In terms of ECVs this is comfortable for Biden. He's taking Pennsylvania and probably Georgia. Nevada too and the latest batch from Arizona looks okay for him to hold on there.
That’s quite bullish - would be 306 in the Electoral College, not even close.
No, not a massively close election really. A PV win by maybe 5m and comfortable in the EC. The super closeness is an illusion fostered by the sequence of the count and the split between mail and in person voting. It's actually an Obama 2012 type affair. But it was quite close. I was hoping for (and expecting) more of an Obama 2008. Ah well, never mind. Like you (I sense) my main feeling is utter relief that Donald Trump is out. I don't get to experience the visceral joy of seeing him crushed, but that doesn't matter so much. He's leaving office. The election was a Referendum on him and by 52/48 the American people said No. This time the 52% got it right. The good guys won this one.
Sums up my own views, Kina.
Be back later. Must congratulate Hyufd and MrEd personally.
The most likely result is 306 - 232. Is that really that close?
MrEd was predicting a Trump win and HYUFD vacillated between a very narrow Trump win and very narrow Biden win.
If the end result is +74 I`ll row back my praise from election night of HYUFD just a touch I think.
Is that fair?
No, it's not. They were much closer than most of us in the teeth of of a tidal wave of pro-Biden over-optimism, some of it from myself.
Hat-tip where hat-tips are due.
Thankyou Peter for your comments, I think it is fair to say I did not get it spot on if it is a solid Biden win rather than neck and neck as I was predicting but certainly did better than those who were predicting a Biden landslide, as did Mr Ed although he was the most heavily predicting a Trump win on here so was not spot on either
Laughable. You were completely and utterly wrong. You were consistent in your belief that Trump would have no problem holding states in the Mid-West, relying heavily on Trafalgar. That turned out to be misplaced, exactly as people said it likely would be.
Trafalgar had Biden winning Wisconsin, Biden has won Wisconsin, I said in my final predictions Biden would win Wisconsin and indeed I also said in my final 269 269 prediction he would win Michigan so went against Trafalgar on that.
Trump is still ahead in the count in Pennsylvania at this stage, so not certain Trafalgar was wrong in their final poll with Trump ahead in PA yet and they did produce polls with Biden ahead there and in Michigan earlier in the campaign. Trafalgar also correctly had Trump ahead in Ohio and Iowa and Trump has won both.
Trafalgar were also far better in the South than most pollsters, having Trump ahead in Texas, North Carolina and Florida consistently when other pollsters had Biden ahead, especially in the latter, Trump has certainly won Florida and Texas and North Carolina likely too, Georgia we will wait and see.
Trafalgar were wrong in Nevada and Arizona where Biden is ahead in both and they had Trump ahead in both in their final polls but I always said they were less reliable in the West in 2016 and likely would be again and so it proved
"A pollster's final poll is all that matters"
That's what you told us HYUFD!
Come on mate, you've done pretty well but don't try to rewrite history!
Looking at the shocking scenes in Arizona and elsewhere, as mobs of Trump supporters seek to stop vote counts and intimidate election workers, I am reminded once more of how many Tories inside and outside the government are desperate for him to win.
It’s very dull watching you flex your moral superiority all the time
Says the man who called women pundits on TV “screeching harpies”.
To be fair to Charles he has been very supportive of Dido Harding and other women in public life who are personal friends of his.
Thats why he supports them it’s actually quite sickening to read his praise for them.
I think the criticism of Dido Harding has been reasonable based on a relatively poor career track record, the failure of TT and T to deliver a 'world beating' service, and the failure to consider adopting the local based model. I'm not so sure that the criticism of Kate Bingham is fair - she has a long career in exactly the field that is required, and as far as I know she has been successful. The criticism seems based on who she is married to.
So was the appointment. Probably. Hence the criticism.
The criticism of Kate Bingham is not about who she is married to. Nor about her CV. Rather there are some questions about the circumstances in which she revealed potentially unpublished price sensitive information to potential investors, why she did that, what advice she received before doing so, if any etc. These are entirely legitimate questions to ask. And we have not, as I understand it, received answers.
The longer this goes on, and the more tricks Trump tries to pull, and the more incitement he indulges in, the stronger the case gets for scrapping the ridiculous electoral college system. If they, you know, just counted the votes and the person with the most votes won, we would have been sure of the result a long time ago. And much harder to claim that someone found several million votes down the back of the sofa to change the result, than the few thousand that might be needed in Pennsylvania.
To keep it at the state level, keep the EC system but appoint the ECvotes proportionately. That way PA would be arguing about 10:10 or 11:9 rather than 20:0 and identifying tipping states would not really be a thing.
The problem with bith these suggestions is, it could after a few electoral cycles break the two party stystem for US president, so is unlikely to be popular with the Reps or the Dems.
Just for a bit of fun, I've divided up the 538 electors to the 50 States plus DC by their 2010 Census population:
California 65 Texas 44 New York 34 Florida 33 Illinois 22 Pennsylvania 22 Ohio 20 Michigan 17 Georgia 17 North Carolina 17 New Jersey 15 Virginia 14 Washington 12 Massachusetts 11 Indiana 11 Arizona 11 Tennessee 11 Missouri 10 Maryland 10 Wisconsin 10 Minnesota 9 Colorado 9 Alabama 8 South Carolina 8 Louisiana 8 Kentucky 8 Oregon 7 Oklahoma 7 Connecticut 6 Iowa 5 Mississippi 5 Arkansas 5 Kansas 5 Utah 5 Nevada 5 New Mexico 4 West Virginia 3 Nebraska 3 Idaho 3 Hawaii 2 Maine 2 New Hampshire 2 Rhode Island 2 Montana 2 Delaware 2 South Dakota 1 Alaska 1 North Dakota 1 Vermont 1 District of Columbia 1 Wyoming 1 USA 538
I think many did call a Biden landslide which obviously didn't happen and some have said they got it wrong which requires courage and respect. Sadly some others have already pretended they didn't make predictions that large or have tried to attack HYUFD and MrEd as being completely wrong when they were a lot closer than many.
I remember the absolute barrage against those of us that predicted a Corbyn victory in 2019 and perhaps that was deserved. However some of those people equally predicted a Biden landslide and so their prediction records are now just as poor (relatively) as mine.
Nice to be back on a level footing!
I'm standing by my predictions and think I was right.
My prediction was that Biden was so far in the lead and turnout was up so high that his odds of winning was 3+ on a pair of dice. That a Biden landslide was more likely than a Trump victory of any margin.
I stand by that. It looks like when all counting is done that it will be a fairly comfortable victory for Biden.
In dice terms it looks like the two dice landed on a 1 and a 4 so below mean expectations for Biden and the 1 gave some false hope to Trump but it was frankly more than enough.
You most certainly attacked HYUFD for his predictions. I recall you mentioning Biden would walk away with it which to be implied a landslide.
Biden has walked away with it. After all the counting is done, Biden is President Elect.
And I criticised HYUFD for treating Trafalgar as the gospel truth because they fluked two states right last time while getting other states wrong last time (but that was getting ignored). The results show that to be correct, Trafalgar are all over the place they are not seers.
Biden has not won a landslide.
I am not surprised you won't admit you got it wrong, as usual hypocrisy on this site knows no bounds.
Except I never said that Biden would win a landslide! I said that Biden would win. No more, no less than that he would win and he has won.
I said that a landslide was more likely than a Trump victory but given I was rating a Trump victory as 3% that doesn't mean that I was predicting there would be a landslide. That's saying there is more than a 3% chance of a landslide.
I have a lot of respect for leantossup.ca - they got the last Canada election and the 2018 midterms virtually spot on. They were also very good with the 2019 UK General Election, they were closer to the result than the domestic prediction sites.
Their current prediction for the US Presidency is very close to my own gut feel. I put it as narrowing to 'snake eyes' (1/36) for Trump to win - they're currently rating it as Biden 97.1%, Trump 2.6%, Tie 0.3%
Average ECV prediction is Biden 385.1, Trump 152.9
HYUFD said Trafalgar were right and that Trump would win.
Trafalgar were wrong. Trump has lost.
And yet people are lauding HYUFD? He was wrong. Yes the counter predictions of landslide were also wrong. Its not an either / or
Trafalgar were closer to the results in many states than the other pollsters.
Not by HYUFD's own chosen metric – which was picking the winner not the shares.
Based on their final polls Trafalgar had Trump winning the EC 301 to 237, currently it is Biden 253 and Trump 213 but if the candidates leading in the remaining states remain in the lead as now it will be Biden 270 and Trump 268 so on that basis Trafalgar would certainly be closer than most polls were in the key states as evidenced by 538 (though I recognise Biden may win PA and Georgia too)
Biden is 2.8% behind in Pennsylvania with 89% in Trump is 2.6% behind in Arizona with 86% in.
PA Timing of results:The remaining mail ballots are expected to favor Democrats. PA
AZ Timing of results:Officials in Maricopa County, which includes Phoenix, said they would release another results update after 9 p.m. Eastern.
Arizona's remaining results are unlikely to be as skewed as Pennsylvania given the greater propensity for mail in voting in Arizona by both sides ?
From what I have been reading is it will still lean Biden so people are expecting a narrowing then a slight Biden gain again. "Experts" feel its going to be close but remain a Biden flip
PA still quite strongly looking to be a Biden win but maybe not quite by the margin talked about yesterday.
For the record, my forecast on PA was Biden winning by 97,480.
I was (quite reasonably) mocked and (rather less reasonably) praised in equal measure on here for my rather silly over-precision.
In any case, happy to stand by the forecast and see where I land!!
Around 100,000 in PA looks about right to me perhaps slightly less, I think the projections of 200,00 etc are way too high
763311 ballots outstanding
Say they split 65/35 for Biden then that gives 496,152 Biden votes 267,158 Trump votes
For Biden +228,994
NYTimes currently has Trump 3,215,969 Biden 3,051,555
Add 228,994 to Biden and you get Trump 3,215,969 Biden 3,280,549
For a 65k win
But 65/35 is far too low. The most conservative split I would give is 70/30 for a Biden +305,323 Mail in margin. Giving an overall winning margin of 141k
If Barnesians predictions are correct, there must be a route for Trump to still get this over the line. It essentially comes down to how confident you are that PA will go Biden. So it might be worth a flutter?
The most likely route is Trump gets Georgia (90%), Arizona (50%) amd Penn (25%) so an 11% chance overall.
Is Arizona really as tight as 50/50? I know Trumps closing and has a chance but that much? If Trump wins I suspect he will not be talking to Murdoch for a while for letting the Fox calling it for Biden stand!
I don't know. I've done two models - one by county and the other by mail-in. One gives Arizona to Biden by 76K, the other to Trump by 70K. So I just went 50/50.
Betfair implies 75/25 for Biden. Putting 75% for Arizona on my model gives Biden an overall 92.5% chance i.e. 1.08.
Looking at the shocking scenes in Arizona and elsewhere, as mobs of Trump supporters seek to stop vote counts and intimidate election workers, I am reminded once more of how many Tories inside and outside the government are desperate for him to win.
It’s very dull watching you flex your moral superiority all the time
Says the man who called women pundits on TV “screeching harpies”.
To be fair to Charles he has been very supportive of Dido Harding and other women in public life who are personal friends of his.
Thats why he supports them it’s actually quite sickening to read his praise for them.
I think the criticism of Dido Harding has been reasonable based on a relatively poor career track record, the failure of TT and T to deliver a 'world beating' service, and the failure to consider adopting the local based model. I'm not so sure that the criticism of Kate Bingham is fair - she has a long career in exactly the field that is required, and as far as I know she has been successful. The criticism seems based on who she is married to.
I'd agree with that.
I really don't know much about Bingham one way or another, but it's not obviously a poor appointment. The cozy networks behind all these appointments and contract do seem to have a great deal in common, though.
On vaccines there are issues of availability and unknown effectiveness which mandate that government takes what are effectively gambles. The issues around testing contracts are very different, and the Randox deals do not look (to put it generously) obvious good value at all.
I think many did call a Biden landslide which obviously didn't happen and some have said they got it wrong which requires courage and respect. Sadly some others have already pretended they didn't make predictions that large or have tried to attack HYUFD and MrEd as being completely wrong when they were a lot closer than many.
I remember the absolute barrage against those of us that predicted a Corbyn victory in 2019 and perhaps that was deserved. However some of those people equally predicted a Biden landslide and so their prediction records are now just as poor (relatively) as mine.
Nice to be back on a level footing!
HYUFD did a very good job on this election. He was absolutely right to take the position he did.
Wut? He endlessly championed Trafalgar (in the face of overwhelming evidence) on the basis of correctly calling states and Trafalgar have turn be out to be incredibly wrong in calling states.
I think people have very different views on HYUFD's analysis as they don't separate the "reading the mood" approach and "modelling from polling" approach. IMO his posts are in the top few percent on here at the former (for elections) and in the bottom quarter for the latter.
Filter out the noise and he is very much worth listening to, but with an understanding it is often not statistically robust.
I don't pretend to be a stats expert, yes if you want that look at others
Yes you do. You made out that you knew more about stats than Nate Silver.
Well at this stage I may well be closer to the final EC result than Nate Silver's final 348 Biden 190 Trump forecast was
You STILL don't understand probabilities, so what's the point of even discussing this?
Nate Silver doesn't *make a prediction*, his model does based on polls. If the polls are wrong, the prediction is wrong.
And of course, Nate Silver didn't forecast 348-190.
He explicitly gave a range of electoral college votes for each candidate, within which he thought there was an 80% probability of the result falling.
People who don't understand probabilities really shouldn't bet.
Yep, and 538 were very careful to point out that:
(a) Even a 10% chance of a Trump win is too large to be ignored (it's probably still higher than that, by the way)
(b) Even a normal-sized polling error like the 2016 one would make states like PA very close, and therefore mean it would take a long time before we knew the result
For anyone who actually visited the site rather than look at linked or tweeted pictures, they did a very good job.
I think many did call a Biden landslide which obviously didn't happen and some have said they got it wrong which requires courage and respect. Sadly some others have already pretended they didn't make predictions that large or have tried to attack HYUFD and MrEd as being completely wrong when they were a lot closer than many.
I remember the absolute barrage against those of us that predicted a Corbyn victory in 2019 and perhaps that was deserved. However some of those people equally predicted a Biden landslide and so their prediction records are now just as poor (relatively) as mine.
Nice to be back on a level footing!
I'm standing by my predictions and think I was right.
My prediction was that Biden was so far in the lead and turnout was up so high that his odds of winning was 3+ on a pair of dice. That a Biden landslide was more likely than a Trump victory of any margin.
I stand by that. It looks like when all counting is done that it will be a fairly comfortable victory for Biden.
In dice terms it looks like the two dice landed on a 1 and a 4 so below mean expectations for Biden and the 1 gave some false hope to Trump but it was frankly more than enough.
You most certainly attacked HYUFD for his predictions. I recall you mentioning Biden would walk away with it which to be implied a landslide.
Biden has walked away with it. After all the counting is done, Biden is President Elect.
And I criticised HYUFD for treating Trafalgar as the gospel truth because they fluked two states right last time while getting other states wrong last time (but that was getting ignored). The results show that to be correct, Trafalgar are all over the place they are not seers.
Biden has not won a landslide.
I am not surprised you won't admit you got it wrong, as usual hypocrisy on this site knows no bounds.
Thanks @CorrectHorseBattery for your kind words earlier. Much appreciated. I see the revisionism is starting in earnest. For that reason alone, I will laugh my ass off if Trump actually does pull this off and wins or, as @HYUFD, predicted, it is a tie (especially as the last would also make me a lot of money, thanks @HYUFD for the tip).
I am well covered for all eventualities, so I don't really care from a betting perspective.
In terms of ECVs this is comfortable for Biden. He's taking Pennsylvania and probably Georgia. Nevada too and the latest batch from Arizona looks okay for him to hold on there.
That’s quite bullish - would be 306 in the Electoral College, not even close.
No, not a massively close election really. A PV win by maybe 5m and comfortable in the EC. The super closeness is an illusion fostered by the sequence of the count and the split between mail and in person voting. It's actually an Obama 2012 type affair. But it was quite close. I was hoping for (and expecting) more of an Obama 2008. Ah well, never mind. Like you (I sense) my main feeling is utter relief that Donald Trump is out. I don't get to experience the visceral joy of seeing him crushed, but that doesn't matter so much. He's leaving office. The election was a Referendum on him and by 52/48 the American people said No. This time the 52% got it right. The good guys won this one.
Sums up my own views, Kina.
Be back later. Must congratulate Hyufd and MrEd personally.
The most likely result is 306 - 232. Is that really that close?
MrEd was predicting a Trump win and HYUFD vacillated between a very narrow Trump win and very narrow Biden win.
If the end result is +74 I`ll row back my praise from election night of HYUFD just a touch I think.
Is that fair?
Yes it is. The actual outcome looks like being a fairly comfortable Biden win. No glory either for the predictors of a landslide (like yours truly) or for predictors of a Trump win or a knife edge scenario. But no egg on face either. And kudos to HYUFD and one or two others for being correct on one very important specific - the polls were underestimating the Trump vote.
It is still not over yet, particularly if Trump holds AZ and flips NV,
No, I'm pretty relaxed now. Still have a material financial interest though. Long of Biden supremacy at 28 for £30 a point. So the difference between him winning by 2 or by 74 is quite significant to my overall net P/L on the event.
HYUFD said Trafalgar were right and that Trump would win.
Trafalgar were wrong. Trump has lost.
And yet people are lauding HYUFD? He was wrong. Yes the counter predictions of landslide were also wrong. Its not an either / or
As pollsters, Trafalgar seem wrong/flawed/dishonest. As pundits, they appear to have lucked out for whatever reason. I'm not convinced by the 'shy Trump' theory, but equally there has to be some explanation for the very large polling misses in the midwest (and it should be noted that Trafalgar's predictions weren't precisely on the mark, either). And they didn't pick up (nor I think did any pollster) the very large moves in the Hispanic vote in some states.
The only pollster who comes out of this with any credit is that lady in Iowa who got the state result very close IMHO. The other polls were both fairly off, one way or the other. They got some contests a bit closer than others but none can claim any credit for being close everywhere.
HYUFD said Trafalgar were right and that Trump would win.
Trafalgar were wrong. Trump has lost.
And yet people are lauding HYUFD? He was wrong. Yes the counter predictions of landslide were also wrong. Its not an either / or
As pollsters, Trafalgar seem wrong/flawed/dishonest. As pundits, they appear to have lucked out for whatever reason. I'm not convinced by the 'shy Trump' theory, but equally there has to be some explanation for the very large polling misses in the midwest (and it should be noted that Trafalgar's predictions weren't precisely on the mark, either). And they didn't pick up (nor I think did any pollster) the very large moves in the Hispanic vote in some states.
This will no doubt be discussed at great length here in due course but my first shot at it would be to reassert that Trafalgar really isn't a proper pollster, however....
The 'shy Trump' theory is partly right but it's a bit more complex than many believe. It seems there's a sizeable number of Trump sympathisers out there who don't get adequately sampled. This may be because they are shy about coming forward or more likely they just aren't the kind of people likely to be polled. Cuban Americans come to mind, and we know how they voted in Miami-Dade, but there would be plenty of others who simply can't be arsed with responding to questionnaires. (And who shall blame them?)
I suspect what Cahaly does it to crib numbers from a regular pollster and then tweak them intuitively to reflect the views of voters he thinks are falling outside the regular pollsters' nets. In two successive Presidential elections this has enabled him to get pretty close to the right result.
So, right outcome for the wrong reason. He has a point, and it should be respected, but if he won't reveal his workings he has to be regarded as an interesting sideshow rather than a true pollster.
Comments
That's what you told us HYUFD!
Come on mate, you've done pretty well but don't try to rewrite history!
California 65
Texas 44
New York 34
Florida 33
Illinois 22
Pennsylvania 22
Ohio 20
Michigan 17
Georgia 17
North Carolina 17
New Jersey 15
Virginia 14
Washington 12
Massachusetts 11
Indiana 11
Arizona 11
Tennessee 11
Missouri 10
Maryland 10
Wisconsin 10
Minnesota 9
Colorado 9
Alabama 8
South Carolina 8
Louisiana 8
Kentucky 8
Oregon 7
Oklahoma 7
Connecticut 6
Iowa 5
Mississippi 5
Arkansas 5
Kansas 5
Utah 5
Nevada 5
New Mexico 4
West Virginia 3
Nebraska 3
Idaho 3
Hawaii 2
Maine 2
New Hampshire 2
Rhode Island 2
Montana 2
Delaware 2
South Dakota 1
Alaska 1
North Dakota 1
Vermont 1
District of Columbia 1
Wyoming 1
USA 538
Remember, this just for a bit of fun!
Deleted - quotes
Say they split 65/35 for Biden then that gives
496,152 Biden votes
267,158 Trump votes
For Biden +228,994
NYTimes currently has
Trump 3,215,969
Biden 3,051,555
Add 228,994 to Biden and you get
Trump 3,215,969
Biden 3,280,549
For a 65k win
But 65/35 is far too low. The most conservative split I would give is 70/30 for a Biden +305,323 Mail in margin. Giving an overall winning margin of 141k
So I just went 50/50.
Betfair implies 75/25 for Biden.
Putting 75% for Arizona on my model gives Biden an overall 92.5% chance i.e. 1.08.
I really don't know much about Bingham one way or another, but it's not obviously a poor appointment. The cozy networks behind all these appointments and contract do seem to have a great deal in common, though.
Note the latest no bid deal:
https://www.theguardian.com/world/2020/nov/04/tory-linked-firm-involved-in-testing-failure-awarded-new-347m-covid-contract
On vaccines there are issues of availability and unknown effectiveness which mandate that government takes what are effectively gambles. The issues around testing contracts are very different, and the Randox deals do not look (to put it generously) obvious good value at all.
(a) Even a 10% chance of a Trump win is too large to be ignored (it's probably still higher than that, by the way)
(b) Even a normal-sized polling error like the 2016 one would make states like PA very close, and therefore mean it would take a long time before we knew the result
For anyone who actually visited the site rather than look at linked or tweeted pictures, they did a very good job.
As pundits, they appear to have lucked out for whatever reason. I'm not convinced by the 'shy Trump' theory, but equally there has to be some explanation for the very large polling misses in the midwest (and it should be noted that Trafalgar's predictions weren't precisely on the mark, either).
And they didn't pick up (nor I think did any pollster) the very large moves in the Hispanic vote in some states.
https://en.wikipedia.org/wiki/Ann_Selzer
It is a model which would be very difficult/impossible to replicate nationwide.
The 'shy Trump' theory is partly right but it's a bit more complex than many believe. It seems there's a sizeable number of Trump sympathisers out there who don't get adequately sampled. This may be because they are shy about coming forward or more likely they just aren't the kind of people likely to be polled. Cuban Americans come to mind, and we know how they voted in Miami-Dade, but there would be plenty of others who simply can't be arsed with responding to questionnaires. (And who shall blame them?)
I suspect what Cahaly does it to crib numbers from a regular pollster and then tweak them intuitively to reflect the views of voters he thinks are falling outside the regular pollsters' nets. In two successive Presidential elections this has enabled him to get pretty close to the right result.
So, right outcome for the wrong reason. He has a point, and it should be respected, but if he won't reveal his workings he has to be regarded as an interesting sideshow rather than a true pollster.