Football rankings 1.1

Long-time reader Jon sent in a different view of the QB data.  He uses a nifty tool in Excel to generate a parallel coordinates plot (also called profile plot) on which pairs of QBs can be highlighted and compared.

This chart exploits the foreground background concept very nicely.  One way to deal with abundant data is to highlight only those bits that matter to the question at hand, and relegating the rest to the background.

The gray lines in the background provide context without grabbing undue attention. He also converted every metric to a scale between 0 and 1, similar to what we did with our version.

The Eli Manning / Philip Rivers comparison shows that both QBs were below average on most of these metrics, with Manning near the bottom of each.

Comments

[Alan De Smet]

A line graph seems wrong here. It suggests that the data is a continous stream, which it clearly isn't. I can't look at the graph and conclude that Manning NYG is probably near 0.3 between td (Touchdown?) and int (Interceptions?). That extrapolated data point is meaningless.

[derek]

Lines are nothing but devices for visually connecting points. The only reason for feeling uncomfortable about them because they "suggest" a stream is because of the pre-conception that lines suggest a stream, which is circular reasoning when you put it that way.

To see why lines work well here as information visuals and other devices don't, simply redesign the view above so that the data markers are not connected by lines. The graph is now "kosher" according to the view of lines as streams, but the information is harder to see.

That can never be right: info vis works well when you can see the information, and badly when you can't. Any style rule that forbids the better visual and mandates the worse one is a toxic style rule, and should be dropped.

The profile plot is based on the work of Alfred Inselberg (where he calls them parallel coordinates plots), and is now, rightly, well-established in the business as info vis tools (e.g. in Spotfire), as discussed in this pdf by Stephen Few.

I'm a former skeptic who is now a convert, since I've seen how useful it is for visualising large numbers of records in large numbers of dimensions. In theory, orthogonal coordinates would be better (i.e. scatter graphs), but we only have two, or a little more, dimensions available to us to use orthogonal coordinates in. Parallel coordinates get around this restriction entirely.

[Jon Peltier]

I've written a tutorial showing how to create a chart like the one Kaiser posted:

http://peltiertech.com/Excel/Charts/ParallelCoord.html

It includes a link to the working chart Kaiser displayed above.

[college football team names]

Protective clothing is an essential feature of modern football, but this makes the players vulnerable to heat stroke. One study reported 12 heat stroke deaths among college and high school football players over a three year period.