Paraphrasing Barry Wellman, "The only people measured more than academics are ball players." I wonder if sports data has ever hit the information visualization community?
This line of thought was motivated through getting stuck in my office working on some research papers over the weekend. National Football League playoffs started on Saturday, and I have a rooting interest in the Washington, D. C. franchise who played that day. I didn't have a radio, and really couldn't afford -- the time or the money -- to subscribe to the online game video. So I pulled up the real time game tracker at CBS SportsLine. It worked great. Not only do score updates get pushed to the browser, but a complete log of every play is recorded, numerous stats compiled, and realtime running commentary from a stringer is provided. Yowsa!!
A realtime viz of all the data generated would be interesting, but it also appears that the URL structure for the entire season is relatively easy to understand. A crawler to grab the entire season's pages could be knocked off in a day or two, and you could have a nice data set of real world data to try and visualize. Of course availability on the Web means the raw data is somewhere to be had, but I'm sure someone (read NFL) will charge you an arm and a leg for it.
Why would anybody want to do this? Major PR and commercial possibilities aside, the pool of folks who would want to participate in rigorous experiments would be pretty good. Plus, you could try out expert vs novice style designs, with a relative plethora of, self-identified ;-), subject matter experts available. And ala Martin Wattenberg's experience with the Baby Name Voyager (PDF), I could see hordes of Monday Morning Quarterback's socially working their way through complex data analysis.
By the by, the Wattenberg paper is a quick, good read.