WHAT IS A GLYPH & WHY IS IT SO IMPORTANT?

I would venture to guess that everyone who has opened this has interacted with some form of data visualization before. Almost everyone has had some kind of training in Microsoft Excel and has probably at the least created a hand-drawn chart on graphing paper in science class. I would also venture to guess that not many people could tell me when modern data visualization was created. To shorten a very long story that we will cover another time, it was invented by a guy named William Playfair in the late 1700’s.

Can you believe that? MODERN data visualization was created 3 centuries ago, the craziest part being that it has changed very little since. We’ve seen many evolutions of how we create these types of graphs-- think, color coding, some minor 3D, heat-maps, box plots, time series line charts, etc. But there has been no real revolution of how we look at data since its inception. You may be asking, why do we need a new way to look at data? Or saying if it ain’t broke don’t fix it-- and there’s validity in that. Charts and graphs do precisely what we want them to… for the most part. They show you in a straightforward sense what's happening with your data-- Sales are increasing, weight is decreasing, these are the categories with the highest percentages, these are the places with the oldest populations, ect., and that's really helpful! But what if you want to know why? What if your “why” lived in the details of your data set and not just the aggregate? Insert the creation of the dashboard.

Screen Shot 2019-04-01 at 9.32.37 AM.png

One chart isn't enough? Just add more! We've had dashboards for years now, and there are lots of different software programs that give you fancy ways to create them. There is a problem with this though, and it grows-- no pun intended-- every single year. Every year as technology gets better, industries evolve, and we collect more data. To analyze it ”efficiently” we identify our main KPI’s and chart them out with the rest of our data. Whenever we ask another question of our data, we create an additional chart for our beloved dashboard. Best case scenario, dashboards now have up to dozens of graphical representation PER initiative, and we’re here to tell you that this isn't making humans any better at analyzing their data!

Maybe the fundamental concept of a dashboard is just not a good one. Hear me out-- If a car can't get you to where you are going, you don't attach another car to it to get you the rest of the way there, right? The best revolutions have come out of looking at doing things in a totally new way. For example, the car was invented because someone decided at a certain point faster horses just didn't make sense anymore. The plane was created because at a certain point ground travel was an inefficient way of going long distances. Thinking along these lines, shouldn’t there also be a revolution for how we’re visualizing data in a time where data overload is almost every industry’s main concern?

Enter the beloved Glyph! Back in the way back, about 20 years ago, a group neuroscientists and the U.S. Government spent many years thinking about this data viz problem. They recognized how inefficient humans were at understanding and sifting through data, which was especially problematic for people like intelligence analysts. Through many millions of dollars of initial research funding, they essentially concluded this: humans are terrible with numbers, not much better with graphs, but are exceptionally good at spotting patterns, trends, and anomalies visually using things like shape, color, and size. They researched this extensively and developed what is now known as a Glyph, named after ancestral civilizations who used pictures to tell stories (Get it? Like hieroglyphics?). Today, our version of a Glyph is a collection of geometric objects mapped to a row in a database like the one you see below.

GlyphEd Deck Update 2019.gif

Everything you see on that Glyph is mapped to different variables in your dataset. From the shapes to the colors, to the size of the individual objects, as well as transparency, rotation, and even animation.

Here’s where you say: Wait… you’ve lost me. How is this any less complicated than a dashboard?

The answer is simple. All of the shapes, colors, and sizes are things that research has shown your brain recognizes and understands significantly more accurate than numbers. Another example: you pull up to an intersection. In front of you is a red octagon on a pole with white letters in a language utterly foreign to you. What do you do? I bet you a dollar that you’d stop! Therein lies the magic of the Glyph. Your mind wasn't mapping that stop command to the words. It was mapping it to the red octagon-- a process that research has shown our brains processes sixty-thousand times faster than basic text alone.

When you map your own data to a Glyph, you’re allowing yourself to see n-dimensions of information at one time. Not 3, 4, or even 8- as many as you want/need to see are all floating in one place. Like placing a halo on top of the Glyph to denote religion or color coding a Glyph red to show its negative impact or even making it scale larger with the severity of the problem, the limits of Glyphs stretch to the limits of your creativity It doesn't stop there though. Now that we have a micro level picture of our data elements, we can circle back to that macro-level view traditional data viz used paved the way. Glyphs are placed in a 3D environment on your screen so you can zoom out from an individual and see hundreds, thousands, even millions of other glyphs all placed in space either correlationally using 3 axes (X/Y/Z), or geospatially using latitude and longitude coordinates over a map.

Screen+Shot+2018-03-01+at+12.20.15+PM.png

Here’s where you say: Well, that all sounds great but if this is the next big revolution in data science why haven't I heard about it?

The truth is, the technology was complicated to develop and expensive to distribute. Only governments and large institutions could use it before now. With advances in general technology and computer graphics processing, the cost has drastically come down. Even allowing us to introduce it through the web. In short, you’re hearing about it now because it’s finally ready for release on the world.

In a short time, Glyphs have helped discover millions of recoverable fraud dollars, identified factors to help keep students in school-- saving universities millions, analyzed millions of esports games to help improve players gameplay, assisted NFL teams better look identify player at risk of injury, found life-saving medical insights, and identified countless other anomalies. Companies are even working on using Glyphs to help understand why AI and Deep learning are making individual decisions. All this because of a revolution in data viz that allows people to explore new, better ways to analyze and make sense of their data.

Imagine what you could do analyzing your, or your organization's data with Glyphs. Could you improve sales by understanding why some products are selling better than others? Could you find the next life-saving insight into medical data by being able to see more variables and their correlations at one time? Could you make your school's students more successful by understanding why they learn the way they do? The possibilities, like the glyphs, are truly limitless.

Think about your data in your mind. What do you see?

And more importantly, what are you missing?

By. The GlyphEd™ Editorial Team

Brad LewisComment