renton

Keep the amusing kickback Jpegs under wraps just now. No numbers yet.

Any tweets or leaks about either ICM/YouGov?

Yup, no surprises there.

It's their first indyRef poll as far as I can tell. wait, no ComRes is only fo rthe South of Scotland. Back down to two bludgeons for us to be hit with tomorrow.

That's a huge assumption though. It could be that a lot of folk who wouldn't have bothered have been galvanised by the very presence of the Yes campaign to step forward and save their beloved union.

Apparently a ComRes poll tonight at 6pm as well....

They don't exclude them per say, but they don't find enough of them, so you have small statistical samples that introduce large margins of error, which get multtiplied when your 50 odd 16-24s get's multiplied by more than a factor of 2 to match the Scottish demographic profile. Survation constantly have that problem with 16-24s as well, so you can see the fluctations in that group from poll to poll vs. the more stable (but still changing) older demographics.

No, different customer though, there is an ICM for the Scotsman a tthe weekend which I assume wil be online. This is for the Guardian, I imagine telephone polling is a bit more expensive and ICM ones have previously been treated a s abit gold standard in this regard. Also due YouGov tonight, which after their last two efforts probably will show a reversion to the mean, if not worse. We should see how much of Westminster's Sturm und drang has filtered through to the public, but even allowing for that, I'd still expect a reversion to the mean. Expect two bad polls for Yes tonight then chaps. Gonna likely be a rough ride.

So the next ICM poll, due tonight is a telephone rather than online poll. Means it'll be hard to get a like for like measurment to previous ICM polls. Secondly, it'll be interesting to see how it stacks up against the only other phone pollster: Ipsos-Mori, traditionally a No friendly pollster - is the 'no friendly' bias of Ipsos based on landline calling meaning they need to substantially upwieght small, statisiticall unrepresnetative samples from certain demographics, or is it something to do with othe rpost processing weighting calculations. my own thoughts would be to brace yourself for another no friendly poll result tonight.

The Sun is deifinitely leaning yes.......

Apparently 97% of the elegible electorate have registered to vote. Assume an 80% turnout on the day: Yes would need 1.7 million votes to win.

The itneresting thing is that Murdoch's tweet about Yes internal polling has Ye su by the same basic margin as No's internal polling - according to the un-named MP that Sky news were tweetnig about yesterday. Rumours and probable bullshit, but there you go.

You do realise that even a poll of polls containing guys like Ipsos that hasn't reported in ages is showing a statistical tie at this point, right?

Survation has a bit of a problem with not having enough 16-24s on their panel, and this needs hugely upweighted, it leads to volatility in their polling (see the recent 7 point siwng to No that no one else saw and subsequently corrected) - if it's true, and I'm not saying it is, to see a huge drop in 16-24s but with the headline staying the same would possibly mean a drift to Yes in other groups, allowing for the uncertainty in that age group due to it's small sub sample size (and therefore greater uncertainty of it's representative nature). Also, every other pollster doesn't show 16-24 as being that No friendly. Might be all bullshit mind, just going on comments I've seen elsewhere, not seen the datasets.

Sceptical as I've not seen the actual data sets but apparently there is another crazy fluctuation in Survation's 16 to 24 group. Last poll was 42/45 Y/N, this one is 32/50... which is subsequently upweighted by a facot of two as they can never seem to get enough 16-24s on their panel.

I'd guess that no one has a clue. Certainly Yes seems to do better in groups who are not regular voters

Yeah,this is the general assumption we've all been making for months, and there is some anecdotal evidence for it in the RIC canvasses,albeit those still showed a lot of DKs

We might.

Figured they'd be ahead in Aberdeen - or Yes might win the city, but lose the North East in general. Dundee no surprises there, Glasgow, north and south Lanarkshire are the real battlegrounds. Guessing that's probably an SNP MP, though. Which dilutes the news worthiness of it, a bit.

Usual jitters, can't be helped. I'm not close enough to the campaign to have an idea of how things look to those out talking to folk. There is anecdotal evidence of folk moving to Yes, but then there could as well be droves of new No voters out there we haven't accounted for in the polls.We are basically in the dark as to how the pollling situation will evolve (for all it's accuracy limitations) with another week of intense No heavy media pressure to come.

Well, yeah, the average is 49 +/-3%, all four recent polls are in statistical tie territory, but as dorlomin says, they won't do very well at picking up on the intentions of the lapsed voters. They could be anywhere. PB and YG have polls out soon, haven't heard of an ICM one coming out yet, though you'd think there would be.

I'm not, I thought, about a month ago or so we'd be going down 45/55, now I think we'dd do better but can see us losing with 48/49%. Bawhair, but enough to stop independence.

It could just be that the Survation and Panelbase polls are more impervious to swings to Yes due to their panels being 'nat heavy' according to Peter Kellner. Panelbase previously didn't match the gap closing seen by ICM and Survation in January 2013. Survation of course saw a pretty massive swing to No, followed by an according swing back to Yes before landing on a NC this time, due to serious upweighting in one of their age groups, fascinating that there is no change whatsoever to Yes or No based on either debate. YouGov, are possibly over sensitive to changes in Labour 2010 voters in particular going to Yes, and that could be a fault with their methodology, but it doesn't explain the corresponding shift in TNS, who have been pretty consistent with their methodology, and not at all volatile, like ICM and Survation have been. Taking the four polls together, you'd put Yes at 49% as an average, but it's hard to correlate the massive differences in trends seen by the pairs of pollsters though.

Need an ICM poll as a tiebreaker.

Well, the Survation field work largely predated the recent media shitstorm. So it's not that, I'd expect to see that backlash in the next YouGov poll if it happens.