MSNBC - Frank Luntz says - if the Prez is not up by at least 3 points in the Polls, he will lose...

[TBBT]

When asked - Frank said that the President has the momentum at the moment in the national polls. However, the race is really close in the Electoral College.

He said that due to increased voter registration and the last remaining undecided voters - who tend to break towards the challenger - he believes that Bush will need at least a 3 point lead in the popular vote (national polls) to win enough battle ground states to win the electoral college.

He states that conventional wisdom says it's one thing to register a lot of new voters, but it’s another to get them to the polls. However, he said he thinks this year is going to be different and therefore he is predicting a record turn out. He says this factor will favor Kerry, making up ground for Kerry by about 0.5%. He says that the remaining undecided voters breaking for the challenger will give Kerry another 2.5%.

[JLS]

Isn't it amazing how many Democrats have repeated this undecideds break for the challenger lie this year? And it sounds reasonable even if the data do not support it.

Anyone who wants to read the data via Gallup can look at the second table in this link:

http://www.dalythoughts.com/Update-05-26-04.htm

Also read Dales discussion before and after the second table.

[Lando Lincoln]

He's wrong. Last minute undecideds break for the incumbent.

Absolutely correct.

Lando

[paltz]

oh please. ghw was 16 pts behind dukakis going into the election in '88 and it was a landslide victory.

[SuziQ]

he will loose...Donald Rumsfeld on the Breck Girl.

ROTFL!! I'd pay real money to see that!

[Hattie]

Actually he means the fraud will help Kerry.

[Brimack34]

Which polls?

[pjd]

These are code words for "You need a cushion to overcome the multiple votes from voter fraud from the Dems"

I think you are right. This is preparation for people to not ask questions when W is ahead by 3pts, but still loses because of Dem vote fraud. They are pre-emptively covering their a__es.

[HawaiianGecko]

	Frank Newport of the Gallup organization says undecides break for the challenger.   Boston, Mass.: From a historical standpoint, where do Bush's numbers need to be going into election day? I've heard that he needs to be above 51% to offset undecides who break for the challenger. If this is the case, it seems his people might be a bit concerned at this point. Barring any major news or gaffes, I suspect we'll see these same polling numbers come Nov. 2nd. Frank Newport: Interesting point. There is a consistent conventional wisdom that "undecides break for the challenger"; i.e., that "what you see is what you get" for an incumbent. The idea is that incumbents are well known, and those who remain undecided have in essence already decided that they are wary of voting for the incumbent.

[jayef]

Not even remotely possible. There is not way that the Democrats can produce 5 million fraudulent votes. Think before you post.

[ambrose]

http://news.bbc.co.uk/1/hi/world/americas/1008046.stm

Monday, 6 November, 2000, 12:21 GMT

How Gore could lose, but win



The electoral college vote could go Gore's way

By Gordon Corera and Steven Cviic in Washington

How can Al Gore lose and win the US presidential election at the same time?

It may sound like a trick question, but with a race this close and a lot of "what ifs" buzzing around, there is one scenario which is causing a lot of excitement and apprehension in political circles.



Bush could lose the electoral college vote

Typically, the workings of the electoral college are left to the real obsessives - but this time round people are paying particularly close attention to it.

Both main candidates know Tuesday's vote will be determined by who wins in a dozen or so marginal states including Florida, Pennsylvania, Michigan, and Missouri.

This is because under the US system each state counts for a certain number of electoral college votes, out of total of 538, and it is these votes which will determine who wins the presidency.

The number of votes per state corresponds roughly to its population. Huge California for example has 54 votes, tiny Delaware has just three, and whoever comes out on top in each state gets all its votes.

Whichever candidate wins the most votes in each state wins all of that state's electoral college members. The other candidates get none.

Number crunching

Republican George W Bush is given a slight lead in the polls, but because of the electoral college system, he could win the popular vote nationwide, and still lose the White House.



Al Gore is hoping to pick up key swing states

Two Columbia University political scientists have calculated Mr Gore could win the 270 electoral college votes needed, as long as he stays only 2.2% or less behind Mr Bush in the popular vote.

This is because Mr Bush's big national vote is piled up in the South, while where Mr Gore leads (including battleground states) he leads by only a small margin.

And at the moment, in many of the states Mr Bush is leading, he is leading by big margins - all the extra votes in these states are, in effect, wasted votes.

So if Al Gore were to win states like Michigan, Florida and Pennsylvania by small margins, he could win the electoral college but lose the popular vote.

Such a result would not be unprecedented. Twice before a candidate has won the popular vote but lost in the electoral college.

-snip-

[jwalsh07]

Have you seen the polls showing Bush etting almost 20% of the black vote? I think it may be real, the question is where is it happeneing? Definitely in the south but I don't know about the midwest. What do you think sensei?

More food for thought. Registration means nothing unless you can turnout a warm body. I don't see these new dem registrants crawling across broken glass to vote for John Kerry. Perhaps this is a case where quality one ups quantity.

[ladyinred]

Don't feel bad, everyone makes mistakes, people just love to give everyone a bad time around here!!!

[jayef]

Uhhh . . . that should be NO WAY! Think before you post challenging someone to think before they post ;-)

[rintense]

That argument simply won't work this year. According to a FNC poll, 66% of likely voters think it is 'upsetting' to change Presidents in a time of war. They view it as unsettling. Those numbers alone point to a swing to Bush with undecideds.

[ambrose]

Year	Race	1 Month Out	Next To Last Poll	Result	Verdict
1936	Incumbent FDR vs. Landon	FDR 51, Landon 44	FDR 54, Landon 43	FDR 61, Landon 37	Broke towards incumbent.
1940	Incumbent FDR vs. Willkie	FDR 51, Willkie 42	FDR 51, Willkie 42	FDR 55, Willkie 45	Broke evenly.
1944	Incumbent FDR vs. Dewey	FDR 47, Dewey 45	FDR 47, Dewey 45	FDR 53, Dewey 46	Broke towards incumbent.
1948	Incumbent Truman vs. Dewey	Dewey 46, Truman 40	Dewey 50, Truman 45	Truman 50, Dewey 45	Broke towards the incumbent.
1952	No incumbents. Democrats the incumbent party. Ike vs. Stevenson	Ike 51, Stevenson 38	Ike 48, Stevenson 39	Ike 55, Stevenson 44	Broke evenly.
1956	Incumbent Ike vs. Stevenson	Ike 51, Stevenson 41	Ike 51, Stevenson 41	Ike 57, Stevenson 42	Broke towards incumbent.
1960	No incumbent President. Incumbent VP Nixon vs. Kennedy	Kennedy 49, Nixon 45	Kennedy 49, Nixon 45	Kennedy 50, Nixon 50	Broke for incumbent VP.
1964	Incumbent LBJ vs. Goldwater	LBJ 64, Goldwater 29	LBJ 64, Goldwater 29	LBJ 61, Goldwater 38	Broke towards challenger.
1968	No incumbents. Democrats the incumbent party. Humphrey vs. Nixon	Nixon 43, Humphrey 31	Nixon 44, Humphrey 36	Nixon 43, Humphrey 43	Broke towards incumbent party.
1972	Incumbent Nixon vs. McGovern	Nixon 60, McGovern 34	Nixon 59, McGovern 36	Nixon 61, McGovern 38	Broke evenly.
1976	Incumbent Ford vs. Carter	Carter 47, Ford 41	Carter 48, Ford 44	Carter 50, Ford 48	Slight break towards incumbent.
1980	Incumbent Carter vs. Reagan	Carter 47, Reagan 39	Carter 47, Reagan 39	Reagan 51, Carter 41	Broke strongly towards challenger. So did some of the decideds.
1984	Incumbent Reagan vs. Mondale	Reagan 58, Mondale 38	Reagan 56, Mondale 39	Reagan 59, Mondale 41	Broke evenly.
1988	No incumbent President. Incumbent VP Bush vs. Dukakis	Bush 49, Dukakis 43	Bush 53, Dukakis 39	Bush 53, Dukakis 46	Broke evenly from a month out. Broke slightly towards challenger from the 2nd to last poll.
1992	Incumbent Bush vs. Clinton	Clinton 47, Bush 29	Clinton 43, Bush 36	Clinton 43, Bush 38	Broke towards incumbent.
1996	Incumbent Clinton vs. Dole	Clinton 48, Dole 39	Clinton 52, Dole 41	Clinton 49, Dole 41	Broke evenly.
2000	No incumbent President. Incumbent VP Gore vs. Bush	Bush 48, Gore 43	Bush 47, Gore 45	Gore 48, Bush 48	Broke towards incumbent VP.

With only three exceptions, the incumbent party's candidate did at least as well as the challenger with the late breaking vote, and usually did a lot better. The three exceptions? Barry Goldwater and Mike Dukakis each made small gains while remaining considerably behind in an impending crushing defeat. Ronald Reagan proved to be the exception to every rule, winning not only the undecideds but also taking away considerable support from Jimmy Carter during the last weeks of the 1980 election.

[HawaiianGecko]

	Frank Newport of the Gallup organization says undecides break for the challenger.   What's the correlation between 66% of likely voters think it is 'upsetting' to change Presidents in a time of war, and fifty years of history saying that undecided voters typical break for the challenger?

[captcanada]

I saw the show. Wait for the transcript. I believe Luntz was saying that if Bush has a 3 point lead going into the election it covers all the voter registration, and cell phone owners not pollable theories, and battleground state dynamics.

Luntz says momentum is clearly with Bush, that he is 4 or so points ahead now...

That the best way to look at polls is to throw out the high, the low and average the remainder to get where the race really is.

[ottothedog]

I'm sure you can find an astrologer w/30 yrs. experience, doesn't mean that either Luntz or the astrologer can call the election. :)

BC 04 / KE 04 probably have a better idea of what the dynamic is as they both have an army of pollsters. The dynamic seems to be that sKerry Edwards are the ones trying to move the polls. The future being in the future, however, we will all have to wait and see.

[AHerald]

Thought you might be interested. FReeper Dales has done a study of this and has a very detailed chart and analysis of this subject over at his web site, dalythoughts.com (you'll find the chart related to how undecideds break toward the incumbents about midway down his page)

From 1936-2000, Dales found the following:

With only three exceptions, the incumbent party's candidate did at least as well as the challenger with the late breaking vote, and usually did a lot better. The three exceptions? Barry Goldwater and Mike Dukakis each made small gains while remaining considerably behind in an impending crushing defeat. Ronald Reagan proved to be the exception to every rule, winning not only the undecideds but also taking away considerable support from Jimmy Carter during the last weeks of the 1980 election.

[Darkwolf377]

"He's wrong. Last minute undecideds break for the incumbent."

From what I've heard, you are correct: In LOCAL races the undecideds go for the challenger, but in presidential they go for the incumbent. Or, I am guessing, they don't vote at all.