May Day! May Day! It's not May's Day, as Google survey reveals 57% of Scottish public prefer independence to Brexit

I'm not sure if the Record have abandoned Survation as their regular pollster, but for whatever reason they've commissioned Google to produce a Scottish political survey, which is apparently demographically representative (albeit on the basis of algorithm-derived 'inferences' rather than definite information).  However, it's not yet clear that it's been politically weighted in the way that would be standard for an online poll conducted by a BPC firm, so we should certainly be very cautious about the results.

What stands out is a question asking whether people would prefer independence within the EU, or to remain in the UK under a Tory government after Brexit.  The result is startlingly decisive -

Independence within EU : 56.7%
Brexit under the Tories : 43.3%

It's important to stress that this is not a "Yes lead", any more than yesterday's widely misreported Panelbase poll was.  There are undoubtedly people out there who would answer "no" to the straight question "Should Scotland be an independent country?" because of misplaced doubts over whether an independent Scotland could really remain in the EU, or over whether staying in the UK really means Tory rule for the foreseeable future.  But that's not to say that these findings are meaningless - they chart a course to how a Yes vote could conceivably be won, if the choice is framed correctly.

There are also voting intention numbers for the general election, which are initially quite hard to make sense of, because (in contrast to the practice in standard polls) the Don't Knows haven't been stripped out.

Westminster voting intentions :

SNP 38.9%
Conservatives 24.6%
Labour 17.8%
Greens 8.4%
Liberal Democrats 6.2%

A rough calculation suggests that the SNP would be on around 40.6% of the vote if Don't Knows were removed - essentially identical to their 41% showing in the recent YouGov poll.  However, the Tories would be on only 25.7% - making this their worst showing of the general election campaign so far.  The SNP's 15% lead over the Tories also equals the record in the Survation poll as the biggest lead of the campaign (compared to 13% with YouGov and 11% with Panelbase).  Far more important, though, is the implausibly high vote for the Greens.  If it can be assumed that a decent chunk of that vote is actually destined for the SNP, this poll is effectively implying a very, very healthy SNP lead.

All of this does of course depend on whether we can trust Google's methodology, and given that it's so untested, I'm not at all convinced that we can.