Sibboleth
Labour
'Sit on my finger, sing in my ear, O littleblood.'
Posts: 13,624
Member is Online
|
Post by Sibboleth on Aug 29, 2020 10:47:05 GMT
Its a bit of a myth that 2016's polls were way out, actually. The final average IIRC was a Clinton lead of 3 points, she scored just below 2. The issue is that state polling - which was much more encouraging for HRC than national polling - was poor. |
|
|
|
Post by curiousliberal on Aug 29, 2020 11:27:59 GMT
Beyond "Not the best" - trash, these days. Polls like Rasmussen weight heavily for Republicans and swing wildly, but they don't engage in the pseudoscience which (it's recently emerged) Trafalgar turned to after 2016. This is an IVR pollster which assumes a shy Trump effect amongst respondents to a computer and, if its methodology is anything like its lead pollster says, assumes said shy Trump vote will be bigger than it was in 2016. Instead of weighting by education or for higher Republican turnout before talking to respondents, they unskew their poll afterwards thanks to this effect, assuming lots of self-declared Biden supporters are lying thanks to "social desirability bias." Despite this they can't be completely disregarded. There record in 2016 was better than some established pollsters, especially in Michigan and the wider MidWest. And yet, Harry Enten at CNN makes the point that if you believe Trafalgar you have to believe that almost all the other pollsters are wrong... and not just wrong, but wildly wrong at the national and state levels (and the national polling was pretty accurate in 2016 and 2018). It would be interesting to hear more about the unkewing that Trafalgar are using - is this the same as in 2016/18 or is this now assuming an even larger douse of 'shy-Trump' voters, for which there isn't a huge amount of evidence. Their record in 2016 was much better than their record in 2018, but they weren't using this particular method in 2016. In my eyes, that's what's moved them from just being an unreliable pollster to being a bit of a wreck. |
|
jamie
Forum Regular
Posts: 4,969 
|
Post by jamie on Aug 29, 2020 12:08:50 GMT
Despite this they can't be completely disregarded. There record in 2016 was better than some established pollsters, especially in Michigan and the wider MidWest. And yet, Harry Enten at CNN makes the point that if you believe Trafalgar you have to believe that almost all the other pollsters are wrong... and not just wrong, but wildly wrong at the national and state levels (and the national polling was pretty accurate in 2016 and 2018). It would be interesting to hear more about the unkewing that Trafalgar are using - is this the same as in 2016/18 or is this now assuming an even larger douse of 'shy-Trump' voters, for which there isn't a huge amount of evidence. The unskewing was added after 2016 so they’re making their polls even more Republican than 2016. Being biased towards Republicans worked well for outlets like them in 2016, but they did very poorly in 2018 when the polling average was very close to accurate. |
|
|
|
Post by adlai52 on Aug 29, 2020 14:28:59 GMT
Its a bit of a myth that 2016's polls were way out, actually. The final average IIRC was a Clinton lead of 3 points, she scored just below 2. True, although the state polls were out by some way. However, since then the bulk of the established pollsters have worked to correct this - and they were pretty accurate in 2018. |
|
|
|
Post by adlai52 on Aug 29, 2020 14:35:19 GMT
Despite this they can't be completely disregarded. There record in 2016 was better than some established pollsters, especially in Michigan and the wider MidWest. And yet, Harry Enten at CNN makes the point that if you believe Trafalgar you have to believe that almost all the other pollsters are wrong... and not just wrong, but wildly wrong at the national and state levels (and the national polling was pretty accurate in 2016 and 2018). It would be interesting to hear more about the unkewing that Trafalgar are using - is this the same as in 2016/18 or is this now assuming an even larger douse of 'shy-Trump' voters, for which there isn't a huge amount of evidence. Their record in 2016 was much better than their record in 2018, but they weren't using this particular method in 2016. In my eyes, that's what's moved them from just being an unreliable pollster to being a bit of a wreck. It might also be a case of a broken clock being right twice a day, if Trafalgar are modelling for an electorate that skews more 'Trumpy' and that coincides with undecided voters in 2016 breaking heavily for Trump then they could have benefited from that. But if they weren't using this method back in 2016 and it's based on the assumption that there is a large contingent of 'shy Trump' voters - which there isn't much evidence for - then it's probably sensible to take their results with a pinch of salt. Suspect that if the Trump campaign and RNC believed these figures we'd be seeing a much larger investment in Michigan and James' Senate campaign. |
|
|
|
Post by curiousliberal on Aug 29, 2020 15:09:46 GMT
Their record in 2016 was much better than their record in 2018, but they weren't using this particular method in 2016. In my eyes, that's what's moved them from just being an unreliable pollster to being a bit of a wreck. It might also be a case of a broken clock being right twice a day, if Trafalgar are modelling for an electorate that skews more 'Trumpy' and that coincides with undecided voters in 2016 breaking heavily for Trump then they could have benefited from that. But if they weren't using this method back in 2016 and it's based on the assumption that there is a large contingent of 'shy Trump' voters - which there isn't much evidence for - then it's probably sensible to take their results with a pinch of salt. Suspect that if the Trump campaign and RNC believed these figures we'd be seeing a much larger investment in Michigan and James' Senate campaign. James' Senate campaign is actually getting investment - he's outraised the incumbent Democrat which is true for no other Senate Republican challenger. However, he's running in an even worse environment than 2018 and Gary Peters - while not necessarily the most popular Senator - isn't going to be caught sleeping like Stabenow was just two years ago. As Trump has apparently repelled MI more than the other "blue wall" states if polling is to be believed (possibly because of his spats with their popular governor), poor Jason Lewis (the Republican candidate in Minnesota) has been largely ignored in favour of helping James despite him not being significantly more favoured and despite Trump going dark in MI while his team begins to finally set its sights on MN (which it should have done a long time ago). |
|
|
|
Michigan
Aug 29, 2020 16:02:26 GMT
via mobile
Post by conservativeestimate on Aug 29, 2020 16:02:26 GMT
It might also be a case of a broken clock being right twice a day, if Trafalgar are modelling for an electorate that skews more 'Trumpy' and that coincides with undecided voters in 2016 breaking heavily for Trump then they could have benefited from that. But if they weren't using this method back in 2016 and it's based on the assumption that there is a large contingent of 'shy Trump' voters - which there isn't much evidence for - then it's probably sensible to take their results with a pinch of salt. Suspect that if the Trump campaign and RNC believed these figures we'd be seeing a much larger investment in Michigan and James' Senate campaign. James' Senate campaign is actually getting investment - he's outraised the incumbent Democrat which is true for no other Senate Republican challenger. However, he's running in an even worse environment than 2018 and Gary Peters - while not necessarily the most popular Senator - isn't going to be caught sleeping like Stabenow was just two years ago. As Trump has apparently repelled MI more than the other "blue wall" states if polling is to be believed (possibly because of his spats with their popular governor), poor Jason Lewis (the Republican candidate in Minnesota) has been largely ignored in favour of helping James despite him not being significantly more favoured and despite Trump going dark in MI while his team begins to finally set its sights on MN (which it should have done a long time ago). I don't agree. In 2018 the Democrats were 9% ahead nationally and James lost by 6% to a three-term incumbent. Stabenow also had more name recognition than Peters does and James had considerably less money and presumably less name recognition than he does now. In my opinion, Gary Peters is more or less a generic Democrat and I think if Trump wins MI, James wins. |
|
|
|
Michigan
Aug 29, 2020 16:09:51 GMT
via mobile
Post by curiousliberal on Aug 29, 2020 16:09:51 GMT
James' Senate campaign is actually getting investment - he's outraised the incumbent Democrat which is true for no other Senate Republican challenger. However, he's running in an even worse environment than 2018 and Gary Peters - while not necessarily the most popular Senator - isn't going to be caught sleeping like Stabenow was just two years ago. As Trump has apparently repelled MI more than the other "blue wall" states if polling is to be believed (possibly because of his spats with their popular governor), poor Jason Lewis (the Republican candidate in Minnesota) has been largely ignored in favour of helping James despite him not being significantly more favoured and despite Trump going dark in MI while his team begins to finally set its sights on MN (which it should have done a long time ago). I don't agree. In 2018 the Democrats were 9% ahead nationally and James lost by 6% to a three-term incumbent. Stabenow also had more name recognition than Peters does and James had considerably less money and presumably less name recognition than he does now. In my opinion, Gary Peters is more or less a generic Democrat and I think if Trump wins Michigan, James also wins. They were 6% ahead nationally when you account for uncontested seats and, ceteris paribus, James had a better chance of breaking out in a James vs the Senator race than having that tied to a more polarised presidential ballot in 2020. Biden is ahead by about 9% now and Trump is (correctly) pulling resources out of Michigan (he can still win it, but there are better targets - I’d say it’s likely D). Peters can also use Stabenow’s narrow victory to draw Democratic fundraising/organisational support, whereas Stabenow was previously seen as a strong incumbent (and thus didn’t campaign too hard, like Bill Nelson). James is by no means DOA, but I’d guess his chances are slightly worse this time. |
|
|
|
Post by conservativeestimate on Aug 29, 2020 16:15:10 GMT
I don't agree. In 2018 the Democrats were 9% ahead nationally and James lost by 6% to a three-term incumbent. Stabenow also had more name recognition than Peters does and James had considerably less money and presumably less name recognition than he does now. In my opinion, Gary Peters is more or less a generic Democrat and I think if Trump wins Michigan, James also wins. They were 6% ahead nationally when you account for uncontested seats and, ceteris paribus, James had a better chance of breaking out in a James vs the Senator race than having that tied to a more polarised presidential ballot in 2020. Biden is ahead by about 9% now and Trump is (correctly) pulling resources out of Michigan (he can still win it, but there are better targets - I’d say it’s likely D). Peters can also use Stabenow’s narrow victory to draw Democratic fundraising/organisational support, whereas Stabenow was previously seen as a strong incumbent (and thus didn’t campaign too hard, like Bill Nelson). James is by no means DOA, but I’d guess his chances are slightly worse this time. In terms of approval, Peters still has 34% saying Don't Know. Whether this hurts him going into the election I don't know, but does it help? |
|
|
|
Post by curiousliberal on Aug 29, 2020 16:17:51 GMT
They were 6% ahead nationally when you account for uncontested seats and, ceteris paribus, James had a better chance of breaking out in a James vs the Senator race than having that tied to a more polarised presidential ballot in 2020. Biden is ahead by about 9% now and Trump is (correctly) pulling resources out of Michigan (he can still win it, but there are better targets - I’d say it’s likely D). Peters can also use Stabenow’s narrow victory to draw Democratic fundraising/organisational support, whereas Stabenow was previously seen as a strong incumbent (and thus didn’t campaign too hard, like Bill Nelson). James is by no means DOA, but I’d guess his chances are slightly worse this time. In terms of approval, Peters still has 34% saying Don't Know. Whether this hurts him going into the election I don't know, but does it help? It definitely hurts to give his opponent more opportunities to define his candidacy and that is why James still has a better chance than Trump. |
|
jamie
Forum Regular
Posts: 4,969
|
Post by jamie on Aug 29, 2020 18:08:52 GMT
Peters is a pretty anonymous Senator but Stabenow did little more campaigning than get her name on the ballot in 2018 so he doesn’t have to do much by Election Day to be a better candidate than her.
|
|
|
|
Post by timmullen1 on Aug 29, 2020 19:50:08 GMT
Stabenow is renowned as a poor fundraiser and poor campaigner, back in 2012 Harry Reid had to read her the riot act and put some of his people into her campaign team as she was underperforming. Peters is a freshman Senator who's not had a high profile during his first term in the Senate; if the national environment favours Republicans he would have a chance, but the Trump campaign seem to be getting increasingly pessimistic about Michigan, and the Republican Senatorial Campaign Committee are playing defence, trying to protect Collins, Tillis, Gardner and McSally in particular, before they can start focussing on picking off Democratic incumbents, and then James is going to be second in line to Tuberville in Alabama.
|
|
|
|
Michigan
Aug 29, 2020 20:00:26 GMT
via mobile
Post by curiousliberal on Aug 29, 2020 20:00:26 GMT
Stabenow is renowned as a poor fundraiser and poor campaigner, back in 2012 Harry Reid had to read her the riot act and put some of his people into her campaign team as she was underperforming. Peters is a freshman Senator who's not had a high profile during his first term in the Senate; if the national environment favours Republicans he would have a chance, but the Trump campaign seem to be getting increasingly pessimistic about Michigan, and the Republican Senatorial Campaign Committee are playing defence, trying to protect Collins, Tillis, Gardner and McSally in particular, before they can start focussing on picking off Democratic incumbents, and then James is going to be second in line to Tuberville in Alabama. That is a suboptimal strategy from the perspective of wanting to retain the Senate. AL is a Republican lock and Gardner and McSally are both in worse positions than James and probably Lewis too (if he had proper fundraising with the help of the RNC), although McSally is at least in a state where the presidential ticket could benefit from more Republican investment. They should be investing in competitive states in which they’re marginally favoured instead of trying to gun for reach blue seats like CO-SEN or pretending AL-SEN is competitive because it has a Democratic incumbent. |
|
|
|
Michigan
Aug 29, 2020 20:33:14 GMT
via mobile
Post by timmullen1 on Aug 29, 2020 20:33:14 GMT
Stabenow is renowned as a poor fundraiser and poor campaigner, back in 2012 Harry Reid had to read her the riot act and put some of his people into her campaign team as she was underperforming. Peters is a freshman Senator who's not had a high profile during his first term in the Senate; if the national environment favours Republicans he would have a chance, but the Trump campaign seem to be getting increasingly pessimistic about Michigan, and the Republican Senatorial Campaign Committee are playing defence, trying to protect Collins, Tillis, Gardner and McSally in particular, before they can start focussing on picking off Democratic incumbents, and then James is going to be second in line to Tuberville in Alabama. That is a suboptimal strategy from the perspective of wanting to retain the Senate. AL is a Republican lock and Gardner and McSally are both in worse positions than James and probably Lewis too (if he had proper fundraising with the help of the RNC), although McSally is at least in a state where the presidential ticket could benefit from more Republican investment. They should be investing in competitive states in which they’re marginally favoured instead of trying to gun for reach blue seats like CO-SEN or pretending AL-SEN is competitive because it has a Democratic incumbent. Don’t ask me why but it seems that it’s convention that you defend your incumbents first when you’re playing defense - I guess name recognition and record of delivery are factors. The National Committee’s don’t tend to get overly involved in anything other than Presidential races although they can steer outside expenditure to where they think it useful. |
|
|
|
Post by curiousliberal on Aug 29, 2020 20:44:58 GMT
That is a suboptimal strategy from the perspective of wanting to retain the Senate. AL is a Republican lock and Gardner and McSally are both in worse positions than James and probably Lewis too (if he had proper fundraising with the help of the RNC), although McSally is at least in a state where the presidential ticket could benefit from more Republican investment. They should be investing in competitive states in which they’re marginally favoured instead of trying to gun for reach blue seats like CO-SEN or pretending AL-SEN is competitive because it has a Democratic incumbent. Don’t ask me why but it seems that it’s convention that you defend your incumbents first when you’re playing defense - I guess name recognition and record of delivery are factors. The National Committee’s don’t tend to get overly involved in anything other than Presidential races although they can steer outside expenditure to where they think it useful. The obvious reasons (from my perspective) seem to be cronyism and deliberate advertising to future potential candidates that the caucus will stick by them to the end. Frankly, the first is bad strategy and the second is a case for politely saying that CO-SEN is a priority while actually spending most NRSC efforts in more competitive territory. |
|
|
|
Michigan
Aug 29, 2020 20:59:11 GMT
via mobile
Post by timmullen1 on Aug 29, 2020 20:59:11 GMT
Don’t ask me why but it seems that it’s convention that you defend your incumbents first when you’re playing defense - I guess name recognition and record of delivery are factors. The National Committee’s don’t tend to get overly involved in anything other than Presidential races although they can steer outside expenditure to where they think it useful. The obvious reasons (from my perspective) seem to be cronyism and deliberate advertising to future potential candidates that the caucus will stick by them to the end. Frankly, the first is bad strategy and the second is a case for politely saying that CO-SEN is a priority while actually spending most NRSC efforts in more competitive territory. I wouldn’t totally disagree, but there does seem to be a quid pro quo on loyalty. The caveat is that the candidate has to deliver, so if Gardner were to fall too far behind on fundraising, or there comes a point where the polling looks terminal they will triage their candidate and switch resources elsewhere. Their defence of Gardner probably lasts up to any debates with Hickenlooper where he’ll try and capitalise on the recent ethical reprimand given to Hickenlooper and his repeated statements that he’d no interest in serving in the Senate. If the polls don’t shift after that the GOP may abandon him. |
|
|
|
Michigan
Sept 1, 2020 20:19:05 GMT
via mobile
Post by thinwhiteduke on Sept 1, 2020 20:19:05 GMT
Michigan
Biden 48%
Trump 44%
Jorgensen 3%
Hawkins 1%
@ppppolls/@progressmich (D) 8/28-29
|
|
|
|
Post by robert1 on Sept 2, 2020 9:02:32 GMT
On Sept 1st 2016 Real Clear Politics polling average in Michigan showed Clinton on 45.0, Trump 36.8
|
|
|
|
Post by conservativeestimate on Sept 2, 2020 9:11:24 GMT
On Sept 1st 2016 Real Clear Politics polling average in Michigan showed Clinton on 45.0, Trump 36.8 Far fewer undecideds this time. And unlike in 2016 voters who dislike both aren't breaking for Trump. |
|
The Bishop
Labour
Down With Factionalism!
Posts: 26,676
|
Post by The Bishop on Sept 2, 2020 10:50:52 GMT
On Sept 1st 2016 Real Clear Politics polling average in Michigan showed Clinton on 45.0, Trump 36.8 As you will surely know the methodologies of most state polls have been significantly changed since then, to take more account of education levels. |
|
