United Kingdom General Elections: December 12th, 2019

[Oryxslayer]

What exactly are they basing this South Cambridgeshire insanity on? Allen’s not running there and the Tories won it by 25 points in 2017 (with the Lib Dems in third).

Same reason Esher and Walton is a tossup: they have data favoring the lib-dems and their targeted campaign style with those demographics in those regions.

[Oryxslayer]

I have some questions about races.

Scotland
- did any seats in Scotland vote leave in the referendum?
- what % of the snp vote voted to leave the European Union. Are they politically homeless with the SNP being so for leaving the uk but staying in EU?

Wales
- Aberconwy has a local council rep for labour who ran in 2017 and reduced the majority from 4-5k to under 1000 votes. The sitting MP has stood down and the tories have a candidate that was introduced in November that lives in Suffolk. Why is the labour candidate not winning?
- Does the remain/tactical vote work with Plaid? I know they did a pack with Liberals and Greens but in races where they get 10% they hurt Labour like Aberconwy.
- Ynys Mon - why is this such a volatile 4 way marginal?

England
- Why are the liberals running candidates in labour seats where they have no chance of winning? (Red wall) Is it true liberals hate labour more than the conservatives and they don’t care who wins so just run the candidate?
- What are the well off remainers who typically vote Tory doing with their vote in the north/midlands? Any Tory remainers moving over?
- why is Loughborough a show me seat? (Predicted virtually every winner). It has a well regarded university but unlike other uni towns) Is Canterbury similar to Loughborough?
- why is the West Midlands so anti-EU? Stoke, Wolverhampton, West Bromwich- labour predicted to lose 10-12 seats here.
- what caused Cornwall seats to swing dramatically to labour? They were always polling very low and in third place and liberals had mps here and it completely changed.

- What happens if Tories win majority and Boris loses in Uxbridge? Who goes to the Queen and asks her to be Prime Minister ?

I'll try to answer these to the best of my ability. Remember, you have to put yourself in someones shoes and try and look at things from their perspective rather than consider just your own.

Scotland

Banff and Buchan voted for Brexit, at least according to my data. It's a constituency of fishing communities in the highlands that wanted out of the EU's fisheries policy. I'm not sure if there is data on SNP/Brexit voters. However, Brexit is not the most important issue in Scotland, if it was then the Tories would be back at 1. The Yes/No polarization is more determinant these days. Now, this offers some data on your second question. Back before 2014, the SNP's best results were in the Highlands, the same area now swinging hard to the Conservatives. This is partially because the collapse of Scottish Labour handed the SNP a base that they actively could court on all fronts: the corridor between Glasgow and Edinburgh. Their old base was never committed to all the SNP proposals, these same regions voted fiercely against independence. So, some of it was death of the old generation, and some of it was the Oil and Fishing industries looking for parochial parties opposed to the Green/Europhilic parties of the strip.

Wales

Your question about Aberconwy is simple: the seat was previously held by a Tory and it voted for Brexit. Frankly, I'm surprised it's even in the same group of seats as the rest of the northern tossups. Now, sometimes there are things unique to candidates that make MRP less believable in a handful of seats. However, the UK has a proud tradition of airlifting in candidates - locals really only matter when the race is close, or you are a Lib-Dem. Plenty of Tory and Labour MPs never cared about their safe seat until the party said 'you've proven your loyalty, go stand there.'

Now, about Plaid. The PC vote is either incredibly inflexible or incredibly volatile depending on the voter. If you identify as Welsh, speak welsh, and this is your heritage, you will nearly always vote Plaid. Then you have the second group who identify as welsh but without the roots, and are more on the PC's to be against Labour. More of these exist at the local level. They will always hop over and vote differently at the Westminster level. this also answers your question about Ynys Mon, it's a place with a respectable amount of Welsh speakers.

England

If you are a national party, you run or cross-endorse as close to 631 candidates as you can get. Simple. A good number are sacrificial lambs who will never get a penny from the central office, but they are there to earn votes and continue to prove you are a national party. Same thing happens in Canada. the Lib-Dem brand in a good number of places is "not Tory" and "not Labour."

Polling seems to suggest there is a bit of a divergence between Northern ans Southern Tory remainers. Northern remainers have an additional dichotomy between them and their working class neighbors to reinforce their Conservative vote, such a divide is less a factor in the south. But the Tory majorities are larger in the south.

Those seats to the west of Birmingham are very in favor of Leave. Also, Boris's most favorable region is the midlands, at least according to YouGov's opinion polling regions. So take that as you will.

[Oryxslayer]

One of the most visible places of improvement for Labour and the biggest dropoff for the conservatives from two weeks ago in Yougov's model was Scotland. However, I would caution inferring anything from this prediction. Why? Well...

The headline here is that 27 seats, nearly half of Scotland, are considered to be Leaning or Tossup. There is only one safe seat of the entire 59 for any party: Aberdeen North for the SNP. Quite a few seats moved multiple catergories and changed ratings entirely between the polls. Now, lets look back to 2017. About half of the miscalled seats in YouGov's model were in Scotland, with the SNP winning 17 more seats in YouGov's model then in reality.

Back before YouGov dropped their first model I warned that they made two mistakes surrounding their successes: they missed a lot of seats in Scotland (sometimes badly), and underpolled parties that camping on 'targeting' rather than 'playing wide.' So, I'm not arguing in favor of any ideological position. Instead, the shear amount of competitive seats should make it incredibly apparent: YouGov has only a limited idea of whats going on North of the border. This isn't really their fault though. We only have one years worth of data on a 4-party Scottish system compared to the decades down south, every seat having three notable parties at minimum messes with weights when compared to two-party contests in England, and tatical voting among unionists is a serious thing whereas it has only ever occurred in limited capacity in at least 500 seats. All models have Scottish problems though, so it isn't YouGov's fault. Less than one hour in to the 2017 BBC broadcast they say that there is an asterisk in regards to Scotland, since so many seats were projected as uncertain and on knife's edges. Hell, BBC's exit poll gave Gordon to the Lib-Dems, a projection that looks LOL tier in hindsight, but Scotland is just uncertain.

So, the same two warnings from two weeks ago apply. Expect YouGov to be undershooting the targeted parties, and they only have a marginally better idea of whats going on in Scotland when compared to us.

[Oryxslayer]

Okay, this is the final YouGov MRP, separated into their Safe/Likely/lean/Toss statistical rating categories. I'll analyze it a bit, since there  is a good amount of data hiding behind the topline.



Compared to the previous YouGov MRP poll, the Tories lost net 11 seats north of Coventry to Labour. However, that topline obscures a nastier number. All 11 still remain close, and the algorithm considers them Lean or Tossups. On labours side of the equation though, 8 seats slid from Likely/Lean/Toss labour significantly towards the tories, and a good number more moved towards team blue outside of the competitive band. This is compared to 5 seats which left the battlefield, though nearly all were urban (like Wirral South) and probably never should have been in the battlefield. Only High peak is a notable movement in labours favor off the field of play. In essence, the potential Tory gains in the north got wider and more accessible.



it's the south and London where things moved in a less favorable fashion for the Conservatives. A whole lot of seats in the commuter shires moved from safe to likely, and plenty of targeted seats got more marginal. it's very well possible that YouGov are underselling potential Tory losses in the South because of how every model misses the Lib-Dems and their 'key targets' campaign style. The last thing I have to add is that YouGov has no demographic variable for Jews even in the large MRP, the community is too small, so just ignore all three Barnet seats. They got 2/3 Barnet's wrong last time in 2017, so this is a repeated offence that really isn't their fault.



I have already mentioned mostly how YouGov has no idea what is going on in Scotland, but I will just summarize it here. There is only one safe seat for any party in the 59, and half are competitive. Last time, nearly half of YouGov's miss-called seats were in Scotland. No model has any idea what is going on because we only have one years worth of party loyalty to work with, it's a 3 or 4 party system, and tactical voting is a realistic thing. Even the BBC exit poll made some LOL tier predictions up here because of how uncertain this whole thing is.



Wales deserves mentioning because of conflicting data sources. On one hand, we have the YouGov Welsh barometer giving us a Lab+3 result. On the other hand the MRP has Lab+9 for their welsh subsample. Both the poll and the model were very accurate in 2017 so we are left with a confusing picture. One thing that is clear is that five northern welsh seats (Aberconwy, Dely, Clwyd S, Vale of Clywd, Wrexham) all seem to be moving as one unit. Oh, and who knows whats going on in Ynys Mon, every model under the sun historically fails to get enough targeted data on PC to accurately project their Westminster vote.

[Oryxslayer]

https://twitter.com/ElectionMapsUK/status/1204797680536498176

YouGov says this seat is a Lab-tilting tossup, this poll suggests Con-tilting. Either way, tossup.

[Oryxslayer]

https://twitter.com/ElectionMapsUK/status/1204842675851137024

Last minute poll shows no overall change.

[Oryxslayer]

https://twitter.com/ElectionMapsUK/status/1204884474938449920

Another shockingly stable poll.

https://twitter.com/britainelects/status/1204884850974609408

One that's less than stable.

[Oryxslayer]

Less than 24 Hours remain until BBC opens the exit poll envelope.

[Oryxslayer]

https://twitter.com/ElectionMapsUK/status/1204914010056994816

Survation with their closing remarks.

[Oryxslayer]

Is there any chance Corbyn loses support of the party after the election? He's headed for another GE loss most likely, but does that matter?

The mans going on 71. If Boris wins a majority than he can't remain on purely because of his expected age in 2024, which is when one has to assume the next GE is. This is ignoring the potential loss of authority. Unless Labour does historically bad though, and Corbyn needs to be kicked to the curb for blame like Miliband, he's in a reasonable short-term position. Labour can't afford to go into chaotic infighting right when the UK is about to reshape her position in world affairs and potentially enter fiscal uncertainty. However, he has lost authority. It's in everyone's best interests therefore that Corbyn announces that he will be 'resigning in the near future.' For Labour generally, it puts them in the position to benefit from Boris's parliament and Brexit. The general public will see not Corbyn's Labour and it's failings, but their idealized (wishfully) potential Labour. For Corbyn and his allies in particular, it gives them time to stack the leadership contest in favor of some preferred successors.

[Oryxslayer]

https://twitter.com/britainelects/status/1205260563888197633

Supposedly very close in Sunderland Central, but no cigar.