YouGov poll is even better for Yes among definite voters

The YouGov datasets were released yesterday afternoon, and as in the last poll from the firm, they show that the No lead is 2% lower among definite voters than it is among the whole sample.

Should Scotland be an independent country? (Definite voters only, Don't Knows excluded)

Yes 44% (+4)
No 56% (-4)

That's another record high for Yes from the No-friendly firm YouGov.  The previous record was 43%, and if memory serves me right that was only reached once. 

If this was Ipsos-Mori, the figures for definite voters would of course be used as the headline numbers, on the basis that there is a strong correlation between the number of people who tell polling firms they are absolutely certain to vote, and the actual turnout in elections.  When you look at it that way, there is now - for the time being at least - a degree of convergence between YouGov and other pollsters, in spite of the notorious "Kellner Correction".  YouGov's turnout-filtered Yes figure of 44% is very similar to ICM's unrounded figure of 44.6% (which admittedly is based on a more complex form of turnout weighting).  Remarkably, all six pollsters are currently showing a Yes vote of at least 42%.

But is there any way the No side can convince themselves that the surge for Yes seen in the three polls over recent days isn't really happening?  Well, with a bit of effort you can construct a case.  Although ICM showed a 2% swing, Yes were starting from an unusually low base in the previous poll, so you could argue that the change was an artifact of the margin of error.  Panelbase have shown a return to an all-time high for Yes that was reached in a previous poll, so you could argue that the slight dip in the interim was also a margin of error effect, and that nothing has really changed.  However, that ignores the suspicion that the net effect of Panelbase's recent methodological changes is No-friendly, and that the new poll is in real terms better for Yes than anything that has gone before.

As for YouGov, although they're showing an outright record high for Yes, you could pray in aid the fact that they have recently made a mildly Yes-friendly methodological change (country of birth weighting), and that on a  pound-for-pound comparison the new poll is no better for Yes than the previous high of 42% in the poll for Channel 4 News in the spring.  The problem with that argument, though, is that there have been no fewer than three YouGov polls in the interim which have left the distinct impression that the 42% finding was an aberration.

So the argument that there has been no change is a bit convoluted, to say the least.  Professor Curtice said yesterday that we cannot rule out the possibility that there has been more movement to Yes.  I'd put it the other way round - the balance of probability is that there has been a swing to Yes, but because ICM's figures weren't quite as good as the others, and because of the uncertainties surrounding the methodological tweaks made by Panelbase and YouGov, we can't completely exclude the possibility of an unchanged position.  What we sure as hell CAN exclude the possibility of, though, is a post-debate boost for No.

Incidentally, someone asked on the previous thread if YouGov ask for party political voting intentions before referendum voting intentions.  The answer is yes, and they also seem to switch back and forth between asking for Westminster voting intention first, and going straight to Holyrood voting intention.  In theory, this is bad practice which could distort the referendum findings.  However, when Panelbase stopped asking for the party political answers first (after being criticised for it by Curtice), it didn't appear to make much difference.

*  *  *

REQUIRED SWINGS

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

Swing required for 2 out of 6 pollsters to show Yes ahead or level : 4.5%

Swing required for 5 out of 6 pollsters to show Yes ahead or level : 6.5%

*  *  *

SCOT GOES POP POLL OF POLLS

MEAN AVERAGE (excluding Don't Knows) :

Yes 43.6% (+0.6)
No 56.4% (-0.6)

MEAN AVERAGE (not excluding Don't Knows) :

Yes 37.8% (+0.5)
No 48.8% (-0.7)

MEDIAN AVERAGE (excluding Don't Knows) :

Yes 42.6% (+0.1)
No 57.4% (-0.1)

(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.)