I’ve already confessed to being a web stats junkie. There has been no improvement in this area since my confession. Rather, as the statistics get more interesting, my addiction grows ever worse. In fact, my web stats have become one of my primary means of finding new and interesting websites. I let them find me. And, naturally, sites linking to aov tend to be cool and interesting.
Every few weeks over the last year, a few oddities gave crept into our referrer logs here at aov. Most peculiar, by far, is http://www.iaea.org or the International Atomic Energy Agency. To be clear, as far as I can tell, the IAEA has never actually had a link to Acts of Volition from their website. It appears, as is sometimes the case, that our web stats have been fooled into thinking that a website is sending visitors our way.
This happens occasionally (some javascript forwarding functions can cause it), but usually from sites that an aov reader visits regularly. Are any of you members of the International Atomic Energy Agency?
The International Atomic Energy Agency is apparently a wing of the United Nations that “serves as the global focal point for nuclear cooperation”. Acts of Volition is all about cooperation, but not so much about atomic energy. We prefer more photogenic energy sources in the hopes that we’ll end up better off than these guys.
If you’re interested in finding others who find us in some way notable, see this list of some of the most common referring pages to actsofvolition.com.
Older versions of some browsers also give bogus referers (Netscape 4.something on either the Mac or PC…can’t remember exactly). When someone types an address into the location bar or selects a bookmark, these busted browsers, instead of sending the proper referer (i.e. nothing), they send the URL that the browser is currently displaying.
That may be where your IAEA referers are coming from…an IAEA employee that has their busted browser’s home page set to iaea.org but visits aov regularly via a bookmark.
Fun fact: referer is a correct misspelling.
and a misspelling too oft learned (after being forgotten) in a coding language of choice.
As for your habits steve, I am becoming more and more entertained by the same phenom… One of my pages links to random sets of weblogs (webloggers are generally Referring site junkies), and thanks to some handy coding I can see that there is a 3/5 ratio of sites that are visited from the list which are revisited at least once by the owner… It’s a mini culture underneath all those logs.. err.. blogs?.
Yeah, jason, I figured it was something like that (my guess was a site with javascript location ‘back’ links – but your explanation sounds more likely). I find the prospect that someone visits both Acts of Volition and the International Atomic Energy Agency site quite amusing.
Jevon’s right – it’s all reciprocal. You link to Google, and they link right back.
Oddly enough, I see the same site as one of my referrers. What does that say about AOV and kalsey.com that for some reason people visit IAEA first?
I’m getting the same thing. And, suitably, I found this discussion via my referrer logs.
I have also heard the browser bug can take effect when someone has a page up in a different window: i.e., our atomic friend keeps a work-related site on hand as s/he surfs blogs, in case his boss strolls by.
Okay, nuke-boy, where are ya?
I’ve got IAEA referers, you’ve got IAEA referers, we’ve all got IAEA referers — I’ve checked all of the logs for all of my sites, and there are IAEA referers in all of them, all with a User Agent of “Mozilla/3.0 (compatible)”. My suspicion is that there’s a browser or robot out there (looks more like a browser — it’s not repeated requests for pages, just one every now and again) that just has it’s referer string hardwired to “http://www.iaea.org”.
An Easter Egg of sorts? 🙂
Ok, somebody explain this one. Here are two hits from my tracker for cfudge.com from Feb 1:
13:16:42 cacheflow.fhwa.dot.gov 169.135.1.5 Mozilla/3.01 (compatible)
13:16:42 cacheflow.fhwa.dot.gov 169.135.1.5 Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0; .NET CLR 1.0.2914) 1152×864 (24 bits)
Exact same time, two seperate hits, two different setups, and one appears to be something to do with .Net ??
An innocent question – does this undermine the validity of all your statistical information? If this tracking is false – which are not?
Fair question Alan. We can be quite sure of the accuracy of most web stats. For example, when I see a list of referring websites, I can check out the referring site and more than 99% of the time, I’ll find a link to my site. Stats confirmed.
Web stats in general tend to be a bit fuzzy though. Terminology is a killer. The word ‘hits’ has been robbed of any meaning it onces had by being used differently in different stats packages. ‘Hits’ tends to mean every single request for a web server (including HTML files, CSS files, and most significantly, image files – meaning that one ‘page request’ to a site like CNN.com would involve about 30 ‘hits’. Beware of people claiming millions of ‘hits’ – that doesn’t take long.
Identifying browsers and operating systems is a bit of voodoo too. Every browser identifies itself to the server with a user agent string. Basically a line of text that is part english, part gobbledegook. Here’s an example:
Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; AT&T CSM6.0; Q312461)
Do actually tell which browser and operating system you are using, you have to be able to extract that MSIE 6.0 is Microsoft Internet Explorer 6.0 and Windows NT 5.1 actually means Windows XP (interesting, eh?). The AT&T CSM part probably has to do with the browser being customized and downloaded from an AT&T ISP.
Not all browsers identify themselves as clearly.
Mozilla/4.7 [en] (Win95; I)
This on is from Netscape 4.7 on Windows 95. Notice that both strings have ‘mozilla’ in them.
I’ve seen lots of different software for parsing these strings into usful information, but all of it requires constant updating when browsers and operating systems upgrate (who could have predicted that Windows NT 4, Windows 2000, and Windows XP would be identified as Windows 4, 5, and 5.1 respectively. It would be nice if there was a centrally available open database of tying these agent strings to their actual OS and browser – rather than a bunch of splintered efforts for each individual stats package.
Does that answer your question, Al?
By the way, it’s funny to see that I’m not the only one being hit by the IAEA (or at least having my stats report it). Thanks for posting everyone.
Ok you answered that question so I will place a bizzare conspiracy theory on the table and preface doing so that I have no belief in it but want to know that it is not possible. The US government wants to know how big the internet is and maybe what is being done on it and by whom. It has a system of determining these things but knows that to use the internet is to leave a trace. It sets the trail as being left by a rather innocuous UN body which when noticed only causes mild flattery and curiosity.
While your theory has the ring of truth (especially the part about ‘mild flattery’), I don’t think we need wacky consipiracy theories when there is an obvious robot alert!
…but don’t you see, Mr. Garrity, the government is run by robots and they are using the internet to find out all about YOU!!!
Island Edition used to get IAEA refs every once and a while.