Join Date: Dec 2005
Location: Philadelphia, PA
Mentioned: 278 Post(s)
Tagged: 0 Thread(s)
Quoted: 6611 Post(s)
Dish Network (Incorrectly) Predicts Clinton Victory using Viewers to Voters Data
When you watch TV, your service provider is watching your viewing habits. Normally, this sort of data would be used to determine advertising rates and factor into whether a show makes the cut or gets canceled. The Viewers to Voters predictive model was co-developed by Dish’s Data Science and Media Sales divisions and is the first project to come out of the company's Wisdom of Crowds initiative. The program seeks to explore verifiable correlations between viewing habits and people's actions in the real world, including those of voters.
With its Viewers to Voters predictive model, pay-TV provider Dish took viewer behavior data, filtered it, processed it, and used it to issue predictions for tomorrow's US presidential election. Based on that methodology, the company is calling for a Clinton victory.
Aside from picking the next president, the model also predicts that Republicans will maintain a majority in the nation's 115th Congress, with Democrats only gaining two seats in the House of Representatives.
Per Dish, the Viewers to Voters model was applied to the 2014 House elections, and its reliability was pegged at 98%.Warren Schlichting, exec. VP of media sales, marketing and programming noted that "With so much focus around national polling, we thought it’d be interesting to see if we could find a correlation between how our customers interact with DISH and how they may vote. We recognize that our call on the distribution of seats in the House may be an outlier. Yet when we tested the model against 2014 House elections, we found that we were able to predict the outcome at a 98 percent reliability point."
The program analyzed the relationship between viewing habits and political affiliation. According to Dish, "The model determined that customers who watched more sports, religious or family-oriented television were more likely to vote Republican. Examples include NFL Football, GEB America’s Giving Hope, Sharing Life and PBS' Sesame Street."
Dish then went on to describe the viewing habits of Dems: "In contrast, customers who watched more series/specials, education, or music-oriented television were more likely to vote for Democrats. Examples include ABC’s Scandal, Discovery’s MythBusters and NBC’s The Voice."
I have a big favor to ask everybody commenting on this thread: Please don't turn this into a partisan discussion. It's against AVS Forum rules to explicitly discuss politics, so please stick with discussing how viewership data and algorithms might be used to predict the behavior of large groups of people. In the context of the election, this post is about TV viewing habits and analytical models and whether applying algorithms to that data can compete with polling. If the conversation goes downhill, I'll be forced to ask the moderators to shut the thread down. Soon, we will know how well this genuinely new approach to predicting election results works. So again, please keep it civil, discuss the topic at hand, and when tomorrow rolls around… vote.
Mark Henninger, Senior Editor at AVS Forum
Last edited by imagic; 11-09-2016 at 03:31 AM.