BSB’s BS Counter?

The Daily Bellwether’s Bill Sloat put on the hazmat suit and the super heavy-duty goggles, and made a journey over to Buckeye State Blog, where a previously unknown internal controversy boiled over at this September 7 post over this August 2006 post:

Doing the math, BSB’s Sitemeter report covered the past 150 days (1,045,225/6,969) preceding the post.

Sloat further found that the then-proprietor of BSB claimed to have had 3 million page views (that BSB comment is here; memo to Bill — Hyperlinks are your friends), or over 20,000 page views per day, presumably during that same 150-day period.

Your truly has an externally-sourced reason to doubt the 1,000,000 and 3,000,000 claims (click on the pic to open it in a separate window):

The above compares visitor traffic at BizzyBlog vs. BSB during the past two years; I changed the URL of Alexa’s canned 3-year result to 2 years and added a few graphical thingies to come up with what you see. Someone more expert than me might be able to make other modifications that will support the points I’m about to make with better visibility.

The red vertical lines represent the beginning and end of the 150-day period involved, as do the two ends of the double arrow. The maroonish line on the graph represents BizzyBlog; the blue line is BSB.

It’s more obvious on the big JPEG you’ll find here, but even the small view makes it pretty clear that BizzyBlog’s visitor traffic exceeded BSB’s during the period involved.

Yours truly happens to have a cockeyed counter (not visible to readers) that, in addition to real hits from real people, also records robotic hits, which come from search engines, advertisers, spammers and others who don’t have a real person sitting in front of a computer at the other end. The robots inflates the statistics I see by a factor of 2 or 3. I can tell you with certainty that although there may have been one or two days with over 7,000 visitors, BizzyBlog did NOT average 6,969 visitors per day during the period in question, even WITH the robots. The unfortunate reality is that my real human visitor traffic during that time period was no more than 10%-15% of what BSB is claiming.

As to page views, gluttons for visual punishment can look at this graph (BizzyBlog is blue, and BSB is maroonish at this graph). If you do, you’ll see that BizzyBlog and BSB are much closer in page views. This makes sense because of BSB’s multiple authors and forum-like design. Even at that, BizzyBlog (the blue one) still appears to edge out BSB. I am absolutely certain that I did not have a single day with 20,000 non-robotic page views, never mind averaging that number during the entire 150-day period (in, my, dreams).

As I understand it, Sitemeter, the counter BSB was using at the time, is, pre-programmed to exclude robotic hits. I don’t know if it can be done, but even if BSB were able to change Sitemeter’s parameters, based on my experience and what Alexa is telling us, I don’t see how BSB, even with robots, gets to almost 7,000 daily visitors or 20,000 daily page views during the time period involved.

I’ll leave the explanation of all of this to BSB’s current and former leaders. In the meantime, my assessment of the matter is a bit stronger than Bill Sloat’s, which is that “the episode — from what is known so far — appears distressing at least, dishonest at most.” Though of course we’ll never know, it sure would nice to know how much of BSB’s more recent growth was built on a foundation of what appear to be serious misrepresentations about its early- and mid-2006 popularity.

The folks at BSB have had plenty of less-than stellar exemplars at higher levels. Just one of them is here at tonight’s post.

______________________________________________

UPDATE: I am told that Alexa’s accuracy is open to debate. Here is what the Wiki entry says (Wiki links and footnotes omitted) –

There is some controversy over how representative Alexa’s user base is of typical Internet behavior. If Alexa’s user base is a fair statistical sample of the internet user population (e.g., a random sample of sufficient size), Alexa’s ranking should be quite accurate. In reality, not much is known about the sample and possible sampling biases. Alexa itself notes several examples. A known source of bias is the self-selecting, opt-in nature of Alexa traffic tracking software installation, but the significance of this bias on rankings is not reported.

Another concern is whether Alexa ratings are easily manipulated. Some webmasters claim that they can significantly improve the Alexa ranking of less popular sites by making them the default page, by exchanging web traffic with other webmasters, and by requiring their users to install the Alexa toolbar; however, such claims are often anecdotal and are offered without statistics or other evidence.

I am not remotely expert in these matters. It’s nevertheless hard to see how the above could explain a site graphing obviously lower than another one in visitors at Alexa — yet somehow generating 7-10 times as much visitor traffic over a five-month period.