Shock Horror - Katie Hopkins States Something Everyone Can Agree With ...

[andy-hughes]

“Pollsters have demonstrated - once again- why they are an utterly defunct source of information.”

You can't argue with that!

[jim360]

TL;DR: The polls aren't nearly as wrong as you, or Katie Hopkins, or anyone else in this thread makes out. You just keep picking up on the bad examples and ignore the rest. No wonder they are always wrong, if you only pick up on the wrong ones, and then exaggerate the mistakes of those.

[naomi24]

andy-hughes at 11:40. For goodness sake give it a rest. Your constant lip-dragging is becoming tedious.

Back to the subject, Jim isn't a pollster. He knows no more about polls than anyone else. He's simply giving his opinion.

[agchristie]

Poll talk is getting testy - could it soon be the final straw? o:)

[ludwig]

What Jim seems to be saying is that polls are perfectly ok, as long as you don't take very much notice of them.

That's basically what I meant by 'they're bollerks'.

[andy-hughes]

Ludwig - //What Jim seems to be saying is that polls are perfectly ok, as long as you don't take very much notice of them. //

No, that is not what jim is saying.

What he is saying is, you have to accept polls within the perameters in which they operate.

If - as happens with high profile polls like this one, pollsters get it wrong, then it is easy to assume that because this poll was wrong, then all polls are wrong, and a waste of time.

Jim's point is that polls are far more accurate than they are inaccurate - and focusing on individual examples offers a false picture of what polls actually do.

[jim360]

No, what I am saying is that, like literally any other numerical measurement obtained by scientific means (and it is here when I can say, with absolute confidence, that I know what I am talking about!), the central value is not the only thing you should care about. Of equal, or in fact greater, important, is the margin of error attached to that figure. In general, whenever a scientist (physics, biology, political science, whatever -- it's all the same thing) quotes something as being of a certain size, eg giving the number 48, what they really mean is something along the lines of "there is approximately a 70% chance that the measurement is between 44 and 51, and a 95% chance that it's between 41 and 55". Ideally, the range should be smaller, but it is always the case that this range exists, and it's therefore always the case that the central value isn't in itself important. Unfortunately, the central value is all anyone outside the field (ie the media, politicians and the wider public) seem to care about. So they say "Hllary Clinton is going to get 47% of the vote", and are then shocked when she gets 45, even though this was well within the 70% range.

You don't need to be a polling expert to know this. Any basic course in statistics covers it for you. And yet, most people seem to lack this basic statistical knowledge, are horrified or thrilled when a prediction is "wrong", and call the entire field into question based on very superficial reasons.

The central values for polling were in the end wrong for Tuesdays election. However, the final prediction by Nate Silver (mentioned above) gave Clinton around a 70% chance of winning, and Trump a 30% chance. 30% events aren't all that rare, and were well-captured by the polling and modelling in the end.

SO, in short, Katie Hopkins is wrong as usual.

[CrapAtCryptics]

Katie Hopkins must be related to Captain Obvious

[jim360]

"Jim isn't a pollster. He knows no more about polls than anyone else. "

The first bit is right. The second bit is a non sequitur, though. I am sure there is plenty I still have to learn about polls and polling, but I have all the same done plenty of research, reading and studying into polls -- more so than the average person, I am fairly sure.

It is of course an opinion I am offering, and you are free to ignore it, but not even to read it is an incredibly arrogant position to take. This isn't the first time you seem determined to project on everyone the same level of ignorance of a topic as you have. And, if you aren't prepared to do your own research into polling, then you are in no position to dismiss so quickly the opinions of anyone who has.

[naomi24]

Jim, //It is of course an opinion I am offering, and you are free to ignore it, but not even to read it is an incredibly arrogant position to take.//

I haven’t dismissed your opinion. I’ve simply said it is too wordy. That’s not arrogance – it’s eye strain.

[anotheoldgit]

jim360

/// SO, in short, Katie Hopkins is wrong as
usual. ///

Although I don't agree that Katie is wrong as usual, I do much prefer your short posts.