Moore double standard on polls

Jay Reiff sent an email out to Richard Moore's supporter list today declaring that 'Bev Perdue got caught again saying one thing, but doing another.' He was referring to Perdue's New York City trip yesterday, and I don't disagree. But Reiff could be accused of doing the same thing he attacked Perdue for in the message.

In the same e-mail he touted yesterday's Civitas numbers showing Moore down by only 5 and used them to declare that 'with two months to go, this race is clearly still wide open.'

It was just two months ago that Reiff sent an email out to the list with a thinly veiled slap at the methodology Civitas uses for its polls:

Many public polls do not screen participants to make sure they are interviewing likely Democratic primary voters or that the total poll sample accurately reflects the demographic and geographic makeup of past elections...Many public polls start by lumping both Democrats and Republicans into one poll. By the time they get to Democrats (let alone likely primary voters) the sample size shrinks dramatically and increases the margin of error. It's not that the poll is wrong per se, just not very precise.

So what Civitas does is flawed, except when it shows Moore closer than everyone else, in which case its results show the race to be 'clearly wide open.' Just like how Richard Moore says the media shouldn't run the results of automated polls, except when one conducted comes out well for his campaign, in which case an email gets sent out about it to the supporters.

It's perfectly normal for campaigns to cherry pick what poll results they tout. It's less normal to issue broadsides throughout the race attacking various polling methodologies and companies. And those criticisms ring pretty hollow when the campaign touts polls conducted using methods it's slammed when the results happen to be better for Moore. It's saying one thing but doing another, and if Jay Reiff thinks it's hypocritical when Perdue does it, I wonder what it is when the Moore campaign does.