|Daily tracking of Obama%-Romney% on Gallup and Rasmussen|
According to the media, the general election has begun. Both Gallup and Rasmussen have started their daily tracking polls, and every Sunday I will be showing the results. Above the 0% means a poll says Obama is leading nationally, which Gallup has considered true for about a week. Rasmussen has had Romney as the leader for most of the past two weeks, though there were a few days this week when both polls agreed Obama was in the lead. Rasmussen says they are polling likely voters, while Gallup only promises there sample is taken from registered voters. A likely voter sample should be more reliable, but Rasmussen has as their basic model a much higher percentage of Republican voters than any of the other major companies.
|Electoral vote split by Confidence of Victory|
Anyone who has not slept through the last three presidential elections knows that it's the electoral college that gives us a president, not the popular vote. Not every state has been polled yet, but using polls from 2012 or the results from 2000, 2004 and 2008, the split of the 538 electors is as follows.
230 Solidly Democratic electors
181 Solidly Republican electors
127 electors "in the mix"
I define a state as being solid if the Confidence of Victory number is over 95% for one party or the other. For my fellow number nerds, I have written out my methodology for 2012.
|Odds of winning using CoV in purple states as of 4/29|
I take the probabilities for the "in play states" created by the Confidence of Victory method and figure out all possible outcomes, assuming the solid states, both Republican and Democratic, will be won by the favored party. As of the polling so far, the states that are between 95% Confidence for Republicans and 95% Confidence for Democrats are:
Missouri, North Carolina, Virginia, Florida, Iowa, Ohio, Michigan, Pennsylvania
Assigning probabilities to the 256 possible outcomes for these eight races, if the election were held now, Obama would have about a 70% chance of victory and Romney about 30%. Of course, these numbers will change over time with new polling data and new stories in the press that favor one side of the other. As time goes by, the pictures that are now bar graphs or pie charts will switch to line graphs showing the changes over time.
For anyone who feels it's far too early to start this sort of thing, I can only agree. In 2008, I started my Sunday Numbers series just after Labor Day and the introduction to the general public of Sarah Palin. This is what the media thinks the public wants, so I as a lowly blogger have decided to go with the current. I will also be including numbers for the Senate races once all the primaries are over.