[r9etb]

Look at the sample size of polls and the control group of accurate polls for other races. If it were just one poll that had her winning then your point would hold merit. We're talking about seven polls, though, from different firms and probably slight variations in methodologies that were all in agreement in having Angle up in this race.

I'm still going to to blame the sampling methods, as opposed to calling out "fraud."

At root, the methodologies aren't that different, really: to collect most (not all) of the data, the polling organizations have all still got to call folks on the phone.

So the question is: is there, in Nevada, some systemic factor that affects who you can call, and who is likely to answer, that was not accounted-for in the polls? For example, is there a "no-call list" law that favors land-lines over cell phones, or vice versa? Is there a particular demographic of voters that somehow are less likely to be reached by telephone?

I'm willing to bet that the answer to the poll discrepancy lies in that sort of phenomenon, rather than fraud.

An interesting and highly relevant example is the famous Literary Digest poll of 1936, that predicted Landon would defeat Roosevelt in a landslide.

The sampling methodology was very carefully constructed, but the problem was that the sample population was selected mainly from automobile registration and phone books. The surveys were mailed out, and results were drawn from those who returned their surveys, and the response rate was low. The combination of these three, led to a disastrously wrong prediction.

Pollsters have been trying to deal with similar problems in recent years. Lots of people refuse to participate, and many hang up when they recognize that tell-tale pause between answering and the time when the pollster hits the line.

[Lexinom]

Why the disparity, then, against the backdrop of the relative accuracy of the polling in other races? Surely they, too, would contain one or more of these additional factors. Why just Nevada? For your theory to hold up we'd need to compare previous Nevada races and see if their outcomes, likewise, deviated widely from the preceding polling data.

My theory: Fraud was indeed committed on a large scale in a few high profile, hand-picked races. For political reasons, those who were most harshly attacked and ridiculed could not be allowed to win. Hence the $1 million cash infusion to Reid race on Monday - too late to buy advertising time. What was that money for, then, and where did it go?