Boris Johnson's pipe dream of British unity has been CRUSHED this morning by a stunning new Sunday Times poll showing support for Scottish independence has SOARED by 2% over the space of just TWENTY-FOUR HOURS

OK, so the headline is a tribute to the hysterical treatment the Express website gave to the statistically insignificant 1% decrease for Yes in the Scot Goes Pop / Panelbase poll on Friday night. (They published no fewer than two articles on the same subject, and Unionist Twitter went nuts over it.)  But there's a serious lesson here about the way mainstream media journalists treat their readers with utter contempt on polling matters.  Having cynically misrepresented the significance of the margin of error noise that flattered No in the Scot Goes Pop poll, will they do the same thing today now that a fresh Panelbase poll, this time for the Sunday Times, is showing what may well just be margin of error noise that flatters Yes, but that on the face of it amounts to a rather impressive 2% increase in support for independence?  Especially as it takes Yes to 49% - a statistical tie with No, and the best result for Yes in any Panelbase poll since April?  I suspect they won't.

Should Scotland be an independent country? (Sunday Times / Panelbase poll, 9th-12th November 2021)

Yes 49% (+2)

No 51% (-2)

(Note: The Sunday Times are reporting it as a 1% increase for Yes, and a 1% decrease for No, but that's because they're pretending that the only Panelbase polls that exist are the ones they commissioned themselves - which in fairness has been standard practice for all newspapers since the dawn of time.  However, even on the basis of the claimed 1% increase, I can't help but raise a smile at the mundane reference to the Yes vote merely "holding up", given the rather more excitable treatment the Express, Blair McDougall, "HappyBritScot", Pamela Nash, Murdo Fraser et al gave to a similarly trivial 1% decrease on Friday night.)

As I've mentioned before, the two Panelbase polls before Friday both had Yes on 48%.  So an entirely plausible interpretation is that Yes has been steady at around 48% for many months now, and that normal sampling variation caused a meaningless dip to 47% on Friday night, and an equally meaningless bump up to 49% this morning.  But there is also an alternative explanation. Whatever you may think of Nicola Sturgeon, no-one can doubt the excellence of her public relations skills, and she's had a particularly good couple of weeks at COP26.  So it's just conceivable there may have been a genuine Yes bounce off the back of the climate summit in Glasgow.

Unlike the indyref numbers, the Westminster voting intention numbers from the Sunday Times poll are almost identical to the Scot Goes Pop poll - with the exception of the fact that Labour have slipped back into third place (perhaps surprising given Boris Johnson's troubles and the trajectory of GB-wide polls).

Scottish voting intentions for the next UK general election (Sunday Times / Panelbase poll):

SNP 48% (-)

Conservatives 21% (-)

Labour 20% (-1)

Liberal Democrats 7% (-)

(Note: Once again, the Sunday Times are reporting completely different percentage changes, because they're pretending Friday's Panelbase poll doesn't exist.  Instead, they're using the much less recent Panelbase poll they commissioned themselves back in September as their baseline.)

Although the SNP haven't made further progress, this is one occasion when a no change poll may be of interest - because as I mentioned on Friday night, the 48% for the SNP in the Scot Goes Pop poll was the highest the party had recorded in any Panelbase poll for around one year.  The fact that the same figure has now been recorded twice in quick succession may increase the chances that there has been a genuine (albeit modest) uptick.

Seats projection (with changes from 2019 general election): SNP 53 (+5), Conservatives 3 (-3), Liberal Democrats 2 (-2), Labour 1 (-)

The above projection has been calculated by Professor John Curtice on behalf of the Sunday Times.  It presumably must be based on the existing boundaries, because the seats add up to 59 - as opposed to 57 on the proposed new boundaries.

There are also Holyrood numbers in the new poll - and for those the Sunday Times actually do list accurate percentage changes, because there aren't any Holyrood results in the Scot Goes Pop poll.

Scottish Parliament constituency ballot:

SNP 47% (+1)

Conservatives 20% (-2)

Labour 19% (+1)

Liberal Democrats 8 % (+1)

Greens 4% (-)

Scottish Parliament regional list ballot:

SNP 41% (+3)

Conservatives 21% (-2)

Labour 18% (-)

Greens 10% (+1)

Liberal Democrats 8% (-)

Others 3% (-1)

The 3% for 'others' includes Alba, who Panelbase have completely stopped offering as an option in their Holyrood polls - which to me seems a very odd decision given the clear evidence that Alba are still registering when respondents are given the chance to express a preference.

Seats projection (with changes from May 2021): SNP 64 (-), Conservatives 26 (-5), Labour 22 (-), Greens 10 (+2), Liberal Democrats 7 (+3)

PRO-INDEPENDENCE PARTIES: 74 seats (57.4%)

ANTI-INDEPENDENCE PARTIES: 55 seats (42.6%)

PRO-INDEPENDENCE MAJORITY OF 19 SEATS

It's a 1% increase but what it really means is 2% 🤣🤣🤣. It's really very sad.

— Areopagitica (@tweetmentcentre) November 14, 2021

Sorry, run that past me again? On what basis is it not a 2% increase? Take your time.

— James Kelly (@JamesKelly) November 14, 2021

Forgive me I was using 'standard practice since the dawn of time'

— Areopagitica (@tweetmentcentre) November 14, 2021

What I love about this is that you guys have been gloating about a 1% decrease in the poll on Friday night - but in order to maintain the fiction that today's poll doesn't show a 2% Yes increase, you have to suddenly pretend Friday's poll never even existed. Schrodinger's poll.

— James Kelly (@JamesKelly) November 14, 2021

I won't need to go fishing this afternoon at this rate

— Areopagitica (@tweetmentcentre) November 14, 2021

Probably better not. On this morning's form, it would be the fish catching you.

— James Kelly (@JamesKelly) November 14, 2021

The Yes vote has gone from 47% on Friday to 49% on Sunday. Does this logically mean there'll be a Yes majority by Tuesday, @murdo_fraser? I think it must do.

Tick-tock. Tick, and if I may say so, TOCK.

— James Kelly (@JamesKelly) November 14, 2021