The larger the sample, the smaller is the margin of error, because, the poll is supposed to be a microcosm, a sample, of the whole.
The largest sample is in fact the final poll, the vote.
Now, looking at it the other way, the larger sample is the most accurate, ie, the final poll, the vote, is the most accurate reflection of the whole.
If one were to take a sample of the whole, it should be all reason reflect the whole very closely, as long as the sample was a mixed random grab of some of the whole.
Why then, does these late samples not only NOT reflect the larger poll, the vote, but so skewed in one direction so as to overturn the vote, the larger sample.
how can the smaller samples, the still left to count votes, not be reflective of the whole, and shake out the same, rather than what they have been doing in CA over and over, literally NOT reflecting or representing the larger sample, but being way off, so as to be more than a few std deviations off the mean?
Right. The main vote was roughly 50/50 but somehow ALL the undercounted votes favor the Dem 80/20 or more. It is mathematically obvious what has to be happening here. I don’t think the Republicans care as long as their seat is safe and they are certainly not going to pitch in and help Trump help the people take Their country back.