{"id":1884,"date":"2012-11-10T11:25:59","date_gmt":"2012-11-10T15:55:59","guid":{"rendered":"http:\/\/www.danielpradilla.info\/blog\/how-to-create-better-charts\/"},"modified":"2013-01-13T05:47:59","modified_gmt":"2013-01-13T10:17:59","slug":"how-to-create-better-charts","status":"publish","type":"post","link":"https:\/\/www.danielpradilla.info\/blog\/how-to-create-better-charts\/","title":{"rendered":"How to create better charts"},"content":{"rendered":"
\"\"<\/a>
nope, this is not a good chart<\/figcaption><\/figure>\n

We’re often tempted to include some kind of accessorizing effect in our charts without knowing that deep down we’re making them impossible to interpret and thus ruining their purpose. A typical case is the 3D pie chart.<\/p>\n

Many people, including executives, have a certain likeness for 3D charts. Somehow they believe that these look \u00e2\u20ac\u0153modern\u00e2\u20ac\u009d. I invite you to take a look at this chart and try to determine which one is the biggest slice:<\/p>\n

\"\"<\/a>
Item A is twice as big as item C. B is the same as D. But you wouldn’t be able to tell that from the chart.<\/figcaption><\/figure>\n

The pie chart is the worst to establish comparisons. Cognitively we’re terrible at estimating areas. If we add to this perspective, we have what Tufte calls Chartjunk<\/a>:<\/p>\n

\n
\n
\n
\n

Too many data presentations, alas, seek to attract and divert attention by means of display apparatus and orna- ment. Chartjunk has come to corrupt all sorts of information exhibits and computer interfaces.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/blockquote>\n

The chart at the beginning of this post is another example of Chartjunk. The designer uses a \u00e2\u20ac\u0153clever\u00e2\u20ac\u009d design to disguise a chart with only 5 data points. The same thing usually happens with infographics, those increasingly common posters whose main goal is to promote a designer and not present information<\/p>\n

\"\"<\/p>\n

In Envisioning Information<\/a>, Tufte introduces a basic information design principle:<\/p>\n

\n
\n
\n
\n

By giving the focus over to data rather than data-containers, these design strategies are transparent and self-effacing in character. Designs so good that they are invisible.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/blockquote>\n

Designs so good that they are invisible. That’s exactly it.<\/p>\n

\"\"<\/a><\/p>\n

<\/h3>\n

Data\/ink<\/h3>\n

Tufte invented a concept that I really like: the \u00e2\u20ac\u0153data\/ink\u00e2\u20ac\u009d ratio<\/b>.\u00a0<\/b>The idea is to maximize that number, meaning:\u00a0 to use the same amount of \u00e2\u20ac\u0153ink\u00e2\u20ac\u009d in order to make a chart comprehensible.<\/p>\n

Look at the following set of charts. Doesn’t it seem like the chart on the right says the same thing, but that the information is displayed better?<\/p>\n

\"\"
Less ink= more clarity<\/figcaption><\/figure>\n

 <\/p>\n

We removed from the left-side chart:<\/p>\n