re: #91 palomino

The article you quote was from the day after the 2008 election, before all votes had been counted, and lists 6.15% as Obama’s winning margin. He actually won by 7.2%; sometimes it helps to wait for all votes to be counted before citing an article.

The larger point remains unchanged: if you take polls from 20 different pollsters, average them out over many months, you’ll usually get a quite accurate advance portrait of what election day will look like. Just as has happened over the last decade plus, during which we’ve had no big election day shockers.

Well, yeah. Margin of error applies to THIS PARTICULAR sample, and whether it lies within the actual population. If MULTIPLE samples lie in the same general area, then stop paying attention to the margin of error, because it doesn’t apply to multiple samples.