New Analytics Tool Can Accurately Predict Polling Results For Which Candidate Has The Most Support
We’re constantly trying to peer into the future to see who’s going to win, just for a little solace perhaps. Back in the day we used to rely on animals making good lucky guesses tea leaves. But now we have big data, massive amounts of information are getting splurged onto the internet every day and various computer scientists are trying to find ways to crunch it into small nuggets of truth that tell us the way the world is going.
Using machine learning, statistical physics of complex networks, percolation theory and natural language processing physicists at The City College of New York Hern´an A. Makse, Alexandre Bovet and Flaviano Morone have developed a tool to scry Twitter users opinions about the American Election. The toolset uses 73 million tweets collected from June 1st to September 1st 2016 to come to a result.
The analytics tool is surprisingly accurate when compared to the New York Times National Polling Average which represents an aggregate of hundreds of independent traditional polls. More importantly they can forecast the results by 6 to 15 days, which proves the importance of Twitter and social media in general as an early warning sign of things to come in polling.
The analysis found that 4.5% of people are listed as undecided, supporting Trump and Hillary in equal measures. These are probably the most sane people considering.
Trump supporters are massively active, and are interacting like crazy. They also hold the majority of hashtags. The majority of users in the SCGC is generally in favor of Donald Trump for most of the time of observation. “However, the situation is reversed, with Clinton being more popular than Trump, when the entire Twitter dataset population is taken into account” the study points out.
Yet, it remains a question whether the results at the whole Twitter population level will be reflected during the real election day with the victory of Clinton as predicted, or, in contrast, the opinion of the SCGC will prevail giving rise to a victorious Trump campaign.
However Hillary supporters are a silent majority, with massive spikes of support flaring up around major events.
“It is as though Trump supporters dominate the campaign machinery inside the most important strong component, yet, this domination does not extend to the whole Twitter electoral population, since Clinton still has more supporters in the population at large.”
So Trump supporters are a passionate and unified group with a massive amount of hype, but this doesn’t translate to a larger group. The study describes the group as a “very cohesive group with large influence at the core of the network”. It’s this networks greatest strength, and if they want to swing things Trumps way, they’re gonna have to figure out how to properly leverage their power.
“Yet, it remains a question whether the results at the whole Twitter population level will be reflected during the real election day with the victory of Clinton as predicted, or, in contrast, the opinion of the SCGC will prevail giving rise to a victorious Trump campaign.”
You can view the whole study here.