
This question is for testing whether you are a human visitor and to prevent automated spam submission.
I would like to add one of my finds (if it's already mentioned in the original post, sorry for the double) - the National Atlas of Japan (1977) [1].
Take a look at the railway traffic statistics [2]. The visualization there must have been painstaking to make.
[1] https://www.gsi.go.jp/atlas/atlas-e-etsuran.html
[2] https://www.gsi.go.jp/atlas/archive/j-atlas-d_e_49.pdf
edit:
The earthquake epicenter visualization is also worth mentioning:
On the subject of great-looking maps I highly recommend the second Austro-Hungarian Military Survey [1] from the 1860s-1870s, some of the details in those maps are truly exceptional (one of my personal projects consists in mapping the forested area of Romania's territory from nowadays using those maps).
A little less accurate but more beautiful is the first Hapsburg Military Survey, from the late 1700s [2], when and if I'll ever get a bigger house I'll definitely hang some prints of those maps on the walls.
[1] https://mapire.eu/en/map/europe-19century-secondsurvey/?laye...
[2] https://mapire.eu/en/map/europe-18century-firstsurvey/?layer...
Those are some beautiful visualizations. How did they do it in 1977, considering modern software like Tableau don't exist? The amount of expertise that goes into this must've been substantial.
In that timeframe, most visualisation was hand-drawn (especially on maps which were all hand drawn).
I have long been fascinated by timeline designs. So this is right up my street. So impressive what these artists of old managed to create without access to computers and design software.
For more modern timeline visualisations, you might be interested in this list of timeline designs I compiled:
https://www.tiki-toki.com/blog/entry/ten-amazing-online-time...
For timelines, the book "Cartographies of Time" [0] has lots of historical examples.
There is also a recent presentation of the design space of timelines in "Timelines Revisited: A Design Space and Considerations for Expressive Storytelling" [1, 2]. This then developed into the Timeline Storyteller tool [3].
[0]: https://papress.com/products/cartographies-of-time
[1]: https://timelinesrevisited.github.io/preprint.pdf
What I have learned about data viz over the years is that it is a very powerful solution to problems, just as Nightingale's diagrams demonstrated the effect of hygiene, which changed how societies and governments respond to epidemics. What I have also learned is that a solution obviates a dynamic someone thinks it is their job to manage, and a visualization that solves a problem is its own problem.
The "best," diagrams show change over time, and provide the presenter with a way to demonstrate how they are the important pivot point that optimizes and drives that change. The "worst," diagrams are the ones that illuminate the problem in such a way that it is no longer difficult, which humiliates the people it was presented to and designed to help.
I recommend using data viz privately, to reason through and solve problems quickly, and then use the time you save for self investment. The real value I think is to use data viz not as a product for productivity, but as an arbitrage tool for leverage.
Do you have any examples of these "worst" diagrams?
WebOWL Ontology tools, most Sankey diagrams, digraphs, radial dendrograms in d3, and pretty much anything that can represent more than a couple of degrees of freedom.