Polls, Entrance, and the Final Result in Iowa

From a Dkos diary:

What did the 2004 polls the week of the Caucuses look like? What did the "entrance poll" prior to the caucusing look like? And what were the final results?

I have pasted the significant portion of the dkos diary below:

A couple of things to notice:

• The entrance poll is similar to the actual results, but even the entrance poll is off by as much as 5% due to the resulting deals within the caucuses when one or more candidates do not meet the minimum threshold.
• Even the most reliable polls will be wrong because voter turnout is hard to project. The most important thing to note are trends, which this chart does not show. Most of the final polls showed movement for Kerry and Edwards, but the amount of movement is not always obvious since there are a large % of undecideds or people who will change their minds in the final day or two.
• Three people can win the 2008 Iowa Caucuses, and the determining factor may well be what happens to the Richardson, Biden, and Dodd supporters.

One thing to take from the 2004 polls, how do the final polls compare to the entrance poll and how does that compare to the final tally? Two candidates had higher entrance poll numbers than their composite final poll numbers, two candidates had lower entrance poll numbers than their composite final poll numbers. The final tally for each of these four candidates continued in the same direction (ignoring Kucinich who is way below the 15% threshold).

Candidates with late momentum may continue to have that momentum once inside.

-Zen Blade