New York Magazine was trying to be sympathetic to the popular polling figures on its own side of the political, but let out a secret in the process: Public Policy Polling cooked the books all along.

The key paragraph:

When I talked to Tom Jensen, PPP’s director, this morning, he was understandably in the mood to gloat. “These supposed polling experts on the conservative side are morons,” Jensen crowed. “Jay Cost” — the Weekly Standard’s polling expert who’d waged a number-crunching war against PPP — “is an idiot.” But Jensen conceded that the secret to PPP’s success was what boiled down to a well informed but still not entirely empirical hunch. “We just projected that African-American, Hispanic, and young voter turnout would be as high in 2012 as it was in 2008, and we weighted our polls accordingly,” he explained. “When you look at polls that succeeded and those that failed that was the difference.” Given the methodological challenges currently confronting pollsters, those hunches are only going to prove more important. “The art part of polling, as opposed to the science part,” Jensen said, “is becoming a bigger and bigger part of the equation in having accurate polls.”

Note the doublethink here: Jensen and New York Magazine are both so very pleased with the election’s results – as well as eager to take digs at Scott Rasmussen and Jay Cost – that they’re both perfectly willing to overlook the fact that Cost was right. In fact, the truth was much further along than Cost, I, or most any other commentator was ever willing to go.

I don’t remember anyone willing to say PPP was actively rigging the polls to reach chosen results, but there it is in black and white. Jensen decided in advance what he wanted the electorate to look like, and so tweaked the numbers until he got what he wanted. This isn’t a whole lot different from what Research 2000 admitted to doing, folks.

In science, it’s not just that you got the answer you wanted. It’s the process that matters. PPP, Nate Silver, and the New York Times – the heroes in New York Magazine’s story –are not practicing science. They’re taking their own beliefs and wrapping them up in a cargo cult.

Maybe that’s why Silver didn’t even try to project the House of Representatives this time. I didn’t have time to do any projections this cycle because I had to take a full time job. Nate Silver is a paid professional with the New York Times. What excuse is there for him?

I don’t get to do much at Unlikely Voter these days, but this was easy. Meet the Create Your Own Voter Model tool, a.k.a. the Unskewer.

Try out your own party identification assumptions!

Obi-Wan: Anakin, Chancellor Palpatine is evil.

Anakin Skywalker: From my point of view, the Jedi are evil.

Obi-Wan: Well, then you are lost.

By request, I’ve updated the Swingometers to include a new feature: Reverse Swingometer! Instead of plugging in a swing in the two-party vote, and getting a result in seats or electoral votes, this new features lets you plug in that result, and get the necessary swing to make that happen!

Enjoy!

Hey look, a post!

While I’m sure everyone involved is so proud of Vanderbilt’s data filtering app for its recent poll of Tennessee showing Barack Obama still losing in one of the two states he ran behind John Kerry in, but the problem is that the details are made less transparent.

What a shame.

Some pollsters put out lots of information. I’m going to single out Public Policy Polling and SurveyUSA. They put out big, beautiful tables full of information. That serves a dual purpose: It not only lets us study what’s going on with the poll to find out what it’s saying about the public, but it also lets us analyze the poll itself to see if we should trust what it’s saying about the public.

The Vanderbilt poll makes it very difficult to try to figure out just what it’s saying about the basic demographic and political makeup of the voting public in Tennessee. Without that information, we have a difficult time evaluating the poll’s validity and its results.

I hope this doesn’t become a trend.

The Washington Post found that among Registered Voters, Tim Kaine and George Allen are tied at 46 in the Virginia Senate race.

Virginia Virtucon’s Riley thinks that’s a bit misleading, though.

I guess in theory the poll could be accurate, but when the poll is apparently projecting a sharp (more than one quarter) drop in the Republican share of the vote, driving it down below 1964 levels, I’m skeptical.

Whether it’s just an outlier, or a greater methodological problem, I can’t say. But I’m not about to call the race tied based on this poll.

For those who doubt or may have forgotten the difference between Registered Voter and Likely Voter polling for some polls, here’s a chart of every CNN/Opinion Research Generic Ballot poll from 2010, showing Republican lead or deficit per poll, with the RV and LV polls separated.

Clear difference, I’d say.

Rather than look at just one state, I thought it might be interesting to see what Swingometer has to say about a national poll, and as it turns out, the most recent national poll is the tracking poll from Rasmussen Reports.

This one is much better news for Mitt Romney than the North Carolina poll was.

The result is Romney 48, Obama 43, which turns out to be a two-party split of Romney 52.7-Obama 47.3 (R+5.4). The actual 2008 popular vote was McCain 45.7, Obama 52.9 (D+7.2). Together, that’s a swing of R+12.6, a large change from 2008.

Asking Swingometer, we find that a swing that large results in a clear win for Mitt Romney. He picks up Colorado, Florida, Indiana, Iowa, Minnesota, New Hampshire, North Carolina, Ohio, Pennsylvania, Virginia, as well as NE-2 and ME-2.

It’s terribly early of course. It’ll be a while before we get other pollsters trying a likely voter model. But while we wait, what else do we have to look at?

North Carolina was President Obama’s narrowest win in 2008. I’ve long thought that the state would be the quickest, easiest pickup for Republicans in 2012. As the final matchup between Barack Obama and Mitt Romney shapes up, early polling begins to confirm that guess.

500 likely voters, MoE 4.5. Rasmussen Reports finds that North Carolina is still pretty close, with a result of Romney 46, Obama 44. Still though, that’s a swing of a few points right off the bat. In a two-party vote, that’s a swing from 50-50 to 51-49, or R+2.

By itself, that swing is not enough to change the final vote though. According to the Swingometer, 2 points only flips Indiana, North Carolina, and Nebraska’s 2nd district. That still results in a 332-206 win for the President.

Mitt Romney needs a larger nationwide swing, or at least a larger swing in close Obama states, in order to be the next President. But, it’s early yet. Republicans will take some time to unify I suspect, and when that happens, I expect North Carolina to be out of reach for the President.