It's almost terrifying to consider the impact that the results of the low-turnout election we'll see tonight could have on our future. Think of the transformative effect of the English local elections earlier this month - by historical standards the Liberal Democrats didn't actually do all that well, but because they were the only one of the main London parties that made gains, the momentum generated was all for them and the political weather was decisively altered. In one fell swoop, Change UK was effectively taken out of the game. In the early 80s, the late and disgraced Liberal MP Cyril Smith argued that his party shouldn't be seeking to cooperate with the SDP, but rather to "strangle them at birth". A few decades on, it looks like the Lib Dems have heeded that advice in a strikingly similar situation.
The biggest impact of a surprise result tonight could be on the prospects for a no deal Brexit. The main reason that no deal has suddenly become acceptable language for Tory leadership candidates to use in polite company is the surge for the Brexit Party in the polls. But if the Brexit Party underperform expectations tonight, and particularly if the Lib Dems do well, the internal mood within the Tories could swing back in the opposite direction. We already know that turnout has risen more in English local authority areas that voted Remain, so that's not an entirely implausible scenario. European elections have a history of throwing up major shocks, perhaps because there's so much scope for differential turnout to come into play. The 1999 outcome, with a clear Conservative victory and several seats each for UKIP and the Greens, stands out as a result that was way outside the parameters of what most commentators thought was possible. I wouldn't be totally amazed if we see a very close run thing in the popular vote between the Brexit Party and the Lib Dems tonight.
I don't know about anyone else, but I'm getting slightly jittery about the Scottish result. It was a bit troubling to hear the estimates that the overall turnout in Scotland, along with the increase in the turnout, was lower than in England. Turnout was never going to be super-high, but a tolerably high figure would probably have suited the SNP better. And you might think that the English pattern of turnout going up more in strongly Remain areas would also apply here, but I'm not so sure. The voting patterns in the EU referendum were so different in Scotland, with socio-economic groups that went heavily to Leave just over the border in north-east England going the opposite way here, partly because pro-Europeanism is so closely associated with support for an independent Scotland. One possibility is that there may be a difference in turnout between leafy, affluent Remain strongholds and the more working-class areas that voted strongly Remain, and if so, that could favour the Lib Dems more than the SNP. The good news, though, is that the D'Hondt formula favours larger parties, and it may be possible for the SNP to take a third seat even if the polls have overstated their support (although they'd still need to exceed the 29% of the vote they took last time).
Here are a few benchmarks to look out for...
2 seats: This would keep the SNP on the same number of seats they've had continuously since 1994, and would technically equal their all-time high.
3 seats: This would beat the SNP's all-time high, and would also be the highest proportion of Scottish seats that any party has won since proportional representation was introduced twenty years ago. (Labour took three seats in 1999, but there were eight Scottish seats back then rather than six.)
29.0% of the vote: This would equal the SNP's performance of five years ago.
29.1% of the vote: This would equal the all-time high for the SNP (and for any other party in Scotland) since proportional representation was introduced.
32.6% of the vote: This would equal the all-time high for the SNP in all European elections. But it's arguably an unfair comparison, because that figure was recorded in 1994 under first-past-the-post, which is a system that incentivises voters to back larger parties.
36.9% of the vote: This would equal the SNP's performance at the last Westminster general election. It's a real 'apples and oranges' comparison, but you can guarantee the media will still attempt it.
I'll also be keeping an eye out for any sign of the ex-YouGov propagandist Peter Kellner on the BBC's results coverage. I've told this story a number of times, but at the last European elections (held just a few months before the indyref) Kellner wrongly announced to viewers that Labour had emerged victorious in Scotland, at a stage in the evening when it was already pretty obvious from the average swing that the SNP were going to hold on for the win. He never corrected that rather convenient error or apologised for it, and it was quite some time before Alex Salmond was able to point out that the SNP were in fact on top in the popular vote.
There has been some talk (mainly from Iain Dale and Daniel Hannan) that the Tories might be wiped out completely across the UK, but I can't see that happening. Even if they have a spectacularly bad night, they should hold on to the odd seat in larger regions.