|
|
Post by bjornhattan on May 26, 2024 8:38:08 GMT
I want to know how they're doing a regression model with a sample size of (presumably) one. The whole point of regression is you need a lot of data; it's something you might do after fifty constituencies declare on election night to see if the swing is bigger in more strongly Tory seats, not something you do when you have no information at all! It could be that they're using census data and previous demographic voting patterns to attempt to improve projections of national swings on to individual seats, which is potentially useful but seeing the methodology would help to determine that. Or, yes, it could be and probably is complete junk. But even that's difficult enough - some areas where census figures diverge from results will likely shift to more what we'd expect, while others won't. For instance, deprived seaside towns like Weston super Mare are less Labour voting than you'd expect and I wouldn't be surprised if the swing there was particularly large, whereas Merseyside is much more Labour voting than it should be based purely on demographics but that divergence will almost certainly remain. |
|
The Bishop
Labour
Down With Factionalism!
Posts: 38,916
Member is Online
|
Post by The Bishop on May 26, 2024 11:53:12 GMT
I can’t see Labour losing this at a time they are in the ascendancy. In 5 or 10 years time maybe. Greens got an absolutely crushing result here in the local elections, though. Genuinely up in the air at the moment, I would say. |
|
|
|
Post by noorderling on May 26, 2024 12:02:38 GMT
I can’t see Labour losing this at a time they are in the ascendancy. In 5 or 10 years time maybe. Greens got an absolutely crushing result here in the local elections, though. Genuinely up in the air at the moment, I would say. Lib Dems used to score up to 70% in some wards in Liverpool and held a thumping majority on the council in the 2000s, but never got near to winning a Westminster seat in the city in that period, even though the party was riding high in the national polls and took seats from Labour elsewhere. Other example Watford, in local elections absolutely safe Lib Dem, including the mayor, but Lib Dems in 3rd in the Westminster election. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 26, 2024 12:08:35 GMT
Greens got an absolutely crushing result here in the local elections, though. Genuinely up in the air at the moment, I would say. Lib Dems used to score up to 70% in some wards in Liverpool and held a thumping majority on the council in the 2000s, but never got near to winning a Westminster seat in the city in that period, even though the party was riding high in the national polls and took seats from Labour elsewhere. Other example Watford, in local elections absolutely safe Lib Dem, including the mayor, but Lib Dems in 3rd in the Westminster election. Liverpool Wavertree is a lot more monolithically Labour (despite LDs winning all the seats there once) than Bristol Central (which as Bristol West went Lib Dem twice). What's happened, I think, is that the Greens have replaced the Lib Dems in some of these urban seats where the yellow team used to poll well before 2010. |
|
|
|
Post by noorderling on May 26, 2024 12:42:22 GMT
Lib Dems used to score up to 70% in some wards in Liverpool and held a thumping majority on the council in the 2000s, but never got near to winning a Westminster seat in the city in that period, even though the party was riding high in the national polls and took seats from Labour elsewhere. Other example Watford, in local elections absolutely safe Lib Dem, including the mayor, but Lib Dems in 3rd in the Westminster election. Liverpool Wavertree is a lot more monolithically Labour (despite LDs winning all the seats there once) than Bristol Central (which as Bristol West went Lib Dem twice). What's happened, I think, is that the Greens have replaced the Lib Dems in some of these urban seats where the yellow team used to poll well before 2010. LibDems only managed to win 8 years in to a Labour government, in the context of the Iraq war, while their national poll numbers were thrice those of the greens now and Labour’s numbers were lower. Also, the Greens have failed in similar, but admittedly less promising circumstances. Think of Sheffield Central in 2017, were they were trounced. They were also certain they were in with a shot in the Isle Of Wight. A Green victory in Bristol Central is not impossible, but imo improbable. Their chances of victory will be much higher in a future election, with Labour national popularity under pressure. By then they will of course have “run” the council for a couple of years, and history shows us that that not always ends well. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 26, 2024 12:47:19 GMT
Greens got an absolutely crushing result here in the local elections, though. Genuinely up in the air at the moment, I would say. Lib Dems used to score up to 70% in some wards in Liverpool and held a thumping majority on the council in the 2000s, but never got near to winning a Westminster seat in the city in that period, even though the party was riding high in the national polls and took seats from Labour elsewhere. Other example Watford, in local elections absolutely safe Lib Dem, including the mayor, but Lib Dems in 3rd in the Westminster election. I was having dinner with two former Bristol City councillors when the local results in Bristol were being declared and they said something very similar. The "anti council" vote that the LDs had for years has gone over to the Greens resulting in the Greens winning the wards the LDs used to. However this one seat where the Greens' national platform has genuine popular appeal too |
|
haroldthepolitician
Labour
'You can be just what you want to be, Just as long as you don't try and compete with me' - Jarvis
Posts: 216 
|
Post by haroldthepolitician on May 28, 2024 18:52:58 GMT
I want to know how they're doing a regression model with a sample size of (presumably) one. The whole point of regression is you need a lot of data; it's something you might do after fifty constituencies declare on election night to see if the swing is bigger in more strongly Tory seats, not something you do when you have no information at all! This account has been tweeting these "regression model" projections for a number of seats, but there is no detail on what methodology they are actually using. I suppose they could be using data from the local elections, but I'd be very sceptical that that would give useful results. I actually asked him, seems its a combination of demographics, national opinion polls AND local election results (though he assures me they don't have much weight except here and waveney valley) - its certainly bold to rely on local elections even in these circumstance, i'd have thought these are the worst 2 cases to use them on. Though my vibe would be Green - 45% Labour - 43% Liberal Democrat - 5% Reform - 4% Conservative - 3% Could well be the conservatives worst performance if im being honest. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 28, 2024 19:10:07 GMT
I guess another point about regression models is that whilst overall a good model can be the best option to predict elections in the absence of actual polling data, they are known to be pretty poor if the outcome is likely to similar to only a small number of previous outcomes from which it is trained upon. Which is inevitably an issue here.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 28, 2024 19:13:43 GMT
It could be that they're using census data and previous demographic voting patterns to attempt to improve projections of national swings on to individual seats, which is potentially useful but seeing the methodology would help to determine that. Or, yes, it could be and probably is complete junk. But even that's difficult enough - some areas where census figures diverge from results will likely shift to more what we'd expect, while others won't. For instance, deprived seaside towns like Weston super Mare are less Labour voting than you'd expect and I wouldn't be surprised if the swing there was particularly large, whereas Merseyside is much more Labour voting than it should be based purely on demographics but that divergence will almost certainly remain. It's a shame 'Scouse not English' responses aren't recorded in Census national identity data. |
|
jamie
Top Poster
Posts: 7,054
|
Post by jamie on May 28, 2024 19:21:22 GMT
It's a shame 'Scouse not English' responses aren't recorded in Census national identity data. I would unironically like an accent question in the census. |
|
|
|
Post by froome on May 29, 2024 6:53:34 GMT
This account has been tweeting these "regression model" projections for a number of seats, but there is no detail on what methodology they are actually using. I suppose they could be using data from the local elections, but I'd be very sceptical that that would give useful results. I actually asked him, seems its a combination of demographics, national opinion polls AND local election results (though he assures me they don't have much weight except here and waveney valley) - its certainly bold to rely on local elections even in these circumstance, i'd have thought these are the worst 2 cases to use them on. Though my vibe would be Green - 45% Labour - 43% Liberal Democrat - 5% Reform - 4% Conservative - 3% Could well be the conservatives worst performance if im being honest. This will certainly be one of the lowest Conservative votes in the country, though I would expect them to beat Reform here, and possibly also the Lib Dems. |
|
cogload
Lib Dem
I jumped in the river and what did I see...
Posts: 9,141
|
Post by cogload on May 31, 2024 17:00:56 GMT
|
|
|
|
Post by John Chanin on May 31, 2024 17:28:55 GMT
|
|
r34t
Non-Aligned
Posts: 1,175
|
Post by r34t on May 31, 2024 17:41:42 GMT
Lib Dems used to score up to 70% in some wards in Liverpool and held a thumping majority on the council in the 2000s, but never got near to winning a Westminster seat in the city in that period, even though the party was riding high in the national polls and took seats from Labour elsewhere. Other example Watford, in local elections absolutely safe Lib Dem, including the mayor, but Lib Dems in 3rd in the Westminster election. I was having dinner with two former Bristol City councillors when the local results in Bristol were being declared and they said something very similar. The "anti council" vote that the LDs had for years has gone over to the Greens resulting in the Greens winning the wards the LDs used to. If you look at a map of Bristol wards in the early 00s & 2024 they are very similar, yellow has turned Green (the exception is the Greens taking the St George wards) The LDs did take Bristol West, slightly different boundaries but the middle class clifton, redland, bishopston areas being the core. However Thangham is a very good fit for the seat & having the local MP as minister for culture will be a big plus for many. |
|
YL
Non-Aligned
Either Labour leaning or Lib Dem leaning but not sure which
Posts: 4,907
|
Post by YL on Jun 3, 2024 17:23:33 GMT
YouGov's latest MRP has the Greens on 50% here to Labour's 37%.
|
|
|
|
Post by jakegb on Jun 3, 2024 18:09:45 GMT
Owen Jones has done a recent profile of this seat via his YouTube channel.
It highlighted the optimism that many Greens felt about taking the seat. Thangham Debbonaire - by contrast - refused Jones an interview and relied on her activists to get rid of him.
|
|
|
|
Post by batman on Jun 3, 2024 21:56:02 GMT
If I were her I would definitely not talk to him. There is no balance in what he says any more, even more so than previously.
|
|
|
|
Post by Devil Wincarnate on Jun 5, 2024 13:51:12 GMT
If I were her I would definitely not talk to him. There is no balance in what he says any more, even more so than previously. He's openly hostile to her and name checked her as someone he wanted to oust with his vanity project. She is within her rights to tell him to fuck off. |
|
|
|
Post by noorderling on Jun 5, 2024 15:41:06 GMT
If I were her I would definitely not talk to him. There is no balance in what he says any more, even more so than previously. He's openly hostile to her and name checked her as someone he wanted to oust with his vanity project. She is within her rights to tell him to fuck off. In the clip he introduces himself as a journalist, while in the past he claimed he was not a journalist but an opinionmaker, so,he did not need to be balanced. |
|
r34t
Non-Aligned
Posts: 1,175
|
Post by r34t on Jun 5, 2024 17:52:12 GMT
He's openly hostile to her and name checked her as someone he wanted to oust with his vanity project. She is within her rights to tell him to fuck off. In the clip he introduces himself as a journalist, while in the past he claimed he was not a journalist but an opinionmaker, so,he did not need to be balanced. Well he's hardly likely to introduce himself as a grifter, which would be more accurate.  |
|
