How can you project a winner w/ 1% of vote in? Not a valid sample amount.

[southerncrone]

Exit polls must stop. With the internet availability, this is influencing voters in the western parts of the country.

Yet another way our elections are improperly influenced.
:puke:

[stray cat]

1. How often have they ever been wrong when they have?

They usually have it right - these are not based on exit polls

[AndrewP]

2. They are usually correct, but your point stands

I also worry about the Western part of the country not voting late when things look bad elsewhere.

[hayu_lol]

3. Its that dammed liberal media...jumping the gun with...

just 1% returns in.

Want to stop exit polls...go to Mail Only Voting like we have in Oregon and Washington.

Other perks: no crowded parking lots, no teabaggers shaking their bags in your face, and no malfunctioning voting machines.

[southerncrone]

6. Sounds good to me,

except having to rely on the USPS! LOL

[ThomCat]

4. 1% can be a very accurate sample size if they have enough people.

Exit polls have, historically been very accurate. If the sample is sufficiently random from throughout the polling area, a very small sample size can be used and have a very small margin of error. (In other words, have a very high level of accuracy.)

It was only since Bush stole both of his elections that they have started altering the exit polls to match the election data. But significantly, that's only in places with ELECTRONIC VOTING.

The discrepancy between exit polls and the electronic polling data is one of they key pieces of evidence of vote tampering BECAUSE exit polls are so accurate.

[Donald Ian Rankin]

5. Depends how that 1% is selected. Often, it's more than enough.

If a reasonably large number of people vote then the standard deviation in a representative 1% sample will be tiny.

The only reason 1% might *not* be enough to reliably predict an election is if you have sampling bias - if that 1% all come from the same few precincts, for example.