Smarter electoral math

[geek tragedy]

The #1 rule on polling when an incumbent is running is that the gap or spread does not matter--what matters is the incumbent's total support.

So, we should not treat a * 46% - Kerry 43% as a bad thing for Kerry.

I propose this rough breakdown of polls based solely on Bush's numbers, on a state by state basis.

46%=Likely Kerry win
47%=Leaning to a Kerry win
48%=Tossup
49%=Leaning to a Bush win
50%=Likely Bush win

The reason? Incumbent presidents almost never exceed the polls in their actual votes. They remain remarkably flat, whereas challengers fluctuate wildly, but almost always pick up support.

Again, this is nothing precise, but it is more useful than focusing on the spread. A Bush lead of 46-43 should be viewed as a likely Kerry win, not a likely Bush win.

This also should become more accurate as we get closer to the election.

[MissouriTeacher]

1. The 50% mark is critical for Bush.

Incumbents do not generally win unless they are polling consistently at the minumum 50%. Obviously Bush has not been.

[johnnyrocket]

2. What's the map look like if you take that into account?

It's got be be a sea of blue!

[geek tragedy]

4. Using average poll numbers for battleground states:

http://www.realclearpolitics.com/bush_vs_kerry_sbys.html

(I know he's a Repub, but his state numbers are more accessible than EV's)

Florida: Probable Bush win
Ohio: Leaning Bush, but within reach for Kerry
Pennsylvania: Easy Kerry win
Michigan: Likely Kerry win
Missouri: Likely Bush win
Wisconsin: Likely Bush win
Minnesota: Likely Kerry win
Iowa: Likely Kerry win
New Hampshire: Likely Kerry win
Nevada: Likely Bush win
West Virginia: Toss up
New Mexico: Likely Kerry win
Oregon: Very likely Kerry win
Colorado: Likely Bush win
Maine: Easy Kerry win
Washington: Easy Kerry win
New Jersey: Easy Kerry win

So, Bush would still be looking pretty good with these exact numbers at this exact time.

However, Kerry's surge in the national polls should be reflected in later state polls.

[Kammer]

3. No

While I agree with your premise, I disagree with your numbers.
46%=Kerry wins big
47%=Easy Kerry win
48%=Likely Kerry win
49%=Dead Heat, another 2000
50%=Bush win

[geek tragedy]

5. Perhaps.

I'm fudging a little bit because there are still quite a few undecideds out there. For now.

As we get closer to 11-2, I would go with your breakdown.

[Kammer]

6. Undecideds

That is why your premise is historically correct. Undecideds overwhelmingly go for the Challenger.

Going back to 1980, Jimmy Carter polled at 42% in the final polls and received 41% of the actual vote.

In 1984, Ronald Reagan polled at 58% and received 59% of the actual vote.

In 1992 George Bush polled at 37% and received 37% of the actual vote.

In 1996 Clinton polled at 51% and received 49% of the vote.

So, only in 1984 did the incumbent receive a higher percentage in the actual vote than they did in the final poll and then only 1% higher.

If you look at independent polls, not the corporate ones, Bush fails to reach that needed goal in many, many battleground states.

It is looking good. Especially after these debates, when you look at how the independents and undecideds are leaning. Bush is appealing to his base and Kerry is appealing to those who are undecided.

Seems a little late in the game to try to shore up the base at the expense of losing those who are undecided (whoever the hell they are).

[geek tragedy]

7. Bush is gambling that he has 49% or so locked up.

He's trying to preserve that group.

If he's wrong, he's screwed.