Found the following site in my travels:

http://www.quantcast.com/rypn.org#

Limitations of the data set not withstanding...

DPK

They missed mentioning that as a group we are chronically short of money.

PC

"They missed mentioning that as a group we are chronically short of money."

With all those college grads? How can that be?

Interesting, nonetheless. Reminds me a bit of the demographics of oh, say, Buick drivers.....

Howard P.
who always lusted after a '36 4-door Century

It's not very accurate when it comes to the groups of people visiting LOL.

I used it to check on a site where I am a moderator and it is so far off base it is comical. Almost everyone was a pauper and had a postgraduate degree!

We maintain statistics for rypn, both the main website and the forums. The cited source is off by about 500% in terms of the number of visitors. The rest of the nonsense is just that: nonsense. There's NO way that they could/can derive this information. They're making it up.

Well the "98% white" would be pretty accurate if it referred to hair color.

PC

Now THAT is funny. Of course, my hair was 98% while by the time I was 25. At least I still have it.

OK, first of all, I totally agree the numbers seem to be way off base in a lot of cases.

However, I'm curious. You say off by 500%... in which direction?

I'm assuming the number they show is unique visitors. So you're either saying you get 500% less, which would be about 850 uniques per month (that sounds low) or that you get 500% more, which would be about 22,000 uniques (that sounds high...)

So I'm wondering which way you think they're off. I'm guessing you think they're low, since I know they're way low for my sites. It's usually about 1/2 to 1/3 though, not 1/5...

Actually, they can do pretty well when you're talking about large groups, such as for sites like CNN and FOX etc. IP numbers are assigned by location, and they correlate those with the neighborhoods etc.

Granted, it's rough and they do a lot of extrapolation based on a small data set. Kind of like the old Neilsen Ratings, where they'd have a small group of people write down what they watched on TV, and then use that to say which shows were most popular.

So to say "they're making it up" isn't really fair. However, I will grant you that while it is based on facts, it's still only a rough approximation at best, and for a smaller site like this one it's just not accurate enough to worry about.

I'd say they're close on the male/female thing, and if anything they over-estimate the female portion. However, to say that over 50% of the site traffic is just folks passing by seems way off, this place seems like it's made up of mostly regulars.

I was a little bit off too. Try 16,250 unique visitors and 65,000 visits per month, on average. And they have NO way to even guess the male to female ratio or anything else in the demographics. They are trying to impute value to what is essentially nonsense.

This is totally conjecture, but if they tracked which ads people clicked on, correlated them with IP addresses, then tracked them going to this site, they may be able to extract gender, age, and financial state using that. It would still be a very rough approximation, but it may be possible.

That's pretty good. You're doing better than I would have guessed. Then again, it's very hard to judge traffic on forums since most folks just read and don't post.



Well, I disagree on that point. By tracking ips and the sites they visit etc, they can get reasonable approximations. If you examine any individual, they may be way off, but if you look at a large group, then it gets a lot closer.

After all, look at the stats for this site:
84% Male, mostly in the 35 to 49 age group...

Sounds fairly accurate to me, especially the gender ratio. Hard for them to hit that number so closely by just guessing... Do you disagree?

I disagree to the extent that they have no access to our statistics, to our logs, or to any other tracking information. So anything, in my opinion, that they publish is crap, for which they wish to be paid.