New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
My EMA averages are Con/Lab/LD/BXP/Green 40.4%/29.0%/14.9%/7.0%/3.1%
The effect of the slight tick up in Labour share is 320/227/32/0/1
[Wait for incoming howls of ridicule]
EDIT: This includes effect of BXP not standing in a number of non-Tory seats.
I think the problem you and anyone else brave enough to put their name to some numbers (I’m not) will have is that some of what’s in there will be sensitive to quite minor movements. If I were you I might express the result as a range by way of “showing your working”.
I for one am extremely grateful for the data you’ve given up for free that I can look through easily and take a view on.
Another poll...another one showing an extended Tory lead. Only 1 of all the polls over the weekend didn't show Tories extending their lead (and that was showing same lead). And that is after an absolutely shit week for the Tories in the media.
But Brexit pulling half their candidates is the dominant factor in that. Represents an increase in Tory vote share in seats they already hold. Not where they need to make gains.
Yes, I'm clutching at straws...
No you're not. It accounts for 2-4% of the increase in Tory share but no increase in Tory seats. (I know it wards off challenges from LDs in about four seats).
The LibDem share is reducing as I predicted because of the targeting. It is good news for remainers as it will be reducing dramatically in non-target seats where LDs will be squeezed by Labour (avoiding a split remainer vote) and increasing dramatically in LD target seats (with a Labour squeeze).
As there are less than 100 LD target seats, this successful strategy leads to a reduction in overall LD share but an increase in number of LD seats.
"Though Farron has been pursuing an 11th-hour strategy to persuade Labour voters to back him in Tory-Lib Dem marginals, Vauxhall is among a small number of Labour-held seats still in play for the Lib Dems. Leaked internal data indicates the party is neck-and-neck in the south-London seat, where it came fourth in 2015, after a backlash against the Labour leave supporter Kate Hoey, who has a 12,000 majority. Vauxhall had one of the highest remain votes in the country – 79%. Local Lib Dems have plastered the constituency with images of Hoey campaigning with the former Ukip leader Nigel Farage. The canvassing data from the final week of the campaign puts the two parties neck-and-neck, with the Lib Dems one point ahead, showing 35% voting for Hoey and 36% for the Lib Dem candidate, George Turner. "
The final result was Hoey 57.3%, LD 20% Majority over 20,000.
Those LibDem neck and neck's don't half require some neck.....
Another poll...another one showing an extended Tory lead. Only 1 of all the polls over the weekend didn't show Tories extending their lead (and that was showing same lead). And that is after an absolutely shit week for the Tories in the media.
But Brexit pulling half their candidates is the dominant factor in that. Represents an increase in Tory vote share in seats they already hold. Not where they need to make gains.
Yes, I'm clutching at straws...
No you're not. It accounts for 2-4% of the increase in Tory share but no increase in Tory seats. (I know it wards off challenges from LDs in about four seats).
The LibDem share is reducing as I predicted because of the targeting. It is good news for remainers as it will be reducing dramatically in non-target seats where LDs will be squeezed by Labour (avoiding a split remainer vote) and increasing dramatically in LD target seats (with a Labour squeeze).
As there are less than 100 LD target seats, this successful strategy leads to a reduction in overall LD share but an increase in number of LD seats.
"Though Farron has been pursuing an 11th-hour strategy to persuade Labour voters to back him in Tory-Lib Dem marginals, Vauxhall is among a small number of Labour-held seats still in play for the Lib Dems. Leaked internal data indicates the party is neck-and-neck in the south-London seat, where it came fourth in 2015, after a backlash against the Labour leave supporter Kate Hoey, who has a 12,000 majority. Vauxhall had one of the highest remain votes in the country – 79%. Local Lib Dems have plastered the constituency with images of Hoey campaigning with the former Ukip leader Nigel Farage. The canvassing data from the final week of the campaign puts the two parties neck-and-neck, with the Lib Dems one point ahead, showing 35% voting for Hoey and 36% for the Lib Dem candidate, George Turner. "
The final result was Hoey 57.3%, LD 20% Majority over 20,000.
Those LibDem neck and neck's don't half require some neck.....
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
Nevertheless, I think 30% is ab absolute minimum they will achieve.
26% if they are lucky
The problem for Lab is that while their vote share has recovered a bit Corbyn's leader ratings are still in the sewer (worse than Farage).
Let's not forget that just after the last election, Corbyn's net rating was 0, rising to +17 after Grenfell Tower. It's now consistently around -42 with all pollsters.
Unless Corbyn's ratings start to improve then I could see Labour underperforming their polling on the day.
My prediction, for what it's worth (very little) is that we (Lib Dems) will finish up with 14% and about 35 seats with a very clumpy vote distribution.
I am in the same ballpark. I think the drop in the LD vote share is tactical voting rather than evenly spread.
Sarah Olney says she is "Building a Brighter Future for Richmond Park and North Kington".
Quite the attention to detail, that one.
I know! I delivered about 400 leaflets with that in the headline and didn't notice it. The life span of most leaflets is the 20 seconds between picking it up off the mat and putting it in the recycling bin.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Labour were above 30 in the previous ICM. It's just one poll, which is not to say it's wrong, but most of the others differ and even with this one the Tory lead is decisive (sorry Barnesian).
Labour is 30 within MoE right in pretty much every poll?
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
Nevertheless, I think 30% is ab absolute minimum they will achieve.
26% if they are lucky
The problem for Lab is that while their vote share has recovered a bit Corbyn's leader ratings are still in the sewer (worse than Farage).
Let's not forget that just after the last election, Corbyn's net rating was 0, rising to +17 after Grenfell Tower. It's now consistently around -42 with all pollsters.
Unless Corbyn's ratings start to improve then I could see Labour underperforming their polling on the day.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
My EMA averages are Con/Lab/LD/BXP/Green 40.4%/29.0%/14.9%/7.0%/3.1%
The effect of the slight tick up in Labour share is 320/227/32/0/1
[Wait for incoming howls of ridicule]
EDIT: This includes effect of BXP not standing in a number of non-Tory seats.
May I ask how your model did in 2017?
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
My EMA averages are Con/Lab/LD/BXP/Green 40.4%/29.0%/14.9%/7.0%/3.1%
The effect of the slight tick up in Labour share is 320/227/32/0/1
[Wait for incoming howls of ridicule]
EDIT: This includes effect of BXP not standing in a number of non-Tory seats.
May I ask how your model did in 2017?
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
Please, please do. It was great fun to have AndyJS's spreadsheet in 2016 and to have spreadsheets like yours and hopefully a few others will be quite entertaining.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
My EMA averages are Con/Lab/LD/BXP/Green 40.4%/29.0%/14.9%/7.0%/3.1%
The effect of the slight tick up in Labour share is 320/227/32/0/1
[Wait for incoming howls of ridicule]
EDIT: This includes effect of BXP not standing in a number of non-Tory seats.
May I ask how your model did in 2017?
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
Is there going to be a GE NoJam prediction contest? Always good to see who is missing their pants when the tide goes out.
Please join us in Peterborough tomorrow, Tuesday 19th November, to canvass with Nigel.
With four weeks to polling day on 12th December, we need to focus our efforts to get as many votes as possible and win seats in Parliament.
This is the only way Boris and the rest of Parliament can be held to account to deliver the clean-break Brexit we voted for. We are so close, and we need your help.
Your support so far has already helped us make huge shifts in politics this year. We’ve changed the Prime Minister, we’ve shaped the political narrative, and now the ‘Establishment’ are doing everything they can to discredit us because they know we represent an end to their career politician gravy train, running at the people’s expense.
Please come and help us canvass. Every effort will make a difference.
Please join us in Peterborough tomorrow, Tuesday 19th November, to canvass with Nigel.
With four weeks to polling day on 12th December, we need to focus our efforts to get as many votes as possible and win seats in Parliament.
This is the only way Boris and the rest of Parliament can be held to account to deliver the clean-break Brexit we voted for. We are so close, and we need your help.
Your support so far has already helped us make huge shifts in politics this year. We’ve changed the Prime Minister, we’ve shaped the political narrative, and now the ‘Establishment’ are doing everything they can to discredit us because they know we represent an end to their career politician gravy train, running at the people’s expense.
Please come and help us canvass. Every effort will make a difference.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
My EMA averages are Con/Lab/LD/BXP/Green 40.4%/29.0%/14.9%/7.0%/3.1%
The effect of the slight tick up in Labour share is 320/227/32/0/1
[Wait for incoming howls of ridicule]
EDIT: This includes effect of BXP not standing in a number of non-Tory seats.
May I ask how your model did in 2017?
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
Please, please do. It was great fun to have AndyJS's spreadsheet in 2016 and to have spreadsheets like yours and hopefully a few others will be quite entertaining.
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
I started to rewrite my model in Python (it was in excel but I can't find it) but decided not to bother because I figured that UNS (or even the weird percentage based variant I used last time) wouldn't work with so much change.
I did think that given we've got three sets of elections on the same boundaries I could work out how much each seat tended to move over or under the UNS to help produce a better model.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
My EMA averages are Con/Lab/LD/BXP/Green 40.4%/29.0%/14.9%/7.0%/3.1%
The effect of the slight tick up in Labour share is 320/227/32/0/1
[Wait for incoming howls of ridicule]
EDIT: This includes effect of BXP not standing in a number of non-Tory seats.
May I ask how your model did in 2017?
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
Please, please do. It was great fun to have AndyJS's spreadsheet in 2016 and to have spreadsheets like yours and hopefully a few others will be quite entertaining.
Hopefully it will be profitable as well!
Indeed, I only used AndyJS's spreadsheet for fun but I know a lot of people made money from it.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
Nevertheless, I think 30% is ab absolute minimum they will achieve.
26% if they are lucky
The problem for Lab is that while their vote share has recovered a bit Corbyn's leader ratings are still in the sewer (worse than Farage).
Let's not forget that just after the last election, Corbyn's net rating was 0, rising to +17 after Grenfell Tower. It's now consistently around -42 with all pollsters.
Unless Corbyn's ratings start to improve then I could see Labour underperforming their polling on the day.
I believe that Corbyn's terrible personal ratings are a fundamental bar to progress, because they reflect that many people will just not countencance lending him their vote. He was always a Marxist - people could forgive that in 2017, because, well, he was never going to get near enough to power to implement it.
But in the last 30 months he has shown himself to be utterly ambivalent about Brexit. And the anti-semitism that was not an issue in 2017 is absolutely forefront to those being asked to put it to one side. They are being asked to defer their buyers' remorse until the Jewish people start leaving Corbyn's Britain. That is too high a price for many to ever pay.
Please join us in Peterborough tomorrow, Tuesday 19th November, to canvass with Nigel.
With four weeks to polling day on 12th December, we need to focus our efforts to get as many votes as possible and win seats in Parliament.
This is the only way Boris and the rest of Parliament can be held to account to deliver the clean-break Brexit we voted for. We are so close, and we need your help.
Your support so far has already helped us make huge shifts in politics this year. We’ve changed the Prime Minister, we’ve shaped the political narrative, and now the ‘Establishment’ are doing everything they can to discredit us because they know we represent an end to their career politician gravy train, running at the people’s expense.
Please come and help us canvass. Every effort will make a difference.
Trump is restrained by Constitution and Congress, Britain Trump with a majority has no such restraints. We have seen this autumn how contemptuous Johnson is of Parliament.
Anyone got any insights or hunches regarding Hartlepool consitituency?
Three-way marginal - very tricky to call. Tice for BXP.
I`m on Labour at 11/10 at the moment but may have this wrong.
Are there any sites giving prices on it?
I would go on the Tories there. BXP less than 15% everywhere.
I wouldn’t mind a bet on Labour 3rd there if I could get it. The noises locally say they are not at all optimistic. But yes, I agree about tories favourites
"Some interesting tables from this poll. - Labour constituencies with majorities under 10%: Lab 44% Tory 33% - Tory constituencies with majorities under 10%: Tory 51% Lab 25% Suggests the Tories are piling up votes in seats they already hold while Labour are hanging on in their seats."
Please join us in Peterborough tomorrow, Tuesday 19th November, to canvass with Nigel.
With four weeks to polling day on 12th December, we need to focus our efforts to get as many votes as possible and win seats in Parliament.
This is the only way Boris and the rest of Parliament can be held to account to deliver the clean-break Brexit we voted for. We are so close, and we need your help.
Your support so far has already helped us make huge shifts in politics this year. We’ve changed the Prime Minister, we’ve shaped the political narrative, and now the ‘Establishment’ are doing everything they can to discredit us because they know we represent an end to their career politician gravy train, running at the people’s expense.
Please come and help us canvass. Every effort will make a difference.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
My EMA averages are Con/Lab/LD/BXP/Green 40.4%/29.0%/14.9%/7.0%/3.1%
The effect of the slight tick up in Labour share is 320/227/32/0/1
[Wait for incoming howls of ridicule]
EDIT: This includes effect of BXP not standing in a number of non-Tory seats.
May I ask how your model did in 2017?
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
Can't you pop some 2017 numbers into it and see how it goes?
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
Nevertheless, I think 30% is ab absolute minimum they will achieve.
26% if they are lucky
The problem for Lab is that while their vote share has recovered a bit Corbyn's leader ratings are still in the sewer (worse than Farage).
Let's not forget that just after the last election, Corbyn's net rating was 0, rising to +17 after Grenfell Tower. It's now consistently around -42 with all pollsters.
Unless Corbyn's ratings start to improve then I could see Labour underperforming their polling on the day.
I believe that Corbyn's terrible personal ratings are a fundamental bar to progress, because they reflect that many people will just not countencance lending him their vote. He was always a Marxist - people could forgive that in 2017, because, well, he was never going to get near enough to power to implement it.
But in the last 30 months he has shown himself to be utterly ambivalent about Brexit. And the anti-semitism that was not an issue in 2017 is absolutely forefront to those being asked to put it to one side. They are being asked to defer their buyers' remorse until the Jewish people start leaving Corbyn's Britain. That is too high a price for many to ever pay.
One of the reasons I have always hated Corbyn - please note, very far from the only reason - is that as far back as 2015 I knew he was an associate of neo-Nazis including Holocaust deniers, and that he had lied about it. Is he personally racist? Don’t know. The evidence suggests he is - certainly he has made racist remarks that may have been very tasteless and stupid attempts at humour - but it’s not conclusive.
What has changed is that it has become so painfully obvious that there is a problem here nobody can ignore it any more. They can either chose to condemn it, or to excuse it.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
Nevertheless, I think 30% is ab absolute minimum they will achieve.
26% if they are lucky
The problem for Lab is that while their vote share has recovered a bit Corbyn's leader ratings are still in the sewer (worse than Farage).
Let's not forget that just after the last election, Corbyn's net rating was 0, rising to +17 after Grenfell Tower. It's now consistently around -42 with all pollsters.
Unless Corbyn's ratings start to improve then I could see Labour underperforming their polling on the day.
I believe that Corbyn's terrible personal ratings are a fundamental bar to progress, because they reflect that many people will just not countencance lending him their vote. He was always a Marxist - people could forgive that in 2017, because, well, he was never going to get near enough to power to implement it.
But in the last 30 months he has shown himself to be utterly ambivalent about Brexit. And the anti-semitism that was not an issue in 2017 is absolutely forefront to those being asked to put it to one side. They are being asked to defer their buyers' remorse until the Jewish people start leaving Corbyn's Britain. That is too high a price for many to ever pay.
One of the reasons I have always hated Corbyn - please note, very far from the only reason - is that as far back as 2015 I knew he was an associate of neo-Nazis including Holocaust deniers, and that he had lied about it. Is he personally racist? Don’t know. The evidence suggests he is - certainly he has made racist remarks that may have been very tasteless and stupid attempts at humour - but it’s not conclusive.
What has changed is that it has become so painfully obvious that there is a problem here nobody can ignore it any more. They can either chose to condemn it, or to excuse it.
Either he's racist or British Jews need to change to understand British irony.
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
I started to rewrite my model in Python (it was in excel but I can't find it) but decided not to bother because I figured that UNS (or even the weird percentage based variant I used last time) wouldn't work with so much change.
I did think that given we've got three sets of elections on the same boundaries I could work out how much each seat tended to move over or under the UNS to help produce a better model.
I might need to start looking at it again.
I think UNS is fine for small changes in share. It's hopeless for large changes and clumping effects.
I have an additive model (UNS) and a multiplicative model.
If LibDem national share goes from say 7% to 14% then the additive model adds 7% to the share in every constituency and the multiplicative model increases it by 100%.
So Constituency A goes from 2% to 9% with additive model and 2% to 4% with multiplicative. Constituency B goes from 20% to 27% with additive model and 20% to 40% with multiplicative.
The multiplicative model I think better reflects clumping i.e. rewarding strong seats and not weak seats.
My main model is now 75% additive and 25% multiplicative but I'm gradually increasing the multiplicative weighting.
Additive or multiplicative makes little difference to the Tory vote projection because the current share is near the 2017 share. It makes a big difference to the Labour and LD projections.
If I make the multiplicative weighting 100% it leaves the Tory seats unchanged but Labour drop another 20 seats and LDs gain another 20. But it is not a straight swap. The Tories lose an extra 20 seats to the LDs and gain an extra 20 seats from Labour.
New ICM poll for @Reuters: Conservatives 42% (+3) Labour 32% (+1) Lib Dems 13% (-2) Brexit Party 5% (-3)
Labour solidly above 30 is concerning
Their polling average is still below 30%.
My EMA averages are Con/Lab/LD/BXP/Green 40.4%/29.0%/14.9%/7.0%/3.1%
The effect of the slight tick up in Labour share is 320/227/32/0/1
[Wait for incoming howls of ridicule]
EDIT: This includes effect of BXP not standing in a number of non-Tory seats.
May I ask how your model did in 2017?
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
Can't you pop some 2017 numbers into it and see how it goes?
I didn't have it then. I'd like to think it would have done rather well, but I would wouldn't I.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
I started to rewrite my model in Python (it was in excel but I can't find it) but decided not to bother because I figured that UNS (or even the weird percentage based variant I used last time) wouldn't work with so much change.
I did think that given we've got three sets of elections on the same boundaries I could work out how much each seat tended to move over or under the UNS to help produce a better model.
I might need to start looking at it again.
I think UNS is fine for small changes in share. It's hopeless for large changes and clumping effects.
I have an additive model (UNS) and a multiplicative model.
If LibDem national share goes from say 7% to 14% then the additive model adds 7% to the share in every constituency and the multiplicative model increases it by 100%.
So Constituency A goes from 2% to 9% with additive model and 2% to 4% with multiplicative. Constituency B goes from 20% to 27% with additive model and 20% to 40% with multiplicative.
The multiplicative model I think better reflects clumping i.e. rewarding strong seats and not weak seats.
My main model is now 75% additive and 25% multiplicative but I'm gradually increasing the multiplicative weighting.
Additive or multiplicative makes little difference to the Tory vote projection because the current share is near the 2017 share. It makes a big difference to the Labour and LD projections.
If I make the multiplicative weighting 100% it leaves the Tory seats unchanged but Labour drop another 20 seats and LDs gain another 20. But it is not a straight swap. The Tories lose an extra 20 seats to the LDs and gain an extra 20 seats from Labour.
Absolutely. Personally I’ll be gobsmacked if a 10% lead doesn’t translate into a majority but it will be interesting to see it Barnesian is right.
It sounds a lot, but if you take the last 40 years of results and plot lead vs majority, actual results can vary from that trendline by miles.
As an average 10% lead would get you just under 100 majority, but Blair 2001 was 90ish above that trendline, and Cameron 2010 was the same amount below ….. so it's not the best guide :-)
I have been looking at the possible demographics of the next Lib Dem parliamentary party, Assuming the 12 seats from 2017 seats are held and they make 28 specific gains there would be 17 men, 23 women and 6 BAME. This includes gains in Bermondsey, Brecon, Cambridge, Ceredigion, Cheadle, Chelsea, Cheltenham, Eastleigh, Fulham, Guildford, Hazel Grove, Hornsey, Leeds NW, Montgomery, N. Cornwall, NE Fife, Richmond, St Albans, St Ives, Hallam, South Cambridgeshire, Sutton and Cheam, Thornbury and Yate, Totnes, Wells, Wimbledon, Winchester. Of course there could be fewer but there is a definite trend away from pale and male.
I have been looking at the possible demographics of the next Lib Dem parliamentary party, Assuming the 12 seats from 2017 seats are held and they make 28 specific gains there would be 17 men, 23 women and 6 BAME. This includes gains in Bermondsey, Brecon, Cambridge, Ceredigion, Cheadle, Chelsea, Cheltenham, Eastleigh, Fulham, Guildford, Hazel Grove, Hornsey, Leeds NW, Montgomery, N. Cornwall, NE Fife, Richmond, St Albans, St Ives, Hallam, South Cambridgeshire, Sutton and Cheam, Thornbury and Yate, Totnes, Wells, Wimbledon, Winchester. Of course there could be fewer but there is a definite trend away from pale and male.
Sorry - should also have included Cities of L and W.
"Some interesting tables from this poll. - Labour constituencies with majorities under 10%: Lab 44% Tory 33% - Tory constituencies with majorities under 10%: Tory 51% Lab 25% Suggests the Tories are piling up votes in seats they already hold while Labour are hanging on in their seats."
I think Ashcroft showed Lib Dems preferring a Corbyn government by a ratio of 2 to 1 . As a normal Labour voter Swinsons constant attacks on Labour are pissing off many of my friends who are ardent Remainers .
If your red line is staying in the EU then everyone knows that will only happen with a Labour government there as at least the biggest party .
Swinson has to be careful to not go too far with her attacks on Corbyn . I said a few weeks ago the Lib Dems will be lucky to poll 12% and I stand by that .
I agree with that. Her constant attacks on Labour irritate me too. But I understand the political necessity of taking that stance to reassure Tory switchers and it certainly won't stop me voting LibDem (or indeed canvassing, delivering, telling, knocking up). She mustn't go too far that she puts off tactical Labour voters.
Her attacks on LAB and Corbyn are important in order to win over CON GE2017 remainers. This is key group.
I've tactically voted LAB for the past two elections but no more. Corbyn failure to deal with LAB's antisemitism is what has turned me.
Bumped into some jolly Lib dem canvassers tonight in Surton and Cheam..they had been down to Cheltenham at the weekend and were very bullish about their chances there..much less so in Sutton.
Anecdote Alert My "Labour till I die" Dad is voting Green. Apparently he spent a long time on the phone to a canvasser (he's a registered Labour supporter so they're allowed to phone canvass him) explaining that he was voting Green (the local Unite to Remain candidate) to protest at their Brexit policy. He's always previously said that voting Labour for general economic and social policy reasons was much more fundamental.
How could any business have him as a spokesperson after that interview. It would be commercial suicide.
Yes, how to explain the transformation of dashing RN pilot with Donny Osmond looks to seedy parasite in the space of a few years? Was it inevitable? Could things have been different if things had been different? Or was the Andrew as he became always there in the Andrew he once was? It's impossible to know. One can only speculate and there seems little point in that.
The randy andy moniker was neither a mistake nor without an understanding of his fancies. Hes just moved from 2nd in line and 'protected' to 8th in line and dispensable. The Epstein contact book and case will unmask a great many more untouchables.
I read that Charles is going to slimline things once he becomes King, with his siblings demoted to second-class royals.
How could any business have him as a spokesperson after that interview. It would be commercial suicide.
Yes, how to explain the transformation of dashing RN pilot with Donny Osmond looks to seedy parasite in the space of a few years? Was it inevitable? Could things have been different if things had been different? Or was the Andrew as he became always there in the Andrew he once was? It's impossible to know. One can only speculate and there seems little point in that.
The randy andy moniker was neither a mistake nor without an understanding of his fancies. Hes just moved from 2nd in line and 'protected' to 8th in line and dispensable. The Epstein contact book and case will unmask a great many more untouchables.
I read that Charles is going to slimline things once he becomes King, with his siblings demoted to second-class royals.
And his only child to be the only other civil lister
Comments
I for one am extremely grateful for the data you’ve given up for free that I can look through easily and take a view on.
Let's not forget that just after the last election, Corbyn's net rating was 0, rising to +17 after Grenfell Tower. It's now consistently around -42 with all pollsters.
Unless Corbyn's ratings start to improve then I could see Labour underperforming their polling on the day.
Here are the latest model predictions. For comparison I show the MRP projections and latest constituency polls (only five of them so far).
https://docs.google.com/spreadsheets/d/1yIHH_ZtcH9w9JF5e8WwYD6QuhOhlVwCO_GboafT6kfc/edit?usp=sharing
Jewish settlements no longer illegal - US
https://www.bbc.co.uk/news/world-middle-east-50468025
A phrase about fools rush in...
His concern is the trend and the actual lead on polling day.
Although this rather uncharacteristically tetchy post from you is rather telling.
I'll have it open in the early hours of Dec 13th so I can monitor how the early results correspond to the model predictions. Later on the 13th I'll either have egg on my face or a big smirk.
I might even publish the latest version here on the 12th.
Please join us in Peterborough tomorrow, Tuesday 19th November, to canvass with Nigel.
With four weeks to polling day on 12th December, we need to focus our efforts to get as many votes as possible and win seats in Parliament.
This is the only way Boris and the rest of Parliament can be held to account to deliver the clean-break Brexit we voted for. We are so close, and we need your help.
Your support so far has already helped us make huge shifts in politics this year. We’ve changed the Prime Minister, we’ve shaped the political narrative, and now the ‘Establishment’ are doing everything they can to discredit us because they know we represent an end to their career politician gravy train, running at the people’s expense.
Please come and help us canvass. Every effort will make a difference.
I did think that given we've got three sets of elections on the same boundaries I could work out how much each seat tended to move over or under the UNS to help produce a better model.
I might need to start looking at it again.
But in the last 30 months he has shown himself to be utterly ambivalent about Brexit. And the anti-semitism that was not an issue in 2017 is absolutely forefront to those being asked to put it to one side. They are being asked to defer their buyers' remorse until the Jewish people start leaving Corbyn's Britain. That is too high a price for many to ever pay.
"Some interesting tables from this poll.
- Labour constituencies with majorities under 10%: Lab 44% Tory 33%
- Tory constituencies with majorities under 10%: Tory 51% Lab 25%
Suggests the Tories are piling up votes in seats they already hold while Labour are hanging on in their seats."
I had a look, I can't see anything like this
What has changed is that it has become so painfully obvious that there is a problem here nobody can ignore it any more. They can either chose to condemn it, or to excuse it.
I have an additive model (UNS) and a multiplicative model.
If LibDem national share goes from say 7% to 14% then the additive model adds 7% to the share in every constituency and the multiplicative model increases it by 100%.
So Constituency A goes from 2% to 9% with additive model and 2% to 4% with multiplicative.
Constituency B goes from 20% to 27% with additive model and 20% to 40% with multiplicative.
The multiplicative model I think better reflects clumping i.e. rewarding strong seats and not weak seats.
My main model is now 75% additive and 25% multiplicative but I'm gradually increasing the multiplicative weighting.
Additive or multiplicative makes little difference to the Tory vote projection because the current share is near the 2017 share. It makes a big difference to the Labour and LD projections.
If I make the multiplicative weighting 100% it leaves the Tory seats unchanged but Labour drop another 20 seats and LDs gain another 20. But it is not a straight swap. The Tories lose an extra 20 seats to the LDs and gain an extra 20 seats from Labour.
Sorry that's a bit technical.
As an average 10% lead would get you just under 100 majority, but Blair 2001 was 90ish above that trendline, and Cameron 2010 was the same amount below ….. so it's not the best guide :-)
This includes gains in Bermondsey, Brecon, Cambridge, Ceredigion, Cheadle, Chelsea, Cheltenham, Eastleigh, Fulham, Guildford, Hazel Grove, Hornsey, Leeds NW, Montgomery, N. Cornwall, NE Fife, Richmond, St Albans, St Ives, Hallam, South Cambridgeshire, Sutton and Cheam, Thornbury and Yate, Totnes, Wells, Wimbledon, Winchester. Of course there could be fewer but there is a definite trend away from pale and male.
My "Labour till I die" Dad is voting Green. Apparently he spent a long time on the phone to a canvasser (he's a registered Labour supporter so they're allowed to phone canvass him) explaining that he was voting Green (the local Unite to Remain candidate) to protest at their Brexit policy. He's always previously said that voting Labour for general economic and social policy reasons was much more fundamental.
Labour are so fucked.