• Welcome to the new Internet Infidels Discussion Board, formerly Talk Freethought.

2020 Election Predictions

To cut through all the noise on the news shows, since I'm especially curious who will win, I just check Betfair. I figure that many of the punters betting many MANY thousands of dollars there will be checking relevant news. Recall that a week ago, Biden's derived win chance was as high as 69%
Swammerdami said:
[about 1:30 am New York]Betfair is still taking bets. Fluctuating rapidly; right now Biden is 44%
Click to expand...
I've been watching that number on and off today. At one point Biden was down to 30%. ::insert sad frightened emoji::

Now [7:30 am new york] he's back up to 63%. Hallellujah.
 
Jimmy Higgins said:
As I continue to mindfuck the stats to death... I stumbled on this.

UofTexas had a poll released yesterday that puts Biden 3 points ahead in Texas. Yeah... that sounds crazy. So I dove into the numbers.

In 2016, the Exit polls indicated Trump won just under 90% of the Republican vote. Which seems small. Clinton won 93% to Trumps 88% of Dem/GOP, respectively (in 2008 it was 89/93 for Obama/McCain). With numbers like that, how did Trump win, well other than Texas having a lot of Republicans? Trump won independents 52 to 38, which seems good... and got him Texas, but McCain won Independents in Texas 62 to 36.

The UofTexas poll? Biden leads independents 52 to 29! :eek: Yeah, if that is accurate, it deserves another :eek:. Now I know what you are thinking... how the fuck can Biden be only 3 pts up if he has Indys by that much? Well, it is still Texas. So we need to dive down a little deeper. Is it possible that that is the case?

NY Times/Siena released a poll back in September. The numbers aren't as rosy as the UofTexas (some branch) poll, but it indicates that Biden had a 40 to 39 lead with 6% going to the Lib and 14% unknown. The result of that poll was a tie. However, within their own party, it was 94% for each candidate. Does this mean Texas might go for Biden? It seems so unlikely, but if a poll indicates high voting percentage from his own party (which reduces an under representation), and an indication that Independents are swinging in Biden's favor, it makes the math get quite interesting. Indeed, it seems impossible, but this is 2020!
Click to expand...


Why would you think exit polls are accurate, when the pre-election polling was not?
 
Should have put money on Arizona as that was paying pretty well for Democrat when the numbers seemed clear it went to Biden off the bat.
 
Daily reminder that Trump tried to steal the election by fucking with the popular vote and encouraged voter intimidation.
 
GenesisNemesis said:
Daily reminder that Trump tried to steal the election by fucking with the popular vote and encouraged voter intimidation.
Click to expand...

Ah, but none of that glorious effort could have borne fruit if not for Uncle Vlad's misinformation campaign working its miracle on the South Floridian Cuban-American gullibles.
 
Betfair now shows Biden as almost 96% to win. That seems a bit low.

I've been so busy checking Betfair over the past several weeks, that I never looked at Fivethirtyeight to see what numbers Nate Silver was showing. What probabilities was he touting during the final week of the campaign?

In another thread there's a discussion about whether polls, a website like Fivethirtyeight, or a betting market like Betfair gives the most reliable probability estimation. How did Nate Silver do? Given the closeness of this election, I'd say any predictions that Biden would win easily have been discredited.
 
Swammerdami said:
Betfair now shows Biden as almost 96% to win. That seems a bit low.

I've been so busy checking Betfair over the past several weeks, that I never looked at Fivethirtyeight to see what numbers Nate Silver was showing. What probabilities was he touting during the final week of the campaign?

In another thread there's a discussion about whether polls, a website like Fivethirtyeight, or a betting market like Betfair gives the most reliable probability estimation. How did Nate Silver do? Given the closeness of this election, I'd say any predictions that Biden would win easily have been discredited.
Click to expand...

A single event tells you nothing about probabilistic predictions.

If Nate Silver says there's a 10% chance of a Trump victory, does a Trump victory show his prediction to be wrong? Of course not - 1 in 10 chances pop up fairly frequently (every tenth time, on average). As we can't have ten shots at the 2020 election, we can't know how good Nate Silver's predictions are for a single election; Only whether they are broadly accurate when viewed over a number of elections.

For what it's worth, Mr Silver said it was unlikely (3:1 or so, IIRC) that Trump would win in 2016; And even less likely (9:1 or thereabouts) in 2020.

Trump won in '16, and appears to have lost in '20. That's not in keeping with Mr Silver being a total idiot; But it's probably not enough data to prove that he's a genius.
 
Nate Silver awards grades (A, B, C) to different pollsters. How well did his grades correlate with pollsters' performance?
bilby said:
Swammerdami said:
Betfair now shows Biden as almost 96% to win. That seems a bit low.

I've been so busy checking Betfair over the past several weeks, that I never looked at Fivethirtyeight to see what numbers Nate Silver was showing. What probabilities was he touting during the final week of the campaign?

In another thread there's a discussion about whether polls, a website like Fivethirtyeight, or a betting market like Betfair gives the most reliable probability estimation. How did Nate Silver do? Given the closeness of this election, I'd say any predictions that Biden would win easily have been discredited.
Click to expand...

A single event tells you nothing about probabilistic predictions.

If Nate Silver says there's a 10% chance of a Trump victory, does a Trump victory show his prediction to be wrong? Of course not - 1 in 10 chances pop up fairly frequently (every tenth time, on average)....

For what it's worth, Mr Silver said [Trump victory] was ... even less likely (9:1 or thereabouts) in 2020.
Click to expand...

Your main point is correct and was mentioned by me in the earlier thread. One of my U.S. patents is in the field of probability estimation, so we can take this discussion to another thread if you wish.

But the U.S. election was NOT like a spin of a roulette wheel. Voters' preferences were strongly entrenched. We now know that the polling had systemic biases(*); some pundits and betters already suspected that; others didn't.

Based on your "(9:1 or thereabouts)" and the election's actual closeness, I'm awarding this round to Betfair.

* - 95% or more of those contacted for polling declined to participate. Surely the pollsters knew that participation was very unlikely to be uncorrelated with other key voter attributes, and that this extremely low participation potentially introduced huge error. Either they were very stupid, or kept mum to avoid admitting their polls were unreliable.
 
My reading of the Nate Silver team is they wound up building a model that demonstrated Biden had many paths to victory while Trump had few paths to victory. Everybody (participants, modelers) seemed to conclude that Pennsylvania was the key in this year's election before the election. That turned out to be the case. While there were some errors in determining attributes of voters by party, the model was robust enough to survive those errors and yield a consistent result.

I give the day to 538. Poll methodologies are only as good as the data going in to them. A method that takes this into account is preferred to one that just goes by those who bet.
 
fromderinside said:
I give the day to 538. Poll methodologies are only as good as the data going in to them. A method that takes this into account is preferred to one that just goes by those who bet.
Click to expand...
So in your opinion, few if any of the people placing bets of $10,000 or much more at Betfair bothered to study the analyses at fivethirtyeight.com ?

Adam Kucharski's The Perfect Bet might give you a higher opinion of some present-day gamblers.
 
Don2 (Don1 Revised) said:
Don2 (Don1 Revised) said:
The Republicans have a lot of structural advantages: gerrymandering, electoral college takeover, higher frequency of disqualified Democrat votes, higher frequency of purging Democrat voters, an amazing propaganda machine that hasn't peaked yet, foreign intervention, hypocrites willing to cheat, conservative justice courts at all levels, an electoral federal map that is biased toward low population states, a sensationalist for-profit media willing to call states for Trump first but urban voting takes longer and people leave after media declarations, and we don't even know yet all the changes that are afoot...

So Trump 272.
Click to expand...

I still stand by this post.

Also, after hearing Trump say he is going to the Supreme Court to stop counting, this just reinforces what I wrote re hypocrisy, cheating, propaganda and courts. I bet they stole some ballots, they will be found, and the courts will block them or a similar scenario.

Anyway, he declared victory. He is pushing the center so what he ends up doing will be normalized.
Click to expand...

I've been going back and forth on this...is his strategy just a ploy to energize his base for senate elections in georgia and then later house elections...or is it something more nefarious and he really intends to take the presidency...is his strategy viable? some days i am saying no and then other days I am looking at how Republicans behave. I have reviewed my post tonight and I agree with everything I wrote. I am putting it in list format:
The Republicans have a lot of structural advantages:
  • gerrymandering
  • electoral college takeover
  • higher frequency of disqualified Democrat votes
  • higher frequency of purging Democrat voters
  • an amazing propaganda machine that hasn't peaked yet
  • foreign intervention
  • hypocrites willing to cheat
  • conservative justice courts at all levels
  • an electoral federal map that is biased toward low population states
  • a sensationalist for-profit media willing to call states for Trump first but urban voting takes longer and people leave after media declarations
  • and we don't even know yet all the changes that are afoot...
Click to expand...

I still 100% agree with these bullet points, both in sum and individually.

I also think that we are still at a serious risk of Trump possibly pulling this off to get 272 votes. Right now as I type this, I think he has a viable path. He just has to make noise and get some states to refuse to certify...enough GOP states so that his other structural advantages kick in. There are many. If it goes to the House for a vote, each state only gets 1 vote...so GOP wins. If it goes to the Supremes, he still wins.

Remember this is a guy who said he could go out on 5th avenue and shoot someone and his supporters would still support him. No, it's worse, his supporters would lie and say the dead guy on the ground tried to kill Trump. This is actually what has been happening with Republicans making up stories about election rigging. They are pointing at Biden saying he's the murderer, not Trump, essentially.

The legitimate election results have been projected honestly to have a winner: Biden. But the story of the cheating and actual, official results are not over...
 
Don2 (Don1 Revised) said:
He just has to make noise and get some states to refuse to certify...
Click to expand...

The legal analyses I have read all concur that there is no legal way to do that - all the states' legislatures already have statues requiring their electors to submit their votes for the candidate who got the most votes. Yes, they can change the statutes but the election is past, and they can't change the law ex post facto to get a result they want.
Obviously this isn't going to stop them from trying, even if the handful of GOP objectors grows to half the Party. But success is far from guaranteed. They are depending on the utter corruption of the judiciary - a goal that they have only partially accomplished at this point.
 
Elixir said:
Don2 (Don1 Revised) said:
He just has to make noise and get some states to refuse to certify...
Click to expand...

The legal analyses I have read all concur that there is no legal way to do that...
Click to expand...
There is no legal way to seize power... you just seize it when you have enough people around you to get away with it. Yesterday, Giuliani blamed everyone but "the Jews" and "aliens" for Trump losing.

The GOP's silence on all of this is damnable.

Ultimately, this is providing a BS narrative to take up in the Electoral College count to deny Michigan and Pennsylvania their votes be counted, that provides Trump with a -36 EV for Biden (Trump doesn't need them, he just needs Biden not to have them). Then it requires one last theft to then ultimately seize the White House in a Congressional Election of the President / Vice President.

Currently, Trump is trying his best to delay Wisconsin by recounting the largest counties.
 
Trump is certainly causing chaos, but I'm sure he will be out of the WH by January 20th, even if he has to be dragged out. And, Georgia is going to certify the election results today.

https://www.ajc.com/politics/georgia-official-set-to-certify-states-vote-as-pence-visits-the-state/Q33XGCE6LVCRTK3B3WQMU656NA/

Raffensperger is scheduled to certify the vote after a manual recount of roughly 5 million ballots validated the initial results of the election. That recount uncovered almost 6,000 ballots that had been overlooked in the first tally, resulting in Trump closing his deficit to Biden by 1,400 votes.

The final count showed Biden received 12,284 more votes than Trump, making it one of the thinnest margins in the nation. Only Arizona, where roughly 11,000 votes divide the two presidential candidates, is closer.
Click to expand...

At least our Republican SOS is an honest person who has not given in to the Trump toadies who have tried to over turn a Democratic election. I just hope that the amount of damage and chaos that Trump and his jerks are creating don't do too much permanent harm to the US.
 
Certification of elector slates does not ensure they are counted in Congress.

Which is pissing me off. This damn election took too long, then absentee ballots took a while to count, and now Trump is doing who knows what putting the outcome of the election in doubt until mid-December.
 
You must log in or register to reply here.
Back
Top Bottom