Sobering news, but a great visualization

I’ve seen a lot of visualizations around COVID-19 and particularly like an approach that John Burn-Murdoch has been developing and improving over the past couple weeks for the Financial Times. Here’s an example from today:

Highlights for me are that this approach:

  • Looks at the raw time series data in a different way, creating cohorts by setting the x-axis origin as the day of the 10th death reported in each location.
  • Represents regions such as cities or states/provinces, which seem more relevant to epidemiology than larger geographies like countries.
  • Focuses on deaths reported instead of confirmed cases which can be influenced by differences in testing so seem less usable for cross-location comparisons.
  • Uses a log scale effectively, complete with effective visual indicators for orientation and a helpful explainer.
  • Is updated daily and supported by an active back-and-forth on Twitter.

The result combines a lot of data, a different perspective, and good design to give more insight into pandemic trajectories than everything else I’ve come across.

There are variants of this chart for other geographies, plus more, on the FT’s coronavirus latest page. Check it out.

Web Seer

A few months ago two data artists, Fernanda Viégas and Martin Wattenberg, released a beautiful visualization tool that uses Google Suggest to create an addictive way to explore the world’s searching. I came across this while looking at their 2003 project, History Flow, about visualizing edits to Wikipedia. The example below shows Google’s suggestions to complete the phrases “are diets…” and “is chocolate…”, with the size of arrows and words showing Google’s count of how many web pages address each completed question.

Continue reading “Web Seer”

Map of Wikipedia usage worldwide

Last week I did a short talk at TED on Wikipedia’s evolving impact. I’ve posted an expanded version of the slides (PDF) and want to use a few blog posts to elaborate on some of the points covered.

First point for this blog post, I was looking for a way to visualize some the great analysis Erik Zachte did recently on the geographic source of traffic to Wikipedia. A trip through Commons pointed me to a slick online mapping tool and Erik was incredibly helpful at providing me with the data I needed. Here’s the map we came up with, which represents average monthly Wikipedia page views per internet user during July, August and September of 2009:

Continue reading “Map of Wikipedia usage worldwide”