Nov 07, 2012

Will the accuracy of election predictions push big data into mainstream consciousness?

It was pretty impressive how closely the presidential election results matched the models. Little kids are probably running around with calculators pretending to be Nate Silver, "Look mommy! I'd a data scientist, and I can predict the future."

Do you think the accuracy of the pre-election predictions (the math and data ones, not the talking head pundit ones), and the surprising amount of ink that generated will make big data a household word?


Somewhat, perhaps, although I don't know if most people will actually use or recognize the terms big data or data analytics.  Nate Silver's Five Thirty Eight blog at the NYT, which you referenced, accounted for something like 20% of the NYT's site traffic on the day before the election, so people are clearly interested in the results.  However, I'm not sure that most people even realize that Silver is not a pollster and that he was "simply" running data from polls through his mathematical model to reach an objective "best guess".  What is more likely is that people who saw how useful analytics can be will want to integrate it into their businesses.  The trick is figuring out what to measure, where to get the data, and finding someone who has the ability to develop models that give meaningful results.  I wouldn't be surprised if some consultants are looking over their shoulders, wondering how much longer they can pretend to read the tea leaves while offering up subjective "solutions" to business clients now that there is an objective alternative on the rise. 

No, though it would certainly be nice if it did. Alas, I don't think most voters will really ever understand or care about it. But they may get used to having the information for each future election, and that's still a pretty good accomplishment.
Answer this