Nov 2020
7:00pm, 17 Nov 2020
2,067 posts
|
Cheg
Yeah a narrow one might not have done the business. Don't suppose anyone has the bookies odds on brexit the date before the referendum?
|
Nov 2020
7:00pm, 17 Nov 2020
6,364 posts
|
Dooogs
Not sure if this belongs in this thread or the trans rights thread but Suzanne Moore's left the Guardian some months after catalysing a massive internal Graun row (with a fairly crass 'gender mostly equals biological sex' article). Julie Bindel and the Spectator outraged, which is a helpful heuristic for who's in the right here... |
|
Nov 2020
7:02pm, 17 Nov 2020
2,068 posts
|
Cheg
google.co.uk
|
|
Nov 2020
7:04pm, 17 Nov 2020
6,365 posts
|
Dooogs
I remember when 538's 73% Hillary likelihood came out four years ago,thinking 'Eek, that's a bit too similar to the Brexit odds a few months back "...
|
|
Nov 2020
7:13pm, 17 Nov 2020
2,069 posts
|
Cheg
Yeah the polls have been shocking, and from the look of that article the bookies did well over Brexit.
|
Nov 2020
7:26pm, 17 Nov 2020
24,101 posts
|
Johnny Blaze
Even the Brexiteers seem quite exhausted by it all, and the guff about the brave new Brexit world seems like whistling in the dark these days. I doubt many of them actually believe it: Tories say it for career preservation. Aside from Patel, who seems positively gleeful about being the politician who is proud to deny people like her parents the ability to migrate here. I'd like to think that we will gradually move back towards the EU again to the point where joining again is a bit of a no-brainer. It may take a prolonged period of national decline though. |
Nov 2020
7:59pm, 17 Nov 2020
12,270 posts
|
rf_fozzy
"the polls have been shocking" Nope. The polls have been about as accurate as they have been historically. The one blemish was the 2016 US Pres election where the polls had a systematic error due to incorrect weighting and *very* late shift in support in the undecides towards Trump that was ignored due to poller herding, in some midwestern states (overlooked was the fact that they were well within the polling error almost everywhere else) and 2020 US Pres election in Florida where they seem to have underestimated cuban hispanic support for Trump (probably incorrect wieghting). Polls for the 2008, 2012, 2014, 2018 elections have all been within the margin of error. In the UK, the 2015, 2017, 2019 General elections have all be within the MoE (see YouGov's MRP model and how close that was) and the Brexit Referendum was well within the MoE. |
|
Nov 2020
8:12pm, 17 Nov 2020
6,366 posts
|
Dooogs
Fozzy and co: as someone who's not a pro statistician, is there anything realistic the pollsters can do to make their individual MoEs smaller? The obvious (if not necessarily feasible) answer is "make the sample sizes bigger" - or does that just magnify the potential sample selection errors? |
|
Nov 2020
8:17pm, 17 Nov 2020
2,070 posts
|
Cheg
You do this every time fozzy. They got remain wrong they got Clinton wrong. Margin of error. Whatever they were wrong. You can't be a little bit pregnant. centreonconstitutionalchange.ac.uk projects.fivethirtyeight.com |
Nov 2020
8:28pm, 17 Nov 2020
37,629 posts
|
Nellers
Predictions are really hard, especially about the future. More or Less did a bit about polling. They're in a tricky spot. Mostly they used to do random telephone polling but these days who answers the phone to a number they don't know? And what's the alternative? Put a poll on any website and it's obviously self selecting. There's got to be a new way to get a representative sample but they don't seem to have found it yet. Saying it was More or Less, might have been Prof Buzzkill actually. |
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 :-)
Jan 2025
-
Economics
Jan 2025
-
Dear Scottish Fetchies
Jan 2023
-
Any economists out there - question
Oct 2022
-
Power and exploitation - please check my sanity
Oct 2018