Hat tip to August J. Pollack for finding this. It's tempting to think of this sort of thing as mere ham-handed marketing with nothing more than the profit motive behind it, a desperate attempt to bring women into a consumer group perceived to be hostile to said women via tired stereotypes. It's easy because, well, it's largely true, but to dwell on that would be to miss the fact that this kind of thinking is surprisingly pervasive at all levels of the gaming community, from the players to the press, and even, to some extent, to the academics.
Richard Cobbett covered this territory more effectively than I ever could with "Writing A 'Girls In Games' Article", an essay that ought to be required reading for anyone attempting to discuss gender and games. Girl Gamer seems to flow from several lines of thought critiqued by Cobbett, specifically points 3, 4, 8 and 9, with the greatest emphasis on point 4.
Thing is, the idea that women, when expressed as as an arithmetic mean, prefer certain genres, modes and features was not arbitrarily pulled from the ether. Statistically, it has some support, and even for those of us who feel that the American faith in statistics is more often religious than scientific in nature, that support is hard to ignore. But even at their best, statistics are only empirical, prone to methodological error, and are not, in and of themselves, predictive. (That's where "theory" comes in. Creationists beware.) Group identities are useful things, but they are ultimately fictions. I like fiction; fiction can be compelling and useful, and you don't have to be a mystic to understand that things that exist subjectively can and often do affect things that exist objectively. To riff a bit on a quote from a dead conservative/libertarian humorist whose name I cannot, at this moment, find, women are only available in units of one. Out here in the really real world, they're not actually a hive mind.
Which brings me to the Sarah Problem. Sarah is not an average or a composite, but an actual human being, made mostly of water, and capable of reflecting on her own existence. While I haven't verified it directly, her name, physical appearance, and the image she projects suggest that she has two X chromosomes. She is, in short, a woman. And the rules we apply to women in the context of their relationship to videogames do not seem to apply to her. She's not big into The Sims or casual games. She isn't turned off by brutal violence or highly sexualized female avatars. (And yes, sports fans, she's straight. That should save a couple of comment writers a minute or two.) She bought her PS2 before I bought mine, and nearly every time I get into a bloodbath like Devil May Cry, God of War or Resistance, she's already bought, played, and usually beaten it.
This would seem to make her something of a statistical outlier, but I can't sign on to the assumption, implicit in many discussions of gender and videogames, that this makes her experience as a gamer or as a woman less valid. Because, well, she exists. She's a friend of mine. And her experience ought to be part of the discussion. Individual experiences matter. In addition, in a large enough sample group--say, people who play videogames--outliers can be comprised of rather large groups, and sometimes the exceptions to the rule are among the most interesting and important.
The need to create a "feminine space" in videogames, however worthwhile that goal might be, has led to an irritating phenomenon I refer to as the gendering of genre: Halo is for boys, The Sims is for girls. Boys like speed, competition and violence, girls like story, personalization and collaboration. And, if you were to take a poll, that's certainly true for some of them. But things like story, personalization and collaboration are important in and of themselves, not because they might be marginally more likely to appeal to women. We're seeing a great expansion in paidia play in nearly all videogame genres now, due to a combination of market demands and the new creative options available due to advancing technology. The development of new genres is a good thing, period. Will these new genres help developers and publishers expand their consumer base? Who cares? Electronic Arts' bottom line really isn't my problem.
The problem with gendering these aspects of gameplay is that actual flesh-and-blood women do occasionally fall on the "masculine" side of the spectrum, and this creates a conflict in our construction of the topic. If violent, ludus-heavy action games are masculine, then Sarah is something of a ludic transvestite. Whereas before she might have been thought of as unfeminine for playing videogames, now she can look forward to being thought of as unfeminine for playing the "wrong" videogames.
More to the point, treating Sarah's play experience as being "masculine," in the sense of being equivalent to the experience of a male playing the same games, collapses her into a group to which she does not belong. That meatsuit she wears influences her consciousness, her sense of identity, and the way she's treated by others, just as my (marginally different) meatsuit does for me. Her experience might very well be different from that of the "usual" gamer, and for research purposes that seems like it might be kind of important. Yes, it's always interesting to think about what games non-gamers might like to play, and a lot of those non-gamers happen to be women. But if there is some social good to be gained from having more women playing videogames (a question I'll not attempt here), it seems like women who already play and like games in multiple well-established, commercially successful genres would be worth listening to as well.
In Arcanum, female avatars are given a +1 bonus to Constitution and a -1 penalty to Strength. Even in a world of elves and dwarves, default status is issued to human males--white ones, judging by the available character portraits. The discussion of the "female problem" in the videogame industry does not have to function along similar lines. So let's all grasp a firm hold of our undergraduate understanding of the difference between "sex" and "gender," and remember that we don't know much about biology, culture and causality, and that demographic data that's true right now might not be for very long. To pretend otherwise, to reify what might be fairly arbitrary taste issues, would be stupid.