Rants tag

Rants, ruminations, and rambling remarks from my mad, muddled, meandering mind.

Friday, June 10, 2016

NBI: Regression Analysis of Posting and Page Views

Instead of Warcraft Movie hotness, my Dear Reader, you are going to get graphs today.

So it's time once again for the Newbie Blogger Initiative. In the spirit of delayed events that seem to be a running theme this year, I have neglected to discuss or promote the NBI until almost a third of the way into the month. I don't participate as much in NBI as I have in the past, for various reasons. But I do think it is worthwhile endeavor, bringing together the MMO blogging community, both old and new.

Statistical Navel Gazing

A little over a month ago, Ravanel Griffon celebrated her fourth blogging anniversary with a bit of analysis. Curious as to the impact of posting on page views (or "hits") Ravanel did some regression analysis, and decided posting more accounted for a little over 40% of any increased visits to her blog.

I was also curious, and so decided to do a little analysis on my own blog.
For the first few months after I started my blog, Google/Blogger was not tracking my pageviews, so that is the reason for the gap on the left end of the orange line. Unlike Ravanel, my two graph lines don't seem to correspond that much. In fact, I haven't even posted this second article of June 2016, and my hits for the month already higher than all but two months going back to June 2015. You may also notice, that I have been very inconsistent with posting over the years, some months, I have been very prolific; others, I've hardly posted at all.
Doing a little Excel magic, I learned a bit about  regression analysis on a scatter plot. I included my whole run from the time Google started recording my pageviews because I actually got a stronger correlation, something Ravanel did not find. The line represents the rough prediction on whether posting more get me more pageviews. The R2 score indicates that increased posting accounts for only about 15% of any increased hits. In fact, the analysis predicts that if I don't post at all, I should still get 4556 hits in the month. I did fail to post anything in February 2015, and got over 6000 hits.

Analysis of My Analysis

Now, I want to point out a few flaws in this sort of analysis. Belghast of Tales of the Aggronaut posted every day for several years, breaking his streak only recently. For him, comparing posts to hits in this way would be futile, the time line chart would probably be better to see whether he is gaining audience. Also, the ebb and flow of pageviews does not always correspond to calendar months. My hits for last month totaled almost 18,000, but most of that was in the second half of the month. That surge has continued into June, and so over the past thirty days, I have gotten almost 30,000 hits. I reached 10,000 hits in thirty days back in November and December of 2012, but didn't crack 10K in a single calendar month until the following August, when I had posted 60 times in two months (July & August 2013), and the next month September 2013 was a longstanding record for me, on the strength of a single post that got shared on Reddit.

My point in illustrating this is to say simply: Post what you want when you want. You will gain a small following of loyal readers, which will increase over time. Posting regularly will help, but is not a guarantee of popularity. Blogging is not a race, it is a series of morning jogs. And like those morning runs, it is mostly something you do for yourself.

[EDIT: Adding my sawtooth hits graph for the past day for Bhagpuss. It's been like this for weeks. The aberration about 9 a.m. is when I first published this post.]



~~~~~~~~
Creative Commons License
This article from I Have Touched the Sky is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. If you repost part or all of the work (for non-commercial purposes), please cite me as the author and include a link back to the blog.

If you are reading this post through RSS or Atom feed—especially more than a couple hours after publication—I encourage you to visit the actual page, as I often make refinements after the fact. The mobile version also loses some of the original character of the piece due to simplified formatting.

4 comments:

  1. Them some mighty charting powers I see :) Is that a blogger thing or did you pull the data into a third party tool?

    ReplyDelete
    Replies
    1. Microsoft Excel. There is a link to a regression analysis tutotial on Youtube in the paragraph just below the scatterplot chart. :)

      Delete
    2. I've long since given up trying to make any sense of page views. Last month was my second-highest ever, with just under 26k views. April was 18k, March 22k, Feb 17k, Jan 22k... this year's graph looks like the teeth of a comb. As far as I'm concerned I've posted roughly on the same frequency all year...

      Delete
    3. I added today's hits in graphic format.

      Delete