Subscribe to DSC Newsletter

Yesterday, there was a meetup in Arlington on Political Campaign Data Science with a panel of four experts to talk about their work. As we have all heard, the 2012 Presidential election was steeped in data like never before, with the Obama campaign in particular using advanced analytical methods to target voters, and with Nate Silver and other polling aggregators providing fascinating insights into the dynamics of the campaign. Sasha Issenberg, author and journalist, summarized the ways that modern campaigns aggregate and use data in every aspect of voter contact. Ken Strasma, practitioner, talked about the value of microtargetting, and explained how statistically customized messaging is valuable in both political and nonpolitical marketing. Alex Lundry, practitioner, dove a little farther into microtargetting, describing how these models are built in practice and Shrayes Ramesh, academic, will spoke about his research into causality and political contributions, and the extent to which money leads or follows policy.

Peter Bruce, president of Statistics.com offers his comments:

Some interesting snippets from the DC Data Science Meetup last night.  Ken Strasma heads a predictive modeling firm that does work for the Democrats. Dressed in a conservative suit & tie, he described to an audience of several hundred how successful micro targeting of voters was not a magic formula or snake oil, but the result of detailed mastery of data, computational power, and algorithms.  Using hundreds of variables and huge amounts of voter-specific data, the goal is to place each voter in one of 9 cells in a 3x3 table where the columns are preference (favor our guy, undecided, favor other guy) and the rows are voting habits (always votes, sometimes votes, never votes).  Only 3 cells are worthy of campaign spending (can you guess which ones?).  Alex Lundry, his counterpart on the Republican side, described how the Romney campaign used visualization techniques to analyze campaign spending and advertising in (almost) real time, and respond accordingly.  His chart for TV advertising in Denver on Nov. 5 showed one curiosity - on Fox, Obama outspent Romney by a huge margin, while other networks were more balanced. Lundry was casually-dressed, and made a strong recruiting pitch to the audience of data scientists, saying the Republicans needed them - desperately.  When asked by moderator Harlan Harris whether, if they had to choose, they would choose better data or better algorithms, Lundry chose algorithms, and Strasma chose data.  Strasma noted that high quality granular data (e.g. a voter telling you who he favors) will usually trump the most sophisticated predictive algorithm.  The exception is likelihood of voting - the non-voters usually lie and say they will vote, and the algorithms know better. 

Views: 547

Tags: algorithms, campaign, data, microtarget, modeling, predictive, statistics.com

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Patrick K Stroh on December 28, 2012 at 5:28am

"9 cells in a 3x3 table" ... sounds complicated; better spin up the Hadoop cluster!

Videos

  • Add Videos
  • View All

© 2020   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service

console.log("HostName");