For a glimpse into the reasons why the prospects for a revival of Republican fortunes anytime soon are dismal, have a look at John Podhoretz’s list of “10 reasons why McCain might win” (via Andrew Sullivan). For those readers who don’t have time to click “read more,” the executive summary is that every single item on Podhoretz’s list contains multiple errors of fact, logic, math, or inductive/probabilistic/statistical reasoning — and rather basic ones at that — all packed into one or two sentences apiece. Let’s go through them.

1) One poll has undecided voters at 14 percent on the last weekend, which means most of them probably really aren’t undecided, that they are either going to stay home or vote preponderantly for McCain and pull McCain across the finish line.

Today’s CBS poll finds 4% undecided, and Obama ahead of McCain 54-41. The ABC/WaPo poll finds 3% of registered voters with “no opinion” (as opposed to voting for “other” or “not voting”), and Obama ahead 53-42; among likely voters it’s 2% with no opinion, and Obama leads 53-44. According to IBD/TIPP, 8.7% are “not sure” and Obama leads those who are 47.9 to 43.4. Diageo/Hotline says 5% are undecided and Obama leads 51-44. Rasmussen leaves room for at most 3% undecided and estimates an Obama lead of 51-46. The Research 2000 people think there are a whopping 1% are undecided and that Obama leads 51-44. Zogby, helpful as ever, estimates the proportion of “others/not sure” at 6.8% — not all of whom (obviously) are not sure — and the overall spread at 49.1 to 44.1. According to Gallup’s three models, the proportions with no preference are 5%, 4%, and 4%, and the Obama leads are 52-41, 52-42, and 52-42.

We’re in an environment in which dozens of polls are released every day. Podhoretz doesn’t know that you can’t cherry pick a single poll to support a pet theory, or else he does know and still did it; either way, a competent editor wouldn’t let him write about polls again. Unfortunately, his editor is John Podhoretz.

Besides the fact that in each of the polls I cited, the proportion of undecided voters is considerably smaller than 14%, particularly astute readers will have noticed that in most of them, Barack Obama wins more than 50% of the vote. That means that even granting Podhoretz’s entirely evidence-free claim that the undecideds will vote “preponderantly” for McCain, there are probably not enough of them to pull him “across the finish line.”

2) Most pollsters are claiming the electorate this year is six to nine points more Democratic than it is Republican. That would be an unprecedented shift from four years ago, when the electorate was evenly divided, 37-37, Republican and Democratic, and a huge shift from two years ago, when it was 37-33 Democratic. A shift of this size didn’t even happen after Watergate.

We can narrow the trouble here down to three possibilities: Podhoretz doesn’t know what ‘unprecedented’ and ‘huge’ mean, he can’t perform basic arithmetic, or he doesn’t know what those words mean and he can’t perform basic arithmetic. (Option 4 is that he takes his readers — perhaps correctly — for idiots.) If Podhoretz’s numbers are correct, the Democrats enjoyed a 4 pt. net advantage in 2006, an improvement of 4 points over 2004. If their advantage this year is 6-9 points, then their net improvent over two years since 2006 is between 2 (that’s 6 minus 4) and 5 (that’s 9 minus 4) points. At the low end, it’s half the increase of the previous cycle; at the high end, it’s fractionally greater. In other words, Podhoretz is the mafia don in the classic Mr. Show sketch who shoots his underlings when they question whether 24 is the highest number.

Meanwhile, the explanation of a larger partisan shift towards the Democrats today than over the post-Watergate period (and again, what period are we talking about? a month following Nixon’s resignation? a year? four years?) is the profound demographic shifts that favor Democrats, just as demographic dynamics from the 60s to the 80s favored Republicans

3) Obama frequently outpolled his final result in primaries, which might have many causes but might also indicate that he has difficulty closing the sale.

Utter nonsense. “On average, Barack Obama overperformed the Pollster.com trendline by 3.3 points on election day.” Read the rest here.

Also, one of the causes of the McCain campaign’s strategic incompetence seems to be that they, like Podhoretz, gave serious credence to the obviously ridiculous idea that primary results are predictive of general election results.

4) The argument in the past two weeks has shifted, such that many undecided voters who are now paying attention are hearing about Obama’s redistributionist tendencies at exactly the right moment for McCain.

Jehovah might know what Podhoretz means by “the argument…has shifted,” because no one else does, including, I’d be willing to wager, Podhoretz. If he means that the polls have shifted in any meaningful way towards McCain, well…they obviously haven’t. If he means that the issue landscape is now more favorable for McCain…it isn’t. (Read the internals on the polls cited above.) If he means McCain has picked an electoral winner by attacking “Obama’s redistributionist tendencies”…in fact, McCain has picked a 58-37 electoral loser. If he means that McCain and Palin’s ludicrous demagogy, not to mention their consultations with Joe the Foreign and Domestic Policy Expert, have succeeded in moving their core partisans and almost no one else to adopt the party line on terms like ‘redistribution’ — though not, to be sure, on the substance of redistributive policy — well, then he’s still on to nothing that implies a McCain victory.

5) The tightening in several daily tracking polls indicates a modest surge on McCain’s part that could continue through the weekend until election day. If he is behind by three or four points right now, a slow and steady move upward could push him past the finish line in first place.

A) Here’s Pollster’s trend line. Here’s RCP’s. Here’s TPM’s. What, you expected Podhoretz to know the difference between an actual trend and statistical noise? B) McCain’s not behind by “three or four points,” therefore C) any move of his will need to be fast and jolting, especially considering the margins Obama has already banked among early voters.

6) In terms of the electoral map, the energy and focus McCain is directing at Pennsylvania could pay huge dividends if he pulls it off. If he prevails there, it might follow that the message will work in Ohio too. And if he wins Pennsylvania and Ohio, he will probably win even if he loses Virginia and Colorado.

Spending your life savings on the lottery, likewise, would pay huge dividends if you pulled off a win. It’s not as if the McCain campaign has any clearly better option than pouring resources into Pennsylvania, though it’s worth noting not just that Pennsylvania is an extremely long shot, but also that it’s a very expensive state to contest, and that every dollar spent there can’t be spent in states like Virginia, Colorado, or Ohio. Now, if there is some major national movement that puts McCain in contention in Pennsylvania, that should probably be enough to win him Ohio. If there isn’t any such national movement, even if the McCain campaign somehow induced an aberrant, trend-bucking result in Pennsylvania, it’s anybody’s guess exactly how that is supposed to influence the Ohio outcome. By osmosis?

The “energy and focus” McCain is directing at Pennsylvania allow Obama to open up a larger advantage in energy, focus, and money in Ohio than he would have otherwise. Let’s assume that Podhoretz already knows this. Well, then he doesn’t know how to add to 270. A universe in which McCain has won Pennsylvania and Ohio but Obama has won Virginia and Colorado is a universe in which Obama has almost surely won New Mexico and is a heavy favorite to win Nevada — which would put him over the top.

7) Early voting numbers are not oceanic by any means, which may indicate the degree of enthusiasm for Obama among new voters is not something new but something entirely of a par with past candidates, like John Kerry. And they show more strength on the Republican side than most people expected.

More raw, unsupported assertion. You’ll be shocked to learn that it is almost entirely false. Try a google news search for “early voting,” which Podhoretz evidently did not do. Here’s a St. Petersburg Times account, which, oddly enough, describes the early voting in Florida as “presag[ing] a tsunami.” Does that qualify as oceanic? As for the margins, today’s CBS poll, like just about every other poll that has measured early voting, finds Obama with a massive lead.

I think, in this case, I can piece together what happened. About a week ago, a RedState blogger badly misinterpreted some figures on still-ongoing early voting in California, a state I promise is not in contention, replicating earlier misinterpretations of figures on Democratic GOTV operations in Ohio. Podhoretz must have somehow internalized the idea that McCain was doing unexpectedly well in early voting nationally. The ordinary procedure, given a data set that seems to imply some remarkable finding, is to apply some sort of sanity test — see whether the data actually say what they appeared to at first glance, see whether they might be distorted in any way, etc. That procedure was not followed.

8 What happened with the Joe the Plumber story is that Obama has now been effectively outed as a liberal, not a moderate; and because liberalism is still less popular than conservatism, that’s not the best place for Obama to be.

What happened with the Joe the Plumber story is that almost nobody cared. As for the idea that “because liberalism is still less popular than conservatism, that’s not the best place for Obama to be,” in addition to being factually confused, this is a use/mention problem. It is certainly true that the label ‘liberal’ is somewhat less popular than the label ‘conservative.’ That has nothing whatsoever to do with substantive voter preferences for liberalism or conservatism — as all the studies of political beliefs in mass democracies since Philip Converse’s original have shown, almost no one has the slightest clue what the substantive tenets of liberal and conservative ideology are. The names ‘liberalism’ and ‘conservatism’ themselves are simply totemic markers of personal and tribal identity. On that score, the unpopularity of the Republicans dwarfs the unpopularity of the abstract notion of liberalism, and just as importantly, the salience of the unpopularity of the GOP dwarfs that of the unpopularity of liberalism. Joe the Plumber, I’m sorry to have to report, is unqualified to clean out figurative septic tanks like the Republican brand.

9) The fire lit under Obama’s young supporters in the winter was largely due to Iraq and his opposition to the war. The stunning decline in violence and the departure of Iraq from the front page has put out the fire, to the extent that, like the young woman who made a sexy video calling herself Obama Girl and then didn’t vote in the New York primary because she went to get a manicure, they might not want to stand on line on Tuesday.

I’ll be brief since I’m now repeating myself. This is pure fantasy, without a scintilla of evidence to support it, and heaps of evidence contradicting it.

10) Hispanic voters, who are always underpolled, know and appreciate McCain from his stance on immigration and will vote for him in larger numbers than anyone anticipates.

What does Podhoretz mean by “Hispanic voters…are always underpolled”? Does he mean there aren’t enough polls of Hispanic voters? I’m not sure how many would constitute enough, but there have been plenty of polls of Hispanic voters, and they have univocally shown Obama with staggering leads among Latinos. Does he mean that pollsters are undercounting the Latino share of the electorate? In that case, Obama’s overall lead is even greater than the current estimate in the polls. Does he mean that pollsters are somehow systematically undersampling McCain-supporting Latinos? That would have nothing to do with Latinos per se, but with a systematic skew of the polls in Obama’s favor. Podhoretz could have spared the effort of writing all those words and just written instead that he thinks the polls are wrong, but has no reason for thinking so apart from the cognitive constraints of dwelling inside an epistemic cocoon. Better still, he could have merely written, “It’s admittedly not the strongest case” and left it at that. Instead, well…you see the result.

Along the same lines as Podhoretz’s latest unintentional tarnishing of the Commentary brand, this RedState post by Dan Perrin predicting a McCain victory made the rounds a few days ago and gave us a view of another room inside the bunker. Daniel Larison has already said most of what needs to be said about it, but let me add one point Daniel missed. Perrin’s seventh and final “reason McCain-Palin are a lock to win” includes this putative datum:

According to the Associated Press, [a Bradley effect] will cost Obama six points at the polls. The AP estimate could be low.

Presumably, Perrin is referring to the notorious AP study from September. Never mind the many methodological issues with that study (chief among them: there is no accounting for points added to Obama’s margin due to any sort of reverse effect); let’s assume the report was on the money. What it claims is that the share of the Democratic vote Obama loses on account of race is already depressed by about 6 points; i.e., whatever racial effect there might be is already priced in to current polls. Which in turn means that Obama’s performance in polls is, by the lights of the AP study, an accurate representation of his electoral standing, at least vis-a-vis racial issues. A Bradley/Wilder effect, by contrast, is the underperformance of black candidates on election day relative to their strength in polls. That is precisely not what the AP report posits. In the header to the post, Erick Erickson identifies Perrin as a “Republican strategist.” I would hope no one has actually paid him money for strategic services.

Lastly, to get a full picture of the GOP cocoon, it’s worth revisiting Robert Stacy McCain’s remarkable post pronouncing excommunication on any would-be Republican who fails to fall in line behind Sarah Palin. It’s too bad McCain doesn’t realize how truly he spoke when he said that “I saw the Republican Party today, standing in line to see Palin at Shippensburg University”; if the Republican Party really is the Sarah Party, it will soon fit into a university gym.

What efforts like Podhoretz’s, Perrin’s, and McCain’s show is that, for a fairly representative cross-section of the Republican pundit class, the problem with what the party has been up to over the last decade is that they haven’t been doing it hard enough. Don’t worry, Republicans; with Sarah carrying the party standard in four years, the real truth about Obama and William Ayers is sure to get out, and then The One will be toast. Surge! Petraeus! White flag of surrender!

As Julian Sanchez says in the comments thread to McCain’s post, “So, the plan for 2012 is to pull 30% of the vote, but be REALLY REALLY excited about it? Good luck with that.”