The Ebb and Flow of Movies: Box Office Receipts 1986 - 2007

The New York Times has posted a visualization of movie grosses from 1986-2007. The technique they use is called ThemeRiver which was developed in 1999 [PDF].

It's an interesting graphic but I do have some problems with it. The actual numbers behind it are not available in the graphic. The Times could have overcome this problem but didn't.

With this technique, instead of the height being relative to a base line (which would have forced some data elements to eclipse one another) the amounts are represented by visible areas. This means there is no simple way to plot the values on the vertical side of the graph.

The New York Times could have overcome this by making the interactive part show the numbers. Instead, it shows a synopsis of the movie and a link to the Times' summary.

The coloration indicates box office gross over time. But the color breaks are a little odd. A color change which would distinguish the $500 mark would make sense in this case. The lightest color designation is superfluous because it is effectively blotted out by its white outline.

While the grosses are calculated in inflation-adjusted dollars, it appears that the weekly box office is mapped in non-adjusted dollars. Just compare the big hits of the 80's to recent hits and you'll see what I mean. On the techie side, a better scroll bar and a zoom feature would be useful.

My only other complaint is more of a quibble. The smooth continuous curves suggest high resolution data that does not exist. These figures are based on weekly box office -- only four or five horizontal chunks per month. While there is an undeniable grooviness to the interpolated curves, revealing the data in its choppy weekly form might have yielded some interesting facts. But it would probably be much harder to read.

OK, all that having been said, it's a very interesting graphic. The visualization technique is intuitive and invites comparisons among the different elements -- search for "The Full Monty" for interesting example.

ThemeRiver reminds me of a more rigorous version of the Genealogy of Pop/Rock Music graphic discussed in Edward Tufte's Visual Explanations. There a lot of imaging techniques being developed and very few of them get much traction. ThemeRiver might just catch on. Being in the New York Times certainly helps.

via infosthetics, flowingdata and boingboing.