from Hacker News

Towards a Theory of Bullshit Visualization [pdf]

by polm23 on 9/26/21, 7:50 AM with 14 comments

  • by spenrose on 9/26/21, 1:59 PM

    "In this unhinged rant, I lay out my suspicion that a lot of visualiza- tions are bullshit: charts that do not have even the common decency to intentionally lie but are totally unconcerned about the state of the world or any practical utility. I suspect that bullshit charts take up a large fraction of the time and attention of actual visualization pro- ducers and consumers, and yet are seemingly absent from academic research into visualization design. ... It is my contention that many (most?) visualizations are bullshit in either the Frankfurt sense of attempting to persuade while being totally disconnected from notions of truths about the world, or in the Graeber sense of being created for no readily apparent purpose and to no real useful end. When they inform, these bullshit charts do so in the most superficial and unsatisfying ways— for instance, a bar chart might tell me that so-and-so many widgets were sold during a particular period of time"

  • by AlbertCory on 9/26/21, 3:33 PM

    So disappointed :) I was hoping this was going to be a visualization OF bullshit, not a rant about bullshit visualizations.

    All seriousness aside, Correll fails to realize that the layout, length, and organization of your text are also forms of visualization, and at those, he fails completely.

    "It is my contention that many (most?) visualizations are bullshit in either the Frankfurt sense of attempting to persuade while being totally disconnected from notions of truths about the world, or in the Graeber sense of being created for no readily apparent purpose and to no real useful end."

    So he's shown us some erudition by throwing in Frankfurt & Graeber. Whoopee. When he mentions Tufte, a pioneer and a giant in his field, he's dismissive.

    All that said, there is some worthwhile stuff in here. He needs to study The Elements of Style, hire an editor, and reduce it to a readable three pages or so, single-column.

  • by raphlinus on 9/26/21, 3:15 PM

    A more sophisticated exploration of this topic is Calling Bullshit[1], by Carl Bergstrom and Jevin West. That's a book, but the linked website also has a lot of freely available resources. It's been taught as a college course - week 6 is data visualization.

    [1]: https://www.callingbullshit.org/

  • by kowlo on 9/26/21, 2:49 PM

    I feel this would have been more suitable as a blog post rather than a two-column layout under a "research" subdomain. I wonder if "Tableau Research" has a rigorous peer-review process for its research publications.

  • by OliverJones on 9/26/21, 11:47 AM

    With respect, this author misunderstands Edward Tufte's use of the term "Chart junk." Chart junk isn't just pointless decoration on charts. It's also misleading or missing scales, deceptive or inept ways of representing data, and so forth.

  • by rjuyal on 9/26/21, 6:41 PM

  • by gavinray on 9/26/21, 6:49 PM

    Thanks for sharing this, gave me a smile and a good chuckle.

    Some other people who are knowledgeable in the domain have posted that there might be some rigor issues.

    But I think a lot of us can empathize with the feeling of "No you idiot, that's not how you use that at all!" when it comes to some area we know more about than the average person.

    I'm pretty sure my car mechanic winces every time I bring my vehicle back into the shop. That's what I pay him for I guess.