11/5/12 election forecasts for 11/6/12 (off-topic, but here they are anyway...)
Ron,
Re.:
twerpedbyjulie said:
Custer my friend if I was going to pick one thing for you to be right about out of all the things you've told me...I'd pick this one by a fucking mile.
I guess you mean my comment that it’s likely Obama will be re-elected. That’s based on Nate Silver’s 538 column in the New York Times, and a web site called “electoral-vote.com” run by an anonymous dude who describes himself as an American professor of computer science working at a university in Germany. He calls himself “The Votemaster.” The current page of his web site, which is updated every day, is here:
ElectoralVote
Background: Nate Silver is a statistical analyst; 538, the title of his column, is — as you may recognize — the electoral vote total for all U.S. states. Ergo, the majority required to win the presidency is (1/2)(538) + 1 = 270 electoral votes. Silver rose to prominence when he correctly called every state but one (Indiana, which voted for Obama by 1%) prior to the 2008 presidential election and, as well, successfully predicted the winners of all Senate elections that year. Conservatives complain about Silver because, with his model based on polling statistics and other relevant factors, he’s been forecasting a probable win for Obama for some time. They forget that he correctly forecast the Republican takeover of the U.S. House of Representatives in 2010.
Silver’s current (5 Nov. 2012) “day before the election” estimates from
Election Forecasts - FiveThirtyEight Blog - NYTimes.com are:
Obama’s most-likely percentage of the popular vote nationwide: 50.9%
Romney’s most-likely percentage of the popular vote nationwide: 48.2%
Most probable no. of electoral votes to Obama: 315
Most probable no. of electoral votes to Romney: 223
Probability of Obama winning re-election: 92%
Probability of Romney winning election: 8%
“The Votemaster’s” current (5 Nov. 2012) “day before the election” estimates are slightly less optimistic for Obama. (He subscribes to 6 nonpartisan polls which he pays for out-of-pocket plus with contributions to his web site. Silver subscribes to a much larger number of nonpartisan polls that are paid for, presumably, by the NY Times.)
According to
ElectoralVote the probable no. of electoral votes to Obama, including from states where he is leading in the polls by 1 to 4 percentage points (within the margin of error), is 294.
The probable no. of electoral votes to Romney, including from states where he is leading in the polls by 1 to 4 percentage points (similarly within the margin of error), is 220.
The number of electoral votes from states that are exactly tied (i.e., 50% / 50%) in the polls, using “The Votemaster’s” method of poll-averaging, is 24.
According to this analysis, if all the states that are exactly tied in the polls vote for Romney, Obama will still win re-election with 294 electoral votes.
When “The Votemaster” eliminates Rasmussen polls from his electoral vote maps and tabulations (click “Rasmussen-free maps” on the left side of his main page) — which Silver has determined are biased by 3 percentage points, on average, toward Republican candidates [in close elections, that’s a lot] — he finds Obama is likely to receive 303 electoral votes, Romney is likely to receive 220 electoral votes (both similar to Silver’s estimates), and 15 electoral votes are “tied in the polls.” Using these numbers, even if Romney wins all the “tied” electoral votes Obama will win re-election with 303 electoral votes.
According to “The Votemaster’s” state-by-state mapping and tabulation of Senate polling results, the Democrats are likely to retain 51 Senate seats, the Republicans are likely to retain 45, and 4 Senate races are exactly tied in the polls. Thus, the Democrats seem likely to retain their thin majority in the Senate with 51 to 55 seats (= 51 to 55% of the 100 Senate seats), depending on the breakdown of the “tied in the polls” Senate races.
The Intrade gamblers give the Republicans a 96% probability of retaining their majority in the U.S. House of Representatives, the Democrats a 72% probability of retaining their majority in the U.S. Senate, Obama a 68% probability of being re-elected President, and Romney a 32% chance of being elected President. See:
Intrade - Home
It’s hard to have any feeling for what the outcome of the House elections is likely to be, because I’m not aware of anyone being willing to put up the money to poll all of the 435 House districts effectively and repeatedly nationwide [which, of course, doesn’t mean it hasn’t been done].)
—Custer
