Winners and losers in the polling world | Arkansas Blog

Winners and losers in the polling world



In Nate we trust.

Nate Silver, the New York Times' election guru, was a big winner last night after a heaping dose of vilification not only from Republicans but from competing mainstream media for his statistical analysis that said there was a high likelihood of a presidential victory by Barack Obama.

Republicans hate facts. Climate facts. Budget facts. Polling facts. Their faith is so great that they are blind to facts and presume others — even those who make their living by the reliability of their fact-based predictions — must be corrupt to voice opposition. Obama's victory and small Democratic gains in the U.S. Senate and House, being facts, won't shake their faith still.

Silver's election morning final prediction, posted here yesterday, called it Obama 315 electoral votes and Romney 223 (rounded). It's currently 303-206, with Florida's 29 still in play in a very tight election (also accurately predicted as very tight by Silver). Apologies from Politico and others may or may not be forthcoming.

May we also take a moment for a horse laugh about Fox News' repugnant Dick Morris, so reliably wrong that you can reliably go opposite his predictions and stand a good chance of being right.

Something else was afoot in Arkansas. I'd be remiss if I didn't mention how badly Republican pollsters and pundits missed the mark on predictions — and how effectively they influenced the rest of the pundit and media class in the runup to the election that a runaway Republican takeover of the legislature was in the offing. For example, one Republic blogger dutifully posted this about the predictions from Keith Emis and Ted Thomas and their Republican consulting/polling group:

GOP pollster Diamond State predicts @ARGOP will take 62 to 64 House seats and 24 to 25 Senate seats

It's look like 51 House seats and 21 Senate seats and that margin included some real squeakers. 20-25 percent exaggeration on the House outcome. 15-20 percent exaggeration on the Senate outcome. Was this done to spin the media and help create a self-fulfilling prophecy? Or was it just bad polling and analysis? Emis insists their polling and prognostication were excellent, no matter what the numbers directly quoted say.

Another Republican political strategist who does his work on taxpayer money in the secretary of state's thoroughly political operation had predicted 63 House and 25 Senate seats. But just ask him. He wasn't wrong either, see. Catch his alibi:

Mostly...think it was a general underestimate of black turnout...nevertheless the Senate is close to my prediction

Get it? Black votes are trick votes and shouldn't be factored into election predictions. A nearly 20 percent miss on a Senate prediction is "close."

A Republican blogger said a "big tidal wave nigh" was "highly likely" — meaning as many as 70 House seats and 25 in the Senate. Uh huh.

Democratic side? The Republican pollsters laughed at Little Rock polltaker Ernie Oakleaf's finding that Herb Rule would carry Pulaski County. He did. Incumbent Republican Rep. Tim Griffin got only 44 percent of the vote in his home county.

House Speaker-designate Darrin Williams, whose historic future leadership as the first black speaker is now probably on the scrap heap of history, had predicted to me a 51-seat Democratic majority in the House. He was two critical votes off. The L.J. Bryant race still being contested, Barbara Graves' run against Allen Kerr and a race in Arkadelphia all figured as potentially part of his narrow winning formula, but all appear to have been lost.

Democratic leaders in the Senate yesterday morning predicted 20 to 21 seats for Republicans. It looks like it will be 21.

I confess that the lack of enthusiasm on the part of Democrats in their predictions, along with the certainty of Republicans who claimed scientific polling support for their predictions, led me to expect a lot worse than what transpired. But this is all cold comfort. There are no moral victories in politics, really. In the Arkansas legislature, a one-vote margin is as good as a 10-vote margin.

Comments (23)

Showing 1-23 of 23

Add a comment

Add a comment