Given the success of the approach at the 2017 election I expect we’ll see several MRP seat models this time round. The first one to emerge however is one constructed by Focaldata, using data from mixed sources, including YouGov, that Best for Britain have used to drive a tactical voting website. It has caused some controversy – particularly on the comment pages of the Guardian – with people arguing over the validity of its recommendations. I won’t get too far into that (vote for whoever the hell you want), but thought it was probably worth making a few comments about MRP itself, considering it will crop up again through the campaign.

What is MRP?

First, we need to understand what MRP is. It stands for multilevel regression and post-stratification, which almost certainly doesn’t help. There is a academic paper by Ben Lauderdale and his colleagues who run the YouGov MRP that explains it in great detail here, however the short version is that it’s a modelling technique aimed at producing robust estimates for small geographic areas from large national samples. In the context of elections, that means coming up with estimates of vote share in single constituencies based on a big national sample.

Using traditional techniques even very, very large samples don’t contain enough respondents to be a good guide to individual seats. If you had a huge sample of 50000, divided by 632 seats it would still give you less than 100 people a seat – which wouldn’t be enough to produce decent data. I’ve seen this as a naive criticism of the Best for Britain MRP model (there are only 70 people per seat!) but in fact that is exactly the problem that MRP is intended to solve.

MRP works by modelling the relationship between demographic and political variables and voting intention (the multilevel regression part), and then applying that to the demographics and political circumstances in each individual constituency (the post-stratification). So in this case, an MRP model would look at how demographics like age, gender and education relate to vote intention, and how that differs based on political variables (Is there an incumbent MP? Is it a remain or leave area?). That model is then applied to the known characteristics each seat. What that means is the projection in an individual seat is not just based upon how respondents in that seat say they would vote, it’s effectively also based on how respondents with the same demographics in seats with similar political circumstances say they would vote.

How well does it work?

At the last election YouGov had an MRP model that performed very well – correctly predicting the hung Parliament and some of the more unusual election results like Canterbury and Kensington. Clearly, given the accuracy of the YouGov model, it is possible to use MRP successfully to produce decent seat level estimates from a big national sample.

Best for Britain’s defence of their tactical recommendations relies heavily on how well the YouGov MRP model did in 2017. However, not all MRP models are necessarily equal. It isn’t one single model, it’s a technique, and it’s possible to do it well or badly. It is not certainly not a magical guarantee of accuracy. If we look back to 2017 the YouGov MRP model got all the attention, but it wasn’t the only MRP model out there. Lord Ashcroft also commissioned an MRP model, but that wrongly predicted a Tory majority. Just as some polls have been more accurate than others in recent years, some MRPs may be more accurate than others.

The things that drive the quality of a MRP model should be the quality of the data that’s going into it, and the quality of the model itself – have those designing it picked demographics and political factors that allow them to accurately model voting intentions? As an external observer however, it is quite hard to judge that. For the YouGov model there is its track record from 2017. From other MRP models, we’re driving a bit blind. We know it is a technique that can be very successful if done well, but we won’t really know if it is being done well until it’s compared to actual election results.

Are tactical voting recommendations based on an MRP model sensible?

In principle, yes. MRP is obviously not perfect or infallible – nothing is – but it is an established technique for producing estimates of support in small geographical areas from a larger national poll. Certainly it should be better than using a crude uniform swing, or just basing recommendations on what the levels of support were at the previous election and assuming nothing has changed since.

In practice, of course, it depends on the quality of the model and the tactical decisions that people make based upon them – I certainly don’t intend to get into that debate, especially since I expect there will be various rival tactical voting sites with different recommendation, and perhaps different aims and motivations.

543 Responses to “MRP models and tactical voting”

1 8 9 10 11
  1. “distributers”

    Should that have been “distributors”? I never know when “ers” or “ors” is the appropriate spelling.

    Is there some rule that I can apply, or is the variation in ending just one of those random things that I would need to learn?

  2. Deltapoll seems to confirm Brexit party vote falling away, without giving any benefit to the Conservatives.

  3. @OldNat

    Random, I fear. Just like opinion polls.

  4. @James E

    I think that some Brexit Party voters may eventually slide to DK or even WNV.

    There was a chunk of Leave voters who hadn’t previously voted, and if miffed enough will go back to not voting.

  5. @MiserableoldGit

    “I think we have to hope for the middle option, with the SNP and (holding noses) the LibDems doing enough to deny Johnson a majority.”

    That’s about the size of it, I think.

    By the way, I like the look of this Charles, John33, MiserableoldGit and Crossbat coalition developing. I’m hopeful that JimJam, Valerie, TonyBGT, Alec possibly other UKPR posters of like mind, find it in their hearts and consciences to join us one day! I sense we’re a broad church and tolerant of returning, maybe recanting, heretics:-)}

    More seriously, while there are many posters with different political views to mine on UKPR who I enjoy reading (and one or two I really don’t!), I’m particularly looking forward to hearing from all you kindred spirits over the next four weeks or so. I think we’re beleaguered and very much on the back foot in this election, but your thoughts, observations, anecdotes, expertise, and just the mere fact that I somehow know you’re out there, hoping, maybe forlornly, for the same sorts of things than me, consoles me greatly.

    I need you all more than you’ll ever know. These are dispiriting and worrying times for those of us on the Left.

    Oh, and I forgot the redoubtable Dr Mibbles. I’m hoping he pops up again soon with a 2017 Canterbury-esque prediction as he alone, astride a barren polling landscape, spots the late Labour surge that starts to rumble circa December 10th!


  6. @Batty

    I would be delighted to bring happy anecdotes, but you try finding happy politics in Arundel & South Downs !

  7. MOG

    :-) Though actually :-(

  8. @Crossbat

    I genuinely don’t think my beliefs are misguided; although that’s not to say I would be that surprised if the Tory’s achieved their majority – I just don’t see it being the most likely outcome. Johnson has already fought his campaign over the last few months with hos various cynical stunts and attempts at chicanery, in order to build up a coalition of support around the leave vote.
    People keep saying that he is no Theresa May and there is no way the Tories will fight a campaign as badly as the last but, from what I can see, it is 2017 revisited. Like TM, he simply isn’t the figure which has been cultivated recently and he can’y get through the election without that reality being exposed. As I said before, the debates will be a hammer blow to his campaign and I cannot see him emerging from them without great harm being caused. Maybe I am wrong: doesn’t feel like it though.

  9. *Tories

  10. Interesting analysis of social media campaign spending on Slugger.

    Donaghy notes “Neither the SNP nor the Brexit Party have bothered with social media advertising so far at this election”,

    Certainly true of the SNP in terms of buying advertising, but they have noted that –

    MORE than one million Facebook users have seen the SNP’s “people-powered” digital campaign over just seven days, the party has revealed.
    The total audience reached is the equivalent of around one-fifth of the population of Scotland thanks to users sharing the material placed online by the Edinburgh-based party, which does not include paid-for adverts.
    Analysis, which is publicly available on the platform, also shows that the SNP’s social media presence is the strongest of Remain parties, surpassing the LibDems in terms of Facebook page following, engagement and reach of posts.

    Just as with personal conversations, canvassing and even leafleting, digital persuasion is more effective when it comes from known individuals, not commercial advertising.

  11. @MOG

    “I would be delighted to bring happy anecdotes, but you try finding happy politics in Arundel & South Downs !”

    Not fertile ground then? Whatever happened to the Socialist Republic of South Downs? Ye Gods, has it turned blue? We are truly doomed!

    I now know where you are, but beyond old Chris down in Bournemouth, the geographical spread of our LOC community is a mystery to me. R&D is an exiled Gooner, I think, and not in London any longer. The others reside I know not where.

    I have a plan. Not far from me, in the Malvern Hills, there stands a beacon called, unsurprisingly, Beacon Hill. Shortly after 10.00pm on the night of December 12th, as Huw Edwards reads out Curtice’s infallible exit poll, I will light a beacon on that very hill. It’s visible for huge distances around and was one of only about twelve beacons required to tell the whole country, from Plymouth to way up North, that the Spanish Armada had arrived.

    My blaze will be extinguished quickly if the exit poll is dire for Labour but, if by some chance the result confounds the polls, I will let it burn throughout the night as a glorious symbol of national renewal. I may invite a few diehards with me to the top of Beacon Hill that night. Either to linger in celebration and fire-stoking, or to help me quickly extinguish the flames. If the latter, we will decamp to the nearby Skinners Arms in Malvern, there to drink ourselves into blissful oblivion!

    So, to all my kindred spirits on UKPR. On the fateful night, as the polls close, gaze towards the Malvern Hills in yonder Worcestershire, there to witness a short or extended blaze. If the fire burns only briefly, think of it like this.

    “The beacons are going out all over Europe, we shall not see them lit again in our life-time”

  12. The election is still in a sort of phoney war phase.

    The publication of the manifesto is crucial.

    In 2017 the dementia tax was disastrous for the Tories, abolishing tuition fees was a triumph for Labour.

  13. @Batty

    In sympathy, I will toddle along the 15 miles to Brighton Kemptown, and set the blaze going on Ditchling Beacon to alert our French friends that some of use are still sane.

  14. Westminster voting intention:

    CON: 39% (+3)
    LAB: 26% (+1)
    LDEM: 17% (-)
    BREX: 10% (-1)

    , 07 – 08 Nov
    Chgs. w/ 06 Nov

  15. Labour going up in all three polls.

  16. @James E

    “Deltapoll seems to confirm Brexit party vote falling away, without giving any benefit to the Conservatives.”

    The Tories have already taken what they needed from the BXP/2015 UKIP vote. They have totally destroyed the Tory leaning BXP vote.

  17. Crossbat

    Count me in. I live 4 miles from the centre of Manchester- constituency voted 78% to Remain. We never hear a peep out of the Conservatives. People I know who live in neighbouring leave-voting constituencies would never vote Johnson- they think he’s a posh southern tw-t.

    I don’t think Corbyn can get a majority, but I don’t know where the seats are that Johnson needs to win to have a working majority.

    Hope springs eternal is my catchphrase. It’s replaced “we will not be leaving the EU without a deal on the 31st October”

  18. @ Old Nat

    Talking of social media, George Soros might have some controversy coming his way next week over funding for the Best for Britain’s tactical voting website.

  19. Can I join your coalition please, CB?

  20. One scenario i am hoping for (but not really expecting) is something like

    305 CON
    230 LAB
    50 SNP
    45 LIB + PC + GRN + IND

    What would happen in this situation? Two defeats for Corbyn should see him resign really. However, the SNP+LIB may support a Labour government so that could save Corbyn, but would LIB want to support Corbyn? If there was to be a temporary leader of a GNU then who would that be? Tom Watson would have been one of the more obvious choices but he is out of play now.

    Such a situation could be complex, there could be agreement that a second referendum* was needed to unlock the current situation, or might we be faced with a new election in the spring (with Corbyn having gone)? What would Johnson do? I think he would try to hang on but it may depend on the reasons behind the CONs not getting a majority – if he was responsible then he would likely be forced out.

    I know we can all dream up what if situations and spend too much time pondering hypothetical situations but for someone like me who is really disappointed with the current situation i think it might be the best i can hope for. As i have written before, unless Johnson really messes up, or people really manage to erode away what confidence people have in his deal, then i can only see the number of conservative seats going up unfortunately.

    *i know that most of the conservatives who may be inclined to support a second referendum are not standing this time and also that some labour MPs will not support a second referendum but continuing deadlock could change minds and maybe public sentiment too.

  21. @Norbold

    Of course, and I shouldn’t have excluded you from my list. It wasn’t exhaustive anyway, but yours was a bad omission on my behalf. We need Hireton, CMJ, R&D etc too, should they wish to join. The tent is big and wide.

    Maybe, one day, LeftyLampton and Amber Starr will return.

    Then we really would have a Stop Boris Campaign!

    (Valerie; your application form is in the post!)


  22. Scotland is one area which interests me, particularly the performance of the conservatives there. There has been an assumption from some quarters that the conservatives are going to lose most of their MPs. However, while Stirling is hyper-marginal and the CONs only won Gordon by 5%, there are then 6 seats which the conservatives won in 2019 by between 6 and 9%. Although UNS on current polls would likely see them lose these seats, the high number of unionist (inclduing conservative) d/ks compared to SNP d/ks makes me think that when those d/ks become do knows, that the swing might be small enough that the conservatives just cling on in most, if not all, of those places.

  23. Bantams

    “Best for Britain” – what an odd title for a “tactical voting” site that has no recommendations for anywhere in NI or Scotland. :-)

    Is Soros being accused of Anglo-centrism, or is whatever controversy you refer to just a result of the forthcoming Sunday Times story from the unpublished report on 9 Russian oligarchs funding the Tories?

    Is the B for B site a Lib Dem front?

    In this part of the world, we are well used to the Lib Dems accessing large wads of cash from elsewhere to target voters in their selected seats.

  24. @MOG

    “In sympathy, I will toddle along the 15 miles to Brighton Kemptown, and set the blaze going on Ditchling Beacon to alert our French friends that some of use are still sane.”

    Good man! Of course, however bad the result on December 12th, they’ll never really put the fire out.

    It will burn on.


  25. With regards to Labour apparently having a good campaign set up in Canterbury. I can believe it as Duffield comes across as a popular MP and the smallest conservative majority elsewhere in Kent is 6000 so, unless there is a really dramatic shift in opinion, Labour are best off trying to defend their only seat rather than go hunting elsewhere. To some consternation, the lib dems selected a candidate who is quite vocal on twitter and has a lot of followers. From a 2015 starting point it may make sense to make a push in Canterbury, but not from a 2017 starting point.

    Living in what is apparently a 3-way marginal on current trends (Battersea), i have had A4 leaflets from lib dems (8 pages) and labour (4 pages) but only a small A5 leaflet from the conservatives (don’t mention the referendum!). This despite living in one of the more conservative areas of the constituency – early days yet, i imagine that labour and the lib dems might both be testing the water to see if they can get their noses in front or whether they should deploy resources elsewhere.

  26. Frosty

    Re Scotland

    You are right to say that the distribution of DKs among 2017 Unionist voters (if they vote) will be important in those Tory seats.

    We won’t have a clearer picture until the anticipated YG MRP data becomes available.

    What matters is which parties they are unsure about voting for, and in which constituencies they are concentrated. Also, if those who had a changed VI 2 weeks ago still have that intention, and again. where they are.

  27. Frosty

    Same question to you that I posed to Davwel.

    Were these leaflets hand delivered by activists, or via Royal Mail/commercial distributors?

  28. @Oldnat, re: Scotland. Yes. Hopefully there will be further polling soon and maybe also a poll focussing on the conservative-held constituencies.

    Re: leaflets. No address on them so i guess activists. Did also get a big anti-brexit leaflet from Sadiq Khan too

  29. Sorry, wasn’t just anti-brexit but that was the top message.

  30. Another coalition application here. Not much prospect for a beacon in West London. Had Keir Starmer at our constituency launch today and I was as impressed with him as I would expect. Clever, organised, lucid, personable and dealing courteously but firmly with someone who threatened the overwhelming mood of unity. A campaign heals wounds and splits, especially this one where the stakes are high and the hill to be climbed so steep.

  31. Frosty

    Others will know better than me, but I think “no address” just means they were pre-election campaign leaflets, not governed by strict rules.

  32. This coalition is turning into a bandwagon :)

  33. Among the politicians talking up how good the campaigning was today, it was nice to see one Labour doorstep canvasser say that it was miserable and people didnt want to talk. With it being cold and wet in many places today i can imagine that not many people would want to stand around with the front door open getting cold and letting the cold in. Also, some of the less devoted canvassers may not want to walk house to house in the cold. Maybe the sign of things to come as we move further into November and December?

  34. If the Coalition aims to keep all parts of the current UK as member states of the EU, and pursuing policies of reducing inequality, I’m in.

  35. Maybe the Russian influence report will be leaked at some point. The Sunday Times has some information today.

  36. @OldNat

    It’s up to Batty, as originator and convenor (or convener!) but I’ll be happy to sign your nomination papers.

  37. MOG

    Thanks – though if Batty sets up a Ltd Co with himself as sole arbiter to run the Coalition, I’m oot!

  38. The three polls are a little contradictory, but on average Labour has pulled back 2 points. All within margin for error of course

  39. @Frosty Canvasing door to door.

    Canvassing will be very difficult in the cold and dark , especially in rural communities where there may be no street lighting. In the last 2 weeks I know of 5 burglaries or attempted burglaries in my local area. One of our near neighbors has CCTV of 2 blokes in balaclarvas with a crow bar routing about their out buildings and I have have 2 separate incidents of suspicious vehicles on our drive.

    The police are great and can get here with in 15 minutes , which considering the rural location is very good.

    God forbid anyone knocking on my door in the dark!

    Those parties that have a good door to door canvasing set up will loose out more than those that don’t

  40. Crossbat11,
    “He’s very much a throwback to 1970s Labour, ”

    Yes, I think thats right. In other words, while much of labour has accepted Thatcherism, he has not. But that isnt a criticism. Cometh the hour cometh the man. He was chosen as leader exactly because he has not accepted the economic and government philosophy currently in place, which is seen to be failing.

    Arguably one lot of people have chased after brexit as their salvation when the current system is seen not to be working, and another lot have chased after Corbynism. The common theme would be the current system is failing.

    He is attacked by the right for a couple of reasons, mostly I guess because he is the leader of labour, and automatically they seek to smear him because that is the top strategy in the playbook. But secondly because he rejects Thatcherism, which is a fundamental attack on the position of the conservative party. Moreover his policies when considered calmy are welcomed by voters. The right dare not let him apply them, because it would fundamentally undermine their own position if a Corbyn government proved successful, which it well might.

    If a labour government took office now which was more left leaning than Blair and was successful, it would likely lock in a shift to the left of British politics for a generation. But that wouldnt be because labour had sold it, but becuse that is the direction the public was wanting to go, but conservatives have been fighting a rearguard action against it for decades.

    ” He’s led from a much more narrow base than he needed to”

    You seem to forget that he started with a much wider spread of political views in his shadow cabinet…but they all resigned. He did try to be inclusive. I dont know whether the blairite/thatcherite senior labour people were right or wrong in the court of public opinoin, but they decided to make this a fight and tear apart labour. If labour fails to win now, it will be quite likely their fault. Telling the public you think your party has the wrong ideas is no way to win an election – whether that is true or false.

    Classically, labour care more about the theory of government, whereas conservatives care more about winning. Maybe thats because labour care more about social justice. While labour is arguing publicly about the theory of how it should govern and so be true to its voters, con are applying the practical solution of how to stay in power longest…. and therefore doing all they can to dilute or prevent Brexit without airing their doubts in public. The only debate con are willing to have in public is what sort of brexit, not whether it needs to be stopped. That doesnt mean they believe in brexit.

  41. Good Morning from a cold Bournemouth East where we will be running in an hour to Bournemouth West and back.

    Does anyone know why Jeremy Corbyn and John McDonnell were unable to make it to the Albert Hall last night? Emily Thornberry was there.
    Lots of media coverage on the Labour spending pledges.

    1992 to 1994 Gordon Brown made himself unpopular with Labour members by clamping down on Conference and shadow cabinet promises about spending.
    He did New Labour’s dirty work.
    I am old enough to remember Labour winning power in 64, 74 and 97.

    Despite Labour gloom today, I am inclined to agree with you that it may well be hard for the Tories to pick up a net rise in seats, enough to form a majority.

  42. Frosty,

    Re Battersea – another example imo of a place where the LDs have no chance or coming from 3rd place to win; it is simply not a 3 way marginal.

    Re Leaflets and canvassing prioritisation; the working premise is that voters in more affluent areas will generally vote without any prompting. Especially if they are Labour supporters who had done well in life from a narrow economic perspective they will turn out without any contact beyond a leaflet or 2.

    This works the other way as well with differential turnout showing that voters in less affluent areas need more prompting and this is especially true on polling day with GTVO operations focusing on self identified supporters in those areas.

    In this GE (like 2017) extra focus will go on identifying early on potential voters for squeezing who may well get a specific communication; the old 2 horse race approach. This is the relevance of those arguably spurious LD polls that will be used on leaflets as they may make Tory MPs more likely to be elected with misdirected resourcing and focus.

    This appears to be most prevalent in London with LD votes off Labour helping the Tories.

    Being in second place in more ‘remain’ seats to be well placed for the rejoin commitment in the next GE seems to be more important to some LDs than stopping Brexit now.

  43. @CHRISLANE1945
    Does anyone know why Jeremy Corbyn and John McDonnell were unable to make it to the Albert Hall last night? Emily Thornberry was there.

    Not a fan of Corbyn but he was visiting the residents and emergency services in Doncaster in relation to the floods. I do hope this year people do not try and politicise the remembrance services this year, it is very distasteful in my eyes

  44. Danny and Crossbatt 11
    Good Morning to you both.
    In the 1970’s you will remember that Labour won two GE’s (narrowly) and lost in 1970 to Heath and to Thatcher in 79.
    The 1968-1970 period was terrible as the trade union ‘movement’ undermined Harold Wilson. When he returned to power in March 1974, he and Jim Callaghan in March 1976 to April 1979 were similarly undermined by the ‘comrades’ of the Left in the PLP who regularly voted with the Tories against Healey and Callaghan in the Commons. Harold also felt very hurt by the civil war over Europe within the Party.
    Tony Benn, who adopted that name in 1969 served in all the Labour Cabinets from 1964 to 1979 but in the 1980’s spent a great deal of time attacking his own party’s record.
    The effects on polling were unhelpful to say the least,

  45. There is, we now know, a clear remain bias on this board, and I think it is showing through on the interpretation of the news and ensuing polls.

    The general sentiment (from myself included) was that Cons haven’t had a great start to the campaign while Labour has done OK, barring the last couple of days. This has given heart to many on here.

    The polls tell a different story. I think the point for remainers is that to win, every part of their unofficial coalition needs to fire. Labour has ticked up towards 30% in most polls, which is an improvement, but still a disaster if replicated on Dec 12th. It’s the other parts of chess board that look terrible for remainers.

    In Scotland, the SNP going full belt on independence has made Cons happy, and the polls seem to justify this feeling. They are up, Lab is down, and it’s now very possible that Cons will retain a few of their seats and Lab lose nearly all of theirs. The idea of 13 Con losses looks unlikely.

    Jo Swinson fails to fire, in my view. The Lib Dems going full belt for revoke seems to have lost them support amongst Cons. It was a terrible move, and this, allied to their insistence that Corbyn is as much the enenmy as Johnson means that once again, the Lib Dems are doing the Conservatives dirty work for them. They are harming Lab deeply in places like London, but don’t seem to be picking up the seats in the Con/Lib Dem battlegrounds.

    Meanwhile, Farage struggles to gain traction, and the BXP appears to be losing support. If this remains until Dec 12th, it removes Labs last hope of salvaging something
    fro this election.

    With polls largely now agreeing that Cons are on or around 40%, the upbeat tone from remainers on here is completely misplaced. Much can yet happen, but currently, Johnson is romping this. Every conceivable dynamic, in every possible direction, is working in favour of Con leavers and away from remainers.

  46. Corbyn snubs Festival of Remebrance

    I’m surprised this isn’t getting a lot of coverage

    Out of the election book of “how to drop %10 in the polls.”

  47. ALEC.
    Good Morning to you; yes, 40% is the magic number. I remember Norman Tebbit saying the same thing in the 1987 Campaign in a panel discussion chaired by Robin Day.

    Has anyone else spotted that the Labour leadership and supporters often attack the Blair legacy as Thatcherite or Tory, they nevertheless keep going back to 2010 as the year when public service provision began to decline?
    Not good for polling, I think, from any Labour perspective.

  48. @Danny.
    Some of what you say is no doubt correct. But if I may? You assert that voters like Corbyn’s policies when he gets a chance to explain them? Not all do! Which is obvious as is all voters did then he would be in power by now. I live in a Tory/Libdem marginal and the general feeling here is one of mistrust of Corbyn for the damage her would do to industry. Much of it around here is defence industry and the feeling, not without fact is that Corbyn essentially would massively downscale our nations defences. Around here that costs jobs and livelihood.
    There is also concerns that generally align to hard left politics that have been alluded to in the past such as the cost of nationalism etc.

    Just a note to you others who are desperate to see a Corbyn 2017 renaissance, I’m not saying that’s not going to happen, but there is no evidence of it yet. Tories are presently holding a 10-12 point lead which puts them in majority.
    As an earlier poster said, the launch of the manifestos will probably be key.

  49. CL1945

    @”I am old enough to remember Labour winning power in 64, 74 and 97.”

    Me too-what a political era the 60s & 70s were.

    Somehow those people seem like giants compared with today’s . Perhaps it is the rose tinted glasses of age ?

  50. NRMM,

    Labour in my seat have suspended campaigning until 1400 as a mark of respect which is a Nationwide action by the party but local Tories while not door knocking still delivering leaflets.

1 8 9 10 11