Hal O'Brien

The New York Times has a very interesting infographic: It takes the top 100 Netflix rentals, tells you their titles, and then maps them by zip code in cities like NY (natch), Boston, Seattle, Minneapolis, Atlanta, etc. In addition, hovering your mouse pointer tells you the top 10 in any given zip code, and also where the current title mapped ranks among the top 50 for the zip code.

Not quite Tufte-level info density, but very high. Also interesting for patterns -- very popular movies are popular everywhere, but in the middle things very much separate by neighborhood. Interesting to see, for example, that frat-comedy The House Bunny plays very well in the UW zip code. Something of a thumbs-up from the portrayed subjects. But there are also titles that play in downtowns and not in the 'burbs, and vice versa.

EDITED TO ADD: Three movies I'd suggest looking at the patterns for to see large contrasts are Paul Blart: Mall Cop, Milk, and the two titles by Tyler Perry. The comments page at the NYT had a great line: "It'd be interesting to see the correlation between areas where Mall Cop coincides with Palin supporters."

Another observation from the NYT comments -- Some people find it very difficult to imagine that just because they behave a certain way for a certain reason, others (including possibly a majority) might do different things for different reasons. That is, the idea of a theory of mind being something widespread is probably a very idealistic concept. (It would also go a long way in explaining why the Golden Rule doesn't work as well as it might.)

Usage tip: The maps can be clicked and dragged to show a somewhat larger area than appears at first glance. So for LA, one can see the San Fernando Valley and most of Orange County, for instance.

So.

Once upon a time, there was the Buckminster Fuller Dymaxion projection. It consists of 20 equal sided triangles, and was long the only map of the world you could get that you could fold into a solid (albeit not exactly a sphere).

While at Pomona College, Andrew Chittick and I sat and thought long and hard about the concepts behind the Dymaxion projection -- perhaps with the aid of Reinheitsgebot compliant products, perhaps not -- and we hit upon an idea:

If 20 sides was good, 32 sides was better. With more sides, you'd come closer to a sphere, and have less distortion at the center of each face. Plus, if you picked the right solid, you'd have an international symbol of unity and good times.

Thus was the O'Brien-Chittick Sakabol projection born, circa 1984 or so.

It being the mid 1980s, though, getting GIS databases and doing the CAD/CAM work would be hard. So it was punted.

I am pleased to discover that a parallel development has been found.

They describe the projection as, "Polygnomonic." Isn't that adorable?