If the sampling is wrong, your poll is wrong.Right, so you can’t actually explain what ‘sampling error’ means. Thought not. Let’s add that to the list of logical statistical terms you don’t understand (but like to use).
You didn’t ‘explain’ why the polls or the samples are wrong, expect by using a circular argument that they must be wrong because they didn’t agree with what you thought.
I think the results of a ‘poll of polls’ (or indeed several aggregate polls) is a teensy-weensy bit more credible.
You are flailing...really...you’re doing it again by trying to deflect this onto me. It’s not about who I think will win, my posts have always been about trying to stop your excruciating attempts at ‘analysis’. I note that whenever you’re called on a particular point (such as why polls on voting may not give the same as a poll on approval), rather than recognise the explanation you tend to simply deflect the discussion- ‘switch targets right’.
I’ve also made it clear about my ‘opinion’ of Trump. However I don’t need to make a prediction;* I’m happy to follow professionals who have a marked interest in getting it right. So I’m happy with 13/100. Or 1:8. Whichever you prefer.
If you stop büllshit attempts at statistical analysis I’ll stop calling you on it.
* which as someone who’s conducted surveys professionally I wouldn’t do unless I had the means to conduct a poll of my own.
If enough polls are sampled wrong, your aggregate is wrong and the outliers will be more pronounced(your seeing that now as we get closer to the day).
If your sampling, aggregate and your mean is also wrong. Then your professionals are wrong and set for another historic shellacking from an angry media.