[ambrose]

Did you read SurveyUSA's REPORT? You have to follow TRENDS. This smear was late breaking. Do the recall margins freefall into election day or did the drop finally stop on Monday? The drop may well have stopped on Monday and the recall wins. But it isn't in the bag by any means.

"Erosion in Arnold Support, But Not at Tipping Point"

"Margin for recall dropped 25 points to 14 points between 10/3, 10/4 and 10/5"

http://www.surveyusacom/2003_Elections/CA031006RecallReplace.pdf

[goldstategop]

Its still double digit by your own figures. If the news were really bad, I'd think Press would rub the Republicans faces in it.

[RoseofTexas]

Fox just reported HISTORICAL RECORD NUMBER at the polls! BIG WOOPI! WHAT ARE THE EXIT POLLS?!! DANG IT!

[EllaMinnow]

ambrose, your link isn't working for me.

[Miss Marple]

Quit trying to get everyone upset. There is no point to any of it except documenting what is happening and getting everyone to turn out.

I was certain they would try to steal this. Whether they are successful or not depends on the turnout, plain and simple.

By the way, I just got a call from a catalog company in California so I asked the young man if he had voted. He was Hispanic, and acted rather befuddled when I asked him (in a non-political, cheery way). He told me he was at work. I told him to remember to vote (but I don't think he will). There is an anecdote at least as good as those interviews on CNN. LOL!

[William McKinley]

Yes, I read the report. The support for the recall went from 61 to 57. A four point drop. Of course, that means that the other side would go up four points. Which it did, and they reported as an 8 point swing. All well and true, and all within the margin of error, all well within the range of what could be statistical noise.

As a matter of fact, the last seven polls SurveyUSA did had values of 62, 58, 57, 61, 62, 61, 57. If the true support was holding steady at, say, 60%, and you took seven samples with a MoE of +/-3, you would expect to get fluctuations like this. In other words, could there be a trend there? Yes, but there is no reason to assume there is. The polling data just doesn't tell us that information.