|
YouGov
May 20, 2022 16:58:43 GMT
Post by bigfatron on May 20, 2022 16:58:43 GMT
Not entirely - there is always the option of erstwhile Tory voters not voting at all.
|
|
|
YouGov
May 20, 2022 21:01:30 GMT
Post by greenhert on May 20, 2022 21:01:30 GMT
Not entirely - there is always the option of erstwhile Tory voters not voting at all. Which is probably what many of them will do in next month's by-elections for a start.
|
|
|
YouGov
May 23, 2022 12:12:02 GMT
via mobile
Post by andrew111 on May 23, 2022 12:12:02 GMT
Interesting. The implication is that the Tories are near their rock-bottom support level. I find that plausible.
31% for the Tories is exceptionally low, suggesting Yougov have found a sample where up to 10% of Tory voters have already switched...
|
|
clyde1998
SNP
Green (E&W) member; SNP supporter
Posts: 1,765
|
Post by clyde1998 on May 24, 2022 0:16:05 GMT
Interesting. The implication is that the Tories are near their rock-bottom support level. I find that plausible.
31% for the Tories is exceptionally low, suggesting Yougov have found a sample where up to 10% of Tory voters have already switched... If all the Conservative voters who consider switching did switch (in the proportion of the parties they would consider voting for), it would be roughly Lab 41, Con 26, LDm 12, Grn 9, SNP 5 (~54 in Scotland), RUK 4, PC 1 (~21 in Wales), Oth 3. The implied range for each party was: Lab 24-46, Con 26-36, LDm 5-21, Grn 4-18, SNP 3-6 (~39-70 in Scotland), RUK 2-5, PC 1-2 (~12-44 in Wales) assuming each party runs a candidates in every seat. What's interesting is the Conservatives have a much smaller range than anyone else (except Reform); that's likely due to left-of-centre voters having more left-of-centre alternatives.
For fun, if everyone who's considering switching did switch: Lab 31, Con 29, LDm 14, Grn 11, SNP 5 (~53 in Scotland), RUK 3, PC 2 (~32 in Wales), Oth 5. Notes:- The approximations for the SNP in Scotland and Plaid in Wales are based on the 2019 turnout figures. Additionally, they'll be subject to a significantly higher inaccuracy than national figures due to rounding of small numbers and sub-national breaks by party not being available.
- This is based on the data in the poll, the true range of each party is probably broader.
- The poll didn't (appear to) ask how those who were presently undecided who they may consider voting for - which seems a big oversight - so the ranges are likely more favourable to the Conservatives, due to 2019 voters of right-of-centre parties being more likely to not give an initial VI in the poll than left-of-centre ones.
- My calculation of the 'headline' figures in the poll was: Lab 38, Con 32, LDm 10, Grn 8, SNP 5 (~53 in Scotland), RUK 3, PC 1 (~18 in Wales), Oth 2.
|
|
cogload
Lib Dem
I jumped in the river and what did I see...
Posts: 7,950
|
YouGov
May 28, 2022 4:56:01 GMT
via mobile
Post by cogload on May 28, 2022 4:56:01 GMT
|
|
The Bishop
Labour
Down With Factionalism!
Posts: 36,309
|
Post by The Bishop on May 28, 2022 10:16:46 GMT
Hmmm, if polling was going to do that it would arguably have done so at the start of the year.
|
|
clyde1998
SNP
Green (E&W) member; SNP supporter
Posts: 1,765
|
Post by clyde1998 on May 28, 2022 10:51:09 GMT
This survey could take the number of letters to 54. I 'love' the article proclaims "Boris Johnson set to lose seat", explaining Labour have a 5pp lead, whilst the article goes on to say "What’s more, no fewer than 25 battleground constituencies are currently forecasted to be won by a margin of less than five percentage points – well inside what we would consider a ‘margin of error’ for an exercise of this nature." And pollsters wonder why people don't trust their results - if the primary observation in a set of data is within the margin of error, it's nowhere near as much of a certainty as you're making out. I'm not a qualified statistician, but describing a result that's within the margin of error as a certainty is actively attempting to mislead people who don't know any better! I'd, obviously, be delighted if the Conservative PM lost their seat at the next election, but this article is being disingenuous. The margin of error for an individual constituency is going to be significant - probably north of 10pp - due to the number of factors that cannot be accurately predicted at a localised level: What is there's a local issue different to the national issue? What is there's a really good/bad local candidate? What if the constituency is heavily targeted? We don't know how good the methodology is either - the amount of times we (on this forum) have seen these MRPs, looked at the Scottish results (as an example) and thought this is unrealistic is basically every MRP analysis. Why? Because we've identified something missing: independence preference in the case of Scottish results, perhaps Welsh language knowledge in Wales, type of constituency (rural/urban, lead parties, etc). MRPs currently can work fairly well when there are only two options, but need serious scrutiny when it comes to multiple options. That said, they're probably more accurate than uniform swing, but they're still subject to huge amount of error. I think MRP models are interesting and should be explored, however I feel we're far from being able to use them in the same way as traditional methods or using them as unquestionable gospel in the way the article is.
|
|
jamie
Top Poster
Posts: 6,809
|
YouGov
May 28, 2022 11:48:25 GMT
Post by jamie on May 28, 2022 11:48:25 GMT
Ynys Mon is apparently a Plaid Cymru gain by 21%. As you do.
|
|
|
YouGov
May 28, 2022 12:58:52 GMT
via mobile
Post by aargauer on May 28, 2022 12:58:52 GMT
Bury north looks dubious in view of the locals.
|
|
Sibboleth
Labour
'Sit on my finger, sing in my ear, O littleblood.'
Posts: 15,233
Member is Online
|
Post by Sibboleth on May 28, 2022 14:08:47 GMT
MRP is basically UNS squeezed through a complex model on a constituency-by-constituency basis that will often reflect assumptions (which is not exactly pejorative: these assumptions may well be right! But they are assumptions) as much as facts and will always have bits that are... screwy. The effect is also often to flatten out both results overall and movement. Not uninteresting, then, but not revelatory either. The commentary attached to the figures does the usual thing of grossly over-stressing statistically insignificant differences (from a model that tends to... oh) and is probably best ignored. Still, none of this is going to be nice reading if you're a Conservative MP sitting on anything other than a very large majority.
|
|
|
Post by batman on May 28, 2022 14:16:31 GMT
Bury north looks dubious in view of the locals. logically you appear to be right, but there are many constituencies where Labour tends to do better in local elections than general ones, but plenty which are the other way round. If Labour does well enough to remove the Tories from power, it's straining credulity that the latter could hold on to Bury North. On the basis of local elections, Labour should do better in East Worthing & Shoreham than Bury North, but it's fairly doubtful that that would happen in a general election.
|
|
|
Post by batman on May 28, 2022 14:18:38 GMT
Ynys Mon is a devil to forecast. Not since 1951 has an incumbent lost there, but never since then has a retiring MP been able to hand the seat on to a colleague from the same party. It would however be surprising if Virginia Crosbie managed to hold it, assuming no boundary changes, in current circumstances, or even anything approaching them. Would Albert Owen have lost it had he stood again - he was only 60? Probably not.
|
|
johng
Labour
Posts: 4,484
|
Post by johng on May 28, 2022 14:39:30 GMT
Ynys Mon is a devil to forecast. Not since 1951 has an incumbent lost there, but never since then has a retiring MP been able to hand the seat on to a colleague from the same party. It would however be surprising if Virginia Crosbie managed to hold it, assuming no boundary changes, in current circumstances, or even anything approaching them. Would Albert Owen have lost it had he stood again - he was only 60? Probably not. Definitely. The seat is totally unpredictable and results often don't make sense. I'm not sure any MRP poll will get it right outside of luck.
I also think it is very unlikely that Crosbie will win reelection.
The Tories stood in every ward in May and didn't come close to winning a single seat on the council. Though you never know.
Mind you, Plaid is totally unpredictable there. They got thousands more votes for their candidate in the 2021 Senedd election than Crosbie got in 2019.
|
|
Merseymike
Independent
Posts: 38,978
Member is Online
|
Post by Merseymike on May 28, 2022 15:58:52 GMT
Bury north looks dubious in view of the locals. logically you appear to be right, but there are many constituencies where Labour tends to do better in local elections than general ones, but plenty which are the other way round. If Labour does well enough to remove the Tories from power, it's straining credulity that the latter could hold on to Bury North. On the basis of local elections, Labour should do better in East Worthing & Shoreham than Bury North, but it's fairly doubtful that that would happen in a general election. And Labour would not win Wirral West and not always Wirral South if local election results indicated national outcomes. When Bill Esterson first won Sefton Central there wasn't a single Labour councillor left. They now hold all but five and two of those were elected as Labour. This year they won every ward.
|
|
|
YouGov
May 28, 2022 16:11:28 GMT
Post by batman on May 28, 2022 16:11:28 GMT
yes indeed. I remember the many predictions that Labour couldn't win in Sefton Central in 2010.
|
|
|
YouGov
May 28, 2022 16:23:12 GMT
via mobile
Post by mattbewilson on May 28, 2022 16:23:12 GMT
There are places that labour did have MPs where there were no labour councillors buy usually they were councillors from another opposition party than the Tories.
P&S looks about right. Good results in Stocksbridge, Ecclesfield and Penistone for Labour this year. But will be close if we do win imo
|
|
Sibboleth
Labour
'Sit on my finger, sing in my ear, O littleblood.'
Posts: 15,233
Member is Online
|
Post by Sibboleth on May 28, 2022 16:32:40 GMT
It's an interesting issue isn't it: General Elections are categorically not applied local elections on a higher turnout, but equally it isn't as if what they have to say is entirely meaningless compared to extrapolations from national data. Working out the balance is part of the appeal of following elections.
|
|
iang
Lib Dem
Posts: 1,453
|
Post by iang on May 29, 2022 6:56:37 GMT
It's always been an issue for Lib Dems. The temptation to get carried away by "winning" parliamentary seats on the basis of local elections, which ignores differential voting and lower turnout. Yardley would be a case in point. We "won" it a couple of weeks ago, but it would be pretty unlikely to come from 8% in the General and win it
|
|
The Bishop
Labour
Down With Factionalism!
Posts: 36,309
|
Post by The Bishop on May 29, 2022 8:45:41 GMT
More to the point perhaps, you "won" it at local level for almost two decades before johnhemming finally triumphed in 2005.
|
|
iang
Lib Dem
Posts: 1,453
|
YouGov
May 29, 2022 8:54:59 GMT
Post by iang on May 29, 2022 8:54:59 GMT
More to the point perhaps, you "won" it at local level for almost two decades before johnhemming finally triumphed in 2005. Yes, and unlike some of our other Labour facing former seats, we still dominate at local council level (7 out of 10 seats and two of those that we didn't win are wards that don't fall entirely within the seat boundaries - because of the rewarding for 2018, at the moment the ward boundaries in Birmingham don't match the Parliamentary ones). So there is much more scope in the medium to long term for us to be competitive again at Parliamentary levels, especially as Brexit loses salience. But it's a leap from that to "we are ahead in the locals so we are going to carry it next time"
|
|