7 ms·
If you look at the last few elections in the US, Britain, Aus etc you will find that the polls have been far removed from reality. Those who own polling compan
by bobbytit 5y ago
If you look at the last few elections in the US, Britain, Aus etc you will find that the polls have been far removed from reality.
Those who own polling companies use them as a tool of influence and propaganda, nothing more.
- jonnycomputer 5y agoDefine far-removed from reality. 538's polling averages on election day put Biden at 51.8% of the vote and Trump at 43.4%. The margins of error were roughly 3 points either way. Biden received 51.3% of the popular vote and Trump 46.9%. https://projects.fivethirtyeight.com/polls/president-general/national/ https://projects.fivethirtyeight.com/polls/president-general...
- bobbytit 5y agoRetrospective averages are used by guys like Nate to memory hole the fact that all the big name polls had Biden ahead by 10 points. It is only the big name polls that are quoted by the media during the actual campaign.
- jonnycomputer 5y agoIt wasn't retrospective. It was a prediction, available on the day of the election, and before. Basically, you are talking nonsense, rationalizing your priors.
- bobbytit 5y agoIt is retrospective now. It is you who is talking nonsense.
- jonnycomputer 5y agoOh come on. Silver's polling averages, and his election predictions, were up a year before the election, and are still up for everyone to see. "Retrospective now" doesn't even make sense. Keeping up his model and his averages on the web isn't "memory holing" anything, it's being radically honest. You can still see his 2016 predictions too!
- bobbytit 5y agoSettle down. You are getting caught up on semantics. My sole point is that the average makes the appalling performance of the pollsters in recent years look not quite as bad. I wish I could use the good performance of others to make me look better too!
- jonnycomputer 5y agoYeah, he won by 5% not 10% points. But if you look at what happened, is that they were pretty accurate about Biden's percent, and underestimated Trump's (e.g. undecideds or shy voters broke for Trump). And again, if your margin of error is 3 points, then a 55-45 could be a 52-48, a 10 pt difference shrinks to a 4 pt difference. In any case, even if the polls were systematically off, and they may have been, that doesn't justify calling them all propaganda, or bs. It isn't even evidence of it. Because, as it happens, polling people is effing hard.
- bobbytit 5y agoThe fact that ALL the 'major' polls were heavily biased towards Biden, and Clinton before him, is clear evidence of corruption. They are not all bullshit. In 2016 the LA Times/usc poll was honest and accurate, so it was shut down for the 2020 election.
- jonnycomputer 5y agoNo, it is not "clear evidence of corruption". Not even close. It's evidence of being wrong. And no, not all the major polls overly favored Biden. At this point, you're just making crap up. Look, there clearly are some "bullshit polls", and there are partisan polls paid for by candidates which tend to be favorable to them, and there a good polls with consistent bias, a kind of "house" effect as a consequence of their sampling methodology, etc. That's why Nate Silver does averages of polls, btw, and weights polls in the average by their accuracy in previous elections.
- bobbytit 5y agoPlease name one single major poll that did not 'overly favour' Biden? I am happy to engage but please do not accuse me of lying.