• Home
  • Blog
  • Olympic Data Visualization Mistakes

Olympic Data Visualization Mistakes


I came across this visualization around the time of the London Olympic games back in the summer of 2012. My first thought was “Cool! A data visualization using the Olympic rings!” But upon closer inspection, I realized there were some problems.

oceaniaeuropeamericasafricaasia from gustavo sousa on Vimeo.



The most glaring issue I noticed, is fact that nothing is labelled. What continents correspond to which colors? I’m pretty sure that Red=Americas, Green=Asia, Yellow=Africa, Black=Europe, and Blue=Oceania but I’m not entirely sure. Also, there are six Olympic participating continents (North America, South America, Europe, Asia, and Oceania) and only five rings. Lumping two together (in this case it’s South and North America) is definitely going to skew the results.

Admittedly, fixing the second issue would make the visualization less Olympic themed; however, the use of circles is problematic. Yes, the rings are great for their symbolism, but, circles are much harder to compare than bars. This visualization is morphing, and so it could have morphed into a bar chart that would provide the viewer with a little more context and make it easier to compare.

Context is extremely important in data visualizations, because, without it, we can’t accurately compare results. For instance, two would be twice the size of one even though the numerical difference between two and one is small. Without a scale, the visualization can be misleading, and the viewer can draw false conclusions.

Speaking of context… where are the numbers in this visualization? If we can’t have a bar chart, can we at least get some numbers to understand the true variance from continent to continent?

There are more contextual issues involving the data that is being presented. Are the “military expenditure”, “gun ownership”, and “homicides” metrics measured per capita? Would it not provide a more accurate picture if all of the metrics were measured according to the same scale?

Looking at this from a data visualization perspective, there are a lot of areas for improvement: labels and legends are missing; two continents are lumped together (we think?) instead of excluding one or representing it another way; circles were used instead of another, clearer visualization; and, finally, a lack of contextual information. Looking at it from an Olympic fanboy perspective… it’s still cool. legend, excluding one of the continents instead of lumping two together, the use of circles, and a lack of contextual information. Looking at it from an Olympic fanboy perspective… it’s still cool.

Dundas Data Visualization, Inc.
500-250 Ferrand Drive
Toronto, ON, Canada
M3C 3G8

North America: 1.800.463.1492
International: 1.416.467.5100

Dundas Support Hours: 7am-6pm, ET, Mon-Fri