A few weeks ago, multiple members of Piraeus attended a one day course put on by Edward Tufte in Seattle. Expectations, and skepticism, amongst the group varied between “Who is that?” to “Tufte is a visualization legend!” He did invent the Sparkline and coined the term: “chart junk”. I anticipated he would emphasize clear and simple displays that weren’t heavy on text or intricate designs; users should be able to rapidly discern obvious conclusions from concise graphics. A few minutes after Tufte started however, I realized those expectations weren’t entirely accurate.

 

Tufte’s Ideas

Tufte emphasizes that creators should do whatever it takes to convey information accurately and efficiently, and there is a moral and ethical obligation to convey the truth as it is understood. As long as those requirements are met, no options are taken off the table.

The classic examples he used were: baseball box scores as shown on ESPN, forecasts from the National Weather Service website, NY Times articles, and Google Maps. Collectively these sources convey complex and varied information to millions of users daily. Baseball box scores and weather forecasts smartly organize key metrics using basic tables, Google Maps is a high resolution intricately labeled drawing, and NY Times articles often haven numerous metrics woven within heavy text and a few pictures. All represent effective visual systems according to Tufte.

 

Thoughts

Big data and analytics are buzz words right now but let’s face it: business has always been about moving information from one place to another and then making decisions based on conviction. Just about any “office job” involves recording, transferring, or pitching information. Those who claim that they shun analytics, and prefer to work off “gut instincts” just ingested information from different sources they preferred and combined that with their past experiences to come to a conclusion. Now that computers perform said tedious work, it is easy to get caught up in the idea that everything should happen automatically – if insight and actions aren’t leaping out of your monitor, something isn’t configured properly.

Popular new visualization tools such as Tableau and PowerBI are designed to allow users to spend as much time as possible tinkering with data, trying out different ideas quickly. Many employing these tools have years of experience in their field, and when they have a gut feeling there is likely some truth there (See Blink, by Malcom Gladwell) but that truth is best uncovered with careful consideration over time (see Thinking, Fast and Slow by Daniel Kahneman).

Once that truth is cobbled together by experts through countless small insights gleaned with the assistance of visualization tools, it is time to take that story to a greater audience. That story shouldn’t be told in the tools designed to streamline the discovery process for experts, but rather using creations free from the limitations of technology, blurring the line between business construct and art.

One of Tufte’s favorite examples was how Galileo described his discovery of Saturn’s rings: he simply made two simple drawings right in line with the text. At presentation’s end, Tufte showed a 1570 printing of Euclid’s Elements by John Dee where he incorporated pop-ups in the text to better describe mathematical theories to readers. Tufte expresses that throughout history many of humanity’s famous ideas were conveyed through complex visualizations that combine text with detailed free form drawings.

Apps such as Paper by 53 make hand drawing much easier, while new frameworks such as Bootstrap and D3 allow for rapid development of interactive visualizations. We will explore practical applications of these technologies in future Piraeus blog posts.

Walking out after the session I was struck by a desire to communicate more ideas through handcrafted visuals. Day to day the tinkering will be done in the amazing visualization tools available like PowerBI and Tableau, but let’s not forget to take a step back to reflect on those discoveries and express them through our own creation.

 

By: Karlen Rothenbueler | Associate, Analytics