Wednesday, April 16, 2008

Visitors to the Delphic Oracle

My view is unchanged that public opinion polls are only marginally better at predicting the future than reading goat entrails. But people study them, obsess over them, and imbue them with magical predictive powers. The latest ABC/Washington Post poll offers an excellent opportunity to watch how people react after a visit to the Oracle.

Daily Kos (418 posts as I type) and TalkLeft (198 posts) are all over this poll like stink on a skunk. TalkLeft is notoriously pro-Hillary with DailyKos equally pro-Barack. The poll give Obama a big margin over Clinton.

The Kossacks have been singing praises at the Oracle of ABC and generally partying like anyone who had been told that his future will be bright and cheery - which is to say as boring as a jolly drunk. The TalkLeftists reaction is far more interesting as they are like a king who has been told he will lose the upcoming war, his capital will be sacked, and his daughters raped. The reactions fall along a predictable spectrum.

The Oracle Rocks (typical DailyKos reaction)
Wow! This is huge.

The Oracle is a lying tramp
These polls have been NOTORIOUSLY wrong. And if they DO believe Obama is the nominee for the GE, it is ONLY because he and all of his minions (that includes MSM and DNC) have been SAYING he is!!!!

The Oracle wants us to lose
The relentless media attacks on Hillary and praise of Obama must be having an effect. Otherwise how could people so seriously misperceive both of them?

The Oracle just...just must be wrong
I am completely shocked so many people are clueless. Two-to-one believe Obama is the stronger GE candidate. Wow.

The Oracle is not an all-powerful seer
I distrust national polls, especially this far out from the GE, simply because they are national, and aren't sensitive enough, eg any big spikes or dips in key states are washed out.

By the by, this last one is right. Reading polling data can be a fun game but it should never be confused with reality.

1 comment:

PoliShifter said...

Polls are used to try to shape public opinion.

Pollsters try to take advantage of 'group think'.

So if a poll comes out that says one candidate is ahead they hope that it will be a self-fulfilling prophecy so that their next poll will only reinforce their findings.

The initial 'group think' on bitter-gate was that it would sink Obama in Penn.

One poll was cited with Hillary 20 points ahead as 'clearly' showing voters reacting to Bitter.

Another poll came out that should no change in the numbers but when the dates were looked at the poll was taken right in the middle of the bitter controversy so the news commentators dismissed it as 'not being reflective of true public sentiment'.

Who in the media and what poll predicted a year ago that Obama would still be a contender? The group think at the time was Hillary was the shoo-in candidate.

I don't think we will know until Penn votes. Even at that Hillary was already expected to win.

trying to handicap it then would be if Hillary wins by more than 8 points then bitter had an effect. If she wins by less than 8 points it had no effect. If the margin is even narrower or Obama wins then we could say bitter backfired.

Polls are used by pundits and politicians to justify their position. They are not used to really get at an objective view of reality as people are led to believe.

Most polls ask loaded questions as it is.