I have searched on GAF, but have been unable to find a thread relating to this article yet. Please lock if old.
-------------------------------------
I found this article by Nick Yee at Quantic Foundry on the female demographics of different types of video games while looking around online. I found the results they present in their article very interesting, so that's why I am making this thread for it. I will be referencing some parts of the article here, but definitely take a look at the full article
here.
Prior Research, Problems, and general statistics
You've probably heard the often-quoted statistic that about half of video gamers are women, illustrating how gaming is now a mainstream activity enjoyed by both men and women. This finding comes from the yearly ESA report, and has fluctuated between 38% and 48% in the past decade—currently estimated at 41% in the 2016 report.
Oddly, this statistic can have the exact opposite intended effect. Some gamers argue that the study bundles gamers across platforms and genres, and is thus unrepresentative of the ”real" PC/console gamers. Or that the apparent gender parity means there are no longer gender biases in game design. A game dev we recently chatted with mentioned that some designers she works with still assume that only 5% of core gamers are women, and that the quoted 41% of women are primarily casual gamers.
By bundling across platforms and genres, this 41% statistic doesn't surface how the percentage of female gamers does or does not vary across genres, or how large that variance is. Using survey data we have from over 270,000 gamers, we've put together some findings on the percentage of female gamers across game genres based on the specific game titles they enjoy playing.
Methodology
The Gamer Motivation Profile allows gamers to take a 5-minute survey to get a personalized report of their gaming motivations, and see how they compare with other gamers. Over 270,000 gamers worldwide have taken this survey. The 12 motivations that are measured in our model were identified via statistical analysis of how gaming motivations cluster together.
Alongside gaming motivations, we also collect data on demographic variables. In our full sample, 18.5% are female gamers. We also ask gamers to list specific game titles/franchises that they enjoy playing (up to 9). By sampling gamers who mention a specific game title/franchise, we can generate a profile of that game's engaged audience.
We decided to take an approach based on specific game titles. In our analysis, we manually picked popular game exemplars to create genre groupings. For each game, we calculated the proportion of its gamers that are female. And then we calculated the genre group average. So for example, when we say High Fantasy MMOs, we specifically mean the group average of gamers who enjoy playing World of Warcraft, Rift, Lord of the Rings Online, EverQuest II, and The Elder Scrolls Online.
On demographics:
Gamers from 191 countries participated in the Gamer Motivation Profile. The countries with the largest number of participants were the United States (125k), Canada (14.8k), United Kingdom (13.9k), Brazil (9.4k), Australia (8.5k), Italy (7.2k), Poland (6.8k), Indonesia (6.6k), Denmark (5.8k), Philippines (5k), Germany (4.9k), Sweden (4.4k), France (3.6k), Singapore (3.6k), Netherlands (3.1k), Spain (2.4k), Turkey (2.3k), Malaysia (2.2k), Russia (2.2k), Chile (2.0k), Norway (1.9k), followed by a long tail of other countries.
The percentages listed refer to the proportion of gamers within each genre that are female. So for example, the 69% for Match 3 games means that of the gamers who mentioned a Match 3 game in the data, 69% of them were female. The 69% does NOT mean that 69% of women play Match 3 games.
This part is important to realise:
the numbers you are about to see are percentages of the full pool of respondents, not of only female respondents. As a result, from the fact that the survey only has 18.5% female respondents, the data likely skews further towards the male side than reality dictates. The respondents percentage is more than twice as low as the estimated number of female gamers, so that is an important limitation of the research.
Edit: I have added some comments about this number from the writer of the article at the bottom of the OP, please read for a better insight into the meaning of the gender split disparity.
On Sample Bias:
Compared with the ESA 2015 factsheet of all gamers, our sample has a far higher proportion of male gamers and the average age is far lower. Also, our sample consists of a higher percentage of core gamers (the top game genre in the ESA sample was social games) with a skew towards PC gamers and RPG gamers.
All the games used for the analysis can be found in the
article itself.
More information about the data used can be accessed
here. You can take a look at the survey itself over
here.
Results
Conclusions and Observations
The genre averages range from 2% to almost 70%. This is a 35-fold difference, and illustrates why an overall statistic for all gamers (ignoring genre) can be misleading and confusing.
The data hints at how none of these numbers are set in stone or represent some kind of hard ceiling. Note that variations within the same genre can be much larger than variations between genres. For example, Dragon Age: Inquisition has double the genre average (48% vs. 26%). In fact, this 48% is higher than the next 5 genre averages. This means that the opportunity to attract female gamers may be a lot larger than what the chart is showing, especially if you're the first ones to figure it out in your genre.
It's also easy to read the genres in the chart and pin the cause solely on gender differences in gaming motivations–e.g., women simply don't like X or Y game mechanic, but there may be a lot more going on. For example, games on the bottom of the chart tend to not have female protagonists, tend to involve playing with strangers online, and tend to have a lot of rapid 3D movement which can lead to motion sickness (which women are more susceptible to). Low female gamer participation in certain genres may be a historical artifact of how motivations and presentation have been bundled together and marketed.
----------------------------------
The research has some clear limitations: First of all, there is a difference in the respondents demographic and the reported gaming demographic. Secondly, the research necessarily picks a select number of game titles to represent a genre. As a result, titles that happen to have higher female or male participation do not influence the overall demographic outlook: a good example is Overwatch, which could give some diversity in gender to the sample of FPS titles. Despite those limitations, it is an interesting look at the gaming demographic and gives us some very useful data points from which to work (and to start new research from). I recommend to everyone that they read the article, since there is much more in the way of specific conclusions to be found there.
Edit: I have asked the author in what way he views the disparity of his sample (18.5% female) versus the ESA report's 41%. He gave me the following response:
Nick Yee said:
This largely depends on how you interpret why the ESA had 41% female gamers while we had 18.5% in our respective samples. The ESA randomly sampled and dialed up US households, but they don't provide the details on who got counted as a gamer–e.g., if playing Solitaire last year counts.
What we do know is their sample is biased towards more casual games. In their 2015 report, their most frequent gamers are most likely to be playing social games (31%) and puzzle games (30%). Note that this is their ”most frequent" gamers. We can assume that their less frequent gamers are even more likely to play casual titles. So a large portion of their sample may not regularly play AAA titles at all.
Our sample is biased towards core gamers–people who would click on something labeled a Gamer Motivation Profile. But when we're drilling down to a specific game (like Diablo III in the Action RPG genre), the 41% in the ESA sample is not a meaningful benchmark because a large portion of that sample was likely not playing a AAA game to begin with.
On the other hand, the ”relatively more active" framing is correct. Based on our data, roughly 18.5% of core gamers are female. And relative to this baseline of core gamers, Action RPGs (at 20%) have a ever-so-slightly higher than expected proportion of female gamers.
He suggests that the sample of the survey skews more towards a core gamer audience, which makes sense if you consider that core gamers are probably more likely to fill in a survey on gaming than casual gamers. He added an interesting data point elsewhere that further strengthens this point:
Nick Yee said:
Each genre we analyzed contained between 3-5 game titles. The median sample size for each game title was 1,184. And the median sample size for each genre was 4,657. That corresponds to 1.7% of the full sample.
The genre with the smallest sample was Casual Puzzles (n = 635), and the genre with the largest sample was Western RPGs (n = 60,275). Most of the other genres were in the 3-6k range.
It would seem that the audience is vastly more interested in WRPGs than in casual puzzles, which once more suggests that this is indeed a representation of a more core gaming audience rather than a general gaming audience.
