I majored in Government and Politics at U of MD, so I do really get into the details of campaigns, polling, and debates. Not directly involved in any though. Just interested. So I think of what might happen.
And I supported McGovern in 1972 and he got wiped in that election, so maybe what do I know? LOL! Good thing I didn't make a career of it. But there ARE a few things about it that I am good at. Well, exremists usually DO lose big...
1. Polling questions - You can write polling questions to get about any response you want. Good pollsters work very hard to stay neutral, bad ones write the questions to get the replies they desire.
Consider 2 questions. One poll asks "Of the 3 leading candidates for Party X, which do you prefer"? The other asks "Of the 3 leading candidates, which would you go vote for"? Those seem about the same, but they aren't. One asks preferences and the other asks about the action of voting. Many people have preferences, but only half the people actually vote.
So pollsters can affect the results by asking questions various ways and by deciding who they poll.
An example. If you ask old people in assisted-living houses who they prefer, you get one result. If you ask college students, you get another. If you ask people who are at home in houses on workdays, you get yet a different result. Same for calling people. Landline telephone owners have different views than smart phone owners. Even landline phone owners can be different. Those who are awake at 8 am are different from those who are awake at midnight.
Even worse is when pollsters ask "leading questions". Like "Candidate X wants to weaken our military forces. Do you think that is good or bad"?
Well, of course no one wants that. A more neutral question would be "Should tax money be applied more to military spending or to economic development"? Even better is a list of things money should be spend on and the pollee chooses.
2. Voting experience - Polling people who routinely vote is different from polling people who normally don't. Both can have strong preferences, but if you don't vote, your preference doesn't matter. And oddly enough, excited groups of supporters who routinely haven't voted in the past will usually not vote when it matters. They just have "others things to do that day". Or they feel uncomfortable doing something new to them (voting).
3. Voting accessibility - It matters how easy it is to vote. I know that from personal experience. When I first moved here, the voting line in the local school was 4 hours long and I saw people just giving up and leaving. They had other things they needed to do. Now, my voting place has no line and I am in and out in 10 minutes. Same number of voters, but more places. That affects voting. When voting is made harder in some places, it affects who can afford the time.
4. Ease of voting - When I vote, no one questions my identity. I could prove who I am easily enough. But not everyone can. Even legitimate voters in some places get challenged (unlike me - older white male). When I was single and in college, I had to drive 60 miles home to vote. But I could and did. So I skipped a half day of classes. But what if you had to take 3 buses for an hour and walk a mile when you had kids to care for or a job that demanded you work all the voting hours? That is the deliberate arrangement in some places.
I could go on, but I hope you see my point. Polls have errors, supporters don't show up, voters have easier or harder times voting. No one knows the results until they are counted.
And pollsters have been famously wrong. Dewey Defeats Truman is most famous, but there are others. We young Baby-Boomers were assured that McGovern would beat Nixon in 1972, but we didn't go and vote like the pollsters thought we would (I did). In 2008, The consensus of 7 polls taken just before the NH primary had
Obama winning by 8.2 percent in the voting. Instead, Clinton beat Obama
by a 39 to 36 percent margin. So the polls were off by a staggering 11
percent. And THOSE were a simple one-on-one poll taken among "likely voters"! And the list goes on, and on, and on...
You just never know about polls. As Yogi Berra said, "It ain't over til it's over"...