Bartle, Gender, and WOW

Friday, October 17th, 2008 | Market Trends

Last week’s article was such a hit that we decided to take your advice and do it again, with WoW.

Side note: Feel free to send a comment and let us know what other titles you’d like to see. Note that the data for a particular title is more powerful the more DNA profiles we have – so if you want us to cover a game that you play, make sure you’ve got all your friends from that game over here as well. That sounds like dreadfully gratuitous pimping, I know, but the fact is I can’t draw any conclusions from a data pool of forty people, three of them female. So if you play a less than mainstream title, encourage your friends to sign up, and spread the word on fan sites!

Let’s get going. First, the class breakdown overall:

Obviously, this is less meaningful than the WAR chart, as there is no way (from the data we used) to tell how many of the warriors are Horde and how many are Alliance, and so on. Still, the spread is pretty even, ranging from 6% Shaman to 16% Hunter. The latter makes sense given the mechanics of many WoW encounters. In a future column, we’ll try to dig out the breakdown by Alliance/Horde for you. All I can say for certain is that my data shows Alliance outnumbering Horde by a hair under 6%.

Next, let’s look at the gender breakdown (Note: Gender of the actual player, NOT the gender of the in-game character):

Here’s where I really want to see the Alliance/Horde breakdown, because I suspect some of this balance has to do with which races can play which classes, as it did with the WAR classes. These results aren’t quite as insanely skewed as WAR, but we can see that women are drastically more likely to choose a priest – and men are three times more likely to choose the warrior class as women are, according to their proportion of the population.

Until we get that Alliance/Horde breakdown by class, we can get a rough idea by seeing the player gender breakdown by race:

Okay, on some level, you know some poor researcher is googling “race gender issues” for a Serious Project on Important Topics, and he’s going to get here and see “elves” and “undead.” Furthermore, he’s going to discover a segment of society that cares a whole heck of a lot more about Taurens and Orcs than anything he’s researching. And I wish I could see that guy’s face.

But back on topic: I’ve grouped the Alliance races first, followed by Horde. Again we see women going for the sexy and the lovely, while almost totally avoiding the less delicate races. Orcs and dwarves are sadly neglected among women. Men’s preferences are less pronounced with less of a range. I note with amusement that a slightly higher proportion of women than men choose to play… trolls.

By the way, due to the fact that men outnumber women in WoW, the percentage of players who play a human (remember, last week I pointed out that while everyone tries the freak races, they tend to gravitate to human before long) is at 38% of the total playerbase, even as it’s 20% of all men and 19% of all women. This is why you really need to trust the person spouting statistics at you – it’s really easy to slant things.

As a side note related to proportion, there are nearly four and a half times as many men playing WoW as women. WAR’s proportion of women is nowhere near that, which makes sense given the marketing choices of both games. However, to be fair, women are more likely to play an MMO after watching a friend or loved one take up the game. They are also less likely to be early adopters, since they tend to form long term connections and are wary of the cost of forming such bonds in an environment with no proof of longevity. We’ll come back and take another look in a few years to see if the proportion of women climbed at all. I suspect it will.

So, how about Bartle? What is the percentage of users with each predominate “type”?

Well, the proportion of users who primarily identify themselves as killers is lower than in WAR, obviously. I find it fascinating that in WoW, Killers and Achievers are almost exactly even. Explorers have a whopping 39.5% of the pie, leaving 13.4% for Socializers – exactly the same percentage is WAR has.

Now, let’s see which Bartle types prefer what classes:

Wow, nothing like the definite preferences we saw with WAR. Most types have several classes they like, with the exception of the Explorer having a decided liking for the Hunter above the other classes. However, you can see that the Hunter appeals to all of the archetypes quite evenly.

Finally, since we’re waiting for next week to break it down exactly, let’s see how Bartle might predict a player’s choice of Horde or Alliance:

Wow. That tells me… nothing. Well, okay, not NOTHING. I see that killers prefer Horde by 15% over Alliance. The other three types have a slight preference for Alliance. The proportion of each type that enjoys a particular race is remarkably even.

Overall, even if I knew nothing about the game or its history, I would look at these charts and say this is a game that has settled into its groove.

What say you?

Edit 10/20/2008 5:45pm ET

Due to a crash of our chart provider we had to regenerate the charts after a significant number of new tests and it yielded slightly different numbers due to a few thousand more tests because of the WoW insider article. We have adjusted the article to match the new data.

Tags:

  • Very cool breakdown. It is cool to see how evenly spread out the game players are. Very impressive on the part of Blizzard's ability to make a balanced game.

    I'm not a WoW player but this charts are still interesting to me. I kind of get a picture of the gameplay just from the above information. Cool stuff.
  • Dude
    You should redo the coloring to match the Blizzard Class Coloring so people can read your charts.
  • /stare
  • Your charts assume that women play female characters and men play male characters. That's a very large presumption to make considering that about half of the female characters in my guild alone are played by men. In fact, this nullifies about half of your article since you actually can't make that assumption. You can say that when people choose to play Night Elves, a large portion of them choose to play female NE's but you can't say that those are women choosing to play females.
  • Solder
    I seem to be missing something. The charts don't seem to be showing up for me. When I open the page in firefox, it doesn't show anything. In IE it shows the little (X) bubbles for pictures that aren't loading. Are the charts broken?
  • nothe
    charts are hosted on another server, looks like it's down (/.ed maybe)
  • We're working on getting the charts back up, the host that they were on just died.
  • Dom
    Could you perhaps edit the article to include the results in tabular form until you manage to get the charts up and running again?
  • Yay charts are back. Go Steve (Fatty Chubs). I guess we crashed chartGizmo. :)
  • Reinforcements
    @ #4 - Malsorian, what makes you think they're going by character gender? I don't see anything about how they gathered this data, though I could have missed it. It seems like what you're saying is REALLY OBVIOUSLY TRUE, and I dunno, I can't believe they'd just assume female avatar = female player. That would be stupid. Anyway, they must have used some sort of poll in order to get the Bartle types, so I'd think they got player gender the same way.
  • finnstoney
    Stats listed below don't match the chart above
    "I find it fascinating that in WoW, Killers and Achievers are almost exactly even. Explorers have a whopping 43% of the pie, leaving 14% for Socializers –" but chart shows Explorer 39.5, Killer 28.7, Achiever 18.5, Socializer 13.4

    Also, Blood Elf are Horde and Night Elf are Alliance...
    "I’ve grouped the Alliance races first, followed by Horde."
    Was the title switched or were they out of order?
  • JibrilBenes
    To the commenter who tried to nullify this whole article:

    I did not see anywhere that he was talking about the sex of the character and not the sex of the player.

    To the writer of the article:

    You did state that you grouped races by Alliance and then Horde, however, Blood Elf is the first race on the left, separated from the other Horde races on the right. Looks like an alphabetical sort from my end.
  • Eddy
    Yeah, that data just looks really wrong considering only 1/5 of players on WoW (20% from my understanding) are female. I imagine rp servers can be a little higher.
  • Slacker
    Your population numbers are misleading. In WoW there are 4 completely different play types that all hate each other; Resulting in 4 very different groups of players.

    On PvE servers 64% play Alliance 36% Horde
    On PvP servers 40% play Alliance 60% Horde
    On RP-PvE servers 60% play Alliance 40% Horde
    On RP-PvP servers 46% play Alliance 54% Horde

    Average them out and you will get 54% Alliance and 46% Horde.

    But on anything but a RP-PvP server odds are you are going to see a major slant to one side or the other.
    (My source is warcraftrealms.com)
  • I just inserted a note about the numbers. Our chart provider crashed, so we had to redo the charts and the queries reflected today's numbers which has produced a surge in Bartle tests... which changed the percentages. So we are adjusting the article to reflect the new charts. Sorry about that. I guess that is what we get for trying to host the charts somewhere else! :)
  • PSC
    The statistician in me would love to see:

    - a couple of paragraphs on how you collected the data. Presumably this is from your DNA profiles, but it's not explicit in the article, and the absolute population sizes aren't indicated
    - some comment on the self-survey bias. Would "more active" gamers be more likely to register with your service? How will this bias the results?
    - I'd be pretty sure that there are some "definite preferences" in your more detailed graphs down the bottom. These more detailed breakdowns might well have some statistically significant properties, depending on population sizes. For instance it's likely that "killers" prefer undead and orcs.
  • Mantis137
    I'd be curious to see how something like this looks when applied to EVE Online. Considering it's a global playerbase on one single server, and with the mass array of possible 'career paths', might be interesting.
  • "the percentage of players who play a human is at 38% of the total playerbase, even as it’s 20% of all men and 19% of all women."

    This doesn't make any sense if you put any thought into it. There's no way the percentage for the population can be larger than the percentage of both its component parts. It looks like you forgot to divide by 2 in there, putting it right about where it belongs, at the same level as the genders.

    Also, the killer-achiever comparison doesn't check out. 18.5% and 18.7% might be almost the same, but 18.5% and 28.7% are very not. And given those flagrant errors, I'm guessing the character gender = player gender accusation is also on the money. Oh well, it was a nice idea.
  • Sundra
    To those saying he could have gathered data elsewhere and not just assuming female toons = female players I ask how would he get such data? Account information is private thus making it theoretically impossible for him to have used the actual account information to come up with the charts.

    So I agree that the data is not accurate since a lot of female characters are played by men, for the exact reasons he attributed those choices to a female choice. Even men find that female dwarves are not very appealing and even men would find it sexier to play, say, a female Night Elf or a female caster character.

    It can't also be discarded that some women might actually choose to play a male character.
  • All of the female/male data is based on if the person playing the game is actually a female or male. It has *nothing* to do with if you are a male or female ingame.
  • Daviesmith
    Malsorian is right on the money. While the character may be female, the player most times is not. Women do play WoW, for sure (my wife does, and so does her best friend) however, this does not mean all female toons have a real girl behind the monitor.
  • IconX
    To #4: Malsorian, and especially #24: Daviesmith (since the answer was given in the post RIGHT BEFORE [#23]) ...

    Look at the note after the 2nd paragraph (just before the chart showing the breakdown of genders & classes):

    "Next, let’s look at the gender breakdown (Note: Gender of the actual player, NOT the gender of the in-game character):"

    For them do to this much research and assume that what you play in-game is your actual gender is utterly absurd. Why do this much research without knowing your population base? Looking at this much research, you should assume they based this on ACTUAL gender; which, in fact, is what they stated in the article itself.
  • IAMSHADOW88
    Very good indeed. I'm very excited to see that the Hunter class had the most percentage, because, if anyone has played WoW before, the Hunter class is the most interesting and basically "da bomb." So, yes very good breakdown on the male and female charts too. I give you my props and keep up the good work. "Hunters will rule WoW till the end of time!" -Feelmypain
    Burning Legion Server
  • grend
    @rolken given that they stated that men outweigh women 5 to 1 in the wow population 20% OF MEN would be a greater % OF THE WHOLE POPULATION, and 19% OF WOMEN would be a much smaller amount than 19% OF THE WHOLE POPULATIOON. geddit?

    @ IAMSHADOW
    hunters. Ive got a few friends that stoped playing altogether because hunters were when played properly just really boring.The perfect hunter rotation at 70 when raiding is completed with one button push.... :insert sad emote of your choosing:

    @ anyone who said "gurr gurr women characters can be men IRL gurr gurr" take the time to read the article before you jump to erroneouse conclutions.

    for Sanya, thanks a lot for putting this together, i found it very interesting and having played myself a lot of your charts agreed with my personal experience e.g. overpopulation of hunters QQ, more RL women (yeah yeah how do i know they'r women) playing certain classes etc.

    Im very very interested to see a confirmed alliance races/horde races break down soon. -Grend draenor
  • Bree
    What I'd like to see in the class/gender breakdowns is something regarding talent spec. For instance, out of the large percentage of women who play druids, how many are resto, how many are boomkin, and how many are feral? Out of those who play priests, how many are holy, disc, or shadow? How many women see themselves as "killers" or "achievers" and does this correspond to their spec choice (i.e. enhancement shaman vs. resto shaman)?

    I think it would also be useful to overlay that set of results onto what races are available for each class, to see if there are any correlations. For example, shaman is the only healing-capable class that lacks a higher percentage of female players. Does this percentage shift between factions? I am a woman and play a female troll shaman (caster dps)-- I know zero women who play female orc and tauren shamans and 1 or 2 other female trolls (one played by a man) that are active in high end raiding guilds on my server. However, I see many, many female draenei shamans on Alliance side, played by both men and women. Is there a higher percentage of people (or women in particular) playing shamans Alliance side, since there is a "pretty" race option?

    And while I'm at it, how does the data change when we separate players based on their preference for PVP or PVE play?
  • Archael
    Soon 50% of the class breakdown will be the class Death Knight and the other 50% will be the other 9 classes combined.
  • ooopinionsss
    How you think when the economic crisis will end? I wish to make statistics of independent opinions!
  • Sleutel
    For the record...

    Female, register as Achiever-Killer on Bartle's, and my main is a Night Elf Protection Warrior (currently my guild's MT). As for those races most women avoid? The first toon I ever rolled was Undead, and the second was an Orc. The big reason my main is Alliance is because of the people I currently play with.

    Good luck predicting ME from these stats... :D
  • Manachi
    Fantastic article - very interesting. Thanks for the effort!
  • Generally I do not post on blogs, but I would like to say that this post really forced me to do so! really nice post.
  • I've recently closed an extensive survey on WoW skills translating to real life, and part of my demographics came up with about 38% of the players as female. Looking at Nick Yee's research from a number of years ago (Daedalus Project), his numbers were around 20% female as I recall. I suspect WoW's popularity and acceptability is bringing in more and more female players to the MMORGs, which as a long-time tabletop gamer myself I find a real joy. (By the way I'm ESAK and have played every race and class. Alliance main is human male mage, Horde main troll shammy female.)
  • I think those statistics about which races are predominant amongst female players is hilarious. My two mains (one each for Horde and Alliance), which were my first choices, are a Dwarf and an Orc, and I have practically zero desire to play a Human or Blood Elf.

    I must be stranger than I thought.
blog comments powered by Disqus

Subscribe

Enter your email to get blog update notifications:

Powered by FeedBurner

Archives