Visualizing the Republican Primaries

by Drew Skau 2 years ago Filed Under: Design

Your political views aside, if you love information visualizations, you should be excited about the Republican primaries. They are a great source of relatively rich data at a decent scale, and hopefully there will be some good work that comes of them. The data has a time dimension, four different candidates, 50 states, 2,286 total delegates with 120 independent delegates not tied to any regions. This is a decent set of numbers with some interesting relationships between the dimensions of the data. There will hopefully be too many visualizations of this data to count, but the most high-profile visualizations (and those that are seen by most people) will come from the major media outlets such as the New York Times, the Wall Street Journal and the Washington Post.

Let’s look at what they’ve done with Republican primaries data so far.

The New York Times has a pretty large visualization department, so really great graphics are to be expected. And they’ve delivered two notable visualizations about the primaries so far. The first is a choropleth map with state-level detail zoomable to county-level.

This gives an okay overview, though the problem is state-level detail isn’t enough to tell you anything. In fact, it can be really misleading. It is possible to zoom in and view individual counties, but then you can’t see the whole U.S. at once. Even at the county levels, it isn’t possible to see how many delegates were won by a candidate. Screen area taken up by a county doesn’t represent the population of the county, and delegates are distributed based on population. The information that really matters is not who won which state, but each candidate’s total delegate count.

This takes us to the next visualization The New York Times has created. A simple bar chart is often most effective, and that seems to be the case here.

This isn’t just a standard bar chart, however: the bars are made up of tiny rectangles, each one representing a delegate. There is a line at the cut-off point, making the bars seem to race toward the “finish line”. In addition to bars for the four remaining candidates, there is a bar representing the number of delegates left to show how far the stragglers are from being forced out of the race. The bars also don’t just show the earned delegates, they also have the total possible number of delegates ghosted out. This is a good way to help show the part-to-whole relationship between earned delegates and the total number of delegates.

So The New York Times has some good visualizations of the data, but they are separated into different pages. What if I wanted to view all of this data at once? The Wall Street Journal has created a great all-in-one dashboard.

This dashboard has many of the same elements of The New York Times visualizations. They have the bar chart with the candidates racing toward the 1,144 delegate point, and the choropleth map. It is nice to be able to see both of these at once. Seeing the map tells you how many states are yet to have their primaries, while seeing the bar chart tells you who is in the lead, who is behind, and what the margins are between them. But there is another timeline visualization at the bottom, with circles positioned at the days when primaries will be held for different states. This is where the real beauty of this visualization comes in. The interactivity in this dashboard lets you explore the different time steps of the primary process. Clicking on a given state shows the delegates from that state in the bar chart, and jumps the timeline to the given point in time. Clicking on the timeline shows the states that were involved in that primary and the total delegates decided at that point in time.

This time component is really nice, but there is no sense of how the candidates have progressed over time. You can only see snapshots of points in time, not a full progression. To get the full progression, we can turn to the Washington Post.

This cumulative stacked bar chart timeline combo is difficult to name because it does so many things, but it is incredibly easy to understand! The horizontal axis is the timeline, with tick marks at each primary. The vertical axis is the number of delegates, with the shaded regions representing the delegates going to each candidate and the total represented by the total stack. As the timeline progresses, the total number increases as delegates get added. There is a small progress bar on the left to show how many delegates have been awarded so far, although this is not really needed because you can see this from comparing the current point to the ending point. Interactivity is really what puts the finishing touches on this chart. Clicking on a candidate in the legend adds the all-important 1,144 delegate line and shows just that candidate’s delegates vs. total delegates in the chart. This shows each candidate’s progression toward nomination.

These are the most notable visualizations of primary data so far. There are hopefully many more to come. It would be great to see some that integrate timelines of some of the major stories and issues along with each candidate’s positions on the issue. It will also be interesting to see how visualizers handle the points where candidates are forced to drop out because they can no longer gain enough delegates to reach the 1,144 mark. Regardless of who wins or loses, primary season is a fun time for visualization fans!

 
Drew Skau is a PhD Computer Science Visualization student at UNCC, with an undergraduate degree in Architecture.