The COVID Charts

Observations on data visualizations of the coronavirus outbreak

The COVID Charts are brief critiques of data visualization and science communication of the coronavirus outbreak. They are not statements about the underlying science or public health policy.

If you would like me to critique a specific chart, get in touch.

latest charts all the charts

The COVID Charts -- Observations on data visualizations of the coronavirus outbreak -- Martin Krzywinski

▲ Inaccurate, sloppy and illegible . A slide from the presentation that explains the goals of community mitigation by comparing the time progress of daily cases with and without intervention. The chart projects that by “flattening the curve” deaths can be reduced from 1.5–2.2 million to 100,000–240,000. (White House Coronavirus Task Force, 31 March 2020).

31 March 2020

Inaccurate, sloppy and illegible

White House Coronavirus Task Force

critique

poster

background

[https://en.wikipedia.org/wiki/White_House_Coronavirus_Task_Force](The Coronavirus Taskforce) uses press briefings to communicate updates, guidelines, and policy changes to the public during the 2020 COVID-19 pandemic in the US.

core message

On charts that are illustrative, shapes, trends and proportions should be reflected as accurately as possible. This is easy to get wrong when not using curves from the source models but rather attempting to depict the data qualitatively.

key guidelines

1. Use an appropriate data generation model even for illustrative graphics.

2. Never deceive the reader by drawing proportions that don't accurately reflect your message.

3. Avoid textures and other chart junk.

critique poster

The COVID Charts -- Observations on data visualizations of the coronavirus outbreak -- Martin Krzywinski / Martin Krzywinski @MKrzywinski mkweb.bcgsc.ca