Facebook as a presidential forecaster? Seriously? Why the heck not? Those who gave any weight to Facebook analytics before the January primaries in Iowa wouldn't have been shocked by Obama's win over Sen. Hillary Clinton. "If people had been taking Obama's Facebook numbers more seriously, they would not have been so surprised when he won the Iowa caucus," said Christine Williams, a professor of government at Bentley University, in the WSJ article.
During the 2006 midterm elections Williams also compared the number of Facebook supporters to the outcome, and came to the same conclusion: that Facebook support did indeed have a significant effect on election results, particularly in the case of open-seat candidates. She explains in this video the predictive model she used. "We found that your number of supporters on Facebook increased as much as 3% your vote share, controlling for all those other things," such as the amount of campaign money raised, party affiliation, and competitiveness of the race, she said.
The number of blog mentions and video views may also be indicative of who's going to win the upcoming election, though that's where we start entering horoscope territory. Says the WSJ: "On some days in recent weeks, Sen. McCain had more blog mentions than Sen. Obama. More people overall have watched the Illinois Democrat's online videos, but on average, Sen. McCain has attracted more viewers per video released by his campaign."
We'll know in a few weeks if our hypothesis tests true. According to Facebook and TubeMogul.com, which has been tracking the number of times official campaign videos have been watched on a daily basis, as of Oct. 17, Obama had 2,168,036 Facebook friends and more than 600,000 YouTube video views. McCain has 581,032 Facebook friends and less than 200,000 YouTube video views.