Posted on 10/21/2008 5:17:15 PM PDT by WatchYourself
The following is an analysis of the 17 state polls adminstered by Rasmussen during the final days of the 2004 election, consisting of mainly "contested" and toss-up states.
Findings:
1)In 17 Polls, Bush Was Underestimated in all 17, Kerry was Overestimated in 4 and correctly predicted in 2. Note that these polls also included a percentage of "undecided" voters. So to be "overestimated" is actually extremely difficult when 5-10% of voters do not even participate with a Bush/Kerry answer. If you remove "no comment" voters from the sample, Kerry's support is overestimated in almost every single state poll, while Bush
2) In all, Kerry was given a proven-to-be-wrong net advantage in 11 of 17 States, Bush was given a net advantage in just 3 states, and the margin of victory was correctly calculated in just 3 states.
3) In the 3 States Bush was given an advantage in, the polling was off by just an average of 2.3%. In the 11 states Kerry was given a miscalculated advantage, he was overestimated by an average of 5%.
4) In all, Bush was Underestimated by an average of 3.5% per state by Rasmussen, while Kerry was Underestimated by just .7%, meaning Kerry was overestimated by about 2.8% per state poll.
5) Final Summary: Kerry had a net polling error advantage of almost 3% state-by-state. 65% of Rasmussen state polls gave Kerry a greater advantage than he actually received, while 17.5% Gave an advantage to Bush, and 17.5% gave no advantage. Kerry was almost 4 times more likely to receive a polling error in his favor than Bush was, and the margin of error was by an average of almost 5%.
Notes: Rasmussen, like many polls, uses an "adjusted" polling average where he determines the percentage of Democrats, Republicans, and Independents that make up his polls. This certainly leads to less volatility in the polls, but Rasmussen also regularly overestimates Democratic support and rarely overestimates support for Republican candidates.
Alabama - Bush Under 10% Kerry Over 5% (Kerry +15)
Arizona - Bush Under 5% Kerry Over 1% (Kerry +6)
Arkansas Bush Under 5% Kerry Over 1% (Kerry +6)
Colorado Bush Under 2% Kerry Under 2% (EVEN)
Florida - Bush Under 2% Kerry - Even (Kerry +2)
Illinois - Bush Under 3% Kerry Under 3% (EVEN)
Louisiana Bush Under 9% Kerry Over 4% (Kerry +13)
Michigan Bush Under 2% Kerry Under 1% (Kerry +1)
Minnesota Bush Under 1% Kerry Under 3% (Bush +2)
Missouri Bush Under 3% Kerry Under 1% (Kerry +2)
Nevada Bush Under 2% Kerry Under 1% (Kerry +1)
New Hampshire Bush Under 2% Kerry Under 1% (Kerry +1)
New Jersey Bush Under 5% Kerry Even (Kerry +5)
New Mexico Bush Under 2% Kerry Under 5% (Bush +3)
Ohio Bush Under 1% Kerry Under 3% (Bush +2)
Pennsylvania Bush Under 2% Kerry Under 2% EVEN
Virginia Bush Under 4% Kerry Under 1% (Kerry +3)
We all need to call our friends and make sure any elderly relatives go vote. Make sure they are going to vote McCain Palin. :-)
I think the fires from the riots will be seen from space.
I wonder if the pollsters will tighten up the numbers before election day to avoid blame.
Of course they will tighten the polls, they are doing that now. Rush has called the polls correctly every step of the way. He has nailed them on the days the race would open up and for whom and when the race would tighten. He is dead on the money when he said they would tighten right about now.
“Today, a friend told me he wasn’t going to vote because he thought Obama has it all locked up.”
Your friend is lucky he has YOU!
But many people won’t have a friend to clue them in, which is the reason for liberal-leaning polls.
Writing a letter to the editor, and/or calling a talk show would reach a lot of voters.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.