I live – and I do math

Having a day from work this week I realized I hadn’t updated this blog in a long time. Guess I have a hard time keeping up with being serious, but I’m going to keep trying!

I wanted to use this space to write an analysis of something I found elsewhere on Blogger, rather than using my casual Livejournal for it.

In this article dated last month: The Game Widow Phenomenon, author Wendy Kays claims to crunch the numbers on how many people consider themselves “Game Widows.” A Game Widow is, as I understand her thesis, a person who has a spouses or significant other but feels that they are less important to them than a video game, video games in general, or the video game industry.

She also invites math whizzes out there to correct her numbers if we see a problem. Well, technically I’m not a math whiz, but I have done some game research so let’s see if I can help this out.

I’ll start at the top, and work my way to down to the bottom, which actually contains the error that struck me as most glaring. She writes:

According to a Blizzard Entertainment press release, there were 10 million people playing WoW as of July 2008. According to Nick Yee’s research, the average MMO player is spending 22 hours a week online. Of those gamers, 30.5% of MMO players are male and dating, 26.2% are female and dating, 33.1% are male and engaged, married or separated, and 60.3% are female and either engaged, married or separated.

That last figure to me seemed somewhat suspect, so I went and looked up Yee’s actual charts about demographics, located here and here. If there are more recent charts that have the 60.3% figure, I couldn’t find them – these charts are dated 2003 so they could definitely be said to be out of date.

At any rate, if we were to use these charts, Yee isn’t saying that 60.3% of MMORPG players are female and engaged, married, separated. He’s saying that, of the subset of MMORPG players that ARE female, 57% percent (the number that I found – 60.3% in her argument) are married. From his study we can conclude that, of the MMORPG players that are female, a higher percent of them are married than the MMORPG players that are male.

If we want to simply make a conclusion as to how many players are married, Yee has that number, or at least that number so far as among his respondents: 36%. If we assume that number has remained steady, that means that out of 10 million WoW subscriptions, there are 3,600,000 game widows.

But wait. We can’t go that far yet. First we need to know how many people Blizzard is calculating as users. Is the 10 million figure actual human beings, or just subscriptions? So many times, MMORPG research falls in to the fallacy of “one person, one account,” which is simply not true (it’s particularly pervasive in research about player avatars, to the point where it bothers me a great deal). A minority of very serious WoW (or MMORPG players in general) players are “double-boxing,” meaning that they have multiple accounts (Two, or sometimes more) for the same person. This can be used to cheat the game or simply to get more alts to play with on your favorite server. We also can’t count the amount of people who are goldfarmers who may have accounts, since a goldfarming company may be counted as 20 or so individual users.

It’s something to consider, but we can possibly assume that double-boxers are outliers, and still arrive at the conclusion that 3.6 million WoW players have spouses. Even so, we can’t come to the conclusion that 3.6 million of these people are game widows. We need to do some more math. We should know how Yee calculates his average number of hours per week: here is a chart. He has it categorized to show that the largest amount of players run between less than 10 hours a week up to 30. Some strong outliers that play the game more often pull the overall average to 22 hours per week.

What information don’t we have? The percent of people who play the game a lot, but are married – there’s no crossover demographics research done, and there would need to be to draw a definite conclusion as to who is a WoW widow. We also don’t have the information about whether or not people who have a significant other do their gaming with them – if they did, that might be felt more as a shared activity than something that was “widowing” the spouse.

If we assume that the demographics of marriage have no effect on the amount of hours played, we can cross over these two charts, and I would argue that if someone plays less than 10 hours a week, they probably aren’t making their spouse a widow for the game. So subtract 26.2% of our original game widow figure, at least, to arrive at 2,656,800 game widows, or half the original figure. This is actually a large estimate because of dual-boxing and other behaviors, but it’s about half her original estimate of people who might be considered WoW Widows.

Moving on:

As of July 2008, Sony had sold 140 million PS2 consoles. According to the Entertainment Software Association, 33% of all gamers are women over the age of 18, and 42% are men over the age of 18. America Online (AOL) and the Associated Press did a poll that showed 33% of gamers are married and have kids. This leaves us with about 19 million female and about 15 million male game widow(er)s worldwide.

We need to know how the ESA defines “a gamer.” Otherwise, don’t these numbers already look surprising? The ESA defines a gamer as basically anyone who plays games. That’s why their demographics for “casual gamers” are so high. Their percent of “online gamers” includes people who only play video poker or Scrabble on Facebook. The ESA also doesn’t look for worldwide figures: it specifically states that its survey is of American Households.

That might be irrelevant to the discussion. You could still feasibly widow your spouse for Scrabble or Peggle. But if you’re trying to make a point based on the overall sales of PS2 consoles, then you’re not crossing the demographics well. We can’t draw any conclusions based on the sale of PS2 consoles and the ESA’s gamer demographics.

We do have the figure in her argument that 33% of gamers are married with kids, and with the ESA’s statistics we can figure out how many people in the USA are gamers. We’ll use their 65% of households figure and try to crunch down: 300 million people in the USA, times .65 is 195 million gaming households. 33 percent of that is 64,350,000 million married with children households that game. High figure, but we can’t draw any conclusion whatsoever as to how many people are “widows” here, because we don’t know the average playtime in these households, and we would need that data to make any argument. Just as many of this is probably parents who play with their kids which would be a positive behavior.

Finally, her last paragraph:

According to the IDGA Developer Demographics Report, 88.5% of game developers are male, and 11.5% are female. According to an IDGA Quality of Life white paper, 61.5% of spouses of game developers say they work too much. We can assume, then, that at least 61.5% of game developers have spouses.

This is the most egregious error. You can’t use the figure “61.5 percent of spouses of game developers say they work too much” and transform it in to “61.5 percent of game developers have spouses.” That’s not equivalent. It would be more correct to assume that 100% of the people interviewed as to whether or not their spouses were overworked, were already spouses of game developers! So you’re basing your figure on stating “100% of game developers are married and 61.5% of these people believe their spouse is overworked.” You can’t make that assumption. You first have to find out how many of the 100,000 estimate are actually married, then take 61.5% of that figure instead.

I think I’ve been more thourough about this but welcome additional comments or corrections.






Leave a Reply

Your email address will not be published. Required fields are marked *