Monday, June 09, 2008

If we simply add up the numbers from the prior post we find there are 2,206,800 pro-Obama results and 1,059,067 pro-Mccain results.

There are alternate explanations for results being skewed, however. Obama may have a larger Internet presence than John Mccain, and the extent of his presence may be tied to the prolonged race for the Democratic Party's nomination. Additionally, Hillary Clinton's loss may have resulted in pro-Mccain or anti-Obama backlash.

However, it is worth noting that if the size of Obama's Internet presence is tied to the prolonged race, Hillary's presence would likely be of a similar size. A quick set of searches gives us the following results:

"clinton for president": 543,000
"will vote for clinton": 46,700
"will vote for hillary clinton": 11,200

The 600,900:2,141,000 Clinton/Obama ratio for these three search terms seems to indicate otherwise. Even Ron Paul has better search stats than Hillary Clinton. It is more likely that Obama appeals to a significantly higher amount of Internet users than Mccain.

Hillary Clinton's Internet/Reality ratio (percentage of search results to percentage of official primary popular votes) is something in the neighborhood of 22/48, whereas Obama's Internet/Reality ratio is more like 78/48. This basically means Clinton's online presence is/was ~50% what her real-world presence is/was, and that Obama's online presence is ~170% what his real-world presence is/was.

Given current polls (as of 6/9/2008), Obama is leading Mccain 47 points to 45 points. This would mean Mccain's Internet/Reality ratio is something like 32/45 if polling is an accurate portrayal of public opinion. This would also place Obama's Internet/Reality ratio at 68/47. Current polling would suggest that pro-Mccain backlash from the Hillary Clinton camp is not very significant.

RealClearPolitics.com has collected statistical data which suggests Internet users favor Obama over Mccain 62.9:35.4.

Now here's the stretch. I will assume RealClearPolitics.com has done limited polling, and assume the search result data to be representative of wider polling. I will also assume current non-Internet polling is limited. Now we can extrapolate what wider non-Internet polling might look like.

Obama
62.9/47 = 68/(?)

Mccain
35.4/45 = 32/(?)

The results would indicate an Obama win of 50.8 points to 40.7 points in the general election.

Now, I'm no statistician. I'm also prone to overlooking very significant details. Don't take any of this as authoritative, as I'm likely to be wrong. But you have to admit... It would be really cool if I'm right.

2 comments:

Jamesly said...

thanks for all the math my uberdearliest friend

Philip said...

As a belated follow-up, the final tally was Obama 52.9%, McCain 45.7%