Nov 2019
11:31am, 5 Nov 2019
9,390 posts
|
larkim
I may be accused of being trite in drawing this analogy. But Rees-Mogg's comments this morning are equally transposed into a Brexit context:- Original quote "And I think if either of us were in a fire, whatever the fire brigade said, we would leave the burning building. It just seems the common sense thing to do." Analogy "And I think if we were in the EU, whatever the (economic / social) experts said, we would want leave the EU. It just seems the common sense thing to do." |
Nov 2019
11:44am, 5 Nov 2019
22,869 posts
|
Johnny Blaze
If the 1st ref is anything to go by, "Labour" won't be campaigning for anything. There will be 2 campaigns - leave and remain. Most Labour big hitters will campaign for Remain, I would guess, unless Corbyn gets a BINO deal. That will be entirely up to the EU. Corbyn will doubtless campaign for "his deal" - whatever it is. The chances of him whipping his MPs to campaign for his deal would appear to be negligible - and the Tories and Farage will refuse to campaign for anything. In other words it will be a huge mess. |
Nov 2019
11:57am, 5 Nov 2019
5,815 posts
|
jda
Messier the better. That’s what I am banking on.
|
Nov 2019
12:01pm, 5 Nov 2019
15,475 posts
|
Chrisull
simbil - I know what they're saying/hedging, but the EU have already said they won't negotiate on those terms, and I think Labour are only not saying "they'd campaign to leave on their deal" so as not to alert their remain supporters. But separately I saw Chris Prosser's thread , it made total sense to me and I agreed with it. Then I see this: Westminster voting intention: CON: 38% (-1) LAB: 25% (-2) LDEM: 16% (-) BREX: 11% (+4) via @YouGov Chgs. w/ 01 Nov And the bit that isn't taken into account is that there are other synergies also -> such as Undecideds to decideds, and decideds to undecided, or smaller party movements such as UKIP -> Brexit, or Lab -> Green. So he can be correct - BUT there can be synergies, such as in the 2015 election, where the Brexit Party could stand, but because of the collapse or movement of other voters to other parties, it could be worse for Labour. Such as a Labour voter thinking, Brexit Party are standing therefore, no risk of Labour not getting in, I can vote Green now. I'm not sure Prosser is factoring that in. In 2015, the Lib Dem vote collapsed and the Tories did better where UKIP were standing. That's what makes me nervous. There are more interconnected things at play than the "obvious". |
Nov 2019
12:21pm, 5 Nov 2019
15,480 posts
|
Chrisull
And bang on queue, new Statesman journo with an example of said: "The thing that would worry me were I a pollster (or a Liberal Democrat) is this: 1) BAME voters more likely to vote Remain 2) BAME voters most loyal bit of Labour vote 3) BAME voters undersampled by polls 4) Could level of Lab to LD defection among Remainers be overestimated?" In which case, in some constituencies where Lib Dem's are coming through (such as Portsmouth South), the Lab vote might be higher,and then it means they just split the Remain vote. |
Nov 2019
3:54pm, 5 Nov 2019
9,223 posts
|
rf_fozzy
If, like Chris, you are worried about the polling, then I point you in the direction of this article and how polling is used to do seat calculations. medium.com The point about the difference between YouGov and Ashcroft predictions for the 2017 election is particularly pertinent. Also a point not included in this article, but that may become important is pollster herding. |
Nov 2019
4:05pm, 5 Nov 2019
22,870 posts
|
Johnny Blaze
I know a bit about regression analysis and one of the key things to remember when building a model is "Bullshit in = bullshit out". And any model is out of date as soon as it is built. Some models might be built on bullshit and others might be more sou dly founded. So, I agree.
|
Nov 2019
4:10pm, 5 Nov 2019
33,061 posts
|
HappyG(rrr)
One of the polling companies recently said that where in the 70s polling was given a tolerance of +/- 1-2 percentage points, in the 90s it was +/- 5-7 percentage point that now they would say it was +/- 15-20 points. Which is next to useless of course. Sorry, that wasn't the polling company, it was an academic who had assessed the accuracy of polling over time.
|
Nov 2019
4:12pm, 5 Nov 2019
22,871 posts
|
Johnny Blaze
The polls do seem to be all over the shop. It strikes me they may be frantically trying to get something more accurate but in the process may be making things worse. The volatility of the political scene can't be helping, particularly with a 4 way split and tactical voting.
|
Nov 2019
4:43pm, 5 Nov 2019
9,400 posts
|
larkim
it does amaze me how small a sample can be statistically valid to give a decent degree of confidence in the outcome. The basic maths of it is that for 60m people, to achieve a 95% confidence level and a +/-3 percent confidence interval, you need just over 1000 samples. The issue polling companies have is obtaining statstically valid samples of 1000 individuals to participate as individuals tend to self select - e.g. perhaps a non-English speaking voter would not answer a call, low wage households may not have a PC to complete surveys, younger people might not take phone calls etc etc. If you can't control for this then your 1000 sample will be biased. And just increasing the sample size doesn't help as some of these factors become entrenched and skew the results. If they could genuinely get 1000 random individuals from across the country to reply honestly to their polling questions they'd get a reliable national share of the vote picture. And of course, that's all before individual constituency results are factored in. |
Useful Links
FE accepts no responsibility for external links. Or anything, really.Related Threads
- Fantasy General Election Jul 2024
- EU Referendum - In or Out? Vote here Aug 2018
- March to Parliament Against Brexit - Sat 2nd July Jun 2016
- EU Referendum Feb 2016
- Ads on Fetch - anyone else getting Leave and Remain?! Feb 2017
- The Environment Thread :-) Oct 2024
- Economics Aug 2023
- Dear Scottish Fetchies Jan 2023
- Any economists out there - question Oct 2022
- Power and exploitation - please check my sanity Oct 2018