f13.net

Though it's not about MMOG's, here is a summary of a recently released report from The NPD Group (http://www.npd.com/) concerning online gamers.

Majority of Gamers Play Online (http://www.emarketer.com/Article.aspx?1003001)



This was a survey of gamers age 13-44. It also included a number that's interesting:

This survey wasn't remarkably new, although one problem with it is the surveys were emailed and there's probably a large population of poor console gamers who don't user email.  This survey again confirms that females are more likely to use the PC than the console for gaming, a point I argued with several people on one of the boards way back who didn't believe it.

Bruce

That's why VCs are putting money into mobile gaming:

http://sanjose.bizjournals.com/sanjose/stories/2004/08/09/story1.html

This one didn't make sense to me, though:



Maybe I'm just out of the online console gaming loop (don't have an adapter for my PS/2) but the above just seems odd.

That's cause women love to play games like these online:

http://zone.msn.com/en/root/cash.htm
http://mirror.worldwinner.com/start.shtml?portal=main&style=wave&partnerID=whome

as well as games like Backgammon (not for money) online.

http://news.gamewinners.com/index.php/news/107/

In my own surveys, I've found out that 98% of potheads play video games, and 75% of gamers are potheads.

And 99% of all messages in ooc channels are posted by potheads mentioning how stoned they are.

75.2% of the time you get a kill on an xbox live game, that person claims to be too stoned to play.

The remainder really are too stoned to bitch about it.

The top prize in PGR2 is a trick TVR that is insanely quick. 99.93% of people who use this car used a cheat to copy somebody else's save. Only 7.4% of those people will admit that they copied a save file. When racing these folks, 75.2% of the time you can easily beat them in a Ferrari Enzo, since they are so stoned that they will ricochet off the walls.

I'd have two questions:

Do respondants self-define 'online' and does popcap count as online gaming?

Are respondants self selected from people who visited your website (and hence are regular internet users anyhow)?

I'd imagine that the respondents were either emailed the survey, or are members of a survey group that gets sent surveys at times, like  Ipsos I-Say (http://www.i-say.com/) or something similar.

Online is more than likely defined for them by the bounds of the survey.

EDIT:



The full report has to be purchased, and I'm quite sure it's high dollar.

NPD has sent me spam email asking to participate in their surveys before.

Bruce

What's their sample size?

How do I know this isn't just pulled out of their ass?

I don't see their questions. This looks about as scientific as the national news phone us surveys.

Methodology
A survey invitation was e-mailed to a representative sample of 15,700 males and females ages 18 to 44, with an over-sampling of males ages 18 to 30 given their high propensity to play video games. An additional 3,300 teens (ages 13 to 17), with an over-sampling of males, were also e-mailed the survey invitation. The survey was fielded on Thursday, June 3, 2004 and closed on Monday, June 21, 2004 with a total of 7,430 responses, yielding a 39% response rate. The survey data is weighted to represent the population of individuals ages 13 to 44 in the US.

Bruce

Surveys are dumb, imo. You cannot get an accurate accounting unless you poll everyone, and even then the results can be easily manipulated. People who make decisions by polls are tools.

How many people here, veteran, knowledgable gamers, have taken part in one of these polls on gaming? I've never seen one.

Bullshit.  Take a stat class.

Although e-mail surveys are the bottom of the barrel as far as accuracy goes.  39% response seems a bit high.

Because everything they teach in school is correct. If you poll people in kentucky, you aren't getting an accurate idea of what people in new york think.

Which is why e-mailing out the survey sucks.  Indeed it sucks worse than a physical mailing.  The best way to do these things is with a phone survey, that is now that almost everyone has a phone.

Statistics is a very well developed field of mathmatics, unfortunately it is usually either accurate or precise but not both.

You do have to perform the survey with a reasonable hypothesis.  Saying you can't survey people in Kentucky and get an accurate idea of what people in New York think is false though, it depends on the topic and how much data you have accumulated that shows where the actual normal and mean are in New York in comparison to Kentucky's.  With enough data points you can show that New York and Kentucky have a correlation thats accurate to within a certain amount and then make deductions based on that information.

When Car Manufacturers decide how many red 2 door coupes to make a year  do you think they poll every household in America to see if they think they will buy a red 2-door coupe next year?  No, they do statistical analysis and build to the numbers suggested.  Even something as nebulous as voting is accurately and precisely modeled, votes are not tallied the night you make them, in fact the real tallies aren't done for weeks after the elections.  Those reports of winners and losers are done through statistical analysis over a few miniscule % of the voting population.  They are very, very accurate the vast overwhelming majority of the time.  Even in the Bush/Gore thing the statistical approximations of the vote only differed from the final counted value by a few hundreds out of a hundred million or more votes.

Statistics is both accurate and precise when used correctly, unfortunately there are way too many people out there using statistics with flawed data to prove incorrect hypothesis so you get the uneducated pooh-poohing the entire branch of mathematics.  It's unfortunate  because a little critical thinking tends to show which sets of numbers can be trusted and which can't but many people are so used to accepting blindly everything they hear from the magic box so when the box touts a set of highly circumspect set of numbers as being accurate and later are proven not to be the masses learn to distrust statistics instead of distrusting the BOX which is the true culprit.

Once again: go take a damn stat class.  Getting appropriate samples sizes and picking samples that are indicitive of the whole are important topics in stats.  Yes, a statistician can screw up his job and come up with stats that are invalid.  An engineer can also build a building badly and have it collapse, and a doctor could kill you, but that doesn't mean any of these things will neccessarily happen (and generally don't).

Edit:  It is important that, as with anything that is being passed as scientific (to include statistics), it's important to cite sources, methodologies, etc, or else it's as useful as used TP.  It's irrelevant for the numbers to be 100% accurate if they were not cited corectly.  This isn't  just "well, they may be off a bit" - it becomes a case of "these are totally irrelevant and without value".

I don't think the point is stats is a meaningless science.  More that it's a perfectly solid rational tool that like anything is susceptible to subjective inputs.  GIGO. Saying "it just has to be used right" is correct but meaningless and is really just another way to say what's really being stated by others, that all to often it is not used right.

Not so if the argument being attacked is "it can't be used right".  The statement "it just has to be used right" ought to go without saying, but it assumes that the "it" can be used with benefit to begin with.

Bear in mind that the information we see from this report is the barest highlights. I'm sure since NPD is selling the report that it goes into a lot more detail, including sourcing, methodology, etc. All the good shit that research people get chubbies over and that puts the rest of us to sleep.

The numbers are interesting in and of themselves. The email survey was more than likely not a typical "spam" survey, but made up of a qualified list of people who have either: 1) purchased a video game online before,  2) signed up for a video game mailing list of some kind, or 3) signed up for surveys before. It wasn't just a random spam email sent out to just anybody with an email address (especially if it's only 15k sent out). 7k is a decent sample size. Hell, I might actually have participated in this survey, since I'm signed up for the Ipsos I-Say thingie and fill out the surveys every once in a while.

No - my point was pure irony (http://dictionary.reference.com/search?q=irony) about Bruce's graphs.

But no - everyone has to prove they are the smartest monkey.

Welcome to the Intraweb.

Thansk I aprekate it.

I got the irony, but I thought it was far more effective as something unstated, so the implication stung HaemishM every time he read the thread rather than giving him a specific post to respond to.

Anyway, if my data was survey-based, this would certainly be a valid point.  But it's not.  That would be like complaining that when the reporter pulls the latest Pentagon budget from their background briefer, he didn't call 100 different beauracrats and derive a statistical average with 2-sigma confidence.

Bruce

Ow, I R STUNG. SO HARD.

It hurts us.

This survey data is more accurate than your chart. Neener.

And my chart's accuracy is greater than 0.

pwned.

Bruce

Here Haemish - I'll beat you to it.

Prove it.

If he tries, I'll edit his post to say "I am a furry" and lock the thread because we all know that will lead us absolutely nowwhere.

Hey, fuckers. Keep your slapfight in one thread, kthx.

Nice avatar, Fur. Lemmiwinks rules.

He is the gerbil king...

Already proven; see previous thread on this subject.

Bruce
PS - I am a furry.

FUCK. I had to read that while I was eating my lunch. Thanks.