1. 0

  2. 3

    That’s a whole lot of words to describe why you should remove words from graphs.

    1. 3

      I will second that Tufte’s Visual Display of Quantitative Information was a life-changing book for me and my work. I’ve also applied it to slide design in my talks and I think it’s been very helpful for reaching the density of information I want to present without overwhelming the audience. Techniques port very nicely. Rather than “avoid distortion” and “small multiples” of graphs, because the audience can’t review/compare slides at their convenience, I have “repetition, emphasizing small changes”. I’ll build up a 5-15 line code sample over 2-5 slides so I can guide them through which parts of the code are ignorable boilerplate and which are the key concept. I’ll leave blank lines so I can add a line or two without any unchanged line moving. When I present a new variation, especially after a digression, I’ll show the original and give the audience a few seconds to confirm that nothing’s changed before I go into the variation. To build consistently towards a final form, my workflow has to go backwards: I start from the final code slides I want to present and then duplicate them, then edit the earlier copy to delete lines, then repeat until I’m back to starting the explanation with a slide of one or two lines.

      1. 1

        Agreed, Tufte’s great. So happy to see his ideas kinda still percolate into the mainstream (like sparkline libraries).