Omgili blog

For the past year and a half we had this question on the front page of Omgili. We followed the hopefuls before anyone knew it’s going to be McCain against Obama, and the Pailin/Biden running mates. In three days we will all know the answer, but in the meantime we can bring our best educated guesses.
Every day new polls predict the victory of Obama over McCain. The problem is that people might not always tell the whole truth about their decision, (We’re all familiar with the Bradley effect) and so the results are certainly not definitive.
I was curious if I can find more clues about the out come of the elections. It is true that people might be too “Politically Correct” to tell the truth, but on online forums, anonymity might uncloak more accurate results.
I ran two sets of queries on Omgili Buzz Graphs – One for Obama vs McCain to see the detailed and overall buzz about the two candidates in the past month, and a second for discussions mentioning only one candidate without the other.

Let’s take a look at the first query graphs:

As you can see, the number of discussions about Barack Obama is indeed higher than about McCain. We can also see that the number of discussions is rising over time as we approach Election Day. I must say that even though the number of discussions is higher for Obama, the margin is quite small (20%-30%) for Obama. I expected a much bigger delta.

The real interesting figure is on the second query. It is only natural that when you speak of one candidate you will also mention the other. I wanted to see how many discussions are taking place around a single candidate Obama or McCain, in other words, who is more “interesting”?

Now these results have a much more distinct trend. First of all, there is MUCH more discussion about Barack Obama than about McCain. Furthermore, as we approach Election day, the number of discussions about Obama is increasing rapidly whereas with McCain the number is decreasing (up to 400%).
Brand Monitoring, and Web 2.0 based research is a very hot topic these days. I am not saying this is a qualified research, but it will sure be interesting to see if there is a correlation between the above graphs and the outcome of the elections.