Compare this to the boxes-and-arrows diagram used in the 4700-word CRS Manual.

What gives the interactive chart its punch is the use of verbs to describe the connections between the elements. Verbs like “contain”, “create”, “perform” are contrasted with “are part of”, “are created by”, “are performed by”. These words identify the relationship between subjects and objects in a much more informative way than lines with arrowheads.

While 2011 support for current common browsers is the most useful view, Irish and Manian have provided data for 2008, 2009, and 2010 as well. In the slideshow below I show a screenshot of each of the four views. It makes a nice animation.

• 
• 
• 
• 

The visual is created with HTML5 and CSS3, so it is best viewed with an current browser. Don’t even bother with MSIE 7.

(via the LinkedIn Web Standards Group)

Each spread is devoted to a single pasta, and explains its geographical origin, its process of manufacture and its etymology – alongside suggestions for minute-perfect preparation.  Next the shape is rendered as an equation and as a diagram that shows every distinctive scrunch, ridge and crimp with loving precision. Finally, a multi-page foldout features a ‘Pasta Family Reunion’ diagram, reassembling all the pasta types and grouping them by their mathematical and geometric properties!

Check out this one for Cavatappi:

Many of the pasta shapes are diagrammed on the Z-axis (a d/D favorite!), showing the delicate shapes in their full undulating glory (view more pasta diagrams on the NY Times site). I’m hungry already.

The key to the masthead is a nice example of information design in itself.

Click through to read the labels.

p.s. Flip Flop Fly Ball creator Craig Robinson has a book out. Good stuff. I’ve linked to him before.

“Mapping spatial information reveals part of human history that otherwise we couldn’t possibly know,” said Anne Kelly Knowles, a geographer at Middlebury College in Vermont. “It enables you to see patterns and information that are literally invisible.” It adds layers of information to a map that can be added or taken off at will in various combinations; the same location can also be viewed back and forth over time at the click of a mouse.

The real joy of this feature is the portfolio of projects that accompanies the main overview. Here, for example, is a section from Ms. Knowles’ viewshed analysis of what General Robert E. Lee could actually see in the Battle of Gettysburg:

The pale ovals represent areas that historians have previously assumed to be visible to Lee. In Ms. Knowles analysis, all the light areas of the map could have been visible, depending on tree lines.

Quote:

The number of colors doubles every 28 years!

This is just the setup. For the real fun, see where Morley went next:

(via Chris Wild’s fabulous How To Be A Retronaut)

Andy Greenwald, writing about HBO’s reuse of character actors in different original series, posted a diagram of “The 66 Busiest Actors on HBO”.  The diagram links actors to each series in which they have made three or more appearances. On the left you might find Roxanne Hart. On the right you might find The Sopranos.

This is a chart of a type. It shows a network, but the assemblage of lines that denotes the network is indecipherable. It’s pickup sticks. (Other network diagrams devolve to spaghetti.)

Partly this is an artifact of organization. The alphabetical list of actors has no meaningful correspondence to the alphabetical list of shows. Imagine if shows were listed chronologically and actors listed in order of first appearance. Then you might see a pattern. Would it be enlightening? I’m not sure. A common problem with network diagrams is that the lines don’t aggregate into meaning. An individual line might tell you something, but only in its connection to a pair of nodes. And if you want to focus on individual nodes — an actor or a show — you don’t need a diagram.

Here, the big picture is not in the lines, but the dots. Scan either list and the diagram quickly informs you of something interesting: Stephen Toblowsky appears in a lot of HBO shows. The Wire employed a lot of actors. But not Stephen Toblowsky (no line).

Roger Doiron of Kitchen Gardeners International has produced this image of what the White House garden would look like “if it were planted to reflect the relative costs of the main crops subsidized by US taxpayers”:

The data is from the Farm Subsidy Database.

…a new interactive graph on The New York Times website invites readers to plot their reactions to two questions: How much of a turning point in the war on terror will Bin Laden’s death represent? (significant to insignificant), and What is your emotional response? (positive to negative).

The format is useful for commenters because they can easily click a square and answer two questions at once, and it’s useful for the casual reader, who can measure the feelings of the crowd at a glance. When you first visit the page, you can click on any square to see others’ comments or to plot your own—or, you can just watch for a few minutes, as I did, as random comments slowly float up and fade out from the mosaic.

To me the format is far more interesting than the opinions. The format shapes the aggregate results.

Given quadrants, there is bias toward adhering to a quadrant.

Given edges there is bias toward approaching the edges.

Given existing dots, I strongly suspect there is bias toward clumping.

Now that I’ve looked at this interactive a few times the other thing that interests me is how it would look as an animation. The Columbia Journalism Review article offers a screen shot taken much earlier than the one above. The patterns are already taking shape.

Randall Munroe, at xkcd, has created one of the more comprehensive attempts to show radiation risk by charting doses in blocks and associating them with specific examples. Depending on color each unit represents one of four values from 0.05 microSeiverts (blue) to 1 Seivert (yellow). A large set of examples in one color becomes a small unit of comparison in the next:

The chart reads in a clockwise circle; better would be a horizontal left-to-right for both data and key. Still, it is a grand effort that repays close reading.