What does a 5% uniform swing from Obama to Romney look like?

[ajb]

A uniform swing from what?

People who voted Obama in 2008, didn't like what he's done thus far and vote for Romney as an alternative.  I'm taking 5% off Obama and moving them toward the Romney from 2008 % results. I use 5% from a recent PPP poll with Obama leading Romney in Nevada, and doing that nationwide.

Well, that PPP NV poll is 51-43, and the 2008 election was 55-43 (to be precise, 55.15-42.65), so that's actually a 4-4.5 point swing away from Obama, not a 5-point. Which makes a difference.

[ajb]

4.5%? Im rounding it up to 5% to make it simplified.

But it is a 4.5% swing, not a 10% swing, which is what seems to have taken hold elsewhere on the thread. Not that there's anything wrong with speculating on any possible swing, but there's no evidence for a 10-point swing in that PPP poll of NV.
While we were at it, it would be fun to speculate on what the election would look like with, say, a uniform 3-point swing towards Obama, as recent polling out of, for example, OH and NC would suggest.

[ajb]

You boys seem to be torturing yourselves a bit. Obama got 53.7% of the two party vote last time. For Romney to win, the uniform swing needs to be about 4.55%, with Romney getting 50.85% of the two party point, at which time he carries Colorado and goes over 269 electoral votes.

But wouldn't a 4.55% swing give Obama about 51.4% of the two-party vote?

[ajb]

You boys seem to be torturing yourselves a bit. Obama got 53.7% of the two party vote last time. For Romney to win, the uniform swing needs to be about 4.55%, with Romney getting 50.85% of the two party point, at which time he carries Colorado and goes over 269 electoral votes.

But wouldn't a 4.55% swing give Obama about 51.4% of the two-party vote?

He didn't mean the actual percentage (as in, "4.55% of 53.7 = 2.41%, 53.7-2.4=51.3%), but rather the Republican amount of the two-party vote (46.3%) with an increase of 4.55 points (46.3 + 4.55 = 50.85%).

Fair enough, but it's important only to use one definition of "swing" at a time. The thread began by looking at the PPP Nevada poll, which suggested that the gap between D and R this time is 4.5 points narrower than in 2008, and that got misinterpreted as being about the Democratic vote going down by 5, and the Republican vote up by 5, which is of course a very, very different thing, and would lead to radically different results nationwide. Calling both things a "5 point swing" causes a lot of confusion.