On "The only poll that counts is on Election Day" and other Self-Deceptions

Apropos of nothing, I’m going to quote something Ace wrote a while back on polls. Anyone who’s familiar with my RS history would know that I’m not writing this to be discouraging, but instead to avoid us repeating the mistake we made in 2006.

Simple; we cannot afford to continue ignoring/dismissing polls with results we don’t like with the same set of excuses. Polls as a whole are useful indicators for letting us know whether we’re headed in the right direction and whether we need to change course or maintain our heading.

Remember that George W. Bush promised us in 2000 that as President, unlike Bill Clinton, he would ignore the polls. And as has always been his wont, he kept his promise. And how. Personally, I thought he meant polls when it came to decision-making, not the thorough refusal to even care about the opinions of his countrymen when it came to their President and his policies and try to bring them around.

We need to stop trying to console ourselves by dismissing polls we don’t like with the same old excuses. Knowing where we are helps us get to where we’re trying to go – we should know if we’re five points behind, not pretend we’re even. Ace laid them (the excuses) all out nicely …

  • Sampling favored Democrats.
  • Polls always favor Democrats, especially weekend polls, because Republicans are out of the house doing stuff.
  • There’s a lot of time left/debates can change everything/a week is a year in an election season/most undecideds or soft leaners only make up their minds 72 hours before the election, or in the voting booth itself.
  • Polls have been wrong before. Dewey defeats Truman. Kerry’s six-hour presidency.
  • The margin of error applies to both candidate’s levels of support. A +/- 3 point margin of error means that either could be plus or minus three points, meaning the whole poll can be off six net points.
  • Every poll will be outside the margin of error 5% of the time.
  • Some other polls show [the Republican] doing better/[the Republican] has internals showing him doing much better.
  • … state polls show some reason for optimism.
  • The Bradley Effect.
  • [The Democrats’] [voters] are unreliable voters.
  • [The Democrat] is effectively “the incumbent” in this election and most undecideds will break 70-30 in favor of [the Republican].
  • The only poll that counts is on [Election Day]

The last one particularly sets my teeth on edge. It’s time to win. Better we see clearly rather than through rose tinted glasses.

My 2 cents.