You know what is funny. You guys are acting as if opinion polls are somehow the reality and votes don't matter. Drawing conclusions like: Romney is getting a lot of the last undecideds.
Opinion polls are only accurate to a certain degree. In Iowa, opinion polls are even less accurate. When they show a close race, know it can go either way.
I don't think Romney can overtake the +62000 early vote margin Obama already has.
But if he does, then he's won the election.
Opinion polls are a window into how people are going to actually vote. That's why they've been used to predict election outcomes for, oh I don't know, over a century. The fact that I even have to explain this to you is mind-boggling.
They are an indicator of how people think, but not an accurate predictor. Particularly not if they fall within the range of the poll sampling error.
Really? They aren't?
http://fivethirtyeight.blogs.nytimes.com/2012/11/04/nov-3-romneys-reason-to-play-for-pennsylvania/#more-37144
Six elections, and the poll average for a state was wrong THREE times. Again, THREE.
I've said this to you Romney guys before. If you argue with this, you are LITERALLY arguing with facts and numbers. Good luck with that.
Yes, I like a good argument. That’s what’s been missing here.
If you take a closer look, you will notice that the closer the polls the less accurate they were in predicting the outcome.
Now, unlike any previous election, this time we have many states that are close.
A small movement last minute could produce completely different results.
I am betting that such a movement will happen. I think that the undecideds will break 75-25 for Romney. The precedent for this is the last somewhat close election (Bush-Kerry 2004). The undecideds broke heavily for Kerry and Bush nearly lost in spite of being ahead by a larger margin in polls.
The closer the polls got the less accurate they were? Where do you see this? Look at 2008, where 29 states received enough poll to draw a good sample size. Of those 29 states, the polling average missed the actual result by 3+ points 9 times. Those states?
AK, AR, AZ, IA, NM, NV, NY, WA, WV
of those, only Nevada was contested, and the MOV for Obama was HUGELY underestimated by every pollster. Everywhere else, it was a blowout.
Now look at the 11 battleground states in the 2008 election
CO, FL, GA, IN, MO, MT, NH, NC, OH, PA, VA... check out the polling averages vs. the result.
The polling average result was within 1.5 pts of the actual in every single one of those states. Why? Because THEY WERE THE BATTLEGROUNDS! They were polled over, and over, and over, and over using a scientific method of polling. Statistics 101, bro. The bigger your sample size, and the more samples you get, there more you are going to approximate the reality of the situation. So, I don't know how your argument holds up. And the CLOSEST of the battlegrounds... Indiana was within .2, Florida was within .3, Missouri was within .5... and the two states polled the most in 2008, VA and OH...
ON THE DOT. As for 2004? How about you go back to that chart. Bush, the incumbent, did better than his polling average in the actual result 18 out of those 24 states. Only six (WI, WA, PA, NY, NV, NM), were better than Kerry than the polling suggested. Where are you getting that undecideds broke for Kerry 75/25? I've never seen that before and the numbers certainly don't back that up... if anything, it would seem that polling underestimated the incumbent.
http://www.college.columbia.edu/cct_archive/nov05/forum.phplook at the chart on that page, I couldn't paste it here. EVERYONE had that race as close as hell. All but 2 had it within 2 pts, some had Kerry winning. Where are you seeing that Kerry took 75% of undecideds?
