Per 538.com: Obama is BACK!

[BlooInBloo]

Phew - it took awhile, but it's worth it!

WOOT! \o/

http://www.fivethirtyeight.com/2008/09/todays-polls-918-obama-regains-lead.html





"On the strength of an abundance of state and national polling, Barack Obama has retaken the lead in our Electoral College projection. Our model now forecasts him to win the election 61.2 percent of the time; it also gives him a slight, half-point advantage in the popular vote. Yesterday, Obama was projected to win the Electoral College just 45 percent of the time, so this is a rather dramatic move upward."

[demokatgurrl]

1. Dang, I'll never figure out statistics- why

is this so different from the one at:

www.electoral-vote.com ???? Can anyone explain this to me?

[BlooInBloo]

2. They do different calculcations. 538.com has historically performed the most accurately....

of all the models I'm aware. There's a detailed explanation on the site somewhere. It's not likely to be of much help to someone who doesn't know a little bit of math, but there'll prolly be a model performance comparison with other models.

[jefferson_dem]

3. Electoral-Vote relies on current polls so it is about the state of the race as of now.

538 uses a more formal mathematical model to project likelihood of success in November, based on current polls and other factors.

[chiefofclarinet]

4. A little more info

I think each map gives something slightly different.

From what I can tell, electoral-vote.com uses the most recent polls (usually only from the most recent day) and makes a map based on them. This one is a snapshot of right now. I don't particularly like this site, especially since they give "ties" out liberally. It also ignores demographics, trends in current voting, the sample size, and the pollsters themselves.

538.com uses the raw data that is the polls, often going back months, and simulates a large number of mock elections using computer software. Each of the polls is weighted by how recent it is, its sample size, which pollster did it, etc. This one gives a map of the current trends. During the post-convention bounce, it went really sour for the Dems because the trends had Obama losing many states. The major problem with this one is that it (normally) takes a lot of time for it turn around.