Political Polling Thread (1 Viewer)

I think this is pretty fascinating. I follow a couple of guys who are in the business, and able to analyze polls by looking at their cross tabs. Apparently, a poll of young people just released shows Biden only leading Trump by 4 points. But the raw results weren’t that, they collect raw results, evaluate their sample, and then “weight” it to correct for sampling errors. (I think I have this more or less correct - happy to be corrected if someone knows better). Evidently this poll released the unweighted results accidentally and then released the weighted results later. Pollers won’t always provide that information, so it is nice to get that.

Another person raising questions about this poll:

I think pollsters way overcorrected their weighting after the 2016 election, even though the polls for that election weren't that far off.

I also think they've lost the capability to gathering a really good diverse sampling from the public, which further impacts their results.

This is all just my opinion, as I have no expertise in polling. It's just what appears to be happening on the surface, but something the pollsters themselves don't want to discuss.

coldseat said:

I think pollsters way overcorrected their weighting after the 2016 election, even though the polls for that election weren't that far off.

I also think they've lost the capability to gathering a really good diverse sampling from the public, which further impacts their results.

This is all just my opinion, as I have no expertise in polling. It's just what appears to be happening on the surface, but something the pollsters themselves don't want to discuss.

Click to expand...

I agree. But the fact that they are obviously off, and have been off since 2016 should be causing them to change their methods. Do some self-scouting, to put it in sports vernacular. They like to think they have some science behind their methods, but it’s hard to see that when they are consistently off in one direction and nothing seems to change. If they were off randomly you could make the case that polls have just lost accuracy. But to be off consistently in one direction is crazy.

Polling is now become more like "betting prognosticators" - if you are right 51% of the time, you are "believable" because your are 2 percentage points above being a loser.

lol

It's funny, back 15 years ago Nate Silver was able to use polling to predict election results quite accurately, but since then they've been less and less reliable, to the point of being almost worthless when compared with actual election outcomes.

This is absolutely a trend at this point. This is from the MI primary tonight.

MT15 said:

Click to expand...

I don't have TwiX, so it's difficult to follow this guy's point. Is he saying the 'red wave' warnings ignored or relied on polls? And where on Earth did anyone get the idea that Murray was vulnerable in WA?

Taurus said:

I don't have TwiX, so it's difficult to follow this guy's point. Is he saying the 'red wave' warnings ignored or relied on polls? And where on Earth did anyone get the idea that Murray was vulnerable in WA?

Click to expand...

He is saying that there were polls that led to all those talking points he listed. There are people trying to go back now and say that the people who were forecasting a “red wave” were ignoring the polls, because the polls were essentially fine. He is saying that is revisionist history. The final average of polls called for Rs to pick up 30-50 seats in the House. The polls were seriously flawed. The final RCP aggregate listed Murray’s race as a toss-up.

Once again. Polls are off.

The trend continues. Polls seem to be over-representing Trump voters.

Two things:

1 - All those Haley voters who say they won't vote for Trump are full of shirt and when the time comes they'll "hold their noses" and pull the lever.
2 - We already know that his post-election reaction does not depend on whether or not he wins the election. He won't be the incumbent this time, so he won't have the levers of power that the sitting president has, but Jan. 6th was mostly just random white dudes with white Fu Manchus and Costas so yea.

MT15 said:

The trend continues. Polls seem to be over-representing Trump voters.

Click to expand...

If you look at it, the places where the polls were off by the most were open primaries. Just comparing the turnout numbers to past primaries, it seems obvious that there were quite a few Democrats in Vermont who voted for Haley. That is something that primary polls can't really account for since they are polling people registered to the relevant party. If you just looked at the Republican vote in Vermont, i suspect it wasn't too far off from what the polls showed.

Some of the states barely had any polls. There was almost no polling in Vermont. Less than 1000 people were polled over the past year.

Also, the margin of error for most polls is around +-3%, so the polls don't have to be exactly right to be right.

samiam5211 said:

Also, the margin of error for most polls is around +-3%, so the polls don't have to be exactly right to be right.

Click to expand...

That’s the whole point of this thread. The polls are consistently off by more than the margin of error. And have been for years.

samiam5211 said:

Just comparing the turnout numbers to past primaries, it seems obvious that there were quite a few Democrats in Vermont who voted for Haley.

Click to expand...

Exit polling in SC showed that Dems voting in the GOP primary were not a significant number. IIRC. There are some, but they seem to be not a huge number.

MT15 said:

That’s the whole point of this thread. The polls are consistently off by more than the margin of error. And have been for years.

Click to expand...

But they aren't really.

Even in the congressional results from 2022, most of them were within the margin of error, it just happened that the elections were decided by less than the margin of error.

I'm talking about the averages of the polls more than individual polls.

Also, 538 is terrible, it isn't really polling, it is Nate Silver guessing.

The RCP average of polls is a pretty good barometer though.

Talking about last nights primaries, there were many states were almost no polling was done, so the one or two polls that were taken of just a couple of hundred people were off, but no one should be looking at those polls and trying to make predictions.