are trying to avenge the kidnapping/murder of a wife/daughter/girlfriend, and how rarely the wife/daughter/girlfriend in question is given any sort of independent character development by the narrative.)
posted by oblique red at 9:10 AM on August 27, 2014 [1 favorite]

Response by poster: All excellent points, guys. Thanks very much. It seems I am after some pretty slippery numbers. I was certainly thinking of percentages, rather than numbers, and I was basically also thinking of NPPs, rather than player characters, since the PCs are naturally going to be given more respect by the game environment. Also, yeah, nadawi is getting at the heart of the question.
posted by Fenriss at 9:10 AM on August 27, 2014

Anecdotally, from having played waaaay too many games over the years:

= The vast majority of enemies are male, and where this is not the case (cf Fallout 3/Fallout New Vegas) it's generally because enemy genders are randomised.

= Sexualised corpses of both genders often result from the 'Havok' physic engine, which can be darkly amusing/grotesque, but is procedural rather than intentional

= A generalised 'kill'em all' amorality is common; cf the underrated Syndicate remake where you can kill or spare civilians of both genders (plus there is an interestingly creepy surveillance scene)

= 'Fridging' or kidnapping of female characters as a way of motivating the male protagonist is very common, almost ubiquitous

= Actual sexual threat or violence is ... extremely uncommon? Mars War Logs had a male-on-male rape threat at the beginning, and the Witcher series portray a world that is sexist as shit, so have a lot of it in the background, but apart from that I dunno if I can recall any.

= Heavily sexualised female models are ubiquitous, though getting less so (why u wear heels into combat, huge-titted female badass!?)

= The most recent Hitman game is comfortably the most squicky gender politics game in recent memory, if you're looking for such.
posted by Sebmojo at 3:15 PM on August 27, 2014