WISCONSIN SURPRISE (RNC internal polls show Feingold 48, Michels 43, from 52/39 4 days ago)

The Corner/NRO ^ | 10/30/04

[Cableguy]

I'm skeptical about Republican Tim Michels beating Democratic senator Russ Feingold in the Wisconsin Senate race, but a friend at the NRSC just sent me tracking numbers that show Feingold's lead slipping from 13 points to 5 points in four days:

10/26 -- Feingold 52 percent, Michels 39 percent
10/27 -- 51-41
10/28 -- 49-41
10/29 -- 48-43

Some Republicans have said all along that Wisconsin is their dark-horse race.

[Cableguy]

Caution: this is RNC internal poll, so probably biased a little (but not a lot) towards Republicans.

[mware]

If it is truly an internal poll, it is not hyped.They need accurate information to go by.

[aft_lizard]

My guess is that internals general will give the candidates best and worst cases, so this could be a worse case.

[flashbunny]

the poll with 39% was way off - because it was another one of those useless university polls.

[Always Right]

This would be hugh if it happened. We can always hope.

[Warlord]

"Caution: this is RNC internal poll, so probably biased a little (but not a lot) towards Republicans."

Actually, the most accurate polls are internal polls taken for strategy purposes and not meant for public disclosure. It would not make sense to bias a nonpublic poll from which a campaign wants good information. I think we'll see more "internal poll" leaks as the final strategies and tactics are put in place.

[pgkdan]

Caution: this is RNC internal poll, so probably biased a little (but not a lot) towards Republicans.

Just the opposite is true. Internal pollong is very sophisticated and very unbiased. If you intend to win you need to know the unvarnished truth. They can dish the spin to the media but the candidate gets the truth.

[codercpc]

I just saw Tim Michels this morning at the Bush rally in Green Bay, and he said that he had just seen some new results that showed them tied!!!(Don't know if it was with the MOE or not)

[Dog Gone]

A 5 pt lead that is accurate going into this weekend probably means that this race is unwinnable for us.

[Ladysmith]

Well, there's an ad out here with Rudy G. endorsing Michels. Maybe that played a factor in it!

[Cableguy]

One of the biggest variable (if they chose to use it) is party ID in a poll. Moving this around really changes numbers. Republican pollsters will fix this at the turnout they think will occur. If they are wrong about the turnout, then the poll will be way off. Dems do the opposite. That is why neutral polls (like Mason-Dixon) or bi-partisan polls (like the Battleground poll) are much better predictors.

[Oblongata]

Also you have to remember the coattails factor. When someone is answereing a poll about senate races they are more likely to say Feingold. When you into a voting booth and are filling out ballots, you are more inclined to vote along party lines. If Bush wins WI, there is a good chance Michels will win as well.

[WinOne4TheGipper]

I was watching Fox News Sunday and Meet the Press last Sunday (I can't remember which one this was on) and either Wallace or Russert pressed the Republican who was there to pick a Senate race they may pick up surprisingly. The Rep. said it was this one. Here's hoping he's right.

[The Old Hoosier]

They also just put $600,000 into this race this week.

[Cableguy]

I like that. Putting money where their mouths are. That means much more than polls.

[Warlord]

"One of the biggest variable (if they chose to use it) is party ID in a poll. Moving this around really changes numbers. Republican pollsters will fix this at the turnout they think will occur. If they are wrong about the turnout, then the poll will be way off. Dems do the opposite. That is why neutral polls (like Mason-Dixon) or bi-partisan polls (like the Battleground poll) are much better predictors."

As you suggest, internal party polls take advantage of information that is not available to the "neutral polls" like Mason-Dixon. This means more precision, not less. Otherwise, the parties would rely on "neutral polls" for strategy and save the millions spent on internal polls.

[Buzz Crutcher]

Michels is an impressive candidate who might be able to pull off an upset if he has enough funding. I just donated a few bucks at www.michelsforsenate.com

[Clintonfatigued]

Is Michels waging an extensive media campaign? Or has Feingold committed a serious gaffe? There must be a reason for the movement.

[Cableguy]

"take advantage of information that is not available"

What would that be?

"This means more precision, not less"

Why would internal polls be more precise?

[StarFan]

I think we'll see more "internal poll" leaks as the final strategies and tactics are put in place.

Amen as they'd help with the conflicting polls we hear hourly.