Less than 5% of the players go to raid, I wish

On top of that, recall that--unless things have changed since I last checked--Nick Yee gets all his data from volunteers.

No, this is incorrect.

On his site he collects data from volunteers.

In the case of Play On, instead, he just uses a better version of the census mod. So he just monitors passively the activities of the characters logged in.

If he used volunteers the study could have been more useful, even if still imprecise.

Yes, he presents things really well. But the presentation is not what actually matters.

Ah, your points make much more sense now, in that case. :P

I'm criticizing the way your studies trick people to think in certain ways.

A survey, a poll or a reasearch are never "wrong" on their own, but in the way they are "read". There's ALWAYS a theory at the base, implicit or explicit. The research is a way to test that theory.

I criticized both this survey about the raid activity and the one about the avatars not because of the studies themselves, but the utility of the conclusions. Why take surveys if they have no concrete use? It's obvious that you aimed to demonstrate something and I criticized that implicit affirmation. The idea that everyone has after reading them.

It's "bullshit" because it has no concrete, solid use and because it favors a misunderstanding. What else is there to criticize in a survey if not its practical or theoretic conclusions?

That survey has implicit affirmations, everyone who reads it would arrive at a wrong conclusion. I was starting to write another attack to Blizzard after I read it, then I figured out the inconsistency and decided to write a defence, instead.

The "harsh" tone is because I don't "build consensus" on this site. I'm not removing legitimacy from your studies, you are already much more in high consideration than me. So I'm not worried about making a notch in your reputation.

Instead, for once, I wanted to doubt about what is usually considered as a truth.

My lord, that is an amazing amount of criticism over what is purely an statistical evaluation that, in the end, ONLY offers very minor conclusions based on the raw data (i.e., statistical fact.

Note that he says:
"Thus, 3.6% of all observed characters spent more than an hour in raid content over the month of January."

He isn't saying:
"Therefore, only 3.9% of all WoW players raid."

There is a BIG difference.

He should be commended for not artificially altering his data with assumptions as you're suggesting, such as your paragraphs which begin with "The truth is different" and "WoW brought the revolution." These paragraphs are formed on pure opinion, and should be suspect by any reader.

The only part of your arguement that I do agree with is the value of the research. Instead of bashing Mr. Yee, why not suggest how to better farm data for a better defined (I didn't say 'accurate' since his article is accurate based on the data sample) model.

Not sure if I agree with your take on this one or rather I believe your inference is incorrect.

1. the data represents accounts and taking a sample of 100 accounts, the data suggests that 3.6 people spent more than 1 hour accessing high end content in World of Warcraft across a specified time period.

2. the data represents avatars, so 3.6 characters accessed that same content for longer than an hour over that time period.

Multi-boxing could be considered a statistical anomoly over the large user base so there is no point in bothering about it however you could provide a +/- percentage to be completist.

Interpretation 1 or 2 doesn't really affect the underlying result because the potential for alts to impact the data exists equally for all users and one unique account is only playing one unique character at any given time.

16.5% of the characters were level 60 and given the requirements of the raid instances, it would probably be fairer to state that of those capable, 30% of World of Warcraft players accessed high end content.

What this data shows me is that Blizzard are creating content of 16.5% of the population of whom only 30% access it at all. And only 21.6% of that targeted population accessed it for any period of time.

-Illaraphaniel

p.s. And actually, you've made the inference that it's unique accounts in your heading, not the data as such as it specifies "characters". But in saying that, one character online per account means the terms are almost interchangable.

Essentially, 95% of all people when accessing World of Warcraft aren't accessing raid content during time spent in game.

p.p.s So you're heading actually skews the results. If you'd have said, "less that 5% of time spent in World of Warcraft is used accessing Raid content" you'd have been much closer to the data. Or even "More than 3/4's of World of Warcraft players can't be bothered using the raid content".

p.p.p.s The differential between total raid users versus raid users who stayed in longer than an hour can easily be attributed to alt swapping in raid for specific usages or for loot in my opinion.