Sunday, November 13, 2005

In honor of post #40, here are some observations and statistics relating to this "posting" activity. In a way, this commemorates post #1 which told you your likelihood of being dead given you lived your entire life in the year 1997 CE (or 1996 or 1992 I don't remember). I didn't know how to add charts, then. Maybe I should revisit that topic.

Notice that here there is no incorporation of completely subjective properties such as "density" (e.g. words per post) or "quality" (e.g.. readability). These are irrelevant and probably interact with one another in unpredictable ways (e.g. high-density posts may tend to be of low quality).

First, how did we get to 40?

As we can see here, we started on January 28th (of 2005 of course) and proceeded to add posts until today, which is November the 13th. The period on the x-axis is 4 weeks, since I couldn't figure out how to mark months.

Now, what about the relative intensity of posting over time? Well, to look at that we see what the frequency of posting is over time. This is like taking a derivative of figure 1, except you do it in Excel instead of knowing how to do derivatives.

Here, we can see that posting comes in spikes, which I will call "spikes". These spikes are then followed by a slow decline to what seems to be a baseline posting rate of around .075 posts per day or "PPD". If you wanted, you could go back through the archives to see what meaningful activity these spikes correspond to, but you probably won't want to do that.

Big, sharp spikes correspond to sequential posts, from one day to the next (i.e. 1 PPD). If I posted twice in one day, that would be 2 PPD, which actually happens once at the beginning of October, but I clipped that value to accentuate other more subtle events.

Notice, however, that posting rate during the summer months was almost uniformly slow, with a little bump around my birthday! Interesting. And there was a big spike toward the end of August, and we all know what was happening then.

If I really wanted to bore you, and if I really wanted to not-do-work, I could do a Fourier analysis of the above frequency plot, and we could talk about that. I would probably need a few years worth of data for that to be worthwhile, though, so you are spared. For now.


  1. Andrew, I have not read your post yet, but I want to throw in an early comment that it pleases me to see graphs in your blog. I don't know why, but seeing those graphs makes me tingly to read what you have to say.

    Okay, now I'm going to read it.

  2. Best. Post. Ever.

    Dammit, now I'm going to have to do some more scientific posts.


  3. What a fantastic post! It was fully of information I did not need and had never considered before, yet was entertained by the learning of. Good work!

  4. It sure was some good old fashioned science sounding stuff you wrote Mr. Andrew. I don't know if I got tingly, but it was facinating as Mr. Spock would say. Facinating...

  5. I love it when Ian comes to visit and post! Hi Ian!

  6. I think I speak for us all when I say that I'm ready for post number 41!!!