To be annoyed that immigration is the main issue for most people in EU debate?

[susanketty]

Whether you are 'in' or 'out', I'm getting tired of immigration being the main issue for people - sure, it will have an impact, but brexit is not going to solve any problems with immigration that people perceive, and immigration fears does not seem to be to me a good enough reason to vote out. And voting in is not going to necessarily lead to a rush of immigrants.

We are not in the Schengen area, we have border control, and EU immigrants make a net contribution to our economy. Brexiteers often say they would like a Norway-esque agreement, which seems to me like all the problems (i.e., free movement of people) with none of the advantages other than being in the trade area (which we are already).

I just think there is more to the debate than immigration and it seems to me like it's been pushed to the front of the agenda to push more people to vote out due to fear.

You are funny.

The latest figures in a poll is the most relevant one to the eventual outcome.

I would care about the rolling average because any one survey of a couple of thousand people, out of a voter population of, say 40 million, will not be very reliable

Statistically a survey of a couple of thousand people has a margin of error of 3% with a confidence level of 95%. What that means is that 19 out of 20 surveys will be within 3% of the correct result. The main problem the pollsters have is getting a truly random sample, hence the need for various adjustments to try to compensate for differences between the sample and the general population, and also for differences between what those polled say they will do and what they will actually do.

Moving averages are meaningless in polling terms. The methodologies used by the various pollsters are different so lumping them together in some way will not produce a better result.

The rolling average is in the 40's, and mostly has been for months

As we approach polling day pollsters will discount "don't knows". If we take them out of the equation the last 12 polls show the result as being somewhere between 60/40 for remain and 52/48 for leave. Eight of these polls suggest that remain will win.

To detect trends it is necessary to compare successive polls from individual pollsters. There is no point comparing an ICM poll with a YouGov poll and trying to spot a trend since the two companies use different methods. Overall the polls suggest that there is a small drift towards remain but the picture isn't very clear.

The one thing that is clear is that phone polls generally show a better position for remain than online polls. Which approach gives the best results is anybody's guess.

"Polls are not indicative any longer for a variety of reasons"

What kind of reasons?

"The 60/40 one if in fact there was one probably excludes don't know/ refuse to say/won't vote. The best I've seen is 55/45 remain, but it excluded these."

See for yourself:

The latest ORB poll puts Remain on 55 per cent and Leave trailing on 42 per cent, among people who definitely intend to vote. Amongst all voters, the Remain campaign now has a 20-point lead, with 58 per cent of voters saying they back the pro-EU campaign.

In March, the polling found that 60 per cent of Conservatives were intending to vote Leave compared to 34 per cent for Remain. Tuesday's poll finds that among Tory voters, 57 per cent now say they will vote to stay in the EU, compared to 40 per cent backing Leave.

"crazily small sample of 1000 (or whatever)"

Market research is not our strong suit, I see.

How many people do you believe are polled in any poll, on any subject? I suggest you look up a few.

I'm still waiting for you answer re why you believe that "polls are not indicative any longer for a variety of reasons".

So that's your "variety of reasons" why polls are not effective "any longer"? That mainstream polls weren't correct last year?

I thought you discovered a statistical error not really .

Your scepticism is perhaps understandable but, needless to say, one badly done poll doesn't mean all polls have stopped working, now and forever.