Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Undefined discussion subject.

SystemSystem Posts: 8,489
edited October 16 in General
Undefined discussion subject.

Read the full story here

«134

Comments

  • tlg86tlg86 Posts: 15,153
    First
  • IanB2IanB2 Posts: 28,733
    Second! The return of the undefined discussion!
  • MikeSmithsonMikeSmithson Posts: 6,319
    Sorry for the technical issues
  • HYUFDHYUFD Posts: 77,818
    edited October 16
    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election in the EC, even if he is likely to lose the national popular vote again.









  • kle4kle4 Posts: 56,679
    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election in the EC, even if he is likely to lose the national popular vote again.









    Damn swing states, always letting people down.
  • PulpstarPulpstar Posts: 61,901
    edited October 16
    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election, even if he is likely to lose the national popular vote again.


    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).
  • HYUFDHYUFD Posts: 77,818
    edited October 16
    Pulpstar said:

    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election, even if he is likely to lose the national popular vote again.


    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).
    Who cares? They forecast Trump would win Michigan and Florida right in 2016 and that is all that matters
  • GallowgateGallowgate Posts: 10,428
    HYUFD said:

    Pulpstar said:

    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election, even if he is likely to lose the national popular vote again.


    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).
    Who cares? They forecast Trump would win Michigan and Florida right in 2016 and that is all that matters
    So what? That doesn’t make them the “gold standard”, that just makes them right in a few states last time. They were wrong in a few too.
  • HYUFD said:

    Pulpstar said:

    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election, even if he is likely to lose the national popular vote again.


    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).
    Who cares? They forecast Trump would win Michigan and Florida right in 2016 and that is all that matters
    So what? That doesn’t make them the “gold standard”, that just makes them right in a few states last time. They were wrong in a few too.
    It's like saying Lord Ashcroft's constituency polls from 2015 were the gold standard because they got most of the Scottish seats right.

    But they were dire in England & Wales.
  • CarnyxCarnyx Posts: 6,867
    FPT - more on the Ellon result (and stressing also it's crap for the LDs and Labour, as one would imagine in a straight battle to the death between the SNP and the Tories in the Brexit and covid era):

    http://ballotbox.scot/ellon-by-election-result
  • AlistairAlistair Posts: 17,100
    edited October 16
    Prior to this election I will produce my forecast for every state.

    I will predict either Biden or Trump to get 100% in each state.

    Come next election I will use those states that I "called correctly" as evidence that I should be listened to in those states in 2024.
  • rcs1000rcs1000 Posts: 33,048
    HYUFD said:

    Pulpstar said:

    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election, even if he is likely to lose the national popular vote again.


    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).
    Who cares? They forecast Trump would win Michigan and Florida right in 2016 and that is all that matters
    With all due respect, it does matter.

    If Trafalgar is a genuine pollster, then we need to take their results seriously.

    If they simply take other pollsters results, and "right shift" then they're like Derren Brown in The System. And were simply lucky. And we should give them no mind at all.
  • HYUFDHYUFD Posts: 77,818

    HYUFD said:

    Pulpstar said:

    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election, even if he is likely to lose the national popular vote again.


    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).
    Who cares? They forecast Trump would win Michigan and Florida right in 2016 and that is all that matters
    So what? That doesn’t make them the “gold standard”, that just makes them right in a few states last time. They were wrong in a few too.
    It's like saying Lord Ashcroft's constituency polls from 2015 were the gold standard because they got most of the Scottish seats right.

    But they were dire in England & Wales.
    It isn't, the only state Trafalgar failed to forecast the winner in 2016 was Nevada, a much better record than most state pollsters then
  • CarnyxCarnyx Posts: 6,867
    Carnyx said:

    FPT - more on the Ellon result (and stressing also it's crap for the LDs and Labour, as one would imagine in a straight battle to the death between the SNP and the Tories in the Brexit and covid era):

    http://ballotbox.scot/ellon-by-election-result

    PS for HYUFD - the Tory vote went down about 10 percentage points. It is not a FPTP seat.
  • Excellent piece of work, Quincel. Thank you. Here's a quick run through of 'other factors'.

    1. Differential turnout. The surprising thing last time was that HS white males turned out in greater numbers than ever before. Who is to say that they will do so again? Conversely, black and hispanic voters didn't show up in the numbers expected. Again, is this likely to repeat in 2020. We just don't know.

    2. Genuine swing. Quite a lot of voters appear to have made up their minds late in 2016 and they broke heavily for Trump. This is unlikely to happen again if only because the number of undecideds is relatively small.

    3. Elasticity. Voters vary enormously from State to State in respect of their propensity to switch votes. Voters in Georgia are relatively 'inelastic', those in Arizona, Florida, and North Carolina less so.

    4. Cheating! Voter suppression, gerrymandering, fraud and intimidation can make a nonsense of any poll.


    As a fellow Biden punter I think only 4. really bothers me because it's unquantifiable, but my democratic instincts lead me to think the impact won't be that great.

    I certainly hope so.
  • kinabalukinabalu Posts: 14,742
    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election in the EC, even if he is likely to lose the national popular vote again.









    That is cherry HARVESTING.
  • FregglesFreggles Posts: 3,441
    HYUFD you are blatantly just cherrypicking polls, generally from lower-tier or biased pollsters.

    For Arizona, 538 added polls from four firms yesterday. Three of them had a Biden lead in all versions of their polls, including Monmouth University (rated A+) where Biden's lead ranged from +2 to +7. You highlighted the only pollster (C/D rated) who found any Trump lead and it's by 1% so MOE from zero.

  • tlg86tlg86 Posts: 15,153
    Excellent header; Trump looks like toast.
  • FoxyFoxy Posts: 18,671
    Carnyx said:

    Carnyx said:

    FPT - more on the Ellon result (and stressing also it's crap for the LDs and Labour, as one would imagine in a straight battle to the death between the SNP and the Tories in the Brexit and covid era):

    http://ballotbox.scot/ellon-by-election-result

    PS for HYUFD - the Tory vote went down about 10 percentage points. It is not a FPTP seat.
    Looking at the results, it looks as if Tory toxicity overrides unionism in SLD and SLAB voters.

    Doomed, the union is doomed...
  • williamglennwilliamglenn Posts: 34,502
  • ydoethurydoethur Posts: 31,862
    tlg86 said:

    Excellent header; Trump looks like toast.

    I’ll take your word for that as a gentleman and a train enthusiast.

    I’ll have to take your word because I can’t see it, so I’ll have to confine myself to awesome punning on random topics.

    Like I usually do, come to think of it.
  • LostPasswordLostPassword Posts: 1,890
    Fabulous post on the US polling. I'm much encouraged. I have also been cheered by Biden winning more TV eyeballs than Trump, not something I thought was remotely possible.
  • IshmaelZIshmaelZ Posts: 3,763
    Freggles said:

    HYUFD you are blatantly just cherrypicking polls, generally from lower-tier or biased pollsters.

    For Arizona, 538 added polls from four firms yesterday. Three of them had a Biden lead in all versions of their polls, including Monmouth University (rated A+) where Biden's lead ranged from +2 to +7. You highlighted the only pollster (C/D rated) who found any Trump lead and it's by 1% so MOE from zero.

    If that is true, it puts Scotch subsample misdemeanours in the shade.
  • dixiedeandixiedean Posts: 8,641
    Surprisingly LGBT friendly lot there.
  • ydoethurydoethur Posts: 31,862
    Freggles said:

    HYUFD you are blatantly just cherrypicking polls, generally from lower-tier or biased pollsters.

    Hyufd’s using dodgy stats?

    I’m shocked. Shocked, I tell you.
  • PulpstarPulpstar Posts: 61,901
    MAGA has lost it's cool from 2016.
  • rcs1000rcs1000 Posts: 33,048
    Alistair said:

    rcs1000 said:

    HYUFD said:

    rcs1000 said:

    I'm calling it:

    Trafalgar is not a real pollster.

    They aren't a company. They have no employees. They have no clients.

    They simply copy and paste other pollsters and add a slight rightward bias.

    Yet if Trump is re elected they would still be the Gold Standard
    I don't believe Trafalgar does polling. I believe they take existing polls and just shift them six points to the right.

    If they were real:

    (1) Where's the corporate entity? Right now, Trafalgar is a website (a dot org) which posts a few claimed polls every now and then. Real businesses are incorporated. Trafalgar is not.

    (2) They claim to use an online panel. Great! That's just like Opinium, YouGov, SurveyMonkey, etc. And you know what all those organisatons have in common? (Apart from actually existing) Yes, you can join their panel. Trafalgar has no way of joining their panel. So where does it come from?

    (3) They would have a business - you know, they'd be commissioned by actual newspapers and corporates to do research. Yet they don't.

    So, this a "pollster" who only does polls every two years (what happens to their panel in between?) and which doesn't exist an as incorporated entity and which appears to have no employees or sources of revenue.

    You can call that gold standard if you like, I call bullshit.
    I thought they claimed to do robo call polling?
    They claim to do both. See the last page of their PDFs.

    Of course, the PDF claims of how they do their polling doesn't match their website.
  • rcs1000rcs1000 Posts: 33,048
    rcs1000 said:

    Alistair said:

    rcs1000 said:

    HYUFD said:

    rcs1000 said:

    I'm calling it:

    Trafalgar is not a real pollster.

    They aren't a company. They have no employees. They have no clients.

    They simply copy and paste other pollsters and add a slight rightward bias.

    Yet if Trump is re elected they would still be the Gold Standard
    I don't believe Trafalgar does polling. I believe they take existing polls and just shift them six points to the right.

    If they were real:

    (1) Where's the corporate entity? Right now, Trafalgar is a website (a dot org) which posts a few claimed polls every now and then. Real businesses are incorporated. Trafalgar is not.

    (2) They claim to use an online panel. Great! That's just like Opinium, YouGov, SurveyMonkey, etc. And you know what all those organisatons have in common? (Apart from actually existing) Yes, you can join their panel. Trafalgar has no way of joining their panel. So where does it come from?

    (3) They would have a business - you know, they'd be commissioned by actual newspapers and corporates to do research. Yet they don't.

    So, this a "pollster" who only does polls every two years (what happens to their panel in between?) and which doesn't exist an as incorporated entity and which appears to have no employees or sources of revenue.

    You can call that gold standard if you like, I call bullshit.
    I thought they claimed to do robo call polling?
    They claim to do both. See the last page of their PDFs.

    Of course, the PDF claims of how they do their polling doesn't match their website.
    Also, their website and email contact details have different domains, which is odd to say the least.
  • PulpstarPulpstar Posts: 61,901
    4 way battle for 3rd place.


    Brexit party might be in with a shot after Boris capitulates.
  • AnabobazinaAnabobazina Posts: 4,659
    Is @HYUFD ‘s analysis simply that you take the best poll for Trump in every state to forecast the outcome? That’s fair enough if so, but for betting purposes I just need to be clear about the cypher.
  • alex_alex_ Posts: 2,953
    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    How did they do in the mid terms (when the other pollsters may even have over-corrected for the Trump effect, as opposed to the Republican effect)?
  • Re: Trafalgar, note that its principle Robert Cahaly is a Republican hack from WAY back.

    Indeed, learned his trade from the late Lee Atwater (remember him) who gave advanced courses in political dirty tricks to a generation of GOP operatives.

    Check out what Cahaly says about himself on Trafalgar website:

    https://www.thetrafalgargroup.org/robert-cahaly-2/

    "As President of the Trafalgar Group, Cahaly’s expertise includes strategic communications, public relations, polling, new/social media utilization, team building/management, federal/state election compliance, campaign development, and GOTV. He is often quoted in national and state media for his political perspective. He is also a sought-after speaker and lecturer on campaign training, issue advocacy, and public relations topics. He has spoken for/to groups ranging from the RNC, state GOPs, and legislative caucuses, to conservative interest groups such as GOPAC, Tea Party Patriots, Tea Party Express, and The Leadership Institute. He has been speaking for over 15 years in 43 states in front of 4000+ candidates, elected officials, and trainees."
  • DougSealDougSeal Posts: 2,861
    HYUFD said:

    Pulpstar said:



    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).

    Who cares? They forecast Trump would win Michigan and Florida right in 2016 and that is all that matters
    No. The previous cycles forecasts don't matter at all. 2016 itself proved that when a number of pundits who called 2008 and 2012 right got egg all over their faces. Sam Wang was considered something of a guru in 2016 having got every election right from 2004 to 2012. He screwed up and you barely hear from him now. You are placing all of your (barely repressed hopes) for a Trump win on once single pollster, Trafalgar, having got two states right in 2016. They almost alone got Wis and MI right (others called Florida for Trump too) but Nevada spectacularly wrong. They also predicted Trump would sweep the under 25 vote.

    The problem you have is this. Many polls (including your beloved Trafalgar who got Nevada wrong by 5 points) got various state polls (not national) wrong in 2016. But not by much. For Trump to win the polling acuracy has to have got significantly worse, not better. The most innacurate of the RCP polling averages in 2016 got the result wrong by around 4 percent. The RCP average for Michigan tonight is 6.7% in Biden's favour. In 2016 RCP had Clinton up by 3.7% in Michigan on the eve of the election. Trump won by a wafer thin 0.3%. That means the polling companies, as a whole, have to be twice as bad as last time. In my view the evidence points to Trafalgar having been lucky.
  • OmniumOmnium Posts: 4,482
    ydoethur said:

    tlg86 said:

    Excellent header; Trump looks like toast.

    I’ll take your word for that as a gentleman and a train enthusiast.

    I’ll have to take your word because I can’t see it, so I’ll have to confine myself to awesome punning on random topics.

    Like I usually do, come to think of it.
    Why is appearing like toast a bad thing? Any ideas as to where this came from?

    I rather like toast.
  • NigelbNigelb Posts: 25,553
    HYUFD said:

    HYUFD said:

    Pulpstar said:

    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election, even if he is likely to lose the national popular vote again.


    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).
    Who cares? They forecast Trump would win Michigan and Florida right in 2016 and that is all that matters
    So what? That doesn’t make them the “gold standard”, that just makes them right in a few states last time. They were wrong in a few too.
    It's like saying Lord Ashcroft's constituency polls from 2015 were the gold standard because they got most of the Scottish seats right.

    But they were dire in England & Wales.
    It isn't, the only state Trafalgar failed to forecast the winner in 2016 was Nevada, a much better record than most state pollsters then
    You’re right, it’s nothing like Ashcroft.
    There is compelling evidence that Ashcroft actually conducts polls.
  • LostPasswordLostPassword Posts: 1,890
    edited October 16
    Pulpstar said:

    4 way battle for 3rd place.
    witter.com/YouGov/status/1317174347447009287

    Brexit party might be in with a shot after Boris capitulates.

    The last time the Brexit Party surged in the polls at the Tories expense it flattered the strength of Labour's position.
  • NigelbNigelb Posts: 25,553
    CatMan said:

    Sorry for the technical issues

    That's not good enough. I pay £0 a month to access this site, and I demand my money's worth.
    Better than that, you can have it back.
  • CatManCatMan Posts: 1,096
    edited October 16
    Rcs suggests Trafalgar polls are dodgy. Next thing you know the site is suffering "technical problems". Coincidence? :#
  • dixiedean said:

    Surprisingly LGBT friendly lot there.
    May be a surprise to (elderly) PBers and others, but my guess is that most of the Putinists cavorting to "YMCA" are NOT aware that it was once (a couple of generations ago) a gay anthem.

    For most people, YMCA is simply a fun old pop song from the days when disco was king.
  • IshmaelZIshmaelZ Posts: 3,763
    Nigelb said:

    HYUFD said:

    HYUFD said:

    Pulpstar said:

    HYUFD said:

    Trump does not even necessarily need to come from behind now, according to some state polls from key swing states in the last few days he is ahead in enough swing states for a narrow re election, even if he is likely to lose the national popular vote again.


    People should know this about Trafalgar's history - Robert Cahaly is the face of the company on twitter.
    https://www.wistv.com/story/13429729/gop-consultant-to-face-sled-arrest-for-illegal-robocalls/
    41-year-old Robert Cahaly faces charges that he purchased, then disseminated computer-generated calls from an automatically dialed announcing service to potential South Carolina voters on September 23.

    According to SLED, the calls were political in nature and were allegedly made without properly disclosing the identity of the originating party to the call recipients, which is in violation of SC Code 16-17-446.

    We previously reported that the Ken Ard's campaign manager was accused of robocalling and that there was a warrant issued for his arrest. In fact, the campaign manager is not accused of robocalling and is not in trouble with the law.

    SLED has determined that the calls were made through a Richland County landline telephone number owned by Gadsden and Greene Strategies, a company owned by Cahaly.

    The calls, according to SLED, were allegedly made to numerous voters in House Districts 26 (Greenville and Pickens), 78 (Richland County), 79 (Kershaw and Richland Counties), 98 (Dorchester County), 108 (Charleston and Georgetown Counties) and 115 (James Island and Folly Beach).
    Who cares? They forecast Trump would win Michigan and Florida right in 2016 and that is all that matters
    So what? That doesn’t make them the “gold standard”, that just makes them right in a few states last time. They were wrong in a few too.
    It's like saying Lord Ashcroft's constituency polls from 2015 were the gold standard because they got most of the Scottish seats right.

    But they were dire in England & Wales.
    It isn't, the only state Trafalgar failed to forecast the winner in 2016 was Nevada, a much better record than most state pollsters then
    You’re right, it’s nothing like Ashcroft.
    There is compelling evidence that Ashcroft actually conducts polls.
    No there isn't, he just rebadges other people's output. He isn't a BPC member.
  • OmniumOmnium Posts: 4,482
    Burnham, once again, seems to have massively overplayed his hand. He's like a god-given lesson in 'overplaying your hand'. I guess some religions may in fact see him as the true God of 'overplaying hands'.

  • Re: Trafalgar & Robert Cahaly, check out this take from SC newspaper:

    https://www.postandcourier.com/free-times/archives/an-ard-rains-gonna-fall-the-political-obituary-of-ken-ard/article_65028b3a-7c55-511b-87c3-7148892a3fd3.html

    "Enter Robert Cahaly, an infamous shadowy political consultant in the vein of the late Lee Atwater. A dozen years ago, The State newspaper’s legendary political reporter Lee Bandy described Cahaly as “a cloak-and-dagger GOP activist who likes to operate in secret.”

    If a Southern Gothic movie were made about the Ken Ard story, it’s doubtful the best casting crew could find a better character to play Ard’s consultant. In describing him, one Ard confidant called Cahaly “an interesting-looking, little, chubby fellow with glasses and bowties — bowtie man — and kind of a weird, unusual kind of personality.”

    And indeed, Cahaly had once even acted out such a role on the silver screen. According to his bio at the Leadership Institute, he played an eccentric Southern pollster in a 2005 independent film called In the Arena, about a fictitious presidential race during the Iowa caucuses.

    In real life, Cahaly is a longtime GOP operative who was a senior aide in the campaigns of former Gov. David Beasley and also worked for Beasley in the governor’s office. If political consulting outfits in South Carolina are the like the Five Families in The Godfather, Cahaly would be a made man."
  • stodgestodge Posts: 7,446
    Evening again all :)

    Excellent piece from @Quincel for which many thanks.

    I try to look at the crosstabs of the various polls (where published) and I simply don't see how the mistakes of 2016 can or will be repeated. The pollsters, unlike many of us, learn from their mistakes and the sampling for this time from those who publish it seems more reasonable.

    Without wishing to get anyone into trouble, I do detect a degree of political-led bias in some of the polling, If you are doing a poll for the "Center for American Greatness" which, it's fair to say, is not wholly sympathetic to Biden, a poll showing Trump doing well or closing in is what you would probably like to produce for your "paymasters".

    The same is doubtless true on the Democrat side.

    "It's the hope that kills you", a wise man once said. The truth is Biden has moved the battle to Iowa, Ohio and Georgia rather than Michigan, Wisconsin and Pennsylvania which (Trafalgar polls notwithstanding) look in the bag for Biden at this time and those 46 EC votes get him over the line if he retains all Hillary's states.

    If the likes of Florida, North Carolina and Arizona all follow (and while there are odd polls still showing a Trump lead in each of those states, the weight of polling suggest Biden is ahead), Biden wins easily.
  • CatMan said:

    Rcs suggests Trafalgar polls are dodgy. Next thing you know the site is suffering "technical problems". Coincidence? :#

    Are you suggesting that a Battle of Trafalgar is underway?

    With RCS as Nelson?
  • LostPasswordLostPassword Posts: 1,890
    alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    Is that a serious suggestion? Is there any particular reason to suspect that's happening?

    In principle opinion polling is one of the defences that a democracy has against an election being stolen by ballot-stuffing, fraudulent counts and similar means.

    If a supposed polling company is intentionally producing poll results to be biased then the intention is to hide the blatant stealing of an election.
  • rcs1000rcs1000 Posts: 33,048
    alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    How did they do in the mid terms (when the other pollsters may even have over-corrected for the Trump effect, as opposed to the Republican effect)?

    They did OK, but not exceptionally. They called FL wrong directionally, and they overstated the gap in favour of the Republican positions by:

    Montana - 4 points
    Texas - 5 points
    Nevada - 3 points

    But they were broadly in line in Arizona and Michigan. Their average error was just under 4 points, which would make them a B- pollster.
  • alex_alex_ Posts: 2,953
    edited October 16

    alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    Is that a serious suggestion? Is there any particular reason to suspect that's happening?

    In principle opinion polling is one of the defences that a democracy has against an election being stolen by ballot-stuffing, fraudulent counts and similar means.

    If a supposed polling company is intentionally producing poll results to be biased then the intention is to hide the blatant stealing of an election.
    They suggestion by others (not myself - i have no idea) on this thread is that Trafalgar is

    1) Not a polling company and doesn't actually do any polls
    2) Produces its "polls" by taking an average of everyone else's and adding 6% to the Trump/Republican percentage.
  • rcs1000rcs1000 Posts: 33,048

    alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    Is that a serious suggestion? Is there any particular reason to suspect that's happening?

    In principle opinion polling is one of the defences that a democracy has against an election being stolen by ballot-stuffing, fraudulent counts and similar means.

    If a supposed polling company is intentionally producing poll results to be biased then the intention is to hide the blatant stealing of an election.
    There's no evidence that Trafalgar is a real, commercial organisation.

    It's not incorporated. It doesn't appear to have any employees or a business address. It claims on its PDFs to use a combination of calls and an on-line panel, yet there's no way to join the panel. It doesn't seem to have any customers (so where does the money come from to do these polls), and is dormant in between elections years. And it's run by a Republican operative.
  • tlg86tlg86 Posts: 15,153
    Omnium said:

    ydoethur said:

    tlg86 said:

    Excellent header; Trump looks like toast.

    I’ll take your word for that as a gentleman and a train enthusiast.

    I’ll have to take your word because I can’t see it, so I’ll have to confine myself to awesome punning on random topics.

    Like I usually do, come to think of it.
    Why is appearing like toast a bad thing? Any ideas as to where this came from?

    I rather like toast.
    I pretty much live off it.
  • MaxPBMaxPB Posts: 22,106
    Possible beheading in Paris.
  • AnabobazinaAnabobazina Posts: 4,659
    @HYUFD
    What is your current EC forecast?
  • rcs1000rcs1000 Posts: 33,048
    I now have Robert Cahaly's mobile phone number.

    Shall I call him and ask him these questions?
  • williamglennwilliamglenn Posts: 34,502
  • alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    Is that a serious suggestion? Is there any particular reason to suspect that's happening?

    In principle opinion polling is one of the defences that a democracy has against an election being stolen by ballot-stuffing, fraudulent counts and similar means.

    If a supposed polling company is intentionally producing poll results to be biased then the intention is to hide the blatant stealing of an election.
    Don't think idea behind Trafalgar's public "polling" is stealing elections. Instead, purpose is to INFLUENCE elections, by assuring Republicans and Rep leaners that they are NOT voting for losers. Which plenty of people do NOT like to do.

    My guess is that they may be giving less skewed versions of their polling (along with cross tabs & details NOT revealed publicly) to Cahaly's GOP paymasters.

    OR (as Robert suggests) they do NOT do any polling at all, just manipulate someone else's numbers.

    Certainly no sensible GOP strategist is relying upon Trafalgar's published numbers.

    Instead, Trumpsky campaign, RNC and other GOP campaigns are doing LOTS of internal polling (just like Biden, DSC, DCCC, etc.). Which is pretty obviously NOT going as well as they'd like & hope.
  • GallowgateGallowgate Posts: 10,428
    edited October 16

    Trafalgar 2016

    (Difference to actual result in brackets).

    Colorado
    Trump 44% (+1)
    Clinton 45% (-3)

    Florida
    Trump 50% (+1)
    Clinton 46% (-2)

    Georgia
    Trump 52% (+2)
    Clinton 45% (-)

    Michigan
    Trump 49% (+1)
    Clinton 47% (-1)

    Nevada
    Trump 50% (+4)
    Clinton 45% (-3)

    North Carolina
    Trump 49% (-1)
    Clinton 44% (-3)

    Ohio
    Trump 49% (-3)
    Clinton 44% (-)

    Pennsylvania
    Trump 48% (-)
    Clinton 47% (-1)

    South Carolina
    Trump 53% (-2)
    Clinton 38% (-3)

    Utah
    Trump 40% (-6)
    Clinton 30% (+2)

    So all over the place really. Yes they may have got the winner right more often than not, but they certainly are not what you’d call “accurate”.
  • NigelbNigelb Posts: 25,553
  • rcs1000 said:

    I now have Robert Cahaly's mobile phone number.

    Shall I call him and ask him these questions?

    Go for it! Though expect you'd have better luck getting straight (or any) answers via a seance with the ghost of Lee Atwater.

  • ydoethurydoethur Posts: 31,862
    Omnium said:

    ydoethur said:

    tlg86 said:

    Excellent header; Trump looks like toast.

    I’ll take your word for that as a gentleman and a train enthusiast.

    I’ll have to take your word because I can’t see it, so I’ll have to confine myself to awesome punning on random topics.

    Like I usually do, come to think of it.
    Why is appearing like toast a bad thing? Any ideas as to where this came from?

    I rather like toast.
    I think it’s a reference to the fact that bread was traditionally only toasted when it was too stale to eat fresh. So it represented the very last stage of a loaf and signified it would soon be gone.
  • rcs1000rcs1000 Posts: 33,048


    Nevada
    Trump 50% (+2)
    Clinton 45% (-1)

    As Clinton won Nevada fairly comfortably, I'm not sure your differences are correct.
  • MaxPBMaxPB Posts: 22,106
    The only response is to make teaching this mandatory in all of France (and Europe) no religion is protected from free speech.
  • GallowgateGallowgate Posts: 10,428
    edited October 16
    rcs1000 said:


    Nevada
    Trump 50% (+2)
    Clinton 45% (-1)

    As Clinton won Nevada fairly comfortably, I'm not sure your differences are correct.
    Apologies. I mixed up Trump and Clinton! Edited.
  • MaxPBMaxPB Posts: 22,106
    Nigelb said:
    I think this is what Mike Coupe was brought in to roll out. I wonder which test they are using, Nanopore or DNA Nudge. The Nanopore test looks like a much better bet to me and if it is successful it should be used for testing at venues, airports and train stations.
  • Richard_NabaviRichard_Nabavi Posts: 26,923
    Superb article, thanks @Quincel.
  • QuincelQuincel Posts: 3,035
    rcs1000 said:

    I now have Robert Cahaly's mobile phone number.

    Shall I call him and ask him these questions?

    I'd be genuinely interested on the results if you did do this.
  • alex_alex_ Posts: 2,953
    How exciting. I just got a "possible exposure notification" on my Covid-19 app.

    Unfortunately before i had finished reading it i got another one advising my "potential exposure" had been analysed and i was safe to carry on as normal. Lucky i was a quick reader or i would have missed it, because i can't seem to find the messages any more!
  • ydoethurydoethur Posts: 31,862
    Nigelb said:

    CatMan said:

    Sorry for the technical issues

    That's not good enough. I pay £0 a month to access this site, and I demand my money's worth.
    Better than that, you can have it back.
    With interest, of course. Maybe 10%?
  • OmniumOmnium Posts: 4,482
    tlg86 said:

    Omnium said:

    ydoethur said:

    tlg86 said:

    Excellent header; Trump looks like toast.

    I’ll take your word for that as a gentleman and a train enthusiast.

    I’ll have to take your word because I can’t see it, so I’ll have to confine myself to awesome punning on random topics.

    Like I usually do, come to think of it.
    Why is appearing like toast a bad thing? Any ideas as to where this came from?

    I rather like toast.
    I pretty much live off it.
    I'd guess some butter on top? (Anything else with it is by-the-by)

    However as a toast fan you have to worry about the bread. A lot of the flashier consumer breads are far too lightweight to stand up to toasting well. The just become dry and empty.

    Cheap bread works almost infallibly well. I've no idea why that should be.
  • DougSealDougSeal Posts: 2,861
    edited October 16
    alex_ said:

    alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    Is that a serious suggestion? Is there any particular reason to suspect that's happening?

    In principle opinion polling is one of the defences that a democracy has against an election being stolen by ballot-stuffing, fraudulent counts and similar means.

    If a supposed polling company is intentionally producing poll results to be biased then the intention is to hide the blatant stealing of an election.
    They suggestion by others (not myself - i have no idea) on this thread is that Trafalgar is

    1) Not a polling company and doesn't actually do any polls
    2) Produces its "polls" by taking an average of everyone else's and adding 6% to the Trump/Republican percentage.
    They essentially admit it but call it weighting.Trafalgar say, simplistically, that the Trump/Republican vote is understated by “shy Trumpers” (another favourite of @HYUFD ) so their methodology is to ask the individual how they think their neighbours will vote. If the respondent says the neighbours will vote a different way to them then they say they’re a “shy” (invariably Trump) voter. They are cagey about the precise methodology.

    What most pollsters got wrong in the rust belt last time was (A) to fail to foresee that undecideds would break for Trump 6 to 1 and (B) not to find enough uneducated voters. As a result Trafalgar struck gold in 2 states by accident. This time (A) their are far fewer undecideds and (B) other pollsters have adjusted their methodology to weight more for education. If there is a systemic polling error this time, which there could well be, it is unlikely to be the same one as last time. People are fighting 2016 over again but it’s a different race with a very different dynamic - notably that the Democratic candidate is far more popular than the last on every measure.

    I want to know what Trafalgar do when there is no election on. How do they make money between cycles and who commissions their other polls?
  • Nigelb said:
    That could provide some welcome relief - let's hope so.

    I wish they would roll this out widely in Universities. I may be starting face-to-face teaching next week, but the virus is spreading through the student population rapidly and some of mine are already in isolation. I have to decide pretty sharply whether to opt out of face-to-face, on grounds of safety. A tough call.

    --AS
  • RobDRobD Posts: 49,021
    alex_ said:

    How exciting. I just got a "possible exposure notification" on my Covid-19 app.

    Unfortunately before i had finished reading it i got another one advising my "potential exposure" had been analysed and i was safe to carry on as normal. Lucky i was a quick reader or i would have missed it, because i can't seem to find the messages any more!

    https://faq.covid19.nhs.uk/article/KA-01252/en-us
  • dodradedodrade Posts: 492
    Not sure how Trump will deal with the humiliation of a Biden landslide not even he will be able to spin away when he could have declared job done and walked away in the summer "undefeated". Unless he genuinely believes he's somehow going to win again and his yes men are too afraid to tell him anything different?
  • NigelbNigelb Posts: 25,553

    Superb article, thanks @Quincel.

    You can see it ?
  • QuincelQuincel Posts: 3,035
    Nigelb said:

    Superb article, thanks @Quincel.

    You can see it ?
    Shh, just say it's great and fall in line.
  • HYUFDHYUFD Posts: 77,818
    Carnyx said:

    Carnyx said:

    FPT - more on the Ellon result (and stressing also it's crap for the LDs and Labour, as one would imagine in a straight battle to the death between the SNP and the Tories in the Brexit and covid era):

    http://ballotbox.scot/ellon-by-election-result

    PS for HYUFD - the Tory vote went down about 10 percentage points. It is not a FPTP seat.
    No, the Tory vote went up on first preferences and Holyrood constituences next year will be decided on FPTP and the list is not decided on preferences either but top up PR
  • OmniumOmnium Posts: 4,482
    ydoethur said:

    Omnium said:

    ydoethur said:

    tlg86 said:

    Excellent header; Trump looks like toast.

    I’ll take your word for that as a gentleman and a train enthusiast.

    I’ll have to take your word because I can’t see it, so I’ll have to confine myself to awesome punning on random topics.

    Like I usually do, come to think of it.
    Why is appearing like toast a bad thing? Any ideas as to where this came from?

    I rather like toast.
    I think it’s a reference to the fact that bread was traditionally only toasted when it was too stale to eat fresh. So it represented the very last stage of a loaf and signified it would soon be gone.
    Thanks. (Hoping in fact that you didn't look it up and it turns out to be wrong, but brilliant)

    I'd imagined 'about to be burned'.

  • alex_alex_ Posts: 2,953
    RobD said:

    alex_ said:

    How exciting. I just got a "possible exposure notification" on my Covid-19 app.

    Unfortunately before i had finished reading it i got another one advising my "potential exposure" had been analysed and i was safe to carry on as normal. Lucky i was a quick reader or i would have missed it, because i can't seem to find the messages any more!

    https://faq.covid19.nhs.uk/article/KA-01252/en-us
    Oh. How unexciting. :(
  • HYUFDHYUFD Posts: 77,818
    edited October 16
    dodrade said:

    Not sure how Trump will deal with the humiliation of a Biden landslide not even he will be able to spin away when he could have declared job done and walked away in the summer "undefeated". Unless he genuinely believes he's somehow going to win again and his yes men are too afraid to tell him anything different?

    Based on the swing state polling I posted earlier Trump may will be re elected, so I would not get too confident and start preparing the champagne for that Biden landslide yet as it may never happen.

    I would have thought 2016 would have taught Democrat supporters victory against Trump is not guaranteed, clearly not in your case
  • NigelbNigelb Posts: 25,553
    alex_ said:

    How exciting. I just got a "possible exposure notification" on my Covid-19 app.

    Unfortunately before i had finished reading it i got another one advising my "potential exposure" had been analysed and i was safe to carry on as normal. Lucky i was a quick reader or i would have missed it, because i can't seem to find the messages any more!

    That is apparently very common, and normal.
    There’s a bug in the program which means these notifications get sent out for every potential exposure however it has been assessed. If you have genuinely been exposed at risk (c. 15mins indoors in proximity, but each case is apparently assessed individually I think), you’ll know about it as you get some massive red flag when you open the app.
  • Beibheirli_CBeibheirli_C Posts: 4,230
    So, once Britain becomes an introverted, isolated laughing stock - what happens to James Bond?
  • OmniumOmnium Posts: 4,482

    So, once Britain becomes an introverted, isolated laughing stock - what happens to James Bond?

    That'll be for just you and he to know.
  • tlg86tlg86 Posts: 15,153
    Omnium said:

    tlg86 said:

    Omnium said:

    ydoethur said:

    tlg86 said:

    Excellent header; Trump looks like toast.

    I’ll take your word for that as a gentleman and a train enthusiast.

    I’ll have to take your word because I can’t see it, so I’ll have to confine myself to awesome punning on random topics.

    Like I usually do, come to think of it.
    Why is appearing like toast a bad thing? Any ideas as to where this came from?

    I rather like toast.
    I pretty much live off it.
    I'd guess some butter on top? (Anything else with it is by-the-by)

    However as a toast fan you have to worry about the bread. A lot of the flashier consumer breads are far too lightweight to stand up to toasting well. The just become dry and empty.

    Cheap bread works almost infallibly well. I've no idea why that should be.
    Flora and marmite. I dread to think how many slices I’ve eaten in my life.
  • NigelbNigelb Posts: 25,553
    ydoethur said:

    Nigelb said:

    CatMan said:

    Sorry for the technical issues

    That's not good enough. I pay £0 a month to access this site, and I demand my money's worth.
    Better than that, you can have it back.
    With interest, of course. Maybe 10%?
    Seems ungenerous.
  • HYUFDHYUFD Posts: 77,818

    Trafalgar 2016

    (Difference to actual result in brackets).

    Colorado
    Trump 44% (+1)
    Clinton 45% (-3)

    Florida
    Trump 50% (+1)
    Clinton 46% (-2)

    Georgia
    Trump 52% (+2)
    Clinton 45% (-)

    Michigan
    Trump 49% (+1)
    Clinton 47% (-1)

    Nevada
    Trump 50% (+4)
    Clinton 45% (-3)

    North Carolina
    Trump 49% (-1)
    Clinton 44% (-3)

    Ohio
    Trump 49% (-3)
    Clinton 44% (-)

    Pennsylvania
    Trump 48% (-)
    Clinton 47% (-1)

    South Carolina
    Trump 53% (-2)
    Clinton 38% (-3)

    Utah
    Trump 40% (-6)
    Clinton 30% (+2)

    So all over the place really. Yes they may have got the winner right more often than not, but they certainly are not what you’d call “accurate”.
    So Trafalgar correctly forecast the winner of every state in that list then, except Nevada.

    The fact they also correctly forecast Clinton would win Colorado shows they are not just a pro Trump pollster in every swing state either
  • Beibheirli_CBeibheirli_C Posts: 4,230

    Sorry for the technical issues

    Technical issues?
  • TheuniondivvieTheuniondivvie Posts: 22,324

    dixiedean said:

    Surprisingly LGBT friendly lot there.
    May be a surprise to (elderly) PBers and others, but my guess is that most of the Putinists cavorting to "YMCA" are NOT aware that it was once (a couple of generations ago) a gay anthem.

    For most people, YMCA is simply a fun old pop song from the days when disco was king.
    M-A-G-A fitting well with the chorus seems to be the main reason it's been picked. Perhaps the cop/military/cowboy/construction worker/biker mix also appeals.
  • IshmaelZIshmaelZ Posts: 3,763
    DougSeal said:

    alex_ said:

    alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    Is that a serious suggestion? Is there any particular reason to suspect that's happening?

    In principle opinion polling is one of the defences that a democracy has against an election being stolen by ballot-stuffing, fraudulent counts and similar means.

    If a supposed polling company is intentionally producing poll results to be biased then the intention is to hide the blatant stealing of an election.
    They suggestion by others (not myself - i have no idea) on this thread is that Trafalgar is

    1) Not a polling company and doesn't actually do any polls
    2) Produces its "polls" by taking an average of everyone else's and adding 6% to the Trump/Republican percentage.
    They essentially admit it but call it weighting.Trafalgar say, simplistically, that the Trump/Republican vote is understated by “shy Trumpers” (another favourite of @HYUFD ) so their methodology is to ask the individual how they think their neighbours will vote. If the respondent says the neighbours will vote a different way to them then they say they’re a “shy” (invariably Trump) voter. They are cagey about the precise methodology.

    What most pollsters got wrong in the rust belt last time was (A) to fail to foresee that undecideds would break for Trump 6 to 1 and (B) not to find enough uneducated voters. As a result Trafalgar struck gold in 2 states by accident. This time (A) their are far fewer undecideds and (B) other pollsters have adjusted their methodology to weight more for education. If there is a systemic polling error this time, which there could well be, it is unlikely to be the same one as last time. People are fighting 2016 over again but it’s a different race with a very different dynamic - notably that the Democratic candidate is far more popular than the last on every measure.

    I want to know what Trafalgar do when there is no election on. How do they make money between cycles and who commissions their other polls?
    "...to fail to foresee that undecideds would break for Trump 6 to 1..."

    That seems to me an extraordinary thing to happen in the absence of a concomitant substantial (though not equally large, obv) move from C to T among those stating an intention. Shy Trumpers is not the worst possible explanation for it...
  • RobDRobD Posts: 49,021
    alex_ said:

    RobD said:

    alex_ said:

    How exciting. I just got a "possible exposure notification" on my Covid-19 app.

    Unfortunately before i had finished reading it i got another one advising my "potential exposure" had been analysed and i was safe to carry on as normal. Lucky i was a quick reader or i would have missed it, because i can't seem to find the messages any more!

    https://faq.covid19.nhs.uk/article/KA-01252/en-us
    Oh. How unexciting. :(
    Disappointing that hasn't been fixed. If something comes up inside the app itself that's when you start worrying.
  • rcs1000rcs1000 Posts: 33,048
    I called his cell, and it rang out and got voicemail.
  • HYUFDHYUFD Posts: 77,818
    edited October 16

    @HYUFD
    What is your current EC forecast?

    Trump 275 Biden 263 as of tonight, Biden picking up Wisconsin and Pennsylvania and NE02 and winning the national popular vote but Trump holding the rest of his 2016 states and being re elected in the EC
  • NigelbNigelb Posts: 25,553
    Quincel said:

    Nigelb said:

    Superb article, thanks @Quincel.

    You can see it ?
    Shh, just say it's great and fall in line.
    You’re right.
    Scintillating analysis.
  • OmniumOmnium Posts: 4,482
    tlg86 said:

    Omnium said:

    tlg86 said:

    Omnium said:

    ydoethur said:

    tlg86 said:

    Excellent header; Trump looks like toast.

    I’ll take your word for that as a gentleman and a train enthusiast.

    I’ll have to take your word because I can’t see it, so I’ll have to confine myself to awesome punning on random topics.

    Like I usually do, come to think of it.
    Why is appearing like toast a bad thing? Any ideas as to where this came from?

    I rather like toast.
    I pretty much live off it.
    I'd guess some butter on top? (Anything else with it is by-the-by)

    However as a toast fan you have to worry about the bread. A lot of the flashier consumer breads are far too lightweight to stand up to toasting well. The just become dry and empty.

    Cheap bread works almost infallibly well. I've no idea why that should be.
    Flora and marmite. I dread to think how many slices I’ve eaten in my life.
    Unless there's a reason not to - butter.

    Have you tried marmite peanut butter? It's like a marmite double-down.
  • HYUFDHYUFD Posts: 77,818
    edited October 16
    alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    How did they do in the mid terms (when the other pollsters may even have over-corrected for the Trump effect, as opposed to the Republican effect)?

    Trump was not on the ballot in the midterms so that is irrelevant in terms of shy Trump effect, there was no shy GOP effect in 2012 or 2014 for example
  • LostPasswordLostPassword Posts: 1,890
    rcs1000 said:

    alex_ said:

    If Trafalgar systematically take other polls and skew them to the Republicans/Trump by 6%, and last time the polls all had a systematic bias to the Democrats of 6% then last time they would have been very successful. If all the other pollsters have now updated their methodology to account for this bias then Trafalgar will look a bit stupid if they've based their assumptions on the same methodology.

    Is that a serious suggestion? Is there any particular reason to suspect that's happening?

    In principle opinion polling is one of the defences that a democracy has against an election being stolen by ballot-stuffing, fraudulent counts and similar means.

    If a supposed polling company is intentionally producing poll results to be biased then the intention is to hide the blatant stealing of an election.
    There's no evidence that Trafalgar is a real, commercial organisation.

    It's not incorporated. It doesn't appear to have any employees or a business address. It claims on its PDFs to use a combination of calls and an on-line panel, yet there's no way to join the panel. It doesn't seem to have any customers (so where does the money come from to do these polls), and is dormant in between elections years. And it's run by a Republican operative.
    That's certainly suspicious. Thanks to others for their comments.
  • rcs1000rcs1000 Posts: 33,048
    HYUFD said:

    Trafalgar 2016

    (Difference to actual result in brackets).

    Colorado
    Trump 44% (+1)
    Clinton 45% (-3)

    Florida
    Trump 50% (+1)
    Clinton 46% (-2)

    Georgia
    Trump 52% (+2)
    Clinton 45% (-)

    Michigan
    Trump 49% (+1)
    Clinton 47% (-1)

    Nevada
    Trump 50% (+4)
    Clinton 45% (-3)

    North Carolina
    Trump 49% (-1)
    Clinton 44% (-3)

    Ohio
    Trump 49% (-3)
    Clinton 44% (-)

    Pennsylvania
    Trump 48% (-)
    Clinton 47% (-1)

    South Carolina
    Trump 53% (-2)
    Clinton 38% (-3)

    Utah
    Trump 40% (-6)
    Clinton 30% (+2)

    So all over the place really. Yes they may have got the winner right more often than not, but they certainly are not what you’d call “accurate”.
    So Trafalgar correctly forecast the winner of every state in that list then, except Nevada.

    The fact they also correctly forecast Clinton would win Colorado shows they are not just a pro Trump pollster in every swing state either
    They overstated Trump's relative position in Colorado, Florida, Georgia, Michigan, Nevada, North Carolina, Pennsylvania and South Carolina.

    But because they called the "not particularly swing" state of Colorado right, "they are not just a pro Trump pollster".

    I actually allege more than that.

    I don't believe they are a pollster at all.
  • rcs1000rcs1000 Posts: 33,048
    rcs1000 said:

    I called his cell, and it rang out and got voicemail.

    I also Tweeted to him with my questions. No response yet.
  • alex_alex_ Posts: 2,953
    rcs1000 said:

    I called his cell, and it rang out and got voicemail.

    Did you leave a message? He's got an opportunity to get the word out to a massive political betting community!
Sign In or Register to comment.