Originally Posted by Botnst

Here's another point that I think argues favorably for some sort of representative voting (allowing that EC is one of many models of representative voting).

First, we have to accept that there is a difference between mistakes and errors.

There's this concept called, "sampling error" in statistics. It is easy to understand in the following way. If I give you a ball bearing and ask you it's circumference, you may start by winding a tape around it and reading-off the marks. Perhaps you'll do it several times and take the average.

Or if you don't have a tape maybe you'll pick-up a ruler and estimate the diameter. Or maybe you'll borrow a micrometer from your neighbor an estimate the diameter. Then you'll multiply by pi and inform me of the circumference.

Each of these methods will give an answer but we will soon realize that there is some slop to the right of the decimal. As our method improves the ditches to the right of the decimal become increasingly stable. But there's still going to be an error out there that nothing we do will eliminate.

That same concept is in play for any method we may use to measure any attribute in the real world. In Florida the sampling error came from the various ways of interpreting "chads." The result is that we do not know and probably cannot know, the exact ballot count in Florida. Or any other state. Now depending on the method of enumeration, this or that state may do better or worse than Florida, but none of them will claim to be absolutely precise and accurate.

So what we do is this. Instead of electing the prez directly, we elect electors. The election of them is certified by the state legislature. This is a check on poor voting technique. We then empower an exact and finite number of them to to vote on a certain date and then count all of the ballots. This is a check on corruption since the chances of corrupting 535 electors without a whiff of suspicions is extremely small.

What is the advantage of electing electors? Well, the counting error is distributed over the number of electors partialed over the several states. This distributes the measurement error (which is an 'honest' source of uncertainty) and mistakes (often a dishonest source of uncertainty) over a large area and under greater scrutiny than can be exacted on a voter population of perhaps 100 million or so.