Smoke me a Kipper, I'll be back for breakfast

I'm increasingly optimistic (but far from complacent) that the "Kippers" and their repulsive lead candidate David Coburn are going to just about fall short of taking a Scottish seat in tomorrow's election - and the icing on the cake is that the SNP look like the party best placed to stop them (although it may yet be Labour or the Tories).  South of the border, though, it's a very different story.  Two eve-of-election polls tonight show that the rest of the UK is on the brink of entering into a veritable Farage à trois -

Opinium European election poll (GB-wide) :

UKIP 32%
Labour 25%
Conservatives 21%
Greens 6%
Liberal Democrats 6%

YouGov European election poll (GB-wide) :

UKIP 27%
Labour 26%
Conservatives 22%
Greens 10%
Liberal Democrats 9%
Others 6%

Although the latter figures look like a virtual dead heat, they're not filtered by certainty to vote (unless YouGov have changed their procedures at the last moment). With Farage's supporters being more motivated than others, it's highly likely that we'll be looking at some kind of UKIP victory at the weekend when the votes are counted. What strikes me about both of these polls, though, is the relatively narrow gap between Labour and the Tories. The traditionally low turnout in Euro-elections means there's a greater chance of an upset, and if the Tories do end up outpolling Labour (even if it's the Tories in second place and Labour in third) it could well have an electrifying effect on the referendum campaign. In the recent ICM poll, the No lead fell to just 3% when respondents were asked to imagine how they would vote if they expected a Tory victory in next year's general election.

As for the race in Scotland, if the SNP outpoll Labour by a significant amount, it will go a long way towards proving that YouGov's Scottish sampling and weighting for GB-wide polls is deeply flawed. Most of the firm's Scottish subsamples for the European election have put Labour in the lead, while their full-scale Scottish poll (which used a different if still dubious weighting procedure) had the SNP in a very narrow lead.

Ideally, if this election is to provide some supporting evidence that Yes-friendly pollsters like Survation are getting it broadly right, we'd be looking to see an SNP vote in the mid-thirties or above. That's admittedly a very tall order - it would be an all-time record (beating the 32.6% share of the vote in 1994), and it would mean doing significantly better than in 2009 when the main opponent was a Brown-led Labour party limping towards general election defeat. An SNP vote in the low thirties or below might be a sign that the methodology of No-friendly pollsters such as YouGov is closer to the mark.  Having said that, the gap between the SNP and Labour will also be an important clue - a sub-30 showing for the SNP would be far less troubling if Labour were trailing several points lower.