What constitutes a Landslide Election?

Usually they say, “take the electoral in a landslide” or “take the popular in a landslide” to be precise. But if they don’t, it is often assumed they are talking about the electoral college because popular is usually much closer even in electoral landslides.

Here’s one description:

Four large states John McCain once seemed well-positioned to win — Virginia, North Carolina, Ohio and Florida — have in recent weeks shifted toward Obama. If Obama were to win those four states — a scenario that would represent a remarkable turn of events — he would likely surpass 350 electoral votes.

I think the idea is that he needs to win by at least 20–30 points or so for a landslide. Something that would be obvious enough not to have to count hanging chads in one state alone. You need 270 electoral votes to win.

For popular vote, according to one commenter on fivethirtyeight.com, he needs 55%+. Which would be ridiculous. And thus the point of calling it a landslide, I suppose.

I think a lot of people are saying “landslide” then self defining what it means when they say it.

empress pixie,
I tried to give you fluve but it stayed zero :-(

So there’s no “rule” as to what a landslide is?

The analysts I hear say a win of 6% or more. That seems like a low bar for a landslide to me.

Wikipedia says it is when a candidate wins by an overwhelming margin. They list examples!

Judi, I think it took even though it didn’t tell you.

@astro, me too.

Also on www.fivethirtyeight.com, they do scenario analyses every night. That is, they take the current data they have and run 10,000 simulations to determine who wins and by what margin, how many times out of 10,000. They say that as of last night’s polling data, Obama wins in a landslide in 5,340 out of 10,000 simulations (and their scenario analysis uses a definition of 375 or more electoral votes). So, I have no idea where they got THAT figure (375+ EVs), but there the only ones I’ve seen so far actually posit a definition, so that’s what I’ve been going with.

@Dale: You win! I went to try and find that definition and could not.

Oh, and in case you’re keeping score at home, Obama wins 9,577 out of 10,000 simulations with the current polling data. That’s not counting an electoral college tie (16 times out of 10,000).

Great EmpressPixie…have Don Pardo tell me what I’ve won!

Also, just thought of this. They express their scenarios on a bell curve graph (they are statitsticians after all). I’m suspecting that statistically speaking, where the mean of 538 EVs would be 269, one standard deviation would put you at 375? It would seem to make a lot of sense, because I thought something like 69–70% of all values exist within one standard deviation of the mean, and if you divide 375 by 538 electoral votes, you get 69.7%.

So I guess mathematically, I would define a landslide election as one where one candidate gets enough EVs to be outside one Standard Deviation of 269 Electoral Votes.

all that math bade my head spin. You get “smart fluve!”

lol, I know I read that, but all I got out of it was that 2/3 seems to be a landslide.

I need to find a way to save questions I want to go back and read when my mind is clear.

Unfortunately, that was about the simplest way I could come up with to explain it. Standard deviations are as I understand a measure of how spread out a set of data are. And they are probably looking at this and saying, if this were random, what would one standard deviation from 269 (1/2 the available Electoral votes, i.e. the midpoint) be. And figuring that as the majority of results in a random bell curve would fall within 1 standard deviation of the mean, if it falls outside one SD, it’s a landslide. That make more sense?

As you can see if the popular vote difference runs to double digits the electoral vote difference is generally greater than 60%. If this holds true this year and the popular vote is Obama 54% McCain 40% then electoral vote could be
Obama 67% McCain 33%

Does that constitute a landslide?

Year Diff % Diff %
Pop voteElec vote
2004.. 2.5…...6.5
2000.. -0.5…..1.0
1996.. 8.5…...50.4
1992.. 5.5…...37.6
1988.. 7.7…...58.6
1984.. 18.2…..95.0
1980.. 11.7,,,,,81.0
1976.. 2.5…...11.5
1972.. 23.0…..93.0
1968.. 0.7…...20.4
1964.. 22.5…..81.0
1960.. 0.2…...15.6
1956.. 15.2…..72.4
1952.. 10.8…..66.4
1948.. 4.5…...21.5
1944.. 7.5…...62.5
1940.. 10.0…..69.5
1936.. 24.0….97.0
1932.. 16.8….77.0
1928.. 17.4….66.6
1924.. 27.2….46.3
1920.. 26.0….52.0
1916.. 3.2…..4.4
1912.. 14.4….65.3
1908.. 8.5…..33.0

http://www.uselectionatlas.org/

BTW a very interesting site

galileogirl – the way the electoral votes break, Obama is not going to need a 14 point spread in the popular vote to get 67% in the electoral vote. Actually, right now, Obama is up by about 8 points and that translates to about 70/30…if he got to a 14 point national lead we’d be seeing Reagan ‘84 numbers.