|
|
Post by Pete Whitehead on Jan 17, 2024 6:17:36 GMT
These are the official (Thrasher & Rallings) ones. I never had a complete set for the whole country
|
|
Harry Hayfield
Green
Cavalier Gentleman (as in 17th century Cavalier)
Posts: 2,922 
|
Post by Harry Hayfield on Jan 17, 2024 7:56:24 GMT
Are Labour really that non existent in North East Scotland and are the Lib Dems really that non existent in the South Wales Valleys (when it was really not that long ago they were second in a lot of them) |
|
xenon
Forum Regular
Posts: 426
|
Post by xenon on Jan 17, 2024 9:31:32 GMT
Are Labour really that non existent in North East Scotland and are the Lib Dems really that non existent in the South Wales Valleys (when it was really not that long ago they were second in a lot of them) Labour have no council seats at all in Aberdeenshire, one in Angus and two in Perthshire so there's practically no local base to speak of. Even before the decline of the past 10 years, they often struggled to break double figures in seats outside Aberdeen and Dundee. |
|
Georg Ebner
Non-Aligned
Roman romantic reactionary Catholic
Posts: 9,867
|
Post by Georg Ebner on Jan 17, 2024 9:52:47 GMT
Differences:  |
|
|
|
Post by Pete Whitehead on Jan 17, 2024 10:20:28 GMT
Differences: My notional for Wimbledon [& Coombe] is on the revised boundaries which are different to the final ones. Not sure what my figures would show for the Wimbledon [& New Malden] version Wolverhampton SE, 'my' figures are actually those provided by Wolverhampton and Walsall councils. I think my own notional ward results would have also had a Conservative lead here. I need to check these things later. |
|
The Bishop
Labour
Down With Factionalism!
Posts: 39,009
|
Post by The Bishop on Jan 17, 2024 13:43:20 GMT
Yup. It’s a mountainous challenge. It is, and Labour getting a majority next time would be an incredible achievement which certainly shouldn't be taken as much for granted as some now do. HOWEVER, this has little or nothing to do with the above factoid given that "national swing" may be a more meaningless concept than in living memory when we next vote. It hasn't been the be all and end all since the end of the postwar "two party system" half a century ago, but was still vaguely applicable as recently as the 2005 and 2010 GEs. Since then things have gone pretty haywire, and that is only likely to be more true next time. As far as Labour is concerned, a double digit rise in its share nationally since 2019 but dropping in dozens of seats (it was 12 in 1997) is distinctly possible. What we can be almost sure of given any significant anti-government swing at all, is that the Tory vote will become less efficient and Labour's majorly more so. |
|
|
|
Post by johnloony on Jan 17, 2024 18:13:27 GMT
What's with the Labour Party in northern Scotland? |
|
|
|
Post by evergreenadam on Jan 17, 2024 20:12:51 GMT
Yup. It’s a mountainous challenge. It is, and Labour getting a majority next time would be an incredible achievement which certainly shouldn't be taken as much for granted as some now do. HOWEVER, this has little or nothing to do with the above factoid given that "national swing" may be a more meaningless concept than in living memory when we next vote. It hasn't been the be all and end all since the end of the postwar "two party system" half a century ago, but was still vaguely applicable as recently as the 2005 and 2010 GEs. Since then things have gone pretty haywire, and that is only likely to be more true next time. As far as Labour is concerned, a double digit rise in its share nationally since 2019 but dropping in dozens of seats (it was 12 in 1997) is distinctly possible. What we can be almost sure of given any significant anti-government swing at all, is that the Tory vote will become less efficient and Labour's majorly more so. About 90% of the reason for the swing required is simply because Labour did so badly at the last election and start from such a low position. The boundary changes and the associated 5% rule and seat redistribution from Wales to England account for the remainder. |
|
|
|
Post by yellowfox on Jan 17, 2024 20:48:39 GMT
What's with the Labour Party in northern Scotland? Always been a very poor area for Labour. If you include Stirling and urban South Fife in the central belt, Labour only won 7 seats north of the Central Belt in 1997 (and only 2 outside of Aberdeen and Dundee cities). Developments since then mean that the Western Isles is there only plausible target seat in the region (Aberdeen North and the new Dundee Central are among the safest SNP seats in the country, Labour have fell to 4th in Inverness and Aberdeen South and the 2nd Dundee seat is getting bits of Angus, which have never voted Labour, added to it). Labour have no councillors in Aberdeenshire, 1 in Angus, 2 in Perth and Kinross, 2 in Highland and 3 in Moray so Labour's problems will be compounded by distinctly lacking a local base on which to build from (only 3 MSPs across the Highlands and North East too). |
|
nyx
Non-Aligned
Posts: 1,046
|
Post by nyx on Jan 17, 2024 21:41:10 GMT
Wolverhampton SE, 'my' figures are actually those provided by Wolverhampton and Walsall councils. I think my own notional ward results would have also had a Conservative lead here. I need to check these things later. Anyone know why Thrasher and Rallings didn't use the actual data available in Wolverhampton? It somewhat reduces the credibility of the whole set of results when we have factual evidence that one is massively out. |
|
stb12
Top Poster
Posts: 8,398
|
Post by stb12 on Jan 17, 2024 23:06:23 GMT
In the first couple of elections to the Scottish Parliament when Labour still largely dominated constituencies the North East and Highlands and Islands were the only regions where they had any list members elected due to a lack of constituency seats
|
|
Georg Ebner
Non-Aligned
Roman romantic reactionary Catholic
Posts: 9,867
|
Post by Georg Ebner on Jan 17, 2024 23:09:03 GMT
Wolverhampton SE, 'my' figures are actually those provided by Wolverhampton and Walsall councils. I think my own notional ward results would have also had a Conservative lead here. I need to check these things later. Anyone know why Thrasher and Rallings didn't use the actual data available in Wolverhampton? It somewhat reduces the credibility of the whole set of results when we have factual evidence that one is massively out. Yet it's honester than pulling a system through and then beginning to alter it wherever wardResults are known, isn't it? Enables us to detect their model's limits. |
|
nyx
Non-Aligned
Posts: 1,046
|
Post by nyx on Jan 17, 2024 23:29:22 GMT
Anyone know why Thrasher and Rallings didn't use the actual data available in Wolverhampton? It somewhat reduces the credibility of the whole set of results when we have factual evidence that one is massively out. Yet it's honester than pulling a system through and then beginning to alter it wherever wardResults are known, isn't it? Enables us to detect their model's limits. I am personally of the viewpoint that where accurate data exists, it should be used... |
|
Georg Ebner
Non-Aligned
Roman romantic reactionary Catholic
Posts: 9,867
|
Post by Georg Ebner on Jan 18, 2024 5:41:19 GMT
Yet it's honester than pulling a system through and then beginning to alter it wherever wardResults are known, isn't it? Enables us to detect their model's limits. I am personally of the viewpoint that where accurate data exists, it should be used... WardResults should be taken into acCount at the beginning, to calibrate the method; but as soon as the model is established not to patch it here and there (at least not without publishing it - but who is willing to do that?). |
|
Georg Ebner
Non-Aligned
Roman romantic reactionary Catholic
Posts: 9,867
|
Post by Georg Ebner on Jan 18, 2024 5:45:22 GMT
Differences: My notional for Wimbledon [& Coombe] is on the revised boundaries which are different to the final ones. Not sure what my figures would show for the Wimbledon [& New Malden] version Wolverhampton SE, 'my' figures are actually those provided by Wolverhampton and Walsall councils. I think my own notional ward results would have also had a Conservative lead here. I need to check these things later. Yes, as written on the top of the spreadSheet, these are Your notionals for the revised version (those for the final one i do not have). Here are the vote-numbers:  |
|
|
|
Post by hullenedge on Jan 18, 2024 7:53:37 GMT
Yet it's honester than pulling a system through and then beginning to alter it wherever wardResults are known, isn't it? Enables us to detect their model's limits. I am personally of the viewpoint that where accurate data exists, it should be used... Of course it should but there are then ripples through neighbouring seats and the model has to be tweaked. As an aside AI will likely calculate the next set of notionals. If we're fortunate AI could create 'ancient' sets of notionals. |
|
|
|
Post by bjornhattan on Jan 18, 2024 8:27:31 GMT
I am personally of the viewpoint that where accurate data exists, it should be used... Of course it should but there are then ripples through neighbouring seats and the model has to be tweaked. As an aside AI will likely calculate the next set of notionals. If we're fortunate AI could create 'ancient' sets of notionals. AI is a rather nebulous term - by a fairly broad definition the approach I use is a form of AI (it is based on Random Forest which is a fairly simple form of Machine Learning). A more sophisticated version which is able to capture localised factors (including how variables affecting results change within the UK, subregional effects such as the Merseyside effect, incumbents' personal votes, and local election results) would certainly be theoretically attainable, and I imagine this would considerably outperform both traditional notionals and demographics only models like mine. |
|
YL
Non-Aligned
Either Labour leaning or Lib Dem leaning but not sure which
Posts: 4,915
|
Post by YL on Jan 18, 2024 8:53:08 GMT
Presumably a statistical model like bjornhattan's would be able to incorporate the Walsall and Wolverhampton ward results, and might even be able to use them to give a bit of information about how similar areas (if there are any outside the Black Country!) voted. Indeed it would also give an estimate for how the Paddock ward split works out. I have added links to the Walsall and Wolverhampton figures to the first post in this thread, as well as adding the Rallings & Thrasher notionals. If anyone can point to any other exact ward figures let me know and I will add those as well. |
|
|
|
Post by carolus on Jan 18, 2024 10:18:23 GMT
Of course it should but there are then ripples through neighbouring seats and the model has to be tweaked. As an aside AI will likely calculate the next set of notionals. If we're fortunate AI could create 'ancient' sets of notionals. AI is a rather nebulous term - by a fairly broad definition the approach I use is a form of AI (it is based on Random Forest which is a fairly simple form of Machine Learning). A more sophisticated version which is able to capture localised factors ( including how variables affecting results change within the UK, subregional effects such as the Merseyside effect, incumbents' personal votes, and local election results) would certainly be theoretically attainable, and I imagine this would considerably outperform both traditional notionals and demographics only models like mine. Presumably this would mostly just require some additional feature engineering, rather than wholesale changes to your model? (I imagine that the local elections stuff would require an additional dataset). |
|
The Bishop
Labour
Down With Factionalism!
Posts: 39,009
|
Post by The Bishop on Jan 18, 2024 13:30:43 GMT
In the first couple of elections to the Scottish Parliament when Labour still largely dominated constituencies the North East and Highlands and Islands were the only regions where they had any list members elected due to a lack of constituency seats It shouldn't be forgotten that at the last GE Labour scored just 18% in Scotland - six points lower than even 2015 and their worst showing there for over a century. In those circumstances, lost deposits in areas they never won even at ScotLab's height are only to be expected really. |
|
