Scotland stands on the brink of independence as Yes vote surges to 47% in historic poll from traditionally No-friendly firm YouGov

Well, it looks like we can stop wondering whether Alex Salmond's victory in the second TV debate had an impact on public opinion. Here are tonight's sensational findings from YouGov -

Should Scotland be an independent country?

Yes 47% (+4)
No 53% (-4)

And when Don't Knows are taken into account, it's...

Yes 42% (+4)
No 48% (-3)

Let's try to put this into some kind of perspective - although given the scale of what appears to be happening, it's difficult to know where to start. Throughout this campaign, YouGov have consistently vied with Ipsos-Mori for the title of the most No-friendly pollster. At one point, just over a year ago, they showed figures of Yes 33%, No 67% - an almost unique example of the London media's beloved "2-1 margin for No" narrative actually becoming reality. Immediately after the SNP's victory in 2011, the President of YouGov, Peter Kellner, made the daft claim that it was literally impossible for Yes to win a referendum that was still over three years away, and since then it's been murderously difficult to escape the impression that he's been (perhaps subconsciously) moulding his firm's methodology to bolster that ludicrous prediction. And, yes, the No-friendly methodology remains in place in tonight's poll - we already have confirmation that there has been no change in methods since the last YouGov poll. So if Yes are on 47% after the so-called Kellner Correction was applied to the figures, the mind boggles as to what the results would be without it. Based on the traditional differential between the various firms, this poll is perfectly consistent with a narrow Yes lead among the more Yes-friendly pollsters such as Survation and Panelbase.

But that last point offers a clue as to why we should still exercise a little caution, because of course the only other published poll to be conducted since the second debate was from Survation, and although it was very good for Yes (an equalling of their all-time best showing) it didn't put them in an outright lead. So it's always possible that the YouGov poll might prove to be an outlier, and that the swing to Yes isn't quite as big as it appears. However, it's very hard to believe that there hasn't been a swing to Yes of some description, and from the comments that have been attributed to Peter Kellner tonight, it appears that even he agrees with that. He claims to have checked that the movement is real and concluded that it is. Doubtless he'll be careful to base that analysis only on the published results, but to my mind it's hugely significant that we know there was another YouGov referendum poll in the field over the last few days - perhaps a Better Together internal poll, perhaps a poll for an academic study. So Kellner will have lots of information to go on in private, and if none of that is causing him to doubt tonight's swing, the conclusion to draw is obvious.

Let's not forget that there was also a poll in the field over recent days conducted by Panelbase, which appeared to be an internal poll for Yes Scotland. I initially assumed that the reason it wasn't passed to the Sunday Herald for publication was that it didn't show any further progress for Yes, but the events of tonight at least raise the possibility that it was actually withheld because the Yes campaign were keen to maintain their underdog status, which appears to be working so well for them. However, that will have to remain an open question for now.

Tonight's timing, of course, could hardly be any better, as this is the first YouGov poll that is partly an exit poll - a significant minority of respondents will have been telling the firm how they have already voted, rather than merely how they plan to vote. So at worst, if YouGov's No-friendly methodology proves to be accurate, it appears that we no longer have to worry that the fact that voting is already underway might put Yes at a significant disadvantage, and leave them needing well over 50.1% of the votes cast on polling day. At best, if YouGov's methodology is painting too pessimistic a picture for Yes, it could be that it's actually a small advantage for us that the referendum has started.

If this poll doesn't prove to be an outlier, probably the one lingering concern might be that we're merely seeing a temporary post-debate bounce. But my gut feeling (for what it's worth) is that Alex Salmond's performance was the equivalent of what we always say about North Sea Oil - it was just a fantastic bonus, not what we rely upon. The greatest amount of persuasion has been done on the ground.

To give a small amount of credit to YouGov, they're one of only two pollsters that have taken the eminently sensible step of introducing country of birth weighting, to correct for the fact that there always seems to be too many English-born people in their samples. But they're also one of only two pollsters (the other is TNS) that do not filter or weight their headline numbers by likelihood to vote. Quite often (but categorically not always) that practice slightly suppresses the reported Yes vote. We'll have to wait for the datasets to see if that has been the case in tonight's poll - last time around the No lead was 2% lower among definite voters than it was in the published numbers.

Finally, I should just give some comparison figures for people who don't fully appreciate just how No-friendly YouGov have always been. Before tonight, their highest ever figure for Yes after Don't Knows are excluded was 43% - and that was only reached in the last poll. That record has been smashed by an almost unbelievable 4%. The previous record for Yes with Don't Knows taken into account was 38%, which again was only reached in the last poll, and which has also been surpassed by 4%. And for three successive YouGov polls between mid-June and early August, Yes were only hovering at 39-40% after Don't Knows were excluded - they're now 7-8% higher than that.

* * *

REQUIRED SWINGS

Swing required for 1 out of 6 pollsters to show Yes ahead or level : 2.0%

Swing required for 3 out of 6 pollsters to show Yes ahead or level : 3.0%

Swing required for 5 out of 6 pollsters to show Yes ahead or level : 4.5%

Swing required for 6 out of 6 pollsters to show Yes ahead or level : 7.0%

* * *

SCOT GOES POP POLL OF POLLS

You won't be surprised to hear that on all counts, tonight's Poll of Polls update sees Yes moving to an all-time high. By far the biggest swing is on the median average, which comes about because YouGov have (albeit almost certainly only temporarily) moved to the Yes-friendly end of the scale.

MEAN AVERAGE (excluding Don't Knows) :

Yes 45.0% (+0.6)
No 55.0% (-0.6)

MEAN AVERAGE (not excluding Don't Knows) :

Yes 39.3% (+0.6)
No 48.0% (-0.5)

MEDIAN AVERAGE (excluding Don't Knows) :

Yes 45.7% (+2.0)
No 54.3% (-2.0)

(The Poll of Polls is based on a rolling average of the most recent poll from each of the pollsters that have been active in the referendum campaign since September 2013, and that adhere to British Polling Council rules. At present, there are six - YouGov, TNS-BMRB, Survation, Panelbase, Ipsos-Mori and ICM. Whenever a new poll is published, it replaces the last poll from the same company in the sample. Changes in the Poll of Polls are generally glacial in nature due to the fact that only a small portion of the sample is updated each time.)