Riley Poll: Here's your roasting, Mike.

Kari Chisholm

On the heels of the release of the latest Riley Report by pollster Mike Riley -- discussed at length by Jeff Alworth here -- Riley appeared this morning on Outlook Portland with Nick Fish.

Asked to make a prediction near the end of the show, here's how he disclaimed:

You know, I'm going to be roasted on BlueOregon. Every time I make a prediction -- if it's not correct, I get completely roasted.

Well, Mike, it's not just that you're making incorrect predictions. None of us have a perfect crystal ball. As they say in sports, there's a reason you play the games.

No, Mike, the reason I'm roasting you here and now is that you're making predictions in the absence of any data. There are lots of armchair pundits out there (and I'll plead guilty to doing making data-less predictions every single day), but there's a higher standard for pollsters. Y'all are the people who are supposed to be data-driven, with a hard focus on the numbers.

What was the prediction Mike made without any data?

The Democratic Senate... I think it's still more likely to be Jeff Merkley. Steve Novick - there's still a lot of life left in him. There may be others that come up.

Now, I happen to agree. But I'm not the guy with a polling operation - offering up analysis that's supposedly based on hard numbers.

Of couse, Nick Fish didn't let him get away with it. He asked:

So, at this point, not having tested Steve Novick, we don't know whether he'd poll ahead of or behind Jeff Merkley, is that right?

To which Riley answered -- with the most honest response of the show:

We honestly don't. In fact, we don't know any more based on these results.

That pretty much says it all, doesn't it Mike? You did a poll, and managed to create confusion where you could have created clarity. Some crystal ball ya got there, buddy.

Oh, and Fish dug further, wanting to know why Riley didn't bother to ask about Novick. Riley's pathetic answer?

Well I had to make a judgement call there. I pay for these questions myself, and so I wanted to ask as few questions as possible. I thought to myself: What's the most likely scenario? With Jeff Merkley, with his prominence as Speaker of the House, he's going to have the greatest name recognition, probably the inside track for the seat to be determined in the primary next year. So I said, between Smith and Novick, or between Smith and Merkley, it's probably going to be Smith and Merkley.

And the best part? Check this out:

I also threw in John Frohnmayer, who is the brother of former Governor Frohnmayer, who was the head of the National Endowment for the Arts in the first Bush Administration. Because he's purported to be a likely candidate as well, with that kind of name familiarity I might as well toss him into the mix.

We're supposed to trust the predictions and punditry of a guy who thought Dave Frohnmayer was the Governor? Sorry, buddy, but Dave lost that race in 1990 to Barbara Roberts.

So, you didn't bother to test Steve Novick - who is in the race and running hard - but you did throw in an undeclared candidate who isn't campaigning? And you also didn't even bother to do a straight Smith/Merkley head-to-head? That's just ridiculous.

And then, to offer up predictions about the U.S. Senate primary, without having asked a single question about it, well that's just not cool. It is, frankly, absurd. It's giving pollsters a bad name.

Next time you go on TV, Mikey, decline to answer the questions you haven't tested. That's your responsibility to the profession.

It may be August, but sing along with me: Pollsters roasting over an open fire...

By Kari Chisholm Aug. 26, 2007