The 2015 exit poll was certainly questioned on the night, with the hat/kilt eating. In fact, the result was even worse for both those parties than the exit poll suggested.
The exit poll is conducted at representative constituencies in every region of the country for a reason
I think they concentrate not on representative constituencies in every region, but on representative marginals, and that the key thing they track is how stated vote on exit compares with the corresponding figures for the exit polls they ran in the same constituencies in previous elections. If my memory is correct on that, there could be a problem this time because the constituencies of greatest interest aren't necessarily the ones where they have built up a history to compare with.
For this reason, I won't be surprised if the exit poll is out by a larger margin this time than we're used to - so I'll be watching the live results as they come in* and comparing with the YouGov MRP projections to try to get an early indication.
Well live scraping the BBC website is working* (and much better scheduled to only look for likely announcements), live building and publishing the spreadsheet to google docs is as well.
Although I capture all the seats / party %'s etc, I won't be putting those in the spreadsheet. At the moment, I plan to just concentrate on Tory / Lab fights, show swing and comparison to MRP (and 2017 result).
* Just hope they don't change the holding pages too much on the night.
How about doing all the seats where the result is expected before (say) 1.30?
I will see. The problem if I start trying to include everything is I have to then have loads of extra columns for all the parties (and not all parties stand in all seats).
At the moment, the spreadsheet is quite concise, you will see the result, the swing, the YouGov prediction...that is what we want to see. Are the Tories winning vs Labour, by how much and is it in line with YouGov MRP.
The exit poll is conducted at representative constituencies in every region of the country for a reason
I think they concentrate not on representative constituencies in every region, but on representative marginals, and that the key thing they track is how stated vote on exit compares with the corresponding figures for the exit polls they ran in the same constituencies in previous elections. If my memory is correct on that, there could be a problem this time because the constituencies of greatest interest aren't necessarily the ones where they have built up a history to compare with.
For this reason, I won't be surprised if the exit poll is out by a larger margin this time than we're used to - so I'll be watching the live results as they come in* and comparing with the YouGov MRP projections to try to get an early indication.
Well live automated scraping the BBC website is working* (and much better scheduled to only look for likely announcements) plus building and publishing the spreadsheet to google docs is as well. So fingers crossed I can hit the button and sit back.
Although I capture all the seats / party %'s etc, I won't be putting those in the spreadsheet. At the moment, I plan to just concentrate on Tory / Lab fights, show swing and comparison to MRP (and 2017 result).
* Just hope they don't change the holding pages too much on the night.
Cant wait to see it!
There won't be much exciting to see. Is a bit like a swan, lots of stuff running on my pc under the water, but there should just be a simple link to a fairly straight forward google doc that will update itself automatically.
The exit poll is conducted at representative constituencies in every region of the country for a reason
I think they concentrate not on representative constituencies in every region, but on representative marginals, and that the key thing they track is how stated vote on exit compares with the corresponding figures for the exit polls they ran in the same constituencies in previous elections. If my memory is correct on that, there could be a problem this time because the constituencies of greatest interest aren't necessarily the ones where they have built up a history to compare with.
For this reason, I won't be surprised if the exit poll is out by a larger margin this time than we're used to - so I'll be watching the live results as they come in* and comparing with the YouGov MRP projections to try to get an early indication.
Well live automated scraping the BBC website is working* (and much better scheduled to only look for likely announcements) plus building and publishing the spreadsheet to google docs is as well. So fingers crossed I can hit the button and sit back.
Although I capture all the seats / party %'s etc, I won't be putting those in the spreadsheet. At the moment, I plan to just concentrate on Tory / Lab fights, show swing and comparison to MRP (and 2017 result).
* Just hope they don't change the holding pages too much on the night.
Cant wait to see it!
There won't be much exciting to see. Is a bit like a swan, lots of stuff running on my pc under the water, but there should just be a simple link to a fairly straight forward google doc that will update itself automatically.
The exit poll is conducted at representative constituencies in every region of the country for a reason
I think they concentrate not on representative constituencies in every region, but on representative marginals, and that the key thing they track is how stated vote on exit compares with the corresponding figures for the exit polls they ran in the same constituencies in previous elections. If my memory is correct on that, there could be a problem this time because the constituencies of greatest interest aren't necessarily the ones where they have built up a history to compare with.
For this reason, I won't be surprised if the exit poll is out by a larger margin this time than we're used to - so I'll be watching the live results as they come in* and comparing with the YouGov MRP projections to try to get an early indication.
Well live scraping the BBC website is working* (and much better scheduled to only look for likely announcements), live building and publishing the spreadsheet to google docs is as well.
Although I capture all the seats / party %'s etc, I won't be putting those in the spreadsheet. At the moment, I plan to just concentrate on Tory / Lab fights, show swing and comparison to MRP (and 2017 result).
* Just hope they don't change the holding pages too much on the night.
How about doing all the seats where the result is expected before (say) 1.30?
I will see. The problem if I start trying to include everything is I have to then have loads of extra columns for all the parties (and not all parties stand in all seats).
At the moment, the spreadsheet is quite concise, you will see the result, the swing, the YouGov prediction...that is what we want to see. Are the Tories winning vs Labour, by how much and is it in line with YouGov MRP.
Sounds great.
If I get time, I might do Tory vs LD seats as a separate sheet. I know there are a number of very close seats in Scotland which if it is super tight could be crucial, but again, just adding more complexity.
But my feeling is, the Tories strategy to win a majority relies on winning Labour seats in the North / Midlands. If that isn't happening they are going to be screwed either way, and I would think YouGov MRP is going to be wrong too.
Of the above just 17 seats have an increase of more than 5%.
Given increasing population that suggests no surge at all in voter registration - it's basically consistent with 2017.
What an indictment of the national press that one independent journalist has done all this work - nobody could be bothered or had even thought of trying to get the data.
Yet we had numerous articles with people getting excited about X00,000 applications on one day etc - which nobody could be bothered to follow up.
I've always been sceptical of those stats. How many people which applied were already on the register, and didn't know/remember they were?
The exit poll is conducted at representative constituencies in every region of the country for a reason
I think they concentrate not on representative constituencies in every region, but on representative marginals, and that the key thing they track is how the stated vote on exit compares with the corresponding figures for the exit polls they ran in the same constituencies in previous elections. If my memory is correct on that, there could be a problem this time because the constituencies of greatest interest aren't necessarily the ones where they have built up a history to compare with.
For this reason, I won't be surprised if the exit poll is out by a larger margin this time than we're used to - so I'll be watching the live results as they come in* and comparing with the YouGov MRP projections to try to get an early indication.
I posted last night about how BBC always report on the night before GE how the result is uncertain, nobody really knows what's going on, there are reports of strange movements in marginals.
Well the other thing that always happens is that on the final day of campaigning they always just report about where the leaders are going and follow them around - they never do much if anything on specific individual policies.
So if there is a late bribe, I think it really needs to be announced now - at the latest this evening to get into tomorrow morning's papers.
You would have thought it would have come Sat / Sun.
It would look soooooo desperate now... but of course it's still possible.
On the other hand, such a public acknowledgement that they're about to be creamed could break the back of their campaign entirely.
I have to say I have found the Labour advent calendar campaign choices slightly confusing. If you are going radical, why not go really populist, with legalize drugs, cancel student debt, overhaul tv sports rights. Planting 2bn trees ain't going to shift a vote.
Strange, isn't it? I feel like I could have written a better manifesto for them, despite not wanting 99% of the things they want.
I’ve wondered whether stuff like legalisation of drugs was vetoed by those wanting to hold the more socially conservative north. They just then forgot to put something appealing in instead.
Overhauling tv sports rights I am sure would have been popular in the North and cost the government nothing. Now it would screw the sports themselves, but bugger that they are all multi-millionaires anyway, right?
I'm not sure it would be that popular. Tell Liverpool fans their club could no longer be able to afford Salah, Firmino, Mane and Van Dijk and see how truly popular that is. Liverpool will vote Labour this election but not because Liverpool fans aren't happy with LFC and the way the Premier League is going at the moment!
Liverpool will vote Labour as it's sees what The Sun says and does the exact opposite as it has done since Hillsborough.
Speaking personally I don't see what The Sun says, never have done and I'm happy with that. Never read The Sun in my life and never will.
By all accounts, Liverpool has not read The Sun since 1989 so how would its denizens know what it says?
Comments
But my feeling is, the Tories strategy to win a majority relies on winning Labour seats in the North / Midlands. If that isn't happening they are going to be screwed either way, and I would think YouGov MRP is going to be wrong too.