October 22, 2011

Sex, Politics, Disasters and Cheating - What I’ve learned in 800 days of the VizWiz blog

It’s hard to believe that I started this blog on August 13, 2009, 800 days ago (yes, I know my first post wasn’t until Aug 17).  The time has absolutely flow by and I’ve learned so much. 

For example, I’ve never shaded an axis in lieu of a color legend until today.  Heck, I never even thought about until I saw a bunch of vizzes on Tableau Public that were using this style. I’ve always included the color legends, but it’s really not necessary on a dual-axis chart like this with only two colors.  You might even call the use of a color legend chart junk in this case.

From what this viz tells me, I should stick to some specific topics if I want to drive traffic:

  1. Sex
  2. Politics
  3. Cheating
  4. Disasters

Does that surprise you?  It doesn’t surprise me since those are topics that generally dominate the news.

image

One final stat of interest: 90% of the total traffic to my blog has occurred since I posted this viz as part of the Flowing Data contest about STDs.  Nathan must be a pretty popular dude.

4 comments:

  1. Andy, I find that using a font color one each axis heading is equally effective and improves data to ink ratio.

    ReplyDelete
  2. I wish I couldn've been at Tableau's conference. From the Tweets, it sounded like a great event.

    And what a great, simple, and effective idea, shading the axis to help the reader figure out what goes with what.

    BTW, what happened in ~ August 2010? Looks like your site went dark, or all your readers were on vacation :>

    Thanks for sharing!

    ReplyDelete
  3. Dan, are you recommending coloring the axis label or do you like the way I did it?

    John, the conference was incredible. I had the honor of speaking, which was a blast. As for August 2010, I didn't even notice there was a lull, good catch. If I recall correctly, I had a problem with the Google stats. I think I had made some changes to my blog template and forgot to turn the tracking back on. Or perhaps I was being boycotted. :-)

    ReplyDelete
  4. I like the data and story, and I would love the chance to remix your chart, because I don't think that a dual axis was employed well. My quick test to see if a dual axis is a good a choice: when the marks from different axes cross, or are above or below the other, is it meaningful?

    ReplyDelete