European Parliament Elections 2019

LuckyLloyd

MacDanger wrote: »

Not really, all bar Ming (~4.3% deviation) were within the margin of error and the average deviation for the top 9 candidates was 2%. Admittedly they seems to be well off on Cuffe but generally decent otherwise

The fact that people choose to ignore the caveat about a poll being based on a sample size and having an associated margin of error is hardly the pollster's fault.

The + / - deviation was 3%

It is not expected that you have numerous candidates at the edges of your margin of error, with some exceeding it.

Fan of Netflix

RTE spent the entire weekend talking about the results on their nonsense Euro poll, now when the real results came in they aren't even broadcasting it. #tvlicense #publicservicebroadcasting

Topgear on Dave

Fan of Netflix wrote: »

RTE spent the entire weekend talking about the results on their nonsense Euro poll, now when the real results came in they aren't even broadcasting it. #tvlicense #publicservicebroadcasting

You dont like "My kitchen rules: Australia"?

dmakc

Fan of Netflix wrote: »

RTE spent the entire weekend talking about the results on their nonsense Euro poll, now when the real results came in they aren't even broadcasting it. #tvlicense #publicservicebroadcasting

Journalists probably read the 3% margin for error as a 97% chance the exit poll stands. The success of Eamon Ryan's 50/1 bet of the century on McHugh being a highlight.

Just Saying

Based on Ming's second preference tallies McHugh will be eliminated before Casey and Smith.At this stage Casey and Smith will be close together approx 15/18 k behind Walsh and there is no indication that either will get enough transfers to close that gap especially as Ming and Carthy will still be short of a quota and therefore absorbing transfers

Eric Cartman

Is anyone doijg a love sheet by first prefs like we had for dublin ?

Eric Cartman

Just Saying wrote: »

Based on Ming's second preference tallies McHugh will be eliminated before Casey and Smith.At this stage Casey and Smith will be close together approx 15/18 k behind Walsh and there is no indication that either will get enough transfers to close that gap especially as Ming and Carthy will still be short of a quota and therefore absorbing transfers

If casey can stay in there when the 2 FF candidates get eliminated then thatll be the decider, if he overtakes walsh then its his seat to lose, if not then its game over.

Just Saying

Regarding the MoE surely it needs better explanation.Using a 3% MoE might be acceptable when a candidate receives 25% of the vote but if a candidate receives 14% having had an exit poll of 10% it is in fact miles out.

I would contend that this represents a deviation of 40% from the poll which is way too much.

I don't think the poll methodology is wrong.I think the polling companies are being told lies.This first arose in the "Shy Tory" polls in 1992 and has latterly manifested itself in the Trump election and the Brexit polls also.

It is where the respondent is afraid to admit a vote that they perceive might not be very popular and align themselves instead with what they think might be a less controversial choice.

MacDanger

LuckyLloyd wrote: »

The + / - deviation was 3%

It is not expected that you have numerous candidates at the edges of your margin of error, with some exceeding it.

It was +/- 4%: https://www.rte.ie/news/2019/0524/1051567-strong-support-for-greens-in-euro-elections-exit-poll/

Just Saying

Gavan Reilly is doing a very good spreadsheet

https://docs.google.com/spreadsheets/d/1YxOV_VKV2ujMhWhKPsQSJHa3KVlQWGPKHfVVipFbnBs/htmlview?usp=sharing&sle=true

blanch152

badtoro wrote: »

From Mings twitter

Prediction after distribution of McGuinness surplus.
Carthy 77619+735=78354
Casey56650+1173=57823
Flanagan 85034+1877=86911
Hannigan 12378+516=12894
Healy Eames 15991+422=16413
McHugh 51019+1501=52520
Rabbitte 30230+719=30949
Smith 42815+1001=43815
Walsh 64500+5835=70335.

Very hard to see beyond Flanagan, Carthy and Walsh on those figures. SF would have to bring transfer toxicity to new levels to lose it.

Very disappointed for McHugh - she is a breath of fresh air.

Christy42

Just Saying wrote: »

Regarding the MoE surely it needs better explanation.Using a 3% MoE might be acceptable when a candidate receives 25% of the vote but if a candidate receives 14% having had an exit poll of 10% it is in fact miles out.

I would contend that this represents a deviation of 40% from the poll which is way too much.

I don't think the poll methodology is wrong.I think the polling companies are being told lies.This first arose in the "Shy Tory" polls in 1992 and has latterly manifested itself in the Trump election and the Brexit polls also.

It is where the respondent is afraid to admit a vote that they perceive might not be very popular and align themselves instead with what they think might be a less controversial choice.

Trump was off by about 1% and Brexit by 2% (with polls trending towards Brexit in the days before the vote). With Trump they did misjudge the voting profiles in certain areas (people who never voted before voted but overall they were fine and even in those locations they were not terrible). Obama /Romney had a bigger error but few noticed but as it just meant the favourite won by more.

The shy voter syndrome is a myth and has never really been shown.

Just Saying

How do you explain the overestimation of the GP vote by a huge amount in both constituencies so?

blanch152

Just Saying wrote: »

How do you explain the overestimation of the GP vote by a huge amount in both constituencies so?

IF you get a representative sample of voters by location, age, gender, previous voting record, wealth, family status, sexual orientation and religion, then 1,000 of them on a particular topic is enough to give you a margin of error of 3%.

IF you miss out on one or more of those criteria, your MOE can be way out.

riffmongous

dublinman1990 wrote: »

I heard from The Guardian that Austrian Chancellor Sebastian Kurz is going to face the sack in a no confidence vote soon.

His OVP party done badly in Austria winning 7 of the country's 18 seats in the EP.

Here is Reuters in Vienna on the story:

Yep gone now. First ever time it's happened. I'm a little surprised the Social Democrats went ahead with it, but at the same time he has to take the blame for going into power with the far right in the first place.. and he's been under constant pressure because of it, as the ibiza scanal is not the first one we've had here. There's been a constant stream of low level incidents, of racism and neo nazi links from members of the far righters, to the point where Kurz's lack of action in response became a meme

Correction on the EP vote though, they didn't do badly, the very opposite, a record vote.

riffmongous

blanch152 wrote: »

IF you get a representative sample of voters by location, age, gender, previous voting record, wealth, family status, sexual orientation and religion, then 1,000 of them on a particular topic is enough to give you a margin of error of 3%.

IF you miss out on one or more of those criteria, your MOE can be way out.

Exactly, it has to be a sampled properly for the MOE to apply

Just Saying

I'm totally familiar with the methodology used and the sample size needed.If any of these factors are out then it is polling company error.

The point I am making is that it cannot be a coincidence that an exit poll rather than a pre election poll has totally estimated the first preference received in 2 constituencies by 1 party ...the Greens.

Coillte_Bhoy

LuckyLloyd wrote: »

The Exit poll was absolutely terrible. Questions to be asked there.

No it wasnt, i havent seen the exact % breakdown but all or most figures within the margin of error i think, the polls do come with the oft used phrase 'a heath warning' im not sure people pay much attention to that warning

Fan of Netflix

Just Saying wrote: »

I'm totally familiar with the methodology used and the sample size needed.If any of these factors are out then it is polling company error.

The point I am making is that it cannot be a coincidence that an exit poll rather than a pre election poll has totally estimated the first preference received in 2 constituencies by 1 party ...the Greens.

Seen a suggestion on twitter Red C didn't do much polling in rural areas. Greens poll bad outside cities and towns.

mrbrianj

with exit polls, its not just who gets asked its also who answers. There are a few variables, for example Green voters may just be slightly more inclined by nature to answer a pollster. And the big one in Irish politics, FF still have a smell of the crash about them, Sf a smell of the RA, and FG a smell of South Dublin private schools- A lot of people will still vote for these parties - but would be kinda embarrassed to admit it!

Sometimes, for the greater good, you have to hold your nose when voting - thank the gods for secret ballots!

Sofa Spud

Just Saying wrote: »

Based on Ming's second preference tallies McHugh will be eliminated before Casey and Smith.At this stage Casey and Smith will be close together approx 15/18 k behind Walsh and there is no indication that either will get enough transfers to close that gap especially as Ming and Carthy will still be short of a quota and therefore absorbing transfers

Gutted for McHugh but getting worried that Casey is in with a real chance - Healy Eames may be FG gene pool, but it's the old school, anti marriage equality gene pool that's less likely to transfer to Walsh and more likely to transfer to Casey, same with the FF vote that's been flagged by polster talking heads on RTE as being more weighted to older, less liberal and more 'foreign welfare cheat' phobic that clicks with Casey's ramblings. Hope I'm completely wrong on this...

Sparko

Fan of Netflix wrote: »

Maybe Matt Carthy would have done better if he confirmed he would sit in the parliament and not give his seat to another Sinn Fein rep. I believe it is Darren O'Rourke from Meath who lost his council seat yesterday. Absurd stuff, there should be a by election. And I say that as someone who gave Carthy a number 2 preference.

Darren O'Rourke scraped through and got the last council seat in his constituency.

Christy42

Fan of Netflix wrote: »

Seen a suggestion on twitter Red C didn't do much polling in rural areas. Greens poll bad outside cities and towns.

You rebalance the polls to account for oversampling from towns (so rural answers would each have a higher weight if you feel like you asked too few of them relative to townies). As long as they asked a decent sample in the rural areas it shouldn't matter if they asked too many people in towns. A simple reweighted deals with it.

Likely what happened is that noticed X said Green a lot and X didn't vote as much as they expected. It would explain why it was off across the board for the greens. (That X may be people from towns or young people or whatever demographic).

is_that_so

Sofa Spud wrote: »

Gutted for McHugh but getting worried that Casey is in with a real chance - Healy Eames may be FG gene pool, but it's the old school, anti marriage equality gene pool that's less likely to transfer to Walsh and more likely to transfer to Casey, same with the FF vote that's been flagged by polster talking heads on RTE as being more weighted to older, less liberal and more 'foreign welfare cheat' phobic that clicks with Casey's ramblings. Hope I'm completely wrong on this...

She just needs to mostly match him. A 12K cushion is quite hard to pull back and there are a lot of assumptions about his transfer-friendliness.

is_that_so

Ireland South declaration unlikely until 6

Squatter

dmakc wrote: »

Wonder if IT will spend tomorrow's paper correcting the crazy green hyperbole they filled today's with.

In fairness to the IT, it also published this:-


Pat Leahy: Nine out of 10 people didn’t vote for the Greens

(no link because the article is behind the IT paywall).

Sofa Spud

is_that_so wrote: »

She just needs to mostly match him. A 12K cushion is quite hard to pull back and there are a lot of assumptions about his transfer-friendliness.

Yeah, hope you're right.

Squatter

Coillte_Bhoy wrote: »

No it wasnt, i havent seen the exact % breakdown but all or most figures within the margin of error i think, the polls do come with the oft used phrase 'a heath warning' im not sure people pay much attention to that warning

Dobbo pointed out to the smug Lisa Chambers last night with barely concealed exasperation that he was "blue in the face" pointing this out to people!

is_that_so

Sofa Spud wrote: »

Yeah, hope you're right.

Last seat in 2014 was won without reaching the quota. It's a pretty common occurrence in the GE as well and positioning at this stage is vital.

AndrewJRenko

Squatter wrote: »

Do you use e-banking? ..............................

Your banking transactions are explicitly tied to you as a person. Your vote CANNOT be tied and traced to you - the secret ballot is a fundamental cornerstone of democracy.