I think the views of the Gogglebox people are more informative at this stage of the election cycle than opinion polls. They slag off Boris too
"Last week, Starmer raised the issue of the 10 p.m. curfew at Prime Minister's Questions. He asked Johnson whether he could provide any scientific evidence to back up the measure. Johnson struggled to do so. To those sitting in the Commons chamber, it was seen as a turning point. With a vote on maintaining the 10 p.m. curfew due this week, there was a chance that Labour could work with rebel Tory MPs to reject the measure. However, not long after the session, Starmer confirmed that his party would support maintaining it. A similar thing happened with the rule of six — a Labour shadow health minister criticised it at the despatch box and then Labour backed it anyway.
On the Labour benches the concern is that Starmer could develop a reputation for never taking a side or fence-sitting. A taste of this came in last month's Gogglebox verdict on his Labour conference speech where the various sofa critics asked if he was even Labour"
Funny, I recall Gogglebox being ignored here when I posted clips about Johnson last year.
It's only funny, in the way you intend it, if it were me who ignored (downplayed?) it, and it wasn't
So do you think clips from last year that showed Johnson being slaughtered by Gogglebox were useful or not? I would tend to think not.
Were they at this stage of the Electoral cycle?
Why should it matter? It's like going to Twitter for your hot takes.
No, the gogglebox crew are much less politically engaged than those on twitter
Then why are their opinions now more relevant than last year? Is it perhaps because they say something you agree with?
I didn't say anything about their opinions from last year!
Do you think their opinions in the run up to the election (not during it, before) are relevant, bearing in mind they were absolutely destroying Johnson?
I didn't see it, so can't really comment one way or the other. But they are slaughtering him now, and I think that is relevant
They were slaughtering Johnson and he won in a landslide. Therefore I think their opinions on Starmer are perhaps not very relevant, to be honest.
Think what you like
That's fine. I was just pointing out hypocrisy from some, when I posted similar videos last year I was told they were irrelevant and not reflective of reality. Now they're pro-Johnson, suddenly silence. Odd that.
Who is pro Johnson? Not the Goggleboxers, they were taking the piss out of him on this weeks show
Who are these "some"? Name names!
Why on Earth are people tagging your post off topic? Silly people.
People HERE, were saying Gogglebox was irrelevant when I posted clips last year of them destroying Johnson. My point is now that there are clips destroying Keir, they're suddenly relevant. I get it, Tories will Tory, Labour will Labour.
I (now) go by polling, it got 2019 right. Gogglebox did not.
But I am the one saying Gogglebox views are relevant, and I was not saying they were irrelevant when they slagged off Boris. They still are slagging off Boris, and I think that is relevant.
So (regretting this already) what on earth are you talking about?
I think the views of the Gogglebox people are more informative at this stage of the election cycle than opinion polls. They slag off Boris too
"Last week, Starmer raised the issue of the 10 p.m. curfew at Prime Minister's Questions. He asked Johnson whether he could provide any scientific evidence to back up the measure. Johnson struggled to do so. To those sitting in the Commons chamber, it was seen as a turning point. With a vote on maintaining the 10 p.m. curfew due this week, there was a chance that Labour could work with rebel Tory MPs to reject the measure. However, not long after the session, Starmer confirmed that his party would support maintaining it. A similar thing happened with the rule of six — a Labour shadow health minister criticised it at the despatch box and then Labour backed it anyway.
On the Labour benches the concern is that Starmer could develop a reputation for never taking a side or fence-sitting. A taste of this came in last month's Gogglebox verdict on his Labour conference speech where the various sofa critics asked if he was even Labour"
Funny, I recall Gogglebox being ignored here when I posted clips about Johnson last year.
It's only funny, in the way you intend it, if it were me who ignored (downplayed?) it, and it wasn't
So do you think clips from last year that showed Johnson being slaughtered by Gogglebox were useful or not? I would tend to think not.
Were they at this stage of the Electoral cycle?
Why should it matter? It's like going to Twitter for your hot takes.
No, the gogglebox crew are much less politically engaged than those on twitter
Then why are their opinions now more relevant than last year? Is it perhaps because they say something you agree with?
I didn't say anything about their opinions from last year!
Do you think their opinions in the run up to the election (not during it, before) are relevant, bearing in mind they were absolutely destroying Johnson?
I didn't see it, so can't really comment one way or the other. But they are slaughtering him now, and I think that is relevant
They were slaughtering Johnson and he won in a landslide. Therefore I think their opinions on Starmer are perhaps not very relevant, to be honest.
Think what you like
That's fine. I was just pointing out hypocrisy from some, when I posted similar videos last year I was told they were irrelevant and not reflective of reality. Now they're pro-Johnson, suddenly silence. Odd that.
Who is pro Johnson? Not the Goggleboxers, they were taking the piss out of him on this weeks show
Who are these "some"? Name names!
Why on Earth are people tagging your post off topic? Silly people.
People HERE, were saying Gogglebox was irrelevant when I posted clips last year of them destroying Johnson. My point is now that there are clips destroying Keir, they're suddenly relevant. I get it, Tories will Tory, Labour will Labour.
I (now) go by polling, it got 2019 right. Gogglebox did not.
But I am the one saying Gogglebox views are relevant, and I was not saying they were irrelevant when they slagged off Boris. They still are slagging off Boris, and I think that is relevant.
So (regretting this already) what on earth are you talking about?
I wasn't saying you alone/specifically, I said when I posted Gogglebox clips here last year, there were people who got very animated about them being wrong about Johnson. They're now silent because they're now slagging off Starmer.
I don't mean to confuse you.
I don't think Gogglebox views are relevant, as 2019 proved.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
I think from the experts, I could see it in their faces, the cases are out of control and getting worse and the only course of action left is a lockdown.
As I said previously, I wish it wasn't the case. But it's now inevitable.
If that is the case, why did they not say ' we need a national lockdown'? You seem desperate to have one for some reason. We are walking a tight-rope - trying to maintain as much of the economy and education as possible, while keeping cases as low as we can, while we wait for the vaccine(s). You want to go hard on the virus - I understand that, but this ignores the evident harms caused by this harshest of all measures. How much mental health trauma, missed cancer diagnosis, business failure are you happy to see for your National Lockdown?
I do think we need a national lockdown but I could see from their faces the data was looking grim and they had surely conceded privately another lockdown is coming.
Why would they not say that? Well not to undermine the PM, which they have been told off for doing before.
We need another lockdown because it's necessary. Cases are totally out of control and the current measures are not working.
Based on what are you claiming that they have already privately conceded it? You may want them to have done so, but that does not mean it actually happened.
I could see it from their faces, that's what I think they've conceded privately. Of course I don't know for sure.
OK, so it's a projection of your own opinion more than anything.
Well it's what I think but you disagree and that's fine, no worries.
My issue isn't with what you think, it's that you are claiming to know what they are thinking privately.
I think privately, they have conceded another lockdown is coming - and I could see that from their faces.
I don't know for sure, neither do you. It's what I think and you disagree. No worries.
I don't think we've ever agreed about anything, it's all in good fun.
- being too complacent about travel from Europe in January and February - moving OAPs into homes from hospitals without testing them - developing the NHS-X app rather than using Apple/Google technology - closing schools at all, or at least without a clear plan to reopen them.
The first let the virus in in the first place (at least with the numbers and speed it came), the second may have cost 10-20k lives, the third has meant we're unprepared for the second wave and the fourth has blighted a year's schooling for a generation.
However, no. 1 accorded with expert advice, so I think it's mainly the other three I'd hold them responsible for.
And afaik there's no evidence that Prime Minister Starmer would have done anything different on any of them.
Nevertheless a government that makes a mistake is accountable for it; saying the other guy would have done the same doesn’t give you a free pass (the Tories would also have invaded Iraq, for example, yet Blair is to blame).
To your list I would add:
- slowing down the testing programme and actively restricting testing as the first wave came to an end; - allowing Cummings to get away with his lies fatally undermining public confidence in the fairness and reasonableness of the first lockdown; - the confusion created by Bozo’s blustering during the first address to the nation including torpedoing his own ‘clear’ five point scale at launch: - the never ending chopping and changing of restrictions imposed on businesses in different sectors and locations, countries moving in an out of quarantine, disrupting lives and business and allowing no-one to plan ahead; - the increasingly obvious fact that lack of a long term plan and political in-fighting right up to Cabinet is hampering our response. - appointing chums like Harding to big jobs they have no qualifications for
I’d also add not extending the Brexit transition in the totally unforeseen circumstances of the virus crisis to remove a further source of uncertainty for business.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
You're again missing the point entirely. The Florida polling average also shows Biden in the lead, so it would have to happen in both Florida and Wisconsin, and Trump would also have to hold Arizona.
Likewise there's also the chance that the polls are wrong in the OTHER direction and Biden has a 6% swing from the average. In which case all bets are off.
Andy C is (possibly unfairly, I must admit) cast as one of the most pro-lockdown voices on PB.
Even he would keep hospitality open.
It does see as if government wants to be seen to "do something", if if that something is not supported by the evidence.
For maximum clarity:
- I think the 10pm curfew has had negligible evidence presented for it and smacks of a compromise of being "seen to do something." I'd eliminate it as part of my package. - Table service only, separated by 2 metres, and a strict limit of people through the door. - Policed by spot checks; any violation sees loss of licence (those who abide by the rules shouldn't be disadvantaged compared to people playing fast and loose) - Targetted direct support (such as a monthly grant equal to their running costs as reported to HMRC for the average of the last two comparable months (eg October 2019 and October 2018) - Monitoring of the situation (ongoing estimates of the infection levels from within each of pubs, restaurants, etc) with a view to closing specifically temporarily if still necessary until the next item comes along - Widespread rolling out of the spit-tests that were being pioneered in June. Pubs and restaurants can have two areas if necessary (those compatible with the above, and one where normal rules apply) subject to taking the spit test, waiting a half hour or however long for the result, and finding a negative outcome (with emphasis to the public that this isn't guaranteed safe - some infectees will get through but the average spread will be way down, and if you can't accept that, don't go in).
I think from the experts, I could see it in their faces, the cases are out of control and getting worse and the only course of action left is a lockdown.
As I said previously, I wish it wasn't the case. But it's now inevitable.
If that is the case, why did they not say ' we need a national lockdown'? You seem desperate to have one for some reason. We are walking a tight-rope - trying to maintain as much of the economy and education as possible, while keeping cases as low as we can, while we wait for the vaccine(s). You want to go hard on the virus - I understand that, but this ignores the evident harms caused by this harshest of all measures. How much mental health trauma, missed cancer diagnosis, business failure are you happy to see for your National Lockdown?
I do think we need a national lockdown but I could see from their faces the data was looking grim and they had surely conceded privately another lockdown is coming.
Why would they not say that? Well not to undermine the PM, which they have been told off for doing before.
We need another lockdown because it's necessary. Cases are totally out of control and the current measures are not working.
Based on what are you claiming that they have already privately conceded it? You may want them to have done so, but that does not mean it actually happened.
I could see it from their faces, that's what I think they've conceded privately. Of course I don't know for sure.
OK, so it's a projection of your own opinion more than anything.
Well it's what I think but you disagree and that's fine, no worries.
My issue isn't with what you think, it's that you are claiming to know what they are thinking privately.
I think privately, they have conceded another lockdown is coming - and I could see that from their faces.
I don't know for sure, neither do you. It's what I think and you disagree. No worries.
I don't think we've ever agreed about anything, it's all in good fun.
I think from their faces they were trying to portray a "this is serious" image.
I repeat for the third or fourth time my question: If a national lockdown is avoided and R comes back down below 1 then will you give credit to the Government for achieving what you seem to think is impossible to achieve?
That's your opinion, I have mine. Just as welcome to air it as you.
Rather than cluttering the board with it, I will leave it there.
I'm not giving an opinion, I'm asking you a question. You think a lockdown is "necessary" don't you?
So I am asking you again: If (big if) a national lockdown is avoided and R comes back down below 1 then will you give credit to the Government for achieving what you seem to think is impossible to achieve?
- being too complacent about travel from Europe in January and February - moving OAPs into homes from hospitals without testing them - developing the NHS-X app rather than using Apple/Google technology - closing schools at all, or at least without a clear plan to reopen them.
The first let the virus in in the first place (at least with the numbers and speed it came), the second may have cost 10-20k lives, the third has meant we're unprepared for the second wave and the fourth has blighted a year's schooling for a generation.
However, no. 1 accorded with expert advice, so I think it's mainly the other three I'd hold them responsible for.
And afaik there's no evidence that Prime Minister Starmer would have done anything different on any of them.
Fair enough take. But politics does not work this way. When things go horribly run under a government the public do not conduct forensic counterfactuals asking themselves if the opposition would have been any better. For example, the financial crash and resulting economic downturn and crisis in the public finances which dominated the GE of 2010. There was no evidence that the situation would have been better or would have been better handled by a Cameron Conservative government rather than Brown's Labour one - the opposite if anything - but this did not prevent a narrative of "Labour's mess" taking root. Similarly here, regardless of how you think Starmer would have performed relative to Johnson, if this pans out as badly as it looks like doing, this government will own it.
I agree. The same thing is true of the ERM in the early 90s. The fact that Brown was if anything more enthusiastic for ERM membership than the Conservatives didn't have any effect on the public mind - the Conservatives still got the blame. People hold governments to a much higher standard than Oppositions. This is justifiable when they have access to more information, such as on military or intelligence matters. But on the economy most statistics are publicly available. I think this virus has fallen somewhere between those two polar cases.
Not really about higher standards. If you are in office, you are responsible for your decisions and actions. Oppositions are only responsible for their words to the extent that any other politician is, which isn’t that far.
Very good article Mike. I suspect there might be a few things there outside, as you said, the hangover fears from 2016, namely:
1. The economic data: As @TSE pointed out, Trump is still ahead on the economy. As I pointed out, Gallup says 56% of Americans feel they are better off than 4 years ago. I think betters are still thinking that voters, at the end of the day will go with their wallets;
2. The potential for a Biden gaffe. Friday for example. Biden's reply to a reporter when asked whether voters deserved to know about what he thought about the packing of the Supreme Court of "no, they don't" is a gift to the Republicans. (Go to 2:28 https://www.youtube.com/watch?v=jWT2_m0F5OQ)
3. The dynamic can change quickly. Maybe a bigger thing is that the dynamic can change very quickly. We saw that with Jeremy Corbyn in 2017 (and it also happened with Ronald Reagan in 1980). Last week, the talk was all of landslides. Now we have polls showing, at the state level, stabilisation.
4. The known unknowns (or unknown ones). Will there be another October surprise? Will Biden get ill? Will more documents be released around the Christopher Steele case etc
Also, thanks for the link. This might be useful as well re NC. Squinting at it, I think that probably close to 80% of Democrats who have returned their ballots voted at the last election vs 75% for the state, so definitely pull forward but also some new voters. Interesting though that the growth rate for request rate for Mail In Ballots is now slowing down fairly chunkily:
On the vote by mail request slowdown in North Carolina, early voting is open soon so I expect a number of people who originally perhaps would have voted by mail have changed their minds and decided they'll do it in person. There's a small risk of catching the virus whilst voting but for I think not a small number of Democrats they feel there is less hi-jinks that can go on if they vote in person. Early voting isn't open yet in many states, will be interesting to see how the numbers carry on when it opens.
Gogglebox is biased towards mouthy exhibitionists and thus is not a representative sample. It is not, however, irrelevant since mouthy exhibitionists have a vote.
Andy C is (possibly unfairly, I must admit) cast as one of the most pro-lockdown voices on PB.
Even he would keep hospitality open.
It does see as if government wants to be seen to "do something", if if that something is not supported by the evidence.
For maximum clarity:
- I think the 10pm curfew has had negligible evidence presented for it and smacks of a compromise of being "seen to do something." I'd eliminate it as part of my package. - Table service only, separated by 2 metres, and a strict limit of people through the door. - Policed by spot checks; any violation sees loss of licence (those who abide by the rules shouldn't be disadvantaged compared to people playing fast and loose) - Targetted direct support (such as a monthly grant equal to their running costs as reported to HMRC for the average of the last two comparable months (eg October 2019 and October 2018) - Monitoring of the situation (ongoing estimates of the infection levels from within each of pubs, restaurants, etc) with a view to closing specifically temporarily if still necessary until the next item comes along - Widespread rolling out of the spit-tests that were being pioneered in June. Pubs and restaurants can have two areas if necessary (those compatible with the above, and one where normal rules apply) subject to taking the spit test, waiting a half hour or however long for the result, and finding a negative outcome (with emphasis to the public that this isn't guaranteed safe - some infectees will get through but the average spread will be way down, and if you can't accept that, don't go in).
Even though we both see the situation from different view points I think our end goal is basically the same. Prevent the economic and social destruction that comes with a second lockdown or whatever the tier 3 measures are.
There are so many things that the government could be doing better. Rapid testing is specifically an area that the government has let perfect be the enemy of good. The Abbott rapid test hits almost 90% accuracy compared to PCR testing, it is a viable solution for wide community and venue testing.
What do you think of moving from isolation to hotel based separation for people who test positive? To me it is close to a silver bullet which will bring the R down quite drastically.
@OGH. Thanks for the header. FWIW, my three halfpenny:
1. Voters are spooked by 2016, despite reassurances from some pollsters they've corrected for 2016 errors in weighting 2. Democrats will always think they might lose even when they are 20 points ahead 3. Given how many states are categorized as toss up or lean atm, the difference between a marginal loss, a small win, a comfortable win and a landslide is not that many % change in actual result, particularly in comparison with the MoE of some of these polls. If there is any correlation in the error of the different pollsters, it could have a huge impact one way or the other 4. Uncertainty is higher than in previous elections - we know fairly well how voters turnout when they have to show up in person or are given the choice to use mail in ballots; we know less well what turnout will be when there are very few physical polling stations, lines may be many hours long on the day, and some are forced to use mail in ballots when they'd rather not 5. Along the same lines of uncertainty - we know that emotions are very high in this election, especially in the younger ethnic demo. Who benefits most from this, and how will this national rage impact younger voter turn out? We don't really know. 6. Tump: his MO is create chaos and unpredictability, spread doubt and confusion.
Then there is the fact that most bettors in a major event like this are not professional bettors like the best on here. Many will be voting on hopes or fears, and others' estimation of odds they are willing to take will reflect loss aversion rather than hard mathematical calculation.
I am still sticking my neck out and predicting a landslide, although I am unsure whether it will be vanilla or Reagan. Best bet is around 125 EC votes for Trump. But I have low confidence in this prediction.
I think from the experts, I could see it in their faces, the cases are out of control and getting worse and the only course of action left is a lockdown.
As I said previously, I wish it wasn't the case. But it's now inevitable.
If that is the case, why did they not say ' we need a national lockdown'? You seem desperate to have one for some reason. We are walking a tight-rope - trying to maintain as much of the economy and education as possible, while keeping cases as low as we can, while we wait for the vaccine(s). You want to go hard on the virus - I understand that, but this ignores the evident harms caused by this harshest of all measures. How much mental health trauma, missed cancer diagnosis, business failure are you happy to see for your National Lockdown?
I do think we need a national lockdown but I could see from their faces the data was looking grim and they had surely conceded privately another lockdown is coming.
Why would they not say that? Well not to undermine the PM, which they have been told off for doing before.
We need another lockdown because it's necessary. Cases are totally out of control and the current measures are not working.
Based on what are you claiming that they have already privately conceded it? You may want them to have done so, but that does not mean it actually happened.
I could see it from their faces, that's what I think they've conceded privately. Of course I don't know for sure.
OK, so it's a projection of your own opinion more than anything.
Well it's what I think but you disagree and that's fine, no worries.
My issue isn't with what you think, it's that you are claiming to know what they are thinking privately.
I think privately, they have conceded another lockdown is coming - and I could see that from their faces.
I don't know for sure, neither do you. It's what I think and you disagree. No worries.
I don't think we've ever agreed about anything, it's all in good fun.
Gogglebox is biased towards mouthy exhibitionists and thus is not a representative sample. It is not, however, irrelevant since mouthy exhibitionists have a vote.
Well one might expect mouthy exhibitionists to be biased towards the current incarnation of the Tories ?
Every day there is a new story (led by comments from the Mayor's office) stating how London numbers are "rising alarmingly" and "new harsh London wide restrictions" are inevitable. And then on almost every occasion the story doesn't justify the hype - whether it be the suggestions about dodgy numbers to do with University students, or, in this case the possibility that local authorities simply taking tough localised action targeting places where rules are being blatantly flouted has been sufficient to bring case numbers down.
The message - we don't have to have crude national measures to get this under control, and relying on individuals to follow "guidance", (but when enforcement of this is almost non-existent) just some tough local measures homing in on problem sites where "super spreader" events are likely to take place. Reward the businesses and individuals that are taking their responsibilities seriously, and go in hard on those that don't.
There's enough evidence whether here or abroad that we can control this thing without completely shutting down the economy. So lets do it.
Sporting have edged their spread up a point - now 316/322. The few weekend polls were sort of ok (ish) for Trump so I wouldn't have expected much movement. The Exchanges tightened a bit, again for no obvious reason, but then they are wildly out of sync with the forecast models so maybe the odds are yielding to rationality, as they tend to do as an event approach start time.
And of course each day that is ticked off moves the dial fractionally towards Biden. That is probably sufficient to explain Nate Silver moving Trump's probable win percent down a point to 13. Trump needs something big, and soon. I don't see where he gets it from. Covid is getting worse again in the USA. It is his worst topic, and it impacts on his best one, the economy.
I am increasingly sure Biden is going to win this. I think the polls are right - Biden by about 10% nationally. States which will play relatively well for Trump are Florida, Virginia, maybe North Carolina. That's not enough. We have spent a lot of time discussing here the possibilty of the polls understating Trump's position. Little attention has been given to the equally likely possibility they are understating Biden's. It is perfectly reasonable, sane even, to assume the polls are right and certainly from a betting viewpoint it would be hard to justify any other stance.
That's my stance, fwiw. Naturally, I'll be quick to report if it changes.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
It's going to be hilarious if Trump manages to win. I really hope he does it.
@kinabalu someone suggests playing Russian Roulette and hands you a revolver with eight chambers, one bullet.
Gun against your temple and you have to shoot yourself once. Would you "clearly" survive? Or would you be favoured to survive?
Don't want to intervene in another of your fights with yourself, but neither. In a race with one horse at 5/1 and nothing else shorter than 10 the horse is a clear favourite but with a 17% implied probability of winning.
On topic, there are two possibilities that could lead to punters giving Trump a better chance than polls suggest.
Firstly, it could be that the polls are systematically quite a long way off. I think Gallowgate's figures are convincing on this, which as I understand it is that even if they are as badly off in all states as they were in the two most inaccurately polled states in 2016 in terms of overstating Clinton and understating Trump, Biden would still win (albeit narrowly). Not only is that, as Gallowgate says, a really extreme assumption made to illustrate the point, it also implies pollsters have learned nothing from 2016. In fact they have corrected (in particular) their undersampling of non-college educated whites, which is a non-trivial shift to Trump in terms of his polling position - the same methodology in 2016 would simply NOT have underestimated Trump (or at least not to the same degree). So the systematic error point is possible (of course) but has to be a low risk.
Secondly, there could be a late surprise that shifts the polls appreciably. There are two problems with this. One is that there is a lot of postal and early voting going on that a suprise cannot now influence - that only increases over time. So the surprise needs to be increasingly large to create a shift in dynamic as October wears on. Secondly, this race has been remarkable in its stability in terms of the narrow corridor of polling for BOTH candidates, and the extent to which polls indicate voters are absolutely certain about their choice. There just aren't a lot of 6/10 likelihood Biden voters who are available to tip into the Trump column in the event of a Biden brain-fart. They are also very well known figures - it's possible some major, game-changing revelation about Biden has been kept hidden over a decades long career that took him to the Vice Presidency, but it isn't likely. No doubt mud will be flung by Fox and so on (and mud thrown back too) but is it likely to be all that sticky?
TL/DR - there remains value in Biden at this stage.
I think the views of the Gogglebox people are more informative at this stage of the election cycle than opinion polls. They slag off Boris too
"Last week, Starmer raised the issue of the 10 p.m. curfew at Prime Minister's Questions. He asked Johnson whether he could provide any scientific evidence to back up the measure. Johnson struggled to do so. To those sitting in the Commons chamber, it was seen as a turning point. With a vote on maintaining the 10 p.m. curfew due this week, there was a chance that Labour could work with rebel Tory MPs to reject the measure. However, not long after the session, Starmer confirmed that his party would support maintaining it. A similar thing happened with the rule of six — a Labour shadow health minister criticised it at the despatch box and then Labour backed it anyway.
On the Labour benches the concern is that Starmer could develop a reputation for never taking a side or fence-sitting. A taste of this came in last month's Gogglebox verdict on his Labour conference speech where the various sofa critics asked if he was even Labour"
Funny, I recall Gogglebox being ignored here when I posted clips about Johnson last year.
It's only funny, in the way you intend it, if it were me who ignored (downplayed?) it, and it wasn't
So do you think clips from last year that showed Johnson being slaughtered by Gogglebox were useful or not? I would tend to think not.
Were they at this stage of the Electoral cycle?
Why should it matter? It's like going to Twitter for your hot takes.
No, the gogglebox crew are much less politically engaged than those on twitter
Then why are their opinions now more relevant than last year? Is it perhaps because they say something you agree with?
I didn't say anything about their opinions from last year!
Do you think their opinions in the run up to the election (not during it, before) are relevant, bearing in mind they were absolutely destroying Johnson?
I didn't see it, so can't really comment one way or the other. But they are slaughtering him now, and I think that is relevant
They were slaughtering Johnson and he won in a landslide. Therefore I think their opinions on Starmer are perhaps not very relevant, to be honest.
Think what you like
That's fine. I was just pointing out hypocrisy from some, when I posted similar videos last year I was told they were irrelevant and not reflective of reality. Now they're pro-Johnson, suddenly silence. Odd that.
Who is pro Johnson? Not the Goggleboxers, they were taking the piss out of him on this weeks show
Who are these "some"? Name names!
Why on Earth are people tagging your post off topic? Silly people.
People HERE, were saying Gogglebox was irrelevant when I posted clips last year of them destroying Johnson. My point is now that there are clips destroying Keir, they're suddenly relevant. I get it, Tories will Tory, Labour will Labour.
I (now) go by polling, it got 2019 right. Gogglebox did not.
But I am the one saying Gogglebox views are relevant, and I was not saying they were irrelevant when they slagged off Boris. They still are slagging off Boris, and I think that is relevant.
So (regretting this already) what on earth are you talking about?
I can see why people are getting frustrated but it's probably 4 years until the next election. I'd rather he slowly build a solid lead as he regains trust rather than something fast and ephemeral based on tomorrow's chip paper.
Very good article Mike. I suspect there might be a few things there outside, as you said, the hangover fears from 2016, namely:
1. The economic data: As @TSE pointed out, Trump is still ahead on the economy. As I pointed out, Gallup says 56% of Americans feel they are better off than 4 years ago. I think betters are still thinking that voters, at the end of the day will go with their wallets;
2. The potential for a Biden gaffe. Friday for example. Biden's reply to a reporter when asked whether voters deserved to know about what he thought about the packing of the Supreme Court of "no, they don't" is a gift to the Republicans. (Go to 2:28 https://www.youtube.com/watch?v=jWT2_m0F5OQ)
3. The dynamic can change quickly. Maybe a bigger thing is that the dynamic can change very quickly. We saw that with Jeremy Corbyn in 2017 (and it also happened with Ronald Reagan in 1980). Last week, the talk was all of landslides. Now we have polls showing, at the state level, stabilisation.
4. The known unknowns (or unknown ones). Will there be another October surprise? Will Biden get ill? Will more documents be released around the Christopher Steele case etc
Also, thanks for the link. This might be useful as well re NC. Squinting at it, I think that probably close to 80% of Democrats who have returned their ballots voted at the last election vs 75% for the state, so definitely pull forward but also some new voters. Interesting though that the growth rate for request rate for Mail In Ballots is now slowing down fairly chunkily:
On the vote by mail request slowdown in North Carolina, early voting is open soon so I expect a number of people who originally perhaps would have voted by mail have changed their minds and decided they'll do it in person. There's a small risk of catching the virus whilst voting but for I think not a small number of Democrats they feel there is less hi-jinks that can go on if they vote in person. Early voting isn't open yet in many states, will be interesting to see how the numbers carry on when it opens.
@ Pulpstar: that is exactly what my wife has decided to do. Initially, she planned to mail her ballot in, but Trump's behaviours and speculations/threats (threculations??) have convinced her to show up in person during early voting, going on the first day and as many times as necessary to find a time when the line is short enough.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
It's going to be hilarious if Trump manages to win. I really hope he does it.
Very good article Mike. I suspect there might be a few things there outside, as you said, the hangover fears from 2016, namely:
1. The economic data: As @TSE pointed out, Trump is still ahead on the economy. As I pointed out, Gallup says 56% of Americans feel they are better off than 4 years ago. I think betters are still thinking that voters, at the end of the day will go with their wallets;
2. The potential for a Biden gaffe. Friday for example. Biden's reply to a reporter when asked whether voters deserved to know about what he thought about the packing of the Supreme Court of "no, they don't" is a gift to the Republicans. (Go to 2:28 https://www.youtube.com/watch?v=jWT2_m0F5OQ)
3. The dynamic can change quickly. Maybe a bigger thing is that the dynamic can change very quickly. We saw that with Jeremy Corbyn in 2017 (and it also happened with Ronald Reagan in 1980). Last week, the talk was all of landslides. Now we have polls showing, at the state level, stabilisation.
4. The known unknowns (or unknown ones). Will there be another October surprise? Will Biden get ill? Will more documents be released around the Christopher Steele case etc
Also, thanks for the link. This might be useful as well re NC. Squinting at it, I think that probably close to 80% of Democrats who have returned their ballots voted at the last election vs 75% for the state, so definitely pull forward but also some new voters. Interesting though that the growth rate for request rate for Mail In Ballots is now slowing down fairly chunkily:
On the vote by mail request slowdown in North Carolina, early voting is open soon so I expect a number of people who originally perhaps would have voted by mail have changed their minds and decided they'll do it in person. There's a small risk of catching the virus whilst voting but for I think not a small number of Democrats they feel there is less hi-jinks that can go on if they vote in person. Early voting isn't open yet in many states, will be interesting to see how the numbers carry on when it opens.
@ Pulpstar: that is exactly what my wife has decided to do. Initially, she planned to mail her ballot in, but Trump's behaviours and speculations/threats (threculations??) have convinced her to show up in person during early voting, going on the first day and as many times as necessary to find a time when the line is short enough.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
It's going to be hilarious if Trump manages to win. I really hope he does it.
Not everyone shares your passion for dangerous thrills!
Gogglebox is biased towards mouthy exhibitionists and thus is not a representative sample. It is not, however, irrelevant since mouthy exhibitionists have a vote.
Well one might expect mouthy exhibitionists to be biased towards the current incarnation of the Tories ?
Yes, I would expect a statistically significant pro "Boris" lean from a Gogglebox panel.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
It's going to be hilarious if Trump manages to win. I really hope he does it.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
It's going to be hilarious if Trump manages to win. I really hope he does it.
Very good article Mike. I suspect there might be a few things there outside, as you said, the hangover fears from 2016, namely:
1. The economic data: As @TSE pointed out, Trump is still ahead on the economy. As I pointed out, Gallup says 56% of Americans feel they are better off than 4 years ago. I think betters are still thinking that voters, at the end of the day will go with their wallets;
2. The potential for a Biden gaffe. Friday for example. Biden's reply to a reporter when asked whether voters deserved to know about what he thought about the packing of the Supreme Court of "no, they don't" is a gift to the Republicans. (Go to 2:28 https://www.youtube.com/watch?v=jWT2_m0F5OQ)
3. The dynamic can change quickly. Maybe a bigger thing is that the dynamic can change very quickly. We saw that with Jeremy Corbyn in 2017 (and it also happened with Ronald Reagan in 1980). Last week, the talk was all of landslides. Now we have polls showing, at the state level, stabilisation.
4. The known unknowns (or unknown ones). Will there be another October surprise? Will Biden get ill? Will more documents be released around the Christopher Steele case etc
Also, thanks for the link. This might be useful as well re NC. Squinting at it, I think that probably close to 80% of Democrats who have returned their ballots voted at the last election vs 75% for the state, so definitely pull forward but also some new voters. Interesting though that the growth rate for request rate for Mail In Ballots is now slowing down fairly chunkily:
On the vote by mail request slowdown in North Carolina, early voting is open soon so I expect a number of people who originally perhaps would have voted by mail have changed their minds and decided they'll do it in person. There's a small risk of catching the virus whilst voting but for I think not a small number of Democrats they feel there is less hi-jinks that can go on if they vote in person. Early voting isn't open yet in many states, will be interesting to see how the numbers carry on when it opens.
@ Pulpstar: that is exactly what my wife has decided to do. Initially, she planned to mail her ballot in, but Trump's behaviours and speculations/threats (threculations??) have convinced her to show up in person during early voting, going on the first day and as many times as necessary to find a time when the line is short enough.
Queue early, queue often?
Just taking Trump's advice ... No, seriously, the expectation is that there will be times when the line to vote in many hours long. As the polling station is only 2 miles from home, a quick drive by to see how long the line is and only stop if it does not look too long.
@OGH. Thanks for the header. FWIW, my three halfpenny:
1. Voters are spooked by 2016, despite reassurances from some pollsters they've corrected for 2016 errors in weighting 2. Democrats will always think they might lose even when they are 20 points ahead 3. Given how many states are categorized as toss up or lean atm, the difference between a marginal loss, a small win, a comfortable win and a landslide is not that many % change in actual result, particularly in comparison with the MoE of some of these polls. If there is any correlation in the error of the different pollsters, it could have a huge impact one way or the other 4. Uncertainty is higher than in previous elections - we know fairly well how voters turnout when they have to show up in person or are given the choice to use mail in ballots; we know less well what turnout will be when there are very few physical polling stations, lines may be many hours long on the day, and some are forced to use mail in ballots when they'd rather not 5. Along the same lines of uncertainty - we know that emotions are very high in this election, especially in the younger ethnic demo. Who benefits most from this, and how will this national rage impact younger voter turn out? We don't really know. 6. Tump: his MO is create chaos and unpredictability, spread doubt and confusion.
Then there is the fact that most bettors in a major event like this are not professional bettors like the best on here. Many will be voting on hopes or fears, and others' estimation of odds they are willing to take will reflect loss aversion rather than hard mathematical calculation.
I am still sticking my neck out and predicting a landslide, although I am unsure whether it will be vanilla or Reagan. Best bet is around 125 EC votes for Trump. But I have low confidence in this prediction.
I agree with pretty well all of this, especially the unusual uncertainty despite very steady polls.
Andy C is (possibly unfairly, I must admit) cast as one of the most pro-lockdown voices on PB.
Even he would keep hospitality open.
It does see as if government wants to be seen to "do something", if if that something is not supported by the evidence.
For maximum clarity:
- I think the 10pm curfew has had negligible evidence presented for it and smacks of a compromise of being "seen to do something." I'd eliminate it as part of my package. - Table service only, separated by 2 metres, and a strict limit of people through the door. - Policed by spot checks; any violation sees loss of licence (those who abide by the rules shouldn't be disadvantaged compared to people playing fast and loose) - Targetted direct support (such as a monthly grant equal to their running costs as reported to HMRC for the average of the last two comparable months (eg October 2019 and October 2018) - Monitoring of the situation (ongoing estimates of the infection levels from within each of pubs, restaurants, etc) with a view to closing specifically temporarily if still necessary until the next item comes along - Widespread rolling out of the spit-tests that were being pioneered in June. Pubs and restaurants can have two areas if necessary (those compatible with the above, and one where normal rules apply) subject to taking the spit test, waiting a half hour or however long for the result, and finding a negative outcome (with emphasis to the public that this isn't guaranteed safe - some infectees will get through but the average spread will be way down, and if you can't accept that, don't go in).
Even though we both see the situation from different view points I think our end goal is basically the same. Prevent the economic and social destruction that comes with a second lockdown or whatever the tier 3 measures are.
There are so many things that the government could be doing better. Rapid testing is specifically an area that the government has let perfect be the enemy of good. The Abbott rapid test hits almost 90% accuracy compared to PCR testing, it is a viable solution for wide community and venue testing.
What do you think of moving from isolation to hotel based separation for people who test positive? To me it is close to a silver bullet which will bring the R down quite drastically.
I honestly don't know. The implications of that if you've got a single-parent family or one with specific caring responsibilities would need to be carefully addressed. It would help a lot with adherence, but anyone pulled into it would need full financial support for the duration.
One thing that's key for me is to find ways to preserve as much as possible of normal life, because that's crucial to getting widespread opt-in and adherence. A second is direct support to people and industries most affected. A third is evidence-based solutions (and avoiding any wishful thinking or cherry-picking, because reality will bite us hard and people will react and overreact. Ex-engineer, ex-military, ex-skydiver, and current microlight pilot talking, there. And to lighten the tone, have a photo from yesterday evening racing the sunset back to my home airfield after a flight over Cheddar Gorge)
I mean, I get that there's an urge to balance the books, but given that we're still in the middle of the greatest natural disaster to affect all of us in our lifetimes, there's some prioritisation here. People shouting that we need to balance the books right now smack of - well, visualise yourself in a disaster movie, in a car trying to outrace a tidal wave. Someone on the back seat is shouting that we should slow down because driving slower is safer, and if we did 56 mph, it'd be more economical.
In the long run, we're going to have to pay for it, yes. We do, however, need to get to the long run first.
How many have been closed down? Reading the thread there was one in that stadium in Wales, but from the tweet it sounds like dozens have been dismantled.
The Nightingale Hospitals really did meet Johnson's "world beating" criteria. A magnificent feat!
@kinabalu someone suggests playing Russian Roulette and hands you a revolver with eight chambers, one bullet.
Gun against your temple and you have to shoot yourself once. Would you "clearly" survive? Or would you be favoured to survive?
I think you've missed my post clarifying but ok we can do this one too -
If I were to be placed in that unfortunate position I would indeed be "clearly favoured" to survive the shot.
Shades of De Niro and Walken here.
Odds of putting a man on Mars by 2035 say Musk 250/1 Nasa 500/1 1000/1 the field. On those quite possible numbers Musk is both clearly favoured to do it, and clearly not going to do it.
TBH 13.4 cases per 100,000 per week is a pretty low chance of meeting someone in your daily life. How low would Burnham have wanted it to go?
Exactly. This is a non-runner. God knows what Van Tam is playing at.
Edit: A more relevant discussion would be whether the North had it less seriously last time and therefore there is less immunity around.
Exactly, London has hit 20% immunity. The R requirement for exponential growth is just a lot higher than everywhere else in the country. Additionally the people who are likely to interact with each other (younger) probably have immunity rates north of 30%. The virus doesn't have as many potential hosts to infect.
I wouldn't be surprised if parts of zone 2 and 3 London have already achieved herd immunity, especially those areas close to transport hubs like Heathrow or St Pancras.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
You're being selective (again). There have been polls that have had him in the lead throughout this cycle, in fact in Ohio the majority of polls have had Trump in the lead throughout, so it is deeply disingenuous to say "back in the lead". So he will have to go against polling in Wisonsin, Florida and Arizona - not just WIsconsin.
But, hey, confirmation bias isn't just a river in Egypt.
Should be quite a few new USA polls out today including two news ones from NYT/Siena for Michigan and Wisconsin.
Of more interest perhaps is also 3 new polls from them coming out later in the week from North and South Carolina aswell as the little polled state of Alaska .
TBH 13.4 cases per 100,000 per week is a pretty low chance of meeting someone in your daily life. How low would Burnham have wanted it to go?
Exactly. This is a non-runner. God knows what Van Tam is playing at.
Edit: A more relevant discussion would be whether the North had it less seriously last time and therefore there is less immunity around.
Exactly, London has hit 20% immunity. The R requirement for exponential growth is just a lot higher than everywhere else in the country. Additionally the people who are likely to interact with each other (younger) probably have immunity rates north of 30%. The virus doesn't have as many potential hosts to infect.
I wouldn't be surprised if parts of zone 2 and 3 London have already achieved herd immunity, especially those areas close to transport hubs like Heathrow or St Pancras.
Heathrow is Zone 6. St Pancras is Zone 1.
The people who work at Heathrow live in and around Acton.
I wondered whether there might be a disproportionate number of Republicans who drop out of the polls between the Registered Voter and Likely Voter screens - shy Trump voters might show up in the polls as Republicans who currently say they might not vote.
I checked the large recent Pew poll - but I can't find the information I was looking for. Darn.
What I did find, which I thought was interesting, was that only 41% would be excited/relieved if Trump was re-elected, compared to 58% who would be disappointed/angry. Would seem like the country has turned against Trump.
But what's this? The same figures for 2016 were 38-60. On this measure Trump has won over some of the American public during the last four years. There is not a greater sense of disgust at Trump than there was in 2016.
The difference is that Biden's scores on this measure are 52-46, which are better than the 47-51 that Clinton received.
Not that much different from 2016 - scroll to 0:15
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
It's going to be hilarious if Trump manages to win. I really hope he does it.
OFF my Winterval card list!
I think @Dura_Ace is something of a Russian specialist ? That clearly encompasses Russian humour.
What lockdown on hospitality? Every pub and eatery in England is free to open from 5am to 10pm if they want to.
When there was a lockdown on hospitality, cases plummeted.
Isn’t he referring to the widely rumoured to be announced shutdown in the north (and maybe London)?
I don't see how any pub/restaurant can be financially viable if people were to actually follow the rules on no household mixing. How many people in pubs are primarily there with only people from their own household? As LadyG pointed out yesterday - why do that (except on special occasions) when you can do the same thing at home for a fraction of the price?
So either the rules are turned a blind eye to (as others suggest is happening all over the place), or the industry will collapse. Within a few weeks they'll be demanding to be put in Tier 3 so they can get the cash. Unless the Tier 2 rules are only to be guidance.
As i suggested above - if you're going to let pubs/restaurants stay open then do so in a way which gives most of them a chance of staying viable. Allow multiple households, but with the quid pro quo that they strictly follow the rules on masks and social distancing. And go in hard, backed up by local authority enforcement powers (NOT the police) on those that don't. And don't enforce through a system of every draconian fines, but by threats to shut down non-complying businesses. In the same way they do with Environmental Health and Licensing breaches.
I would close pubs and restaurants down completely, if nothing else.
Why? What evidence is there that pubs and restaurants being open has lead to the increase in cases?
JPMorgan analyzed data from 30 million Chase cardholders and Johns Hopkins University’s case tracker and found that higher restaurant spending in a state predicted a rise in new infections there three weeks later.
In-person restaurant spending was “particularly predictive.”
Conversely, higher spending at supermarkets predicted a slower spread of the virus.
The idea of overturning it is extraordinarily unpopular among women voters.
Why on earth would that be? Do they imagine what happens to their bodies is their business and nobody else's?
Of course. But it does demonstrate just how unrepresentative is Amy Coney Barrett, and the folly of forcing through her SC appointment before the election.
- being too complacent about travel from Europe in January and February - moving OAPs into homes from hospitals without testing them - developing the NHS-X app rather than using Apple/Google technology - closing schools at all, or at least without a clear plan to reopen them.
The first let the virus in in the first place (at least with the numbers and speed it came), the second may have cost 10-20k lives, the third has meant we're unprepared for the second wave and the fourth has blighted a year's schooling for a generation.
However, no. 1 accorded with expert advice, so I think it's mainly the other three I'd hold them responsible for.
And afaik there's no evidence that Prime Minister Starmer would have done anything different on any of them.
Reopening universities looks likes a disaster if you live in a uni town.
No, reopening universities isn't a disaster. We need to see it as an opportunity for young healthy people to acquire herd immunity before Christmas. Universities and students should be supported to stay isolated properly until herd immunity is achieved. It's the perfect setting, especially campus universities. It's almost like a large version of flu camp. I've been told that some universities think that their cohorts are reaching 40-50% immunity already after just a few weeks, when that hits 66% the virus will simply run out of hosts. Aiui early numbers at universities suggest around 80% of infections are asymptomatic (based on random sampling of swabs) and this is currently the best way for students to get back to normal.
It's damn scary for people who teach, clean, or serve food in Universities - and their vulnerable relatives. (I don't think "proper isolation" really exists, and it definitely doesn't when the slow-witted vice chancellors want to continue face-to-face teaching.) I personally have a tough call to make in this regard.
I'm a little skeptical of asymptomatic percentages at this stage. Early data tends to overestimate.
Pennsylvania Clinton 48.9 (47.5) Trump 45.2 (48.2)
Wisconsin Clinton 49.6 (46.5) Trump 44.3 (47.2)
So let's take the worst case scenario for Biden - a Michigan style error on Trump (+3.3%) and a Wisconsin style error on Biden (-3.1%). Polling average first, calculated error in brackets.
The numbers do not lie. Even if the worst possible polling deviations from 2016 are applied to these swing states, Biden still wins.
And as a follow up to this, as far as I'm aware 538 is already applying a dampener on Biden's figures - at least on a national level. Likewise the possibility of polls being 3% out in both directions at the same time in every state is miniscule.
So on the same error as 2016 Trump wins Florida and Biden wins Pennsylvania and Wisconsin by 0.1%, on that basis it could still be a nail biting night for Democrats
Yes but that kind of error from the polls only occurred in Wisconsin. The chance of the same error happening in EVERY state is tiny. It didn't even happen in 2016.
Trump may only need it to happen in Wisconsin, with some polls already showing him back ahead in Florida and Ohio and Arizona only Wisconsin would be enough for him to be re elected even if he loses Pennsylvania and Michigan
It's going to be hilarious if Trump manages to win. I really hope he does it.
Why?
He is obviously more entertaining than Biden. There is almost no substantive policy difference between them as they are both venal scum who exclusively serve the interests of capital and the 1%. So we might as well have the comedy option.
I would close pubs and restaurants down completely, if nothing else.
Why? What evidence is there that pubs and restaurants being open has lead to the increase in cases?
JPMorgan analyzed data from 30 million Chase cardholders and Johns Hopkins University’s case tracker and found that higher restaurant spending in a state predicted a rise in new infections there three weeks later.
In-person restaurant spending was “particularly predictive.”
Conversely, higher spending at supermarkets predicted a slower spread of the virus.
Flaw being that restaurant spending probably correlates with all sorts of other social activity. Most obviously where there is a ‘lockdown’ or where people are voluntarily avoiding social mingling, restaurant spending will be low.
So "food based" pubs to stay open on Merseyside. Boozers to shut. Spoons stays. Makes zero sense and reveals a class bias imho.
Aren't pretty much all pubs "food based" these days? Or is this just - edgy cocktail bars in town must close, but everything else can stay open?
No. The majority of pubs don't serve food.
Sadly, down here in the deep south they nearly all do. It's increasingly difficult to go for a pint without having to suffer the smell of deep frying and tables laden with cutlery, menus and other rubbish. A great shame. Maybe I need to move back up t'north.
On topic, there are two possibilities that could lead to punters giving Trump a better chance than polls suggest.
Firstly, it could be that the polls are systematically quite a long way off. I think Gallowgate's figures are convincing on this, which as I understand it is that even if they are as badly off in all states as they were in the two most inaccurately polled states in 2016 in terms of overstating Clinton and understating Trump, Biden would still win (albeit narrowly). Not only is that, as Gallowgate says, a really extreme assumption made to illustrate the point, it also implies pollsters have learned nothing from 2016. In fact they have corrected (in particular) their undersampling of non-college educated whites, which is a non-trivial shift to Trump in terms of his polling position - the same methodology in 2016 would simply NOT have underestimated Trump (or at least not to the same degree). So the systematic error point is possible (of course) but has to be a low risk.
Secondly, there could be a late surprise that shifts the polls appreciably. There are two problems with this. One is that there is a lot of postal and early voting going on that a suprise cannot now influence - that only increases over time. So the surprise needs to be increasingly large to create a shift in dynamic as October wears on. Secondly, this race has been remarkable in its stability in terms of the narrow corridor of polling for BOTH candidates, and the extent to which polls indicate voters are absolutely certain about their choice. There just aren't a lot of 6/10 likelihood Biden voters who are available to tip into the Trump column in the event of a Biden brain-fart. They are also very well known figures - it's possible some major, game-changing revelation about Biden has been kept hidden over a decades long career that took him to the Vice Presidency, but it isn't likely. No doubt mud will be flung by Fox and so on (and mud thrown back too) but is it likely to be all that sticky?
TL/DR - there remains value in Biden at this stage.
There is one other possibility, which is something I had vaguely thought about for several months, is impossible to prove but seems to be cropping up anecdotally in places, namely that people are deliberately lying to the pollsters (just for clarity - I am not baking it into my assumptions or believe it but I raise it).
Funnily enough, it was the Corbynites who got me thinking about this with their constant trumpeting of Twitter polls as proof that their policies were popular and that JC had much more support than the media presented. If that was the case, why would political groups not look to manipulate polling? Polls have become increasingly part of the political weaponry arsenal, seen as weapons to fire up the base, persuade people that "your" view is the right one etc. In fact, you see it with the double digit leads for Biden essentially a call saying the race is over (so if you are Republican, don't even bother turning out).
If you look at US polls, either national or state, they tend to rely on small sample sizes. Many have claimed that doesn't matter because, if the pollsters get the right mix, the results should be accurate. But what happens if, as was in a reply to a New York Times article, the respondent tells the pollster that they are "a 18 year old Latina Republican" when they are nothing of the sort? Or you get 20 people who subscribe to a poll and claim they are 2016 Trump voters but now hate his guts and switch? On a 1000-sized national poll, it doesn't take much manipulation to change the picture.
What I don't know is what checks and balances the pollsters have in place to stop this sort of manipulation. If they go through a list of registered voters that has lists of name, place etc, then presumably it is very hard to do (you have the name, location, maybe DOB?). But if the respondent self-certifies, what checks are in place?
I'm not using this as a justification for why the polls show one way, I am genuinely interested in how this could be wrong and why it wouldn't happen in practice?
Comments
So (regretting this already) what on earth are you talking about?
I don't mean to confuse you.
I don't think Gogglebox views are relevant, as 2019 proved.
We can leave it there, let's agree to disagree.
To your list I would add:
- slowing down the testing programme and actively restricting testing as the first wave came to an end;
- allowing Cummings to get away with his lies fatally undermining public confidence in the fairness and reasonableness of the first lockdown;
- the confusion created by Bozo’s blustering during the first address to the nation including torpedoing his own ‘clear’ five point scale at launch:
- the never ending chopping and changing of restrictions imposed on businesses in different sectors and locations, countries moving in an out of quarantine, disrupting lives and business and allowing no-one to plan ahead;
- the increasingly obvious fact that lack of a long term plan and political in-fighting right up to Cabinet is hampering our response.
- appointing chums like Harding to big jobs they have no qualifications for
I’d also add not extending the Brexit transition in the totally unforeseen circumstances of the virus crisis to remove a further source of uncertainty for business.
Likewise there's also the chance that the polls are wrong in the OTHER direction and Biden has a 6% swing from the average. In which case all bets are off.
Next years GCSE and A levels will go ahead with most exams delayed by 3 weeks
Time for a "clearly" surely?
- I think the 10pm curfew has had negligible evidence presented for it and smacks of a compromise of being "seen to do something." I'd eliminate it as part of my package.
- Table service only, separated by 2 metres, and a strict limit of people through the door.
- Policed by spot checks; any violation sees loss of licence (those who abide by the rules shouldn't be disadvantaged compared to people playing fast and loose)
- Targetted direct support (such as a monthly grant equal to their running costs as reported to HMRC for the average of the last two comparable months (eg October 2019 and October 2018)
- Monitoring of the situation (ongoing estimates of the infection levels from within each of pubs, restaurants, etc) with a view to closing specifically temporarily if still necessary until the next item comes along
- Widespread rolling out of the spit-tests that were being pioneered in June. Pubs and restaurants can have two areas if necessary (those compatible with the above, and one where normal rules apply) subject to taking the spit test, waiting a half hour or however long for the result, and finding a negative outcome (with emphasis to the public that this isn't guaranteed safe - some infectees will get through but the average spread will be way down, and if you can't accept that, don't go in).
So I am asking you again: If (big if) a national lockdown is avoided and R comes back down below 1 then will you give credit to the Government for achieving what you seem to think is impossible to achieve?
6.00pm on national tv
The "draw" the simulation is currently throwing up is very odd.
Why not 95% significance? Why should 13% be considered "clearly"?
Not exactly the same thing but if you were to roll a dice would you say the result was "clearly" going to be 2+?
Early voting isn't open yet in many states, will be interesting to see how the numbers carry on when it opens.
Gogglebox is biased towards mouthy exhibitionists and thus is not a representative sample. It is not, however, irrelevant since mouthy exhibitionists have a vote.
Gun against your temple and you have to shoot yourself once. Would you "clearly" survive? Or would you be favoured to survive?
When there was a lockdown on hospitality, cases plummeted.
There are so many things that the government could be doing better. Rapid testing is specifically an area that the government has let perfect be the enemy of good. The Abbott rapid test hits almost 90% accuracy compared to PCR testing, it is a viable solution for wide community and venue testing.
What do you think of moving from isolation to hotel based separation for people who test positive? To me it is close to a silver bullet which will bring the R down quite drastically.
1. Voters are spooked by 2016, despite reassurances from some pollsters they've corrected for 2016 errors in weighting
2. Democrats will always think they might lose even when they are 20 points ahead
3. Given how many states are categorized as toss up or lean atm, the difference between a marginal loss, a small win, a comfortable win and a landslide is not that many % change in actual result, particularly in comparison with the MoE of some of these polls. If there is any correlation in the error of the different pollsters, it could have a huge impact one way or the other
4. Uncertainty is higher than in previous elections - we know fairly well how voters turnout when they have to show up in person or are given the choice to use mail in ballots; we know less well what turnout will be when there are very few physical polling stations, lines may be many hours long on the day, and some are forced to use mail in ballots when they'd rather not
5. Along the same lines of uncertainty - we know that emotions are very high in this election, especially in the younger ethnic demo. Who benefits most from this, and how will this national rage impact younger voter turn out? We don't really know.
6. Tump: his MO is create chaos and unpredictability, spread doubt and confusion.
Then there is the fact that most bettors in a major event like this are not professional bettors like the best on here. Many will be voting on hopes or fears, and others' estimation of odds they are willing to take will reflect loss aversion rather than hard mathematical calculation.
I am still sticking my neck out and predicting a landslide, although I am unsure whether it will be vanilla or Reagan. Best bet is around 125 EC votes for Trump. But I have low confidence in this prediction.
https://www.standard.co.uk/news/london/london-coronavirus-cases-latest-figures-tier-two-lockdown-a4568481.html
Every day there is a new story (led by comments from the Mayor's office) stating how London numbers are "rising alarmingly" and "new harsh London wide restrictions" are inevitable. And then on almost every occasion the story doesn't justify the hype - whether it be the suggestions about dodgy numbers to do with University students, or, in this case the possibility that local authorities simply taking tough localised action targeting places where rules are being blatantly flouted has been sufficient to bring case numbers down.
The message - we don't have to have crude national measures to get this under control, and relying on individuals to follow "guidance", (but when enforcement of this is almost non-existent) just some tough local measures homing in on problem sites where "super spreader" events are likely to take place. Reward the businesses and individuals that are taking their responsibilities seriously, and go in hard on those that don't.
There's enough evidence whether here or abroad that we can control this thing without completely shutting down the economy. So lets do it.
And of course each day that is ticked off moves the dial fractionally towards Biden. That is probably sufficient to explain Nate Silver moving Trump's probable win percent down a point to 13. Trump needs something big, and soon. I don't see where he gets it from. Covid is getting worse again in the USA. It is his worst topic, and it impacts on his best one, the economy.
I am increasingly sure Biden is going to win this. I think the polls are right - Biden by about 10% nationally. States which will play relatively well for Trump are Florida, Virginia, maybe North Carolina. That's not enough.
We have spent a lot of time discussing here the possibilty of the polls understating Trump's position. Little attention has been given to the equally likely possibility they are understating Biden's. It is perfectly reasonable, sane even, to assume the polls are right and certainly from a betting viewpoint it would be hard to justify any other stance.
That's my stance, fwiw. Naturally, I'll be quick to report if it changes.
OTOH, if we were assessing a 3 or higher I would go with the unadorned "favoured".
But it's a language thing so there is some subjectivity.
Firstly, it could be that the polls are systematically quite a long way off. I think Gallowgate's figures are convincing on this, which as I understand it is that even if they are as badly off in all states as they were in the two most inaccurately polled states in 2016 in terms of overstating Clinton and understating Trump, Biden would still win (albeit narrowly). Not only is that, as Gallowgate says, a really extreme assumption made to illustrate the point, it also implies pollsters have learned nothing from 2016. In fact they have corrected (in particular) their undersampling of non-college educated whites, which is a non-trivial shift to Trump in terms of his polling position - the same methodology in 2016 would simply NOT have underestimated Trump (or at least not to the same degree). So the systematic error point is possible (of course) but has to be a low risk.
Secondly, there could be a late surprise that shifts the polls appreciably. There are two problems with this. One is that there is a lot of postal and early voting going on that a suprise cannot now influence - that only increases over time. So the surprise needs to be increasingly large to create a shift in dynamic as October wears on. Secondly, this race has been remarkable in its stability in terms of the narrow corridor of polling for BOTH candidates, and the extent to which polls indicate voters are absolutely certain about their choice. There just aren't a lot of 6/10 likelihood Biden voters who are available to tip into the Trump column in the event of a Biden brain-fart. They are also very well known figures - it's possible some major, game-changing revelation about Biden has been kept hidden over a decades long career that took him to the Vice Presidency, but it isn't likely. No doubt mud will be flung by Fox and so on (and mud thrown back too) but is it likely to be all that sticky?
TL/DR - there remains value in Biden at this stage.
Yay! We are average.
If I were to be placed in that unfortunate position I would indeed be "clearly favoured" to survive the shot.
Shades of De Niro and Walken here.
https://thehill.com/regulation/court-battles/520574-6-in-10-say-supreme-court-should-uphold-roe-v-wade-poll
The idea of overturning it is extraordinarily unpopular among women voters.
So Cornyn is "favoured" (88%) to win Texas over Hegar (12%) whereas Hyde-Smith (91%) is "clearly favoured" to win Mississippi over Espy (9%).
There's no science in the terminology, but that's just their cut-off.
One thing that's key for me is to find ways to preserve as much as possible of normal life, because that's crucial to getting widespread opt-in and adherence. A second is direct support to people and industries most affected. A third is evidence-based solutions (and avoiding any wishful thinking or cherry-picking, because reality will bite us hard and people will react and overreact. Ex-engineer, ex-military, ex-skydiver, and current microlight pilot talking, there. And to lighten the tone, have a photo from yesterday evening racing the sunset back to my home airfield after a flight over Cheddar Gorge)
I mean, I get that there's an urge to balance the books, but given that we're still in the middle of the greatest natural disaster to affect all of us in our lifetimes, there's some prioritisation here. People shouting that we need to balance the books right now smack of - well, visualise yourself in a disaster movie, in a car trying to outrace a tidal wave. Someone on the back seat is shouting that we should slow down because driving slower is safer, and if we did 56 mph, it'd be more economical.
In the long run, we're going to have to pay for it, yes. We do, however, need to get to the long run first.
So what would you advocate then? Closing everything down?
Let us know.
But, hey, confirmation bias isn't just a river in Egypt.
I have heard the Royal Logistics Corp will be helping and have been activated.
Of more interest perhaps is also 3 new polls from them coming out later in the week from North and South Carolina aswell as the little polled state of Alaska .
https://www.youtube.com/watch?v=9mYVi7WHyiU
That clearly encompasses Russian humour.
She talks more shite on Covid-19 than Karol Sikora.
Edit - She also retweets Alistair Hames shite as well.
Spoons stays.
Makes zero sense and reveals a class bias imho.
So either the rules are turned a blind eye to (as others suggest is happening all over the place), or the industry will collapse. Within a few weeks they'll be demanding to be put in Tier 3 so they can get the cash. Unless the Tier 2 rules are only to be guidance.
As i suggested above - if you're going to let pubs/restaurants stay open then do so in a way which gives most of them a chance of staying viable. Allow multiple households, but with the quid pro quo that they strictly follow the rules on masks and social distancing. And go in hard, backed up by local authority enforcement powers (NOT the police) on those that don't. And don't enforce through a system of every draconian fines, but by threats to shut down non-complying businesses. In the same way they do with Environmental Health and Licensing breaches.
In-person restaurant spending was “particularly predictive.”
Conversely, higher spending at supermarkets predicted a slower spread of the virus.
https://www.cnbc.com/2020/06/26/this-chart-shows-the-link-between-restaurant-spending-and-new-coronavirus-cases.html
Furrin but persuasive.
But it does demonstrate just how unrepresentative is Amy Coney Barrett, and the folly of forcing through her SC appointment before the election.
https://twitter.com/realDonaldTrump/status/1315632997601771520
https://twitter.com/realDonaldTrump/status/1315633237667057666
I'm a little skeptical of asymptomatic percentages at this stage. Early data tends to overestimate.
--AS
3 good summaries. Agree with all. My "as we speak" prediction is Biden by 7 in the PV and by 140 in the EC. But with a non-trivial hope for bigger.
Rather than closing any Supremacy I've put 3 hedges on. Biden to win by less than 110 EC. Trump to hold Florida. Trump to get in the range 270/329 EC.
But all for quite small amounts relatively speaking, so my best financial outcome remains (by miles) a landslide against Trump, the bigger the better.
Funnily enough, it was the Corbynites who got me thinking about this with their constant trumpeting of Twitter polls as proof that their policies were popular and that JC had much more support than the media presented. If that was the case, why would political groups not look to manipulate polling? Polls have become increasingly part of the political weaponry arsenal, seen as weapons to fire up the base, persuade people that "your" view is the right one etc. In fact, you see it with the double digit leads for Biden essentially a call saying the race is over (so if you are Republican, don't even bother turning out).
If you look at US polls, either national or state, they tend to rely on small sample sizes. Many have claimed that doesn't matter because, if the pollsters get the right mix, the results should be accurate. But what happens if, as was in a reply to a New York Times article, the respondent tells the pollster that they are "a 18 year old Latina Republican" when they are nothing of the sort? Or you get 20 people who subscribe to a poll and claim they are 2016 Trump voters but now hate his guts and switch? On a 1000-sized national poll, it doesn't take much manipulation to change the picture.
What I don't know is what checks and balances the pollsters have in place to stop this sort of manipulation. If they go through a list of registered voters that has lists of name, place etc, then presumably it is very hard to do (you have the name, location, maybe DOB?). But if the respondent self-certifies, what checks are in place?
I'm not using this as a justification for why the polls show one way, I am genuinely interested in how this could be wrong and why it wouldn't happen in practice?