Democrats: The polls are skewed!

Oh, my.

Democrats have a new message in the 2014 race for the Senate: Don’t trust the polls.

The party is stoking skepticism in the final stretch of the midterm campaign, providing a mirror image of conservative complaints in 2012 about “skewed” polls in the presidential race between President Obama and Republican Mitt Romney.

Democrats who do not want their party faithful to lose hope — particularly in a midterm election that will be largely decided on voter turnout — are taking aim at the pollsters, arguing that they are underestimating the party’s chances in November.

First off, let me give the Democrats all the response they really deserve to this: HAHAHAHAHAHAHAHAHA.

Ahem.

…but, speaking dispassionately, there are several things to consider about polling. First off, there is the minor problem that polling works, except when it doesn’t – and you don’t always know when it doesn’t. And that the pollster that got everything right last cycle is highly unlikely to get everything right this cycle, or even mostly right. And that there is absolutely no such thing as a hard and fast rule about polls in the polling business: there are merely rules about polls that have not yet made a smoking crater in the ground. And, lastly, of course: even at their absolute best there’s always a measurable, attainable chance that any given poll is simply wrong*.

None of which should reassure Democrats, of course. Because while there are no hard and fast rules about polls, there is at least one hard and fast rule about people who talk about polls: the ones that build their strategy around complaining about the methodology or the sample size or the question choices or the order of the questions are the ones who are really, really worried that they’re about to lose. Usually, for good reason.

:brightly, and in sing-song:

Not my prob-lem!

Via

#unskew! https://t.co/WSVy2G2Nwd

— Just Karl (@justkarl) October 13, 2014

(Image via Shutterstock)

Moe Lane (crosspost)

*In the roleplaying game world, we call this ‘rolling a critical fumble.’ Turns out that the odds of rolling a 1 on a 20-sided die is pretty much the same as the odds that a given poll got a result outside the margin of error (5%). Roll enough times, and it’ll probably happen.