Lies, Damn Lies, and Statistics.

The Truth Laid Bear

Someone mentioned last month that I was GunBlog #84 at The Truth Laid Bear.
(A "Links only" ... no original content ... wbsite.)

I found that mildly interesting, and ignored it.

Later, I formed the practice of revisiting the TLB specifically to check my unofficial rating, and noticed that nothing changed, no matter how much activity I actually had on this website.

Eventually I investigated the basis on which these ratings were formed, and I discovered two factors of which I was not formerly cognizant:

1. At least four of the Gunblog websites which were rated higher than Cogito Ergo Geek had not posted new material in over a month!
2. The ratings were based on "SITEMETER" ratings, and whatever the criteria for being included in the list, the ratings were not updated unless you were subscribed to the SITEMETER service.

Because I didn't subscribe to the SITEMETER service, I found it curious that I was rated at all, and even more curious that there were Gunblog websites with even lower ratings!

There must be factors of which I unaware, which affect these ratings.


Since the 'Basic' SITEMETER service was available as a free subscription, I subscribed four days ago.

The first day my 'metrics' rating jumped from 0% to 4000%.
The second day it dropped to 2000%.
The third day it dropped again, to 1200%. And today, it sits at 1100%.

(My "Inbound Links Score" remains unavailable, due to an undefined "Error in loading data". I suspect that it's because nobody has linked to any of my posts in the past four days, but I'm not certain.)

The ratings are based on the number of links from OTHER bloggers, plus the number of posts per day, plus the number of links to posts receiving traffic. I think. Okay, I don't really know the algorithm, but that covers all of the factors. My current rating is based on links + traffic over the past 30 days, and until 4 days ago SITEMETER hadn't recorded any posts. Which is to say ... these ratings are bogus and will remain so for the next 30 days (less the 4 days I have been subscribed to the SITEMETER service.)

I realize that this sounds like a vanity blog at this time, but there is a point to be made here:

Statistics mean very little, except in the broadest terms. For example, there are many gunblogs which are written by much more skilled, and more interesting authors; they are at the top of the list, and they have more readers in a day than I have in a month. Or sometimes in a year! And that's an accurate relative evaluation.

I'll keep my SITEMETER subscription, if only to see what happens at the Traction Control ratings sidebar as my traffic statistics are continually updated.

But when YOU read statistics (about ... for example ... gun control and the effects of new gun control laws on crime), I hope you remember that 'statistics' are just numbers. Due to many factors, including data-gathering techniques, they may be not only misleading but actually wrong in the sense that they actually portray complex socio-political situations.

The people who provide this data won't tell you about the details of either the data gathering or the evaluation of significant events. But you can never be far wrong if you assume that the statistics are like politicians' promises.

NOTE: I don't mean to criticize the statistical evaluation offered at Truth Laid Bear. This is a 'just for fun' unofficial poll, and nobody is going to take it very seriously. I understand that TLB is only offering access to some random statistics and there is no effort to enforce rigid adherence to data-collection paradigms. I appreciate that TLB has included me in the poll, and I hope that I remain in the database.

I only mention this to prove a point without assigning nefarious intentions: in this world, there are three things we need to remember, and that when we evaluate even the most benign evaluation of date, there are three and only three categories EVER presented.

Lies, Damn Lies, and Statistics.