0 votes

How do you project a winner with < 10% in?

Am I a dumbass?

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

because Sheeple are very predictable.

A lot like Sheep actually. Bahhh,,,

if 10% of the precincts contain

a large enough number of people than they can call it. It is mathematically justifiable.

Its all based on statistics

It looks like black magic but it really isn't. Provided that the reporting precincts behave like a random sampling (this is the big IF) then you can predict an election with much less than 10%.

I recently took a graduate course in statistical experimentation. Based on your sample size and the variance from precinct to precinct, a confidence limit can be placed on the predicted outcome. Depending on how tight the confidence interval is, you can call the election with a certain probability of accuracy.

There is more to it... probably more than you want to know (it was more than I wanted to know;).

10% is a HUGE sample if it were random

10% is a huge sample and if it were random, would be more than enough to predict the outcome unless it was really close.

The problem is, that smaller precincts probably report faster. To the degree that this makes a difference because of demographic differences, it can be unreliable to predict based on them.


What do you think? http://consequeries.com/

It's the media's game they use all the time! Propaganda!

It's just to keep everyone off balance and lead the people into believing these are the people who SHOULD win - plants that seed!

depends where the votes are coming from

that is how it is projected, he must be wining rural and urban areas by almost same percentage

This baffles me too by my

This baffles me too by my #'s there are still 600,000 votes left. It jsut don't make sense.

To maybe sway undecideds

on the way to the polls? Alot of Americans dont vote their concience. Almost like standing on the floor at the NYSE- like wanting to follow the market trends maybe. Thats the only advantage I can figure.

many factors

including exit poll results, currently what the results are showing and also looking at where these results are coming from, precinct by precinct.