If you were a bookie... Polls and bets on the 2004 elections

Hm, I dunno. Hope its not cause they have some inside information. Polls for Ohio have actually been OK for Kerry lately - I mean, relative to the huge fluctuations between polls. It's losing WI, IA and NM I was getting worried about now. Kerry can win OH but lose those three states and still lose (if he doesnt win FL).

Nimh
If they can keep the republicans dirty tricks at bay. Kerry will win Fla.

I doubt it. Wanna bet?

sitting just on the other side of the border (the northern border !), i hear more and more analysts say that is becoming increasingly difficult to predict the outcome of the election. i feel that there are simply too many unknowns that may impact the outcome. it has been pointed out by some analysts that polling firms have a great deal of difficulty to correctly read the "mind" of the voter. one analyst pointed out that most polling is still done by telephone, but that people having only a cellphone are not included in the polling process (?). i also watched a group of so-called "undecided" voters being questioned after watching one of the debates. while the majority gave the debate to kerry, when further questioned they thought they might vote for bush after all. i wonder if it is really possible for the pollsters to take all these variables into consideration (one of the encouraging news is that there are apparently dozens of lawyers ready to challenge the election outcome no matter who wins - that should provide a lot of fodder for the news !). hbg

Nimh
What do you want to bet on? Keeping the republicans dirty tricks at bay or a Kerry win?

Polling Report recycled a fifteen-year old article to remind us of something that is sure to give a Democrat new hope:

Incumbent Races: Closer Than They Appear

Basically, the argument is that in an incumbent race the undecideds tend to massively fall to the challenger. Consequently, while the final pre-election polls usually get the percentage for the incumbent right, the challenger mostly does distinctly better than the final poll predicted. Rule of thumb therefore is that if the incumbent polls less than 50%, he is likely to be in trouble. Hence: "If a poll is a mirror of public opinion, think of an incumbent poll as one in which objects are closer than they appear."

Said conclusions were derived from an analysis of 155 races, ranging from mayoral to Congress elections. I can think of Presidential races where things didnt happen that way though. Bush Sr, for example, was defeated in his run for reelection by a narrower margin than the final pre-election polls had indicated, so there the undecideds apparently broke for the incumbent.

Au - that Kerry won't win Florida. But here's hoping, though ...

I can't wait till it's over and find out whether we are on the way to recovery or the road to disaster.

I think we will stay on the road to recovery....unless Kerry is selected.

Larry
How can you be so consistantly incorrect. It must be a gift.

Good evening, yall. Thanks hamburger in Canada for your comments; I think we are all getting weary of polls. This election will probably come down to a 1 or 2% difference in the popular vote nationwide. Decisive will be how the popular vote in a few critical states turns out because the electoral votes from those states will determine who the winner is.
Polls have a disclosed "margin of error." Analysts and reporters should mention that but they don't because in an election this close that would make all polls meaningless. Do you agree, nimh?
Nimh and I have touched on the cell-phone thing you mentioned. To my mind it is not a big thing but it might exclude a lot of young people from being polled. Are young people who don't have regular phones prone to vote more towards one party or the other? I don't know. But they are not in the poll.
The big, big, really big variable in this election will be the huge numbers of newly registered voters. Double digit increases in potential voters in some states. The polls tend to ignore them because, when the pollster calls, she/he is looking for an opinion from those of us who are Likely Voters, folks who have trudged or raced to the balloting place on a regular basis and who will do it again on Nov 2nd. The pollsters can't predict how the hundreds and hundreds of thousands of new voters will vote...if they show up. Agreed, Nimh?
Forget the polls--8 days to go. Thanks. -rjb-

Walter, that's the one!!! Glad it's one of your favorites - I've emailed it to everyone I know LOL.

Over at Slate, an interesting argument about how both candidates seem to be doing better the more the campaign is NOT about them:

Second Best - Should Democrats worry if Kerry noses ahead of Bush?

Today's Zogby poll has Bush at 48% and Kerry at 46%. Both gaining a point compared to yesterday.

With an eye to the Polling Report analysis I linked in above, there's good news and bad news in the internals.

Good news: "Zogby said only one in five of the undecided voters felt Bush deserved a second term in the White House. "These are the most important people in America today. Well over 90 percent of undecideds tell us they're going to vote. And only 20 percent say the president deserves to be re-elected," he said."

These numbers confirm the PR analysis: the undecideds are likely to break against the incumbent. In this case, it appears, quite massively, and on a high turnout.

Bad news: "The number of undecided voters in the poll fell to 4 percent. [..] Bush reached 50 percent in the last of the three days of polling on Saturday that went into the survey."

The idea in the PR analysis being that if an incumbent is at or over 50%, he's quite safe. The undecideds may break to Kerry massively, but if there's only 4% of them left and Bush gets 50% even without them, that doesn't do us much good.

Link

"Zogby said only one in five of the undecided voters felt Bush deserved a second term in the White House"

I might have responded with the majority to the question as well.

But, I voted for Bush anyway, because the challenger deserves it less, IMO.

Probably other independent undecideds share that opinion. But, of course, each is free to choose based on their own value judgments.

I imagine things will "break" a bit more distinctly one way or the other startin' around the upcomin' mid-week, and continue whichever trend is shown through The Election. I also imagine there'll be a few politically impactful headlines startin' with Monday ... the down-and-dirty real meanness is about to begin.

A note, going into the 2000 election, the final polls averaged about a 2.167 pro-Bush spread, though they were sorta all over the place. I'll conjecture should the spread next weekend/Monday be significantly greater one way or the other than that, the die pretty much has been cast.

I dunno timber, one thing I keep coming back to is whether the pollsters' methods will hold up with all of the ambiguity and volatility in the voting populace. Lots and lots of new voters registered, all kinds of stuff going on to try to affect influence that one way or the other (fun stuff like throwing out Dem registrations), not to mention the cell phone thing/ whether they are getting a good sample at all by only calling land lines.

As in, if the pollsters are as accurate as they were in 2000, I agree -- but I wonder if they can possibly be as accurate with so many variables.

I've seen opinions both for and against the cell phone thing being relevant (poll-wise). Did this come up in 2000? I can't remember. Are there that many more cell phone users now than then?

OT: Andrew Sullivan said in an interview this morning that Michigan is trending Bush (he seemed a bit worried). I can't find any info on this. Any Michiganders/Michiganians (lol) out there?

All this polling stuff means nothing. Nader will win in a landslide.

He should you know