carolus
Lib Dem
Posts: 4,386
Member is Online
|
Post by carolus on Dec 29, 2021 19:23:45 GMT
That's a fair point about the squeeze question (although probably still helps a bit). But if you decide you want your model to take into account tactical voting I don't think you actually need to do it constituency by constituency in that way.
I can think of a few different ways to proceed with it: - bin your constituencies e.g. LD vs Con seats, Lab vs SNP targets, and so on and so forth. That probably gets you a vaguely useable amount of responses per category
- use historic data and so on to build a tactical/squeeze model - investigate how vote shares typically change in two party marginals, and so on - if you have enough historic national data then you could probably construct some demographic "ease of squeeze", although I'd imagine you'll find by far the most important feature is which party was second last time. - the whole hog, give up on the actual "MRP" part, and just throw everything into an MLP or some other more powerful model
But the fundamental point is just that none of this is implausible, and that there are ways to do it - if your stated objective is to provide seat by seat predictions you have to be doing something to deal with tactical voting, otherwise you really aren't doing anything of merit at all.
I should note - it's entirely possible that either the modellers are trying to do this and it has gone wrong somehow, or there is something major I am missing.
Everybody seems to be overlooking this quote from the Opinium survey, which I imagine holds just as true for Survation: “The seat predictions are based on a uniform swing, that is, they take no account of local factors, tactical voting or boundary changes and should be regarded as a broad guide, not a precise projection.” I'm not sure that's the case - the Opinium seat projections (the ones for the con leadership candidates?) didn't claim to be based on an MRP. The survation and focaldata ones are MRPs - the purpose of which is to be able to predict seat-by-seat. Some more details: One thing that struck me reading both of those is that they talk a lot about predicting "opinion", and a lot less directly about predicting election outcomes. (Survation at least suggest they've used it for GE on their page). I suspect this ties into bjornhattan's point upthread about them being more suited for opinion modelling than election modelling itself.
|
|
|
Survation
Dec 29, 2021 19:44:27 GMT
via mobile
Post by andrew111 on Dec 29, 2021 19:44:27 GMT
Everybody seems to be overlooking this quote from the Opinium survey, which I imagine holds just as true for Survation: “The seat predictions are based on a uniform swing, that is, they take no account of local factors, tactical voting or boundary changes and should be regarded as a broad guide, not a precise projection.” I'm not sure that's the case - the Opinium seat projections (the ones for the con leadership candidates?) didn't claim to be based on an MRP. The survation and focaldata ones are MRPs - the purpose of which is to be able to predict seat-by-seat. Some more details: One thing that struck me reading both of those is that they talk a lot about predicting "opinion", and a lot less directly about predicting election outcomes. (Survation at least suggest they've used it for GE on their page). I suspect this ties into bjornhattan's point upthread about them being more suited for opinion modelling than election modelling itself. Well, voting intention is a headline activity, a loss leader to get them noticed. But their bread and butter is buying intention!
|
|
|
Post by woollyliberal on Jan 4, 2022 20:51:08 GMT
|
|
|
Survation
Jan 4, 2022 20:52:22 GMT
via mobile
Post by jollyroger93 on Jan 4, 2022 20:52:22 GMT
Jesus Christ It was take in the middle of December for god sake.
|
|
|
Post by woollyliberal on Jan 4, 2022 21:06:09 GMT
My mistake. I saw changes since 9th Dec and thought it was newer than the 10th.
|
|
|
Post by bigfatron on Jan 4, 2022 22:15:36 GMT
My mistake. I saw changes since 9th Dec and thought it was newer than the 10th. Also Labour were on 39%, not 29% ;-)
|
|
Jack
Reform Party
Posts: 8,079
|
Post by Jack on Jan 21, 2022 16:03:32 GMT
|
|
clyde1998
SNP
Green (E&W) member; SNP supporter
Posts: 1,765
|
Post by clyde1998 on Jan 22, 2022 7:44:22 GMT
There's an interesting question in the dataset about whether you'd consider voting for each of the four main parties (Con, Lab, LDm, Grn) among those not giving a VI of that party: the Conservatives held 62% of their 2019 vote in the poll, including undecided voters, and only 38% of those who aren't planning on voting Tory say they'd consider voting Conservative now.
|
|
andrewp
Non-Aligned
Posts: 8,767
Member is Online
|
Post by andrewp on Jan 27, 2022 14:35:17 GMT
|
|
batman
Labour
Posts: 8,327
Member is Online
|
Post by batman on Jan 27, 2022 15:43:14 GMT
Amazing though it may seem to some, I reckon that there are a few switching back to the Tories because, basically, they feel sorry for Johnson.
|
|
batman
Labour
Posts: 8,327
Member is Online
|
Post by batman on Jan 27, 2022 15:44:06 GMT
although 43% was probably too high a figure for Labour in their previous poll anyway.
|
|
|
Survation
Jan 27, 2022 16:09:00 GMT
via mobile
Post by bjornhattan on Jan 27, 2022 16:09:00 GMT
Amazing though it may seem to some, I reckon that there are a few switching back to the Tories because, basically, they feel sorry for Johnson. I also wonder if there are some who think the whole media coverage is getting too puritanical. After all, it seems like they're focusing on cakes and wine than the actual principles involved. Most people don't really care about this alleged drinking culture in Number 10. And while many people are annoyed at these parties, it's the fact there were parties and the subsequent obfuscation that irritates them, not whether or not they were particularly racy.
|
|
Sibboleth
Labour
'Sit on my finger, sing in my ear, O littleblood.'
Posts: 15,233
|
Post by Sibboleth on Jan 27, 2022 16:23:19 GMT
Always useful to remember that most movement between individual polls is statistical noise. Better off trying to look for any general patterns.
|
|
|
Post by mattbewilson on Jan 27, 2022 16:32:18 GMT
I think it's more that people are just reverting to type
|
|
|
Survation
Jan 27, 2022 21:08:28 GMT
via mobile
Post by andrew111 on Jan 27, 2022 21:08:28 GMT
The polling average on Wikipedia basically has the Tories very stable on 32% since mid Dec (N Shrops, give or take), with small uptick this week, Labour still moving up but currently on 40%, Lib Dems stable on 11%, Greens down then up on ~ 5%, and Reform losing a couple of %
The Tory vote stopped going down before Christmas
|
|
|
Post by Davıd Boothroyd on Feb 5, 2022 23:49:58 GMT
Hmm. Something going on here, and I suspect it's some anti-Johnson Conservatives making up poll findings in a way they thought wouldn't be officially denied.
|
|
The Bishop
Labour
Down With Factionalism!
Posts: 36,309
|
Post by The Bishop on Feb 6, 2022 11:37:22 GMT
And now Focaldata (as they pointed out one word, not two) have also claimed the cited polling is fictitious. A rum business indeed.
|
|
|
Post by andrew111 on Feb 6, 2022 12:10:15 GMT
The Times used to be a proper newspaper that checked stories before printing (admittedly a bit of a distant memory!)
|
|
YL
Non-Aligned
Either Labour leaning or Lib Dem leaning but not sure which
Posts: 4,209
Member is Online
|
Post by YL on Feb 6, 2022 12:50:30 GMT
The Times used to be a proper newspaper that checked stories before printing (admittedly a bit of a distant memory!) Didn't they use to sack journalists who invented quotes?
|
|
batman
Labour
Posts: 8,327
Member is Online
|
Post by batman on Feb 6, 2022 12:59:17 GMT
Rigging elections at least has a clear purpose to it. Inventing polls seems to me to be an incredibly pointless exercise, even more so if you get caught doing it
|
|