The designers and developers had one hour to create compelling stories from raw data through informational graphics.
The event, held at WHYY and run by Plymouth Meeting consulting firm Acumen Analytics, was called Data Viz Wars. The competitors, who ranged from Temple students to professional programmers from Conshohocken agency Think Brownstone and Center City ecommerce analytics startup RJMetrics, were given access to the data 48 hours in advance, and could choose one of two sets of data: NCAA wins or earthquakes measured around the world.
The results spanned a range from maps measuring strength and frequency of tectonic disturbances to a color-coded dendrogram of each 2014 March Madness game. See the RJMetrics visualization, which won “Best Pro Team,” here.
Here are the top three things we learned from the teams who participated:
- Pick a metric, any metric. When handling large data sets, it can be tempting to sift through every aspect of it. But the best visualizations were the ones that isolated one or two metrics (ex: time of earthquakes, location) and focused on delivering some insight to users. One team almost didn’t come up with a final product because, according to them, they weren’t sure of what they wanted to say.
- Leave “breadcrumbs” in your design. Make your interactive piece easy enough to use without your explanation every time. The guys from Think Brownstone did that with their visualization of NCAA teams: the teams were represented by circles; the more wins per tournament, the larger the circles. By tracking the teams over years, you can see the changes in success based on changes in circle size.
- You’ll always want more time with the data. The programmers from RJMetrics — Ben Garvey, who built a comprehensive city budget visualization, and Austin Lopez — built a dendrogram that shows the wins and losses of each round of this year’s tournament — but they also wanted to highlight point differentials, in order to show users which games were “tougher” and more “interesting to watch.” The Think Brownstone team felt the same: “If we had an extra hour,” one of them said, “we would have been able to do so much more with the data.”