Good, so you admit they hire external polling companies.
Proving you post 1 nonsense. I'm happy to answer your remaining questions, but in terms of post 1 itself you really just conceded that there is nothing unusual in a campaign chair discussing poll design.
Now, 2 questions:
1. Why would they oversample polls??
2. How can a poll be fair if an oversampled poll shows a clear lead for Hillary, while a poll thats not oversampled (like IBD, LATimes...etc) shows Trump either tied or with a slight lead??
They don't oversample on purpose so your first question is silly. They all try and design a poll that is going to be accurate and fair. They don't all succeed.
Take the LA Times as an example. The LA Times did not intend to create a poll that skewed Republican. They just got unlucky in their sample.
Tracking polls like the LA Times pick their participants once a year and keep calling those same individuals all year long. The idea is to more accurately measure changes in opinion. If those same people shift towards Clinton you know it's an actual shift and not just random sample error.
However the people in the LA Times survey group have included more Republicans than the national average. It looks like they just got unlucky with their sample set. They have a well designed poll but they got a shitty sample.
They are particularly skewed with black voters. Their methodology adjusted the weight of each answer to try and match the national demographic. There are fewer young blacks in their sample than the national average so they multiply the vote of the blacks they do have. They also got fewer younger voters than they should do they overweight the young people in their sample as well to correct that.
Unfortunately the LA Times survey group includes only one young black male from Illinois, and he is sure that he is a Republican. His answers are given many times the weight of other people's because he is the only Illinois black respondent and also of one of only a few young blacks in the LA Times survey. And they keep calling him. And he keeps saying he loves Trump.
His answer alone skews their poll result towards the Republicans by about a percent. He isn't really representative because most blacks are Democrats and he's a Republican. But in the LA Times poll he's representing young blacks and Illinois blacks.
The LA Times didn't do that on purpose. They just got unlucky. And now they are either stuck with him for the rest of the year, or they interfere unscientifically in their methodology and remove him. They have chosen to stick with their methodology.
Does that make their poll useless? No. They still do s great job in capturing shifts in opinion. While their poll returns an absolute number that is too far to the Republican side they are better than other polls at showing you which direction sentiment is moving.
So that's a detailed explanation of why the LA Times poll is skewed. It's not rigged. The bias wasn't intentional, but designing a poll is hard and sometimes despite best intentions you get a skewed result.
