Will research get the election right?

Of all the recent changes in marketing research I’ve been particularly fascinated by those that are pushing the limits of traditional research by combining data sets. With today being Election Day in the United States, I thought it would be interesting to take a look at how this is being done to predict the outcome of the 2012 election.

I recently came across www.FiveThirtyEight.com, which is a polling aggregation site with a blog created by Nate Silver. The site takes its name from the number of electors in the United States Electoral College.

Silver first gained public recognition for developing PECOTA (player empirical comparison and optimization test algorithm), a system for forecasting the performance and career development of Major League Baseball players. PECOTA projects the future performance of hitters and pitchers by relying on matching a given current player to a set of “comparable” players whose past performance can serve as a guide to how the given current player is likely to perform in the future. In addition, PECOTA calculates a range of probable performance levels rather than a single a value such as batting average.

Silver then applied these same principles to the 2008 presidential election, weighting each poll based on the pollster’s historical track record, sample size and recentness of the poll but also supplementing this data with additional information available in polls from “similar” states. He carried this approach one step further by also factoring national polling trends into the estimates for a given state. The idea is that this additional data, when applied correctly, would help to fill the gaps in information about the trends in a given state. It was especially useful in states for which there were few if any polls. (Additional aspects of the methodology are described in a detailed FAQ on FiveThirtyEight.com.) Using this methodology Silver correctly predicted the winner of 49 of the 50 states in the 2008 election, missing only one state which had a winning vote margin of 1%. Even more impressive, he predicted all 35 of the Senate races.

So will Nate Silver be successful again this year? Check out his predictions.

Hopefully we will know the answer before January 20th!

This entry was posted in Market Research in the News, Political Polling, Research Blogs and Communities, Research Industry Trends. Bookmark the permalink.

Comments are closed.