GanjSmokr: HotWingConspiracy: Lt. Cheese Weasel: UColorado has hit the last 7 elections on the nosey.

LOL no

It was created this year and has predicted nothing.

The poll has accurately predicted every presidential election since it was developed in 1980. It is unique in that it employs factors outside of state economic indicators to predict the next president.

The shiatty blog is wrong. Follow the links back to CU and you'll see the model was only built this year and uses historical data available only since 1980. They fit that data to the elections since then and low and behold matched the results. They haven't predicted anything.

amiable: Considering this "predictor" only has 8 data points and is based primarily on retrospective analysis (fit-to-model).

Nate Silver: I saw their paper and I think there are glaring problems with their methodology...it's false advertising to claim CU model has predicted the last 8 elections right. It's a new model. Hasn't predicted anything yet.

Why do the conservatives insist on deluding themselves this way? Isn't it obvious that this prediction is a major outlier? Is it that important to swaddle yourself in thoughts of a comfortable outcome that you're willing to risk major disappointment?

I don't know the answer to any of these questions, but this prediliction towards bubble building is exactly why I no longer want to put you people in charge of anything. Not the dog pound, let alone the presidency.

CampusReform.org is designed to provide conservative activists with the resources, networking capabilities, and skills they need to revolutionize the struggle against leftist bias and abuse on college campuses.

Created to give conservatives powerful new weapons in their fight for the hearts and minds of the next generation of citizens, politicians, and members of the media, CampusReform.org facilitates the establishment of conservative student networks and supports their development as a powerful voice of activism on their campuses. It makes available new opportunities for groups' interaction with alumni, parents, faculty, and other members of the broader community interested in taking a stand for conservative principles on America's college campuses.

Connecting up-to-date communications technologies to a principled stand for limited government, the free market, national defense, and traditional values, CampusReform.org makes possible a new generation of student activism to identify, expose, and combat the radical left now.


Why is everything a war with Conservatives?

Lt. Cheese Weasel: UColorado has hit the last 7 elections on the nosey.

LOL no

It was created this year and has predicted nothing.

theorellior: Cletus C.: A squandered pile of stimulus dollars.

Just because your state didn't know what to do with extra funds didn't mean the stimulus was squandered. I saw lots of "shovel-ready" highway projects around Massachusetts and other New England states during 2009 and 2010.

The 40 year old dilapidated highway I call my commute to work is getting a complete facelift with expanded lanes. I am happy.

Cletus C.: A squandered pile of stimulus dollars.

Just because your state didn't know what to do with extra funds didn't mean the stimulus was squandered. I saw lots of "shovel-ready" highway projects around Massachusetts and other New England states during 2009 and 2010.

WombatControl: qorkfiend: Mathematically possible, sure. Plausible, no. I don't see Romney sweeping New Hampshire, Virginia, Florida, Iowa, and Colorado while losing Ohio.

Although if you look at the polls, it's at least plausible. You have more favorable polls for Romney in all of those states than you do in Ohio.

In the end, you are probably going to be right. If Romney is going to be able to win those states, he's likely to win in Ohio as well. Consequently, if Romney is going to lose those states, he's likely to lose Ohio as well.

The question will be weather Romney's uptick in the polls is a sign that the race is truly shifting or not. Had Romney's numbers settled back down and Obama regained the lead, I'd say that Obama would win. But what's happening is that you're seeing Obama stuck in the same band he's always been at - 47-49%. Which suggests that Romney has an opening, and the polls are showing that 6% of uncommitted voters swinging towards Romney.

Neither candidate will likely get less that 47% of the vote - it's that 6% that will decide who is President in 2013.

What numbers haven't "settled back down"? Obama's back on the climb in the poll aggregators.

Why didn't this blogger go with the same model applied state by state? It predicts Romney with 330 EVs, which is still ridiculous, but at least relevant to how we actually elect presidents.

WombatControl: qorkfiend: WombatControl: Romney probably, but not absolutely, needs to win Ohio to become President.

I'd like to see the electoral scenario where Romney loses Ohio but wins the election.

Take the 2008 states plus IN and NC (Romney's baseline at the moment).

Scenario 1: Baseline + NH, VA, FL, IA, CO, NV = 273 Romney EVs. (This one is possible, although NV is an unlikely state for Romney.)

Scenario 2: Baseline + NH, VA, FL, IA, CO, WI = 277 Romney EVs. (Oddly enough, WI is looking better for Romney than NV is - remember that WI only barely went for Kerry in 2004, so it is becoming more of a swing state).

I wouldn't put odds on either of those happening, although both are possible.

Mathematically possible, sure. Plausible, no. I don't see Romney sweeping New Hampshire, Virginia, Florida, Iowa, and Colorado while losing Ohio.

qorkfiend: WombatControl: Romney probably, but not absolutely, needs to win Ohio to become President.

I'd like to see the electoral scenario where Romney loses Ohio but wins the election.


The only realistic shot he has is flipping Iowa and Wisconsin, which is why the NBC/Marist polling was particularly bad for Romney yesterday. Those were large samples too, 1000+ LVs. By contrast, Rasmussan only used 500 LVs in their last Wisconsin poll (Obama +2).

DubyaHater: Maybe it's misguided, but I have my confidence in Obama back. I watch one poll, Fivethirtyeight because I can't take the ups and downs of the 30 polls published everyday. Key swing states, Colorado and Virginia, are trending back towards Obama. Ohio and Wisconsin are staying blue. I wouldn't worry about Pennsylvania swinging towards a Republican.
Unless Obama says in the third debate that he personally ordered the attack in Benghazi, or the November jobs report takes a huge dump, I doubt we'll see major movement in the polls.
Hell, I even donated to the Obama campaign last night, and I never do shiat like that. All I can is, have confidence.

There has been a state of equilibrium in this race of Obama up +2. There have been short term swings depending on the news cycles involving debates, conventions, "legitimate rape", 47%, etc. that have temporarily bounced the trends away from the +2 baseline of homeostasis, but eventually over time it gradually begins to shift back to that point. We're seeing it again currently. If there's anything that' been remarkable about this particular election is how stasis it's been towards this +2 trendline over the long term.

Per Nate Silver ...

So far this cycle, betting on the polling average converging to Obama +2 has pretty much always been right.

In visual form from Sam Wang's Meta analysis....

WombatControl: As much as I would love to believe that Romney has a 77% chance of winning, there's no way in hell that's true.

But the people who blindly say that the swing state polls are good news for Obama. When you start seeing the race in the key swing states tightening, that's not good news for an incumbent President. The only good polls for Obama recently have been the NBC/Marist ones, that have been heavily skewed towards the Democrats. Otherwise, we're seeing a pro-Romney trend in OH, CO, FL, VA, and even in states like IA, NV, NH, and WI. Hell, there was even a poll showing Romney ahead in PA! (Susquahenna). Now, I don't buy that Romney is really ahead in PA, but if that race is close that means that Obama is going to have to start playing defense more and more.

Romney probably, but not absolutely, needs to win Ohio to become President. And right now President Obama has a slim lead there, but enough to give him the edge. But if Ohio keeps moving in the pro-Romney direction, Obama is likely to lose.

Right now the odds are shifting towards Romney - to be fair I'd have to give the President a slight edge to win, but that edge keeps getting smaller and smaller.


Wow, almost your entire post is false.

TripSixes: It would be delicious if Obama won the electoral college but not the popular vote. Then there might be a bipartisan effort to get rid of the electoral college.

This would absolutely be the best outcome to come out of this election. It's amazing to me that people still stand by an antiquated election method in which the votes of 38 out of 50 states are deemed completely irrelevant right from the outset, including the voters in our largest urban areas as our nation gets more and more urbanized.

I can just see the results of a Romney win and it pisses my face off

Does everything in his power to hamper economic growth, but recession is ending so economy finishes correcting itself. Romney gets all the credit. Obama added to worst presidents list with Carter.

Pisses my farking face off.

Or...as it says in all those financial disclosure statements, "Past performance is not an indicator of future results."

You can take your cue from the pop-up ad that appeared with TFA: "Be the first to know about more leftist abuses on college campuses. Sign up here."

Yeah...that's an unbiased, impartial source. Riiiiiiiiigggtt!

Mercutio74: Wow... 8 correct predictions. With a sample size that large the predictive value is impressive indeed....

No, zero predictions. Basically they took 8 data points and fit an equation that would fit all data points. Then they take that equation and see what it says about this election. It would have to get several elections right before you would want to start talking about predictive power. But the fact that it is so far off (no one thinks there is anywhere near a 77 percent chance of a Romney victory) suggests to me that it won't get it's string of successes. Most likely after this election they will add this election and come up with a new equation which will get the next one wrong. It's important to stress that this "model" has already shown itself to be wrong, because even if Romney wins they put the odds way too high.

Hmm, who to believe... man with numbers, or man with assertions of gut feelings?

Berry noted his model has never been wrong at predicting the outcome of a presidential election.

"For the last eight presidential elections, this model has correctly predicted the winner," he said.

So, did it predict Bush would win in 2000? He sure as hell didn't win the popular vote.

Now, it is certainly possible that Romney will win the popular vote. But the fact that this "model" thinks that is 77% likely means it is deeply flawed, regardless of past performance. A good model would not give a 77 percent chance to something that is probably around 45 percent likely to occur. If you don't see this, you are engaging in magical thinking, and you might as well go buy Romney betting futures for a "bargain".

TripSixes: It would be delicious if Obama won the electoral college but not the popular vote. Then there might be a bipartisan effort to get rid of the electoral college.


Ok, delicious except for having to swim through the morass of morons.

I fear that we'd hear much about "second ammendment solutions" should that come to pass.

Nate Silver is currently showing Obama with a 70.1% chance of winning. Who am I supposed to believe?

I think I'll go with the nerdy Jewish kid. Jews are good at that math stuff.

It would be delicious if Obama won the electoral college but not the popular vote. Then there might be a bipartisan effort to get rid of the electoral college.


Ok, delicious except for having to swim through the morass of morons.

You know, if Romney wins, even though the economy has been improving steadily for months now, they'll credit him. Think about that.