Edward Tufte introduced Slopegraphs in his 1983 book “The Visual Display of Quantitative Information“.
Two decades later, he developed the Sparklines, and they became an instant hit.
Interestingly, almost nobody paid any attention to the Slopegraphs until a few years ago, even though they are extremely elegant and created by the man himself, Edward Tufte, the master of information visualization. Maybe the reason is because Tufte originally called the new graph “Table Graphic”, and it wasn’t until 2011 that he renamed it to Slopegraph. Slopegraphs are also a bit more difficult to draw on most software, for example Excel.
Basically, a slopegraph is a line graph that eliminates and ignores all data points between the start and the end data point.
Consider the table below:
The table shows the percentage of population in five selected Nato and eight selected non-Nato countries with positive view of the U.S. These are annual data from 2002 to 2014. For some countries, data are missing for some years, and these are shown in the table as blank cells.
So what was happening in 2002? In short, George W. Bush was serving his first term as the U.S President, and the country, following the events of 9/11, was together with allied nations involved in a war in Afghanistan. Vladimir Putin had been in the office for two years. In 2014, Barack Obama was the U.S. President for the second period, the U.S. – Russian relationships had deteriorated during the Ukraine crisis in the same year. Putin was still the Russian President.
In between those year, a lot had happened in the world.
You need to spend some time to see patterns in the table. The data come from 13 countries over 13 years, and the countries are not ranked by start or end values, or alphabetically.
So why would you choose the table? Well, if you are interested in following one or all countries year by year, then this is the tool to go to. It would be difficult to see the trends, but you get the specific values per country per year. If you are interested in seeing the trend or the pattern, then you must opt for a graph.
But which graph?
Because we are looking at changes over time, a line graph as below, should be helping us, but it´s not.
The trend is even more difficult to see in a line graph, because there are simply too many countries in the data categories. Personally, in a case like this, I would break down the line graph into four or more graphs, each showing up to four countries at a time.
But if I am not interested in all those eleven years between 2002 and 2014, and want to jump directly from the start point to the end point, then a slopegraph is an obvious choice.
In a slopegraph, the first thing that you automatically get, is a ranking for both the start and the end point. This is not unimportant. Thereby, it instantly forwards you a message. All categories are labeled, and the values are listed. And you clearly see the people in which country are more favorable towards to the U.S. compared to in 2002.
Also, because I color-coded the Nato and non-Nato countries, you can easily separate them from each other.
At a second glance, there is more information in the slopegraph that grabs your attention more quickly than in the table or the line graph. Look at the block of the countries of Germany, Russia, France and Mexico in 2002. All four lye close to each other in a range between 60-64%. Then follow each of them to 2014. France is now ranked as number two overall, up to 75%, whereas Mexico stays on the same level. Germany drops to 51%, but that´s nothing compared to Russia.
You can also see that the two Latin American countries of Mexico and Argentina, and the Philippines, are the only countries with no major change between 2002-2014. The changes in other countries are very or somehow considerable.
Slopegraphs are especially effective when you want to replace two pie graphs and replace them with one single graph.
Often, slopegraphs work very well with more detailed tables, which tend to “complete” them. When creating slopegraphs, you need to very clear. Make the graph tall if you are using many categories. Always include the category name and value on both side.