I think there's so much bad polling and it drives so much of the coverage that it's kind of a depressing topic for me. I think that the ABC News/Washington Post poll and the Wall Street Journal/NBC News poll are the gold standard right now of what we have.

It's not too shocking to see a media guy defend media-funded polls as the "gold standard". But are they really?

There are two ways people gauge a poll's accuracy. The first is to simply take the margin of the poll's final results and compare them to the margin of the final results. So it a poll says 48-41, and the results are 53-46, the poll was supposedly "spot-on" because they both had margins of seven. I think that's a stupid way to gauge accuracy. A poll could similarly guess the results as 12-5, and be equally "accurate".

I think the best way to gauge a poll's accuracy is to take the totals for each candidate, calculate how far off the poll was, and then add them all up together. So in the example above with the poll saying 48-41 and the final results being 53-46, the results would be 10 points off (53-48 and 46-41).

With that standard in mind, here are the current accuracy rankings, based on the current national popular vote spread of 52.7-45.9, and each polls final result:

So ABC/Wapo and NBC/WSJ weren't among the most accurate. Although to be honest, polling was pretty darn good this cycle. None of these results are really that far off the mark. I'll also assume Halperin isn't talking about state-level polling, because NBC used Mason-Dixon this year, which uncharacteristically stunk up the joint, while the Wall Street Journal skipped state-level polling this year after having funded those horrible Zogby internet polls in 2006.

So what else is there? It's true that ABC/WaPo and NBC/WSJ have the money to fund elaborate polls with extensive crosstabs. But so does Pew, and its final numbers were second-best. In fact, it's hard to argue that given the breath of results, that Pew isn't the gold standard. CBS News and Democracy Corps polls also had extensive crosstabs available, all of which helped paint a more complete picture of the electorate and its leanings.

So what the hell is Halperin talking about? Well, he worked at ABC News before his current gig at Time, and his conservative buddies all work at the Wall Street Journal. Throw in a dash of the media trying to keep itself relevant by hyping its own products above all others, and that probably covers it. Because on the merits, if we're going to start talking "gold standard", and if we're going to try and define that by combining accuracy with breadth of results, then Pew would most certainly win. (And Rasmussen wouldn't be far behind.)

As to "bad polls driving too much of the coverage", there was one major culprit -- Zogby. And who was pushing Zogby's polling when it pretended to show a close race? Drudge. The guy that rules Halperin's world.

Meanwhile, John Heilemann follows up Halperin with a more substantive comment (and one clearly rued by Halperin) -- that Nate Silver kicked ass this year. Media companies may want to pretend that their products are the "gold standard", but it took a web-creation like Nate Silver to make sense of all that data and provide readers (and traditional media reporters) with numbers-based analysis of what was important, as well as offering a master clinic in how to read and analyze polls.

Halperin has long depended on his BFF Drudge to help steer the national conversation toward his beloved Republican buddies. So while he likely applauded Drudge's hyping of one-day Zogby samples (out of a three-day rolling average tracker) to create the impression that McCain was still competitive, a lot fewer people cared. All one had to do was head on over to FiveThirtyEight.com to see that, in fact, the numbers continued to paint a grim picture for McCain.