Boards.ie uses cookies. By continuing to browse this site you are agreeing to our use of cookies. Click here to find out more x
Post Reply  
 
 
Thread Tools Search this Thread
30-01-2018, 15:34   #1
Zzippy
Moderator
 
Zzippy's Avatar
 
Join Date: Jun 2007
Posts: 12,751
Mod: Rugby, ZTest
Off Topic Thread 4.0

Thread continued from here

The usual rules apply. Non-rugby talk here. Consider this a social place, but like a pub, things like religion and politics can be divisive, so don't say things here you wouldn't say to someone's face in the pub. Sober
Zzippy is offline  
Advertisement
30-01-2018, 15:39   #2
Venjur
Registered User
 
Venjur's Avatar
 
Join Date: Dec 2004
Posts: 15,521
Quote:
Originally Posted by Zzippy View Post
so don't say things here you wouldn't say to someone's face in the pub. Sober
Zzippy has a nice arse.
Venjur is offline  
30-01-2018, 15:41   #3
Zzippy
Moderator
 
Zzippy's Avatar
 
Join Date: Jun 2007
Posts: 12,751
Mod: Rugby, ZTest
Quote:
Originally Posted by Venjur View Post
Zzippy has a nice arse.
Now that's what I'm talking about... just keep your hands to yourself!
Zzippy is offline  
30-01-2018, 15:41   #4
stephen_n
Registered User
 
stephen_n's Avatar
 
Join Date: Apr 2006
Posts: 9,303
Quote:
Originally Posted by sydthebeat View Post

not correct either

the MRBI poll in march 2015 had the yes vote at 74 % ... it passed at 62%, proving people turn more conservative when closer to the vote.
That doesn’t prove anything except polls are inaccurate. How often do polls predict accurately the outcome? If they did Brexit wouldn’t have happened, Hilary would be president and Theresa May would have a big majority. Attributing it to a swing towards conservatism may be a good narrative for polling companies but it isn’t a factual based one.

Up till last night the grey area existed, now it doesn’t so that supposed 15% undecided will probably shrink rapidly.

Unless the opinion polls are being extrapolated from previously gathered data, though I’m sure that never happens
stephen_n is offline  
30-01-2018, 15:43   #5
irishbucsfan
Registered User
 
Join Date: Aug 2007
Posts: 36,768
MRBI did another poll last week, I believe. Would be interesting to see what it'd be at now that the referendum has actually been announced, the numbers actually matter suddenly!
irishbucsfan is offline  
Advertisement
30-01-2018, 16:10   #6
sydthebeat
Subscriber
 
sydthebeat's Avatar
 
Join Date: Nov 2007
Posts: 25,891
Quote:
Originally Posted by stephen_n View Post
That doesn’t prove anything except polls are inaccurate.
sorry but the trend expressed was that voters tend to vote more conservatively when closer to the voting time... that has been borne out many times in many issues for it to actually be seen as a trend.

you said
Quote:
The notion of people voting more conservatively when it came time to vote was mooted for the marriage referendum too, but didn’t seem to happen
.

i clearly showed that actually... thats exactly what happened.

No one is questioning whether polls are inaccurate or not, they are, purely by definition.... so im not sure why you posted the above.

the reason why i brought up the trend of voters voting more conservative is because the last MRBI poll, which suggested 62% would be in favour of repeal, will mean that the actual voting number will most likely be less than that therefore the 'pro repeal' side shouldnt be resting on their laurels
sydthebeat is offline  
30-01-2018, 16:26   #7
Podge_irl
Registered User
 
Podge_irl's Avatar
 
Join Date: Oct 2007
Posts: 16,449
Quote:
Originally Posted by stephen_n View Post
That doesn’t prove anything except polls are inaccurate. How often do polls predict accurately the outcome? If they did Brexit wouldn’t have happened, Hilary would be president and Theresa May would have a big majority. Attributing it to a swing towards conservatism may be a good narrative for polling companies but it isn’t a factual based one.
Didn't plenty of polls show Brexit was a serious possibility but people figured the voting would be different?
Podge_irl is offline  
30-01-2018, 16:34   #8
ClanofLams
Registered User
 
ClanofLams's Avatar
 
Join Date: Nov 2015
Posts: 3,071
Quote:
Originally Posted by Podge_irl View Post
Didn't plenty of polls show Brexit was a serious possibility but people figured the voting would be different?
Yep and similarly, the presidential election polling gave trump around a 33% chance of winning according to fivethirtyeight, a stats website, before the election. The same site was giving Romney a 7% chance in 2012. So Trumps victory wasn’t as shocking as portrayed.

General election polling in the UK has a poor track record historically I think I remember reading, missed the conservative romp in ‘15 too.
ClanofLams is offline  
30-01-2018, 16:36   #9
Interested Observer
Registered User
 
Join Date: Jun 2014
Posts: 9,245
Quote:
Originally Posted by Podge_irl View Post
Didn't plenty of polls show Brexit was a serious possibility but people figured the voting would be different?
Yes. And things tightened up massively as the vote approached. Red = Leave, Green = Remain.

Interested Observer is offline  
Advertisement
30-01-2018, 17:13   #10
Neil3030
Moderator
 
Neil3030's Avatar
 
Join Date: Mar 2001
Posts: 13,712
Send a message via ICQ to Neil3030
Slightly misleading graph - the mean values came closer together, but the variance (particularly in Leave prediction) increased, just by looking at it. Thus, you can retrospectively say the aggregate of the polls fell on or around the eventual vote, but at the time, predicting the result from the polls was not so easy.
Neil3030 is offline  
30-01-2018, 17:31   #11
Podge_irl
Registered User
 
Podge_irl's Avatar
 
Join Date: Oct 2007
Posts: 16,449
Predicting to would have been difficult sure but to call Brexit unforeseen based on polls is utterly incorrect.
Podge_irl is offline  
30-01-2018, 17:36   #12
Interested Observer
Registered User
 
Join Date: Jun 2014
Posts: 9,245
Quote:
Originally Posted by Neil3030 View Post
Slightly misleading graph - the mean values came closer together, but the variance (particularly in Leave prediction) increased, just by looking at it. Thus, you can retrospectively say the aggregate of the polls fell on or around the eventual vote, but at the time, predicting the result from the polls was not so easy.
I think that's the point really. They don't clearly show Remain to win.
Interested Observer is offline  
30-01-2018, 17:47   #13
stephen_n
Registered User
 
stephen_n's Avatar
 
Join Date: Apr 2006
Posts: 9,303
Quote:
Originally Posted by sydthebeat View Post
sorry but the trend expressed was that voters tend to vote more conservatively when closer to the voting time... that has been borne out many times in many issues for it to actually be seen as a trend.

you said
.

i clearly showed that actually... thats exactly what happened.

No one is questioning whether polls are inaccurate or not, they are, purely by definition.... so im not sure why you posted the above.

the reason why i brought up the trend of voters voting more conservative is because the last MRBI poll, which suggested 62% would be in favour of repeal, will mean that the actual voting number will most likely be less than that therefore the 'pro repeal' side shouldnt be resting on their laurels
You are not clearly showing that people voted more conservatively though. You are showing that the outcome of the vote was more conservative than predicted by the polls. That is not the same thing. The trend you are talking about can be explained by the poor samples used. MRBI polls sometimes use no more than 1000 subjects (Can't find the sample size for the current one), from varying age and gender demographics. They are predominately collected in major population centres, for ease of collection. This will obviously have an impact on the liberal/conservative split. Given that there is a far more conservative viewpoint in general outside the major urban centres. I'm not entirely sure if MRBI make adjustments for that, can't see anything in there posted results to suggest they do.

I would say 62% is squeaky bum time for the pro-repeal campaign as they would be lucky to get 52-55% based on that, given previous polls v outcomes.
I don't see this being anything but a close call in the end.
stephen_n is offline  
30-01-2018, 17:49   #14
Neil3030
Moderator
 
Neil3030's Avatar
 
Join Date: Mar 2001
Posts: 13,712
Send a message via ICQ to Neil3030
Quote:
Originally Posted by Podge_irl View Post
Predicting to would have been difficult sure but to call Brexit unforeseen based on polls is utterly incorrect.
Quote:
Originally Posted by Interested Observer View Post
I think that's the point really. They don't clearly show Remain to win.
Ah ok, I get you now. Agreed.

Variance in polling will continue until they can adjust to the shift in media and technology usage. For e.g., are telephone polls really a reliable assay of voter behaviour these days?
Neil3030 is offline  
30-01-2018, 17:50   #15
Interested Observer
Registered User
 
Join Date: Jun 2014
Posts: 9,245
It's well established that for a large population (i.e. all the adults in Ireland) that 1100 is enough to get within a +- 3% margin of error, which is why around that sample size is used very frequently.

Urban v rural or other factors could be an issue.
Interested Observer is offline  
Post Reply

Quick Reply
Message:
Remove Text Formatting
Bold
Italic
Underline

Insert Image
Wrap [QUOTE] tags around selected text
 
Decrease Size
Increase Size
Please sign up or log in to join the discussion

Thread Tools Search this Thread
Search this Thread:

Advanced Search



Share Tweet