Am listening to Sir John Curtice @whatukthinks talking about the election with @anandMenon1 @UKandEU - clear it's a #brexit election /1

-80% remainers/leavers voting for parties that suit their cause.

-CONS gained ground amongst leave voters; LAB gained among remain voters. “All of the dynamics within camp” /2

Be amazing if the @YouGov poll tonight showed anything other than Tory majority, Q is how big.

BUT sign of narrowing in recent weeks...1/2pc

BUT resisting any parallel with 2017 and the 'labour surge'. /3

Turnout: "boring technical reason" why turnout is going up, is that election is to be fought on new register, which has less dead folk on it! /4

also in camp that says 1) attachment to leave/remain cause creates "levels of attachment" to a political cause not seen since 1960s 2) British social survey has long asked 'how interested in politics are you?', about 30% always says 'interested' but since 2016, that's been 40% /5

Reason Tories ahead is NOT because there's a majorit in favour of leaving...polls show narrow for Remain, but it's the "differential concentration" that matters. The Tory vote has become much more of a Brexit vote; on Remain side, the vote is divided (hence need for tac voting)/6

Leave tactical voting is pretty much fixed/maxed out...so where are Remain voters...well NOT MANY people vote tactically. These are people who hate A, but are agnostic on party B/C, and this is a rare psychological situation. Most who hate A, love B. /7

So, compare with 1997, when tactical voting was significant.

Two things 1) strong 'Tories out' sentiment, united people cross-party...and most Remain voters want @BorisJohnson not to win. To that extent the motivation is there.

BUT in 97, Lab and LD's were "playing footsie"/8

Two parties were open about fact Blair wd sign a coalition deal, and Ashdown wd agree.

Not this time. Swinson thinks Corbyn as useless as Johnson..so relationship nothing like as good.

So Q is which force is more important? /9

So if 1997 levels of tactical voting were replicated you would knock 20 seats of Tory tally...which wouldn't be enough to stop @BorisJohnson on current polls. /10

So now explaining how/why "MRP" works... this is tonight @YouGov poll coming out at 10pm that was accurate in 2017.....it's based on a huge 100k+ data set, which enables accurate geographical weighting that can be replicated across EACH constitutency /11

In 2017 @YouGov was one of two organsiation that used this, they said early it was going to be hung parliament - were abused roundly - and where proved right. So it's taken on big status...BUT in 2017 there was 2nd MRP exercised done for Ashcroft that showed majority of 60! /12

So it's not a magic bullet...

"If the data is crap, the MRP result will be crap"

And Ashcroft data wasn't as good as @YouGov data /13

@whatukthinks On the impact of polling on politics/elections.

"At the end of the day, politicians will be politicians and if they don't like the evidence, they will ignore it."

Is UK more polarised than ever? asks @anandMenon1

Yes, says Curtice, in sense that the #brexit solutions are so polarised; and so, sectorally in society, are leave/remain constituencies. /15

What's gonna happen in the election?

There's a 66% chance Tories get a majority...but 12pts is not to say Tories are 'home and dry'. /16

That's all folks...and we learned that Sir John prefers Steak and Ale pie over boeuf bourguignon 17/ENDS