In a former post I wrote about the art exhibition "Bending Reality" that took place at Schloss Dagstuhl. The exhibition was part of a workshop about curved graph drawings. I am happy to announce that the organizers have set up a webpage that contains all the exhibits (including the nice metro maps of Maxwell Roberts). One of my favorites is the "Flowerpot" by Günter Rote.
Currently I am at a workshop in Dagstuhl about Drawing Maps and Graphs with Curves. Part of this workshop is the exhibition Bending Reality: Where arc and science meet, which is, well, about drawings of maps and graphs with curved edges. I will write about the exhibition in a different post. Today I concentrate on a series of beautiful drawings of metro network maps by Maxwell J. Roberts, which were shown at the exhibition. Max also gave an interesting talk about the ideas and the motivation behind his drawings. Often one has to distinguish between artistic design criteria and usability. I am mostly interested in these drawings as art, and there were quite a few exceptional drawings presented.
Since the workshop is about curved arcs, also most of the metro maps were using curved edges, quite to the contrary to the prominent octolinear design rule. Here is an example of the network map of Paris. The network in Paris is one of the densest in the world. At the first glimpse I thought that the curved version is a bit chaotic due to the absence of parallel lines. However I find it absolutely plausible, that it is easier to navigate compared to the official Paris map. Maybe most important, the ring formed by the lines 2 and 6 is not very visible in the official map, whereas it is the crucial feature in the curved map. Max claimed that people can navigate on the curved map 50% faster. Still half of the people would prefer the traditional map.
I am a big fan of using circular arcs in drawings of plane graphs. Right now contradicting studies have been published whether circular-arc drawings or straight-line drawings are more appealing. (In my opinion it heavily depends on the quality of the circular-arc drawing.)
A group from the TU Eindhoven (Arthur van Goethem, Wouter Meulemans, Bettina Speckmann, Jo Wood) is currently conducting a new online user study. If you would like to participate visit this link. (It will take 10 minutes.)
I close this post with a circular-arc drawing I made for the graph of the dodecahedron which uses the minimum number of arcs.