Polls can be poor predictors of political outcomes

[angrychair]

If 2016 should have taught us anything it should have taught us that. Yes, that election was heavily influenced by Russian interference and gerrymandering but that is precisely my point: polling data didn't reflect that and we were unprepared.

We cannot afford to draw on polling data and make poor assumptions that may or may not be true.

There has been a lot of talk about the Iowa caucus and polling but historically it has been a very poor indicator of outcomes.

# of times since 1972 the winner of Dem Iowa caucus became the Nominee (out of 10 due to not counting 2nd term attempts of Clinton and Obama with no primary opponent of consequence):
5

# of times since 1972 the winner of Dem Iowa caucus also won the GE (out of 10, not counting 2nd term attempts of Clinton and Obama with no primary opponent of consequence):
1

[SWBTATTReg]

1. Good points. Some tend to worship polls too much/or rely on them excessively...nt

[left-of-center2012]

2. Opinion may differ ...

Opinion may differ depending how one's candidate is doing in the polls.

[padfun]

3. The polls were pretty accurate in the last few elections.

Polls don't tell you who will win electoral votes. They show peoples votes. And the sum of the polls had Hillary ahead by 2.9% while she won by 2.1% and more undecideds went to Trump so that explains the .8 difference.

Polls are only accurate if they have a huge data set so you might see many outliers, but when counting all polls together, you get a fairly clear picture of where votes stand.

[angrychair]

7. Actually that is what polls, ultimately, do

As least they are supposed too.

What I'm seeing is a shocking rewrite of history from election night itself.

Leading up to election night, polls showed Clinton winning. Exit polls showed a lean toward Clinton.

While I'm by no means Nate Silver but I am pretty politically aware and educated. I was monitoring dozens of sources that night and the strong consensus was that Clinton had it in the bag. I worked my ass off all the way through election morning.

The consensus was wrong.

I don't want to wake up November 4th to find out they were wrong again because people trusted polls more than what people were actually saying and ignoring the significance of Russian intelligence influence peddling.

[LibFarmer]

4. Every season ... supporters of a candidate not fairing well in the polls

have similar posts like the OP. So very predictable.

[angrychair]

6. So is your response

I didn't mention candidates or people. You did. My OP was about polling and that Iowa is the beginning, not the end, of the primary season.

I knew that certain sensitivities would be touched and some would become defensive. I put the OP in this forum because of that.

The best result is that Democrats win the GE and hold the House and take the Senate.

While I support senator Harris, my goal is for a Democrat to win.

I guess the same cannot be said for everybody but as someone recently said "so very predictable"

[Drunken Irishman]

5. I'm not sure of your point.

1) National polls were pretty accurate in 2016. They had Hillary with about a 2+ point lead and she won by that amount. The state polling was generally accurate beyond three states that were under-polled and failed to pick up the total shift of support Trump recieved. Even still, the polls indicated a shift was happening as the a couple of polls in PA and MI even had Trump up. Regardless, much of the shift was within the MOE.

2) It does no one any good to go all the way back to 1972 to prove your point about Iowa. The primary game has changed so much with the internet that what happened in 1972 or even 1992 is pretty irrelevant. And even 1992 isn't a good example as Iowa was not in play since Tom Harkin was running. So, other candidates didn't contest it. Even still, since 1992, every winner of the Iowa Caucus has gone on to win the nomination:

1996: Bill Clinton
2000: Al Gore
2004: John Kerry
2008: Barack Obama
2012: Barack Obama
2016: Hillary Clinton

Extending it back to 72 skews the results. It's clear Iowa plays a massive role in the success of a candidate. Maybe it's different this go around but I'm hard pressed to think that Kerry or Obama win the nomination if they lose Iowa in their respective primaries.

3) I don't understand this point. Iowa is almost irreverent to it since there's only going to be one Democratic primary winner. Fact is, Barack Obama is the only Democrat, period, to win a general election as a non-incumbent in the last 25 or so years (Clinton before that).