Guernsey Press

This guy averaged the flag of every country to come up with one for the whole world

The project created a mean, median, and modal flag for every continent by amalgamating current designs digitally.

Published

A flag of the world which averages out features found in current designs from around the globe has been created.

The end result shows muted colours with a three-by-three grid clearly visible – created by the plethora of flags which have three vertical or horizontal stripes.

It has been created by a software engineer who made the average design – also known as the “mean” – by digitally stacking existing flags.

The mean version of World flags (Udzu)

The engineer, who uses the Reddit handle Udzu, digitally analysed the designs of the 195 flags of countries recognised by the UN as members plus observers for the project.

But he didn’t just make one flag with his data.

The modal version, which looks at the most common colour at any given point on a flag, resulted in a design which was very red, with a blurry white rectangle extending from its centre and a smudge of blue.

The modal version of world flags (Udzu)
The modal version of world flags (Udzu)

The whole project only took a few hours, Udzu told Press Association.

To start, he pulled in the flags of the world, as used on Wikimedia and made sure all the flags were in the same proportions.

Then he standardised the colours in the case of the mean and modal flags and converted to greyscale for the median flags.

“The mean corresponds to blending all the flags together, a bit like a projector,” Udzu explained.

The mean version of North American flags (Udzu)
The mean version of North American flags (Udzu)

“The median is a bit less intuitive. It roughly corresponds to the most typical brightness at each location.”

The mean version of European flags (Udzu)
The mean version of European flags (Udzu)

In a second version, Udzu used the same methods but weighted the flags for a country’s population and area.

The source code for the project has been shared on Github.

Sorry, we are not accepting comments on this article.