Historigraph
An advanced new form of visual to help you understand complexity.
What is it?
A historigraph demystifies complex events in a way that conventional data visualizations or text cannot do, allowing you to rapidly understand situations that you might otherwise find overwhelming.
How it works
In short, a historigraph lets you read history like maps let you read space.
There are two conventional ways to show change over time. One is a line graph, and the other is a time series. The line graph reduces complex phenomena to a single variable, which does not facilitate understanding. The time series requires the reader to keep track of changes between iterations while filtering out the rest, which is beyond what human perception can handle.
The historigraph takes lots of related but non-scalar variables, such as the states or countries on a map, and projects them onto a single axis in a way that the distance between any two variables on the axis reflects how closely related those variables are. For instance, two US states that are close to each other on a map will be close to each other on X axis, as seen above.
This turns time into a sort of map. Phenomena that occur across space and time become visible. Your visual cortex can process these things much more easily than if you to read about them.
Background
While working for Seattle Sport Sciences on AI-powered video analysis software, I had to create tools for understanding the "thinking" of the AI because our program had users who were both smart and very physically-minded. They weren't going to take on faith what a computer program told them. This meant they had to work in a partnership with the AI.
My solution to this challenge was to take the numerous ineffable variables the AI identified, and visualize their fluctuations in a scrolling heatmap. But arranging the variables in arbitrary orders produced visual noise. I realized that algorithmically sorting them for relatedness worked to surface meaningful patterns.
In testing, users could almost "watch the video" by viewing the feed, allowing them to see patterns and hidden events they might not have noticed otherwise. Once the patterns were noticed in the data visualization, the users could actually see them on the video.
I wanted to create a generalized application of this technique that can be used even for situations without huge data sets, such as historical data, and thus was born the historigraph.
Business applications
The historigraph was successfully implemented with Seattle Sport Sciences, resulting in a product that has sold to several European soccer franchises. The examples below are "consumer" versions that depict history in order to demonstrate the principles.
News media can use the historigraph to keep viewers informed of the big picture rather than relying on anecdotal stories.
Site traffic and content can be tracked over time to identify patterns and events.
See examples
PUT LINK HERE