How Data Viz Swayed the Sprint to the White House

The first U.S. Presidential Debate of the 2016 election created a lot of media buzz and even more massive amounts of data points that voters in so-called battleground states may find interesting.

More than 84 million people watched the historic discussion between Democratic nominee Hillary Clinton and Republican nominee Donald Trump making it the most-viewed presidential debate in history. In addition to television and cable coverage, the 90-minute event was also streamed live on the Internet, allowing an even larger audience to view and participate in the national dialogue.

We though it would be good to show how some of the statistics and data generated from the contest could be visualized. What we found was an opportunity for both major party campaigns to capitalize on the volume and variety of Twitter conversations, especially in the 11 battleground states as identified by news media outlet, Politico.

Tweet Map

The graphic below shows the origin of Twitter traffic dedicated to the topic of the debate using a range of hashtags (more on that later). Our breakdown of the top 10 most active states suggests California, Texas and New York dominated the Twittersphere with their commentary. Interestingly, the swing states of Florida, Pennsylvania, North Carolina and Ohio were also active on the social media platform.

But what about those battleground states with lower than average Twitter traffic? Colorado, Iowa, Michigan, Nevada, New Hampshire, Virginia and Wisconsin did not pull in the same volume of Twitter traffic as other battleground states. Data scientists might recommend that Republicans and Democrats should spend more money in those states to help get their message out. Or, perhaps the candidates could focus on the Eastern seaboard with the main swing states of Ohio, Pennsylvania, North Carolina and Florida.

Hashing Out the Details

So how did we breakdown and identify the activity? Using the following established hashtags, our Analytics team determined that viewers with Twitter accounts used the following hashtags most often:

#Debates2016

#Debates

#debates

#debatenight

Because of the similarities in the wording and capitalization, it’s obvious that a third of all Tweets were spotlighting the Presidential Debate in streams of 140-characters.

What was far more interesting were the other hashtags also being used alongside the variations of #Debates2016. The remaining most used hashtags we found were as follows:

#Trump2016

#Trump

#OmmitedHillaryQuestions

#MakeAmericaGreatAgain

#MAGA (a shortened version of “Make America Great Again”)

#ImWithHer

#election

#SlipperingHillary

#NeverHillary

#MT (a common hashtag that means modified Tweet)

#narcissist

#vote

#tcot (shorthand for “top conservatives on twitter”)

#p2 (progressive-leaning hashtag as a response to tcot)

#HillaryClinton

#DNC (short for Democratic National Committee)

#AMJoy

#gop

The majority of these hashtags were created to support Donald Trump (#MakeAmericaGreatAgain) or discredit Hillary Clinton (#OmmitedHillaryQuestions). When combined with the previous visualization, a political strategist might suggest focusing on these tags in the specific states mentioned above.

The Best Words:

As Secretary Clinton noted in the debates, “Words matter when you run for president. And they really matter when you are president,” our Analytics team also did a breakdown of the candidate’s conversations. The word clouds built below were based on words that were repeated 10 times or more during the debate.

We of course omitted articles of speech (a, an, the) and other short words that would be common to a conversation but hold no significance to the overall message. The only exception to this rule was the word “no” which was used 11 times by Secretary Clinton and 30 times by Mr. Trump.

When combined with the data gleaned from the first two visualizations, a narrative for a political strategist fully emerges. Engaging with candidates in battleground states could be accomplished on Twitter with specific hashtags and phrases and then implemented into the social media strategies of specific states for optimal effect. If nothing else, the results show where funding should and could be spent.

These are examples of how to use data and more to the point, visual analytics to see a little more clarity in our world. It can be applied to nearly all points of a business. At Oracle, we are helping our customers realize the benefits of visual analysis, thereby helping their business grow.

So what other insights can be gained by visualizing data? For more information on Oracle Data Visualization, visit http://www.oracle.com/goto/datavisualization. Also, you can request a free trial here.

(Editor: Special thanks to Chris Garcia for assembling the visualizations)