Syndicate content

visualized

Seven things I learned about data visualization

Tariq Khokhar's picture

Last week, the World Bank Data team descended on New York City for Visualized - a two day event exploring the “evolution of communication at the intersection of big data, storytelling and design.”

It was awesome.

Here are seven things I learned:

1) Iteration is the path to perfection

By now you’ve heard of Nate Silver - the statistician behind FiveThirtyEight and a near-perfect prediction of the 2012 US elections. What you may have missed is the best interactive graphic of the year - the New York Times’ “Paths to the White House” built with Mike Bostock’s D3:

 Shan Carter from the NYT graphics team showed how newspapers have struggled to represent the potential scenarios and actual outcomes of US elections ever since the late 19th century. His team eventually came up with the graphic above, but see how many revisions they went through to get there:

That’s 257 revisions. As early as version 15, you can see the core idea. At version 81, it looks almost done, but it takes another 157 revisions and that extra attention to detail, high production values and pride in your work to be at the top of your game like this.

Lesson: Iterate and aim high: editors are your friends, they’ll make your work stand out. Also: this is the benchmark for what a good data visualization looks like - if you can’t honestly say what you’re doing is at least this good, iterate.