• hansl@lemmy.ml
        link
        fedilink
        English
        arrow-up
        9
        ·
        1 year ago

        There’s nothing particularly pretty about this graph. It’s basically an infographic that has two data points and a bunch of lines.

        My 3 year olds made prettier graphs from “join things on the left with things on the right”.

        In the spirit of this comm, the data itself is secondary to the graph itself. So replace the text with gibberish and ask yourself if it’s still a nice graph. It isn’t particularly beautiful, no. You can feel strongly about the topic, but that’s not important.

        • inspxtr@lemmy.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          1 year ago

          what would you have done differently to communicate the data then? assuming the numbers are correct.

          • haohao@lemm.ee
            link
            fedilink
            arrow-up
            6
            arrow-down
            1
            ·
            1 year ago

            Use more than two data points. It is possible that online dating was huge in 2017, but not before or after it. It’s also possible that online dating was much bigger in 2015, and went down in 2017. This graph paints a fake tendency.

          • hansl@lemmy.ml
            link
            fedilink
            English
            arrow-up
            3
            ·
            edit-2
            1 year ago

            As haohao said, more data would make for more interesting lines. Also, since the data should add up to 100%, maybe use a stack graph? Don’t use straight lines. I would also try to experiment with pivoting the data; show evolution over time of a single trend (in multiple graphs). Merge a bunch of low percent items into “other” to clean up.

            Just ideas, making a great looking graph is mostly art.