Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search


Silent3's Journal
Silent3's Journal
January 22, 2023

A poll doesn't have to be an accurate predictor of electoral victory to provide useful information

DISCLAIMER: I personally think Biden has been a great President, I want him to run, and I think he can win. Anyone who would misinterpret this post as anti-Biden, or as support for any particular pollster, isn't as good at reading between the line as they might imagine themselves to be.

A poll doesn't have to be an accurate predictor of electoral victory to provide useful information

I'm really, REALLY tired of the knee-jerk reactions around here when people post polls.

"I don't believe in polls anymore!"
"That pollster is a right-wing hack!"
"They're just trying to manipulate us!"

You don't have to, however, "believe" that a poll is as good as a magic crystal ball to extract rough information from it.

You should understand that many polls are a form of advertising for marketing services -- it does a pollster no good to get elections wrong over and over and over again by baking in political bias. That tells the world that they're survey techniques are crap, and cuts the prices they can charge paying customers.

Even if some pollsters are merely political vanity projects, I've never heard of any solid studies that prove either the assumption that bad polls discourage turnout, that good polls cause complacency, or, if either effect is real, that the two effects don't come close to canceling each other out anyway. So people should stop acting as if either of those things are true unless they can back it up with more evidence than simply believing these supposed effects are "common sense".

What so many people aren't appreciating is that even if a poll isn't accurate enough to tell you who's likely to win a tight race, even a poll that's off by 10% gives you a rough idea of where the electorate stands.

Take the Harvard/Harris poll someone recently posted, which shows many more people favorably inclined toward DeSantis than Biden.

Of course that poll doesn't mean shit at this point for who will win is 2024. But the poll DOESN'T HAVE TO be supremely accurate or a good electoral predictor to tell you something that's very useful to know.

What it tells you is we have an enormous problem in this country because an overwhelming majority of voters haven't learned from the Trump experience that anyone even vaguely like Trump is dangerously unacceptable. Biden SHOULD beat Trump or DeSantis or any of these MAGA-ish monsters by so many points that the crappiest poll would show Biden way ahead.

Either too many Americans are so ignorant and out-of-touch as to not understand the important differences between Biden and these clowns, or they're nasty enough to actually understand what Trump and DeSantis are, and to want them to win because of that.

You can learn THAT from the Harvard/Harris poll. It's both sad and important to know it.

This post is a repost of any earlier thread. I got locked out of responding to my own thread, so I deleted that one to start again fresh. There was nothing about this OP itself that was cited.

Profile Information

Gender: Male
Hometown: New Hampshire
Home country: USA
Member since: Sun Oct 3, 2004, 04:16 PM
Number of posts: 15,187
Latest Discussions»Silent3's Journal