By the way, official tally has it at 9.2. http://www.electionreturns.state.pa...n.aspx?FunctionID=13&ElectionID=27&OfficeID=1
Uh, has it? Obama doesn't usually cut huge Clinton leads in these states? What was the difference between the candidates in Pennsylvania polls after Texas and Ohio, or even before that? I bet the average wasn't under ten percent.
OK, here we go... assuming the margin of 9.2 holds... The predictions: Major - 16 FranchiseBlade - 12 Nolen - 12 wrath_of_khan - 7 (and a few post later, 5 or 6) Trader_Jorge - "tight" rimrocker - 6 weslinder - 20 bnb - agrees with rimrocker, so - 6 jgreen91 - 8 A3PO - 10 - but after the first exit polls were known. Therefore, even though it was a good guess, a disqualification is in order. That leaves us with jgreen91 as the winner. Congrats.
Jgreen is indeed the winner, but I will go on to North Carolina, and Indiana, and... YEEEEARRRRHHHHHHHAAHHHHHHGGGGHHHHH!!!
Fine, how about Texas? Ohio? California? Almost every state where both have spent significant time campaigning, the result is the same. Big cities and college towns go to Obama, while the overwhelming majority of rural and suburban counties go Clinton.
It's a two-way street. Clinton has held 15-20 point leads in extremely early polling due to name visibility, but Obama has shrunk those leads to 5 and under (and occasionally leading) in almost every poll in Pennsylvania and Ohio with a week to go, and in Texas he took the lead in almost every poll with a week or two to go. But in the final days, the late deciders have broken for Clinton almost every time -- a trend continued yesterday. The theory TJ referenced was this: He's right, because evidence shows that isn't the case. If it was, Obama would peak on the day of the election, instead of peaking a week or two before and then falling back when late deciders go with Clinton.
They should just make a Obama Hillary ticket and be done with whole thing! I wouldn't even have to vote on may 6 then.
True. This shows us that Hillary can't hold a lead, and Obama can't win the final late deciders in any convincing fashion. Maybe it's a draw.
The only real conclusion here is that Obama can't win big states or blue collar votes. He can only win primaries dominated by black voters or in the absence of black voters, states that use caucuses and can take advantage of the younger voters or unemployed voters.
Has Clinton ever increased a lead she has had in the state before both started campaigning there? If she's led by 15-20, what happens to that lead during the campaign, overall? You went from saying the theory was "awful" to a "two-way street." Thanks...
I have the distinct honor of being most incorrect. That goes hand-in-hand with my early predictions of Clinton wrapping up the Democratic race by now, and the Republican race going all the way to the convention.
Those early leads are artificial based on factors very unrelated to the campaign and the issues, and I really wish polling companies wouldn't issue those polls. They're a completely unreliable predictor of the race going forward and only distort the reality of what has gone on in the actual campaigns on the issues. In almost every state where both have campaigned, Clinton has had a significantly better final vote than her numbers showed with 1-2 weeks to go. That, in and of itself, disproves the theory that Obama does better in a state the more time he spends there. And I haven't changed my opinion on anything, but thanks for inserting words into my mouth. The theory is awful because it's a two-way street. It's a cause and effect relationship. The data in the final 1-2 weeks of states where both have spent significant time and resources campaigning simply does not match the theory. Quite frankly, the overall trend is irrelevant when you have more specific sets of data. Other than to push an agenda, for what objective reason would you exclude more specific numbers and more specific trends? To push a sweeping, generalization theory like Obama only improves in states the more time he spends there, you need to have all the data points supporting you, not just a couple of broad numbers and calling it a trend while ignoring the specific nuts and bolts. The data doesn't support the theory.
Thirty Million (+) voters turning out for the primaries -- this is outstanding news for the Democrats heading into the general election.
She counts? This is so ridiculous. First of all, Clinton leads if you include all six of those states, including the estimates for the four you listed. http://www.realclearpolitics.com/epolls/2008/president/democratic_vote_count.html But more importantly, she hasn't said one thing about it -- it was a freakin' post on a message board from Trader Jorge, yet automatically it gets assumed as Clinton's position. This is what frustrates me so much. I like Obama and I like Clinton, and this desire by both groups of supporters to immediately assume the absolute worst regarding the other candidate without evidence is absurd.
Significantly (!!!1!1) better? She won by nine points. What was the average of polls on RCP a week to two weeks ago? What was the average of polls on RCP in Texas? Yes, it does, because the difference is still closer than it was before both entered the state and spent time there. I'm not excluding anything, silly goose. However, you're excluding the first four weeks or more of this campaign (even though Clinton didn't really "significantly" do better than the polling). You only want to count the final days. I'm looking at all of it. You're also excluding events like "Goolsbee-gate" in Ohio or "Bitter-gate" in Pennsylvania. Those events,which occurred near the end, may have stalled or slightly reversed the trend. Where else has Clinton significantly outperformed the final days' polls? California? There was a lot of early voting and it wasn't a one or two state primary. Where else? Did she campaign at all in Wisconsin? Objectivity... What about the general election polls from SurveyUSA?
Hillary can still win it! <object width="425" height="355"><param name="movie" value="http://www.youtube.com/v/uBGyuYKlxIg"></param><param name="wmode" value="transparent"></param><embed src="http://www.youtube.com/v/uBGyuYKlxIg" type="application/x-shockwave-flash" wmode="transparent" width="425" height="355"></embed></object>
From RCP, five polls were conducted in Pennsylvania between the 14th and 17th, approximately a week before the election. Here are the results: Rasmussen 04/17 - 04/17 730 LV 47 44 Clinton +3.0 Zogby Tracking 04/16 - 04/17 602 LV 47 43 Clinton +4.0 Zogby Tracking 04/15 - 04/16 601 LV 45 44 Clinton +1.0 PPP (D) 04/14 - 04/15 1095 LV 42 45 Obama +3.0 Rasmussen 04/14 - 04/14 741 LV 50 41 Clinton +9.0 That averages out to a Clinton +2.8 lead. Yes, winning by 9-10 points after averaging between 2-3 a week before is a statistically significant turn. From RCP, here are the 14 polls taken the week before the March 4 primary in Texas. ARG* 02/29 - 03/01 600 LV 47 47 Tie Reuters/CSpan/Zogby 02/28 - 03/01 736 LV 43 47 Obama +4.0 WFAA/Belo Tracking 02/28 - 03/01 730 LV 46 46 Tie Reuters/CSpan/Zogby 02/27 - 02/29 708 LV 43 45 Obama +2.0 M-D/Star-Telegram 02/27 - 02/29 625 LV 45 46 Obama +1.0 ARG* 02/27 - 02/28 600 LV 44 51 Obama +7.0 Reuters/CSpan/Zogby 02/26 - 02/28 704 LV 42 48 Obama +6.0 FOX News 02/26 - 02/28 600 LV 45 48 Obama +3.0 WFAA/Belo Tracking 02/26 - 02/28 721 LV 46 45 Clinton +1.0 Rasmussen 02/27 - 02/27 503 LV 44 48 Obama +4.0 InsiderAdvantage 02/27 - 02/27 591 LV 47 43 Clinton +4.0 WFAA/Belo Tracking 02/25 - 02/27 735 LV 45 46 Obama +1.0 WFAA/Belo Tracking 02/24 - 02/26 741 LV 46 45 Clinton +1.0 InsiderAdvantage 02/25 - 02/25 592 LV 46 47 Obama +1.0 That averages out to a +1.6 point lead for Obama. She ended up winning by 3.5 points, more than a five percentage point swing in the final days -- again, statistically significant. That's a ridiculously overly simplistic view. Why would you take only a couple of data points to measure when you have data points at every segment along the way to paint a more accurate picture? The Clinton campaign was ridiculously underprepared in February. The reason I'm looking at the final days is that's all that's necessary to disprove the theory! The theory is that Obama rises in popularity proportionately with the time he spends in a given state. If that were the case, there would be a continued steady rise with an increased presence. I have shown that is not the case. The candidates are much better defined now as opposed to a majority of the primary contests, especially with the opposing labels of Republican and Democrat in such hypothetical polls. Not the same scenario.
