21 comments

[ 5.0 ms ] story [ 44.6 ms ] thread
Where ARE the referrals coming from? And why do people view porn after looking at photos of friends?
There are only a limited number of things typical people regularly do online, I'd be willing to bet the list for a large chunk is something like

1. Email 2. Facebook 3. Porn

In which case you have a 1 in 3 chance that porn follows Facebook.

no, 1 in 2 chance. unless you suppose they can go to facebook after going to facebook.
Well you have the permutations,

(email, Facebook, porn)

(email, porn, Facebook)

(porn, email, Facebook)

(Facebook, email, porn)

(Facebook, porn, email)

(porn, Facebook, email)

That is 3! or 6 total possibilities, 2 of which involved porn immediately following Facebook. which is 1 in 3... ...How do you figure 1 in 2?

On the other hand, you've disregarded the possibility of continuing to browse after doing three things (or stopping prior to three things). If browsing is an endless stream, then facebook has to be followed by one of the other two, which is the angle he was working from. Depends how you frame the question and what assumptions you make; it's ill-defined right now.

Edit: On another note, in his place I would have sanitized my own history; a screenshot showing redtube and tube8 in the "visited" style is tacky, to say the least.

I'm sure there's an innocent explanation! Maybe he was performing some "hands-on" research? :-P
Hidden mid-post: Note: for those wondering how these were chosen, I used related links on Alexa to find some big sites

Suuuure :)

(comment deleted)
How are these statistics obtained?
To a first approximation:

  return rand();
It's Alexa toolbar data and therefore pretty unreliable.
Short answer: no

Lets take a look at the clickstreams for some non-porn websites:

- http://www.alexa.com/siteinfo/google.com#clickstream

- http://www.alexa.com/siteinfo/aol.com#clickstream

- http://www.alexa.com/siteinfo/wikipedia.org#clickstream

- http://www.alexa.com/siteinfo/news.ycombinator.com#clickstre...

Facebook is near the top of all of those lists. To me, that looks like people are browsing Facebook and then browsing to these other sites: it doesn't imply that all of the traffic is based on direct link referrals.

The data is also skewed by the fact that it comes from Alexa (and thus the Alexa Toolbar).

So they are not relying on the http header or document.referrer? Seems a little silly...
Well, it appears that they're trying to track something more general than that. Whether that more general data is valuable or not is a different question. :-P
Slightly longer answer: collected only from Alexa Toolbar users, and...

>Which sites did users visit immediately preceding youporn.com?

Not "where did users come from". Not "referrers". Not "how did they get here". Just "what were they doing right before they came".

Facebook is at the top because Facebook accounts for an enormous amount of internet traffic. No other reason. And it doesn't even remotely imply people are finding youporn through links on Facebook.

Supporting evidence: downstream stats put FB at 10.47% and Google at 7.53%. "Downstream" being defined as:

>Where do visitors go after leaving youporn.com?

I highly doubt they're following links on youporn to get to Facebook. Or Google search results. They're just browsing both sites at nearly the same time.

edit: left a slightly-modified comment on the blog.

...which is still somewhat disturbing :P
No disagreement there :)

Perhaps more disturbing than the upstream: the downstream is noticeably higher in favor of Facebook. So, after people have browsed their porn, they go to Facebook. Looking for...?

I'm guessing it's just a matter of probability. Many people only use the web for email, porn and Facebook, and on top of that, Facebook is one of the most frequented sites on the net. If people are visiting porn sites, the odds are they were either at Google or Facebook immediately prior.
I'd guess the answer to the article's headline is "yes," if not only because of Facebook's size. Being so big, it makes sense that it's one of the largest referrers to anything.

I mean, India has almost as many English speakers as the US. These things happen when the population is a significant portion of the universe.

tldr;

My eye captured the 3rd row which shows google.fr is the 3rd (nearly 7%, while others are jut a bit more than 8%).

Wonder what are French habits in the subject which makes google.fr claiming that high.