Journal

I don’t put a whole lot of stock in Alexa ratings…

25 Comments

I don’t put a whole lot of stock in Alexa ratings… unless, of course, they happen to be MY Alexa rating, and they happen to be “up and to the right.”

Like this one…

I’m not sure what’s going on. Webalyzer and Google Analytics both show growth, but not as sharply. I suspect that a group of spyware-illiterati all discovered my strip recently, and have been consuming the archives.

It’s interesting that you can SEE the “dip” in the graph at the time I started plugging Google’s “Switch to Firefox” campaign. Apparently a bunch of my Alexa-infected readers got Google-infected instead.

Ah, statistics. Just because they’re facts doesn’t mean they can’t also be lies.

–Howard

25 thoughts on “I don’t put a whole lot of stock in Alexa ratings…”

  4. I love statistics. They can tell you so much and point out things that you wouldn’t notice otherwise. But you are quite right on the “Just because they’re facts doesn’t mean they can’t also be lies” line. For statistics to be any real worth you need to know a huge amount of the context that surrounds them. Still, they are quite interesting.

  6. So does the “daily reach” thing imply that out of every million Alexa users, around 100 visit schlockmercenary.com on a given day, recently? (Yeah, counting spyware/robots/etc as “users”…)

    1. Right, so I actually went and checked, and that is what it means.

      The page-views statistic supports your hypothesis, about archive consumption at the very least – your daily page views per visiting user is over twice your three-month average. *waves hands*

  7. I was a Statistics major at BYU for a time, and one of the favorite sayings in the department was, “If you torture the data long enough, it will confess to anything you want it to!” Ah, Statistics… 😀

    1. That is an excellent quip, and needs to find its way into my quote file. Is there a specific attribution for it, or should I just go with “anonymous”?

      1. I don’t know who said it originally, so I’d just go with “anonymous.” The other version is, “it will tell you anything you want it to.” Either one works.

      2. Eminently Quotable

        Quoted from a following post by howardtayler: “Using statistics to get at the truth is like using a picture of a girl in a bikini to determine whether or not the two of you will hit it off.”

        Now there is a quote for your file better than the tired old “If you torture the data long enough …” that has made the rounds so often it’s bikini has worn thin!

    1. Actually, that’s kind of counter-intuitive. Once you’ve seen ANY fully-developed female naked, seeing one in a bikini is only different from seeing that same female naked in two ways:

      1) She’s less likely to be embarrassed this way.
      2) The details you have to fill in are probably being supported or contained, changing their shape somewhat.

      This is why a bikini is essentially “nudity,” and why a good evening gown can be so much more enticing/alluring. After all, with the gown, the bits you have to imagine actually require some imagination.

      In short, I disagree.

      –Howard

      1. Actually, I agree more with you, in an apples and oranges kind of way.

        In the case of ladies, some things are often better left to imagination.

        However, in statistics, we’re actually looking for as much actual information as possible, so we aren’t required to use imagination.

        1. Actually, in statistics you’re using information to generate an outline from which the rest of the picture can be imagined. This is fine, except that there are about a million details the stats won’t have covered.

          In that regard your comparison might be better stated as follows:

          “Using statistics to get at the truth is like using a picture of a girl in a bikini to determine whether or not the two of you will hit it off.”

          –Howard

  9. You might find Mint to be a nice stats analysis tool.

    Send me an e-mail (ceejayoz@gmail.com) if you want a demo, as I think the official site demo is disabled due to popularity at the moment.

    1. Yup. Not interested in killing time on it at this point — especially since Schlock doesn’t lend itself well to the necessary reformatting. But really it all comes down to “opportunity cost.” There are a dozen things I can do to make more money and reach more people than I can make/reach with Clickwheel, and every last one of them costs less time.

      –Howard

  11. You are almost at the same level as two years ago, when you were encouraging people to visit the site in IE with Alexa installed…

    1. Yup. Things fell off VERY sharply after that, and with my purpose accomplished I let them. I yanked the Alexa widget from the site with the redesign this last October, so there’s really nothing artificially inflating the stats at this point.

      The Google Toolbar push may have artificially DEflated Alexa stats, but there’s really no way to tell.

      –Howard

Comments are closed.