News:

Welcome to the new (and now only) Fora!

Main Menu

Why Impotus will resign

Started by nebo113, July 01, 2020, 12:06:35 PM

Previous topic - Next topic

mahagonny

After reading this I am convinced he's going for a second term. He's not wrong about one thing. He is a fighter at heart.

https://www.realclearpolitics.com/articles/2020/07/08/in_interview_trump_vows_to_counter_the_lefts_culture_war_143653.html

Sun_Worshiper

Quote from: mahagonny on July 08, 2020, 06:41:52 AM
After reading this I am convinced he's going for a second term. He's not wrong about one thing. He is a fighter at heart.

https://www.realclearpolitics.com/articles/2020/07/08/in_interview_trump_vows_to_counter_the_lefts_culture_war_143653.html

It is fun to speculate, but I don't know why anyone would seriously think he's going to withdrawal.  Look, Trump is in a tough spot, but he's not out of it.  Incumbents have lots of advantages in politics and things could change between now and November.  Democrats would probably win if the election were today, but it isn't, so they need to be careful not lose momentum or create voter apathy by declaring victory (I knew people in 2016 who stayed home or voted 3rd party because they were sure Clinton would win anyway).

Myword

Yes, and Gore, Kerry and Hilary all thought they'd win, too. Reminds me of the story of the tortoise and the hare. The tortoise won.

Or another analogy, Trump is an ace pitcher who will strike out Biden on curve balls.

Sun_Worshiper

Quote from: Myword on July 08, 2020, 08:35:54 AM
Yes, and Gore, Kerry and Hilary all thought they'd win, too. Reminds me of the story of the tortoise and the hare. The tortoise won.

Or another analogy, Trump is an ace pitcher who will strike out Biden on curve balls.

Yes, although Trump is far from an ace pitcher.

mahagonny

#34
Quote from: Myword on July 08, 2020, 08:35:54 AM
Yes, and Gore, Kerry and Hilary all thought they'd win, too. Reminds me of the story of the tortoise and the hare. The tortoise won.

Or another analogy, Trump is an ace pitcher who will strike out Biden on curve balls.

Dukakis. The last republican to win who was predicted to was Reagan. The last landslide democratic win was 1964. Republican landslide wins in '72, '80, '84, almost again in '88. It's not my hope that Trump wins again, but it does look like being underestimated in presidential elections has been a pattern for the republicans.

bopper

Trump is a Narcissist...he has a grandiose view of himself. Others might resign in this situation, but Trump has no shame.
He also has some kind of dementia or something and is decompensating at an alarming rate...but his enablers (children, wife) don't want to give up the power they have.

writingprof

I definitely think that Biden will win, probably in an electoral college landslide.  On the other hand, if I were a Trump supporter, I definitely wouldn't admit it to a pollster in this political climate.  Food for thought.

Sun_Worshiper

If you just follow the data, we're in for a landslide 2020: Polls put Biden up in swing states and far enough up in the popular vote that he should win electoral college; Democrats had landslide year in 2018; Democrats have cleaned up in special elections; Trump unpopularity is highest of any president in my lifetime; Republicans way down in congressional polls (comparable to 2018); economy is in terrible shape and Trump handling of pandemic and race relations poll terribly.

But, again, a lot can happen between now and the election, and it is also likely that Trump's numbers will come up a little as the election approaches and people go to their corners.  Democrats should feel pretty good right now, but they should probably keep that feeling to themselves and campaign as though it is a close race.


writingprof

Quote from: nebo113 on July 10, 2020, 06:13:50 AM
Thoughts?

https://www.foxnews.com/politics/professor-doubles-down-on-prediction-model-showing-trump-having-91-percent-chance-of-winning-election-despite-polls

I'm always awake to the possibility that I am just dumb.  But I have never understood these percentage-based predictions.  Trump has a 91% chance of winning?  Not a 90% chance?  Not a 92% chance?  Since a model like that is not testable after the fact (even if Trump wins, how do we know that he had a 91% chance of winning?), it's not clear to me how the model's veracity can ever be proven.

jimbogumbo

Quote from: writingprof on July 10, 2020, 06:25:01 AM
Quote from: nebo113 on July 10, 2020, 06:13:50 AM
Thoughts?

https://www.foxnews.com/politics/professor-doubles-down-on-prediction-model-showing-trump-having-91-percent-chance-of-winning-election-despite-polls

I'm always awake to the possibility that I am just dumb.  But I have never understood these percentage-based predictions.  Trump has a 91% chance of winning?  Not a 90% chance?  Not a 92% chance?  Since a model like that is not testable after the fact (even if Trump wins, how do we know that he had a 91% chance of winning?), it's not clear to me how the model's veracity can ever be proven.

Completely agree, but I'll take it a step further. The veracity CAN'T be proven. The pollsters who make such claims are the political equivalents of Jean Dixon.

Sun_Worshiper

Quote from: nebo113 on July 10, 2020, 06:13:50 AM
Thoughts?

https://www.foxnews.com/politics/professor-doubles-down-on-prediction-model-showing-trump-having-91-percent-chance-of-winning-election-despite-polls

Interesting, but reductionist and also raises some questions: What about a situation where one primary has 17 or 18 candidates and the other has one incumbent?  With the second primaries being in a red state, does it really matter how well Biden did? 

Bottom line: These predictive models are interesting and fun to look at, but should be taken with a lot of of salt.


ergative

Quote from: writingprof on July 10, 2020, 06:25:01 AM
Quote from: nebo113 on July 10, 2020, 06:13:50 AM
Thoughts?

https://www.foxnews.com/politics/professor-doubles-down-on-prediction-model-showing-trump-having-91-percent-chance-of-winning-election-despite-polls

I'm always awake to the possibility that I am just dumb.  But I have never understood these percentage-based predictions.  Trump has a 91% chance of winning?  Not a 90% chance?  Not a 92% chance?  Since a model like that is not testable after the fact (even if Trump wins, how do we know that he had a 91% chance of winning?), it's not clear to me how the model's veracity can ever be proven.

Fivethirtyeight have talked about how to interpret these numbers quite a lot. It works best if you use your model to make predictions about a lot of races---like, the congressional races and governor races as well as presidential for the last ten years. They went back and looked at how many of the races with an X% chance for a given candidate actually had that candidate won. So for those races in which their model predicted a 25% chance of Susie Stevenson winning, Susie won about 25% of the time.  Here's their work, including downloadable data that you can check for yourself: https://projects.fivethirtyeight.com/checking-our-work/us-house-elections/ (the link goes to house elections, but they have lots of separate forecast models available.)

writingprof

Quote from: ergative on July 10, 2020, 07:59:53 AM
Quote from: writingprof on July 10, 2020, 06:25:01 AM
Quote from: nebo113 on July 10, 2020, 06:13:50 AM
Thoughts?

https://www.foxnews.com/politics/professor-doubles-down-on-prediction-model-showing-trump-having-91-percent-chance-of-winning-election-despite-polls

I'm always awake to the possibility that I am just dumb.  But I have never understood these percentage-based predictions.  Trump has a 91% chance of winning?  Not a 90% chance?  Not a 92% chance?  Since a model like that is not testable after the fact (even if Trump wins, how do we know that he had a 91% chance of winning?), it's not clear to me how the model's veracity can ever be proven.

Fivethirtyeight have talked about how to interpret these numbers quite a lot. It works best if you use your model to make predictions about a lot of races---like, the congressional races and governor races as well as presidential for the last ten years. They went back and looked at how many of the races with an X% chance for a given candidate actually had that candidate won. So for those races in which their model predicted a 25% chance of Susie Stevenson winning, Susie won about 25% of the time.  Here's their work, including downloadable data that you can check for yourself: https://projects.fivethirtyeight.com/checking-our-work/us-house-elections/ (the link goes to house elections, but they have lots of separate forecast models available.)

Thank you. This is helpful.

jimbogumbo

538 is good- if you check Nate Silver went above and beyond in 2016 to explain why Trump actually had a "good" chance as compared to the networks' silly statements.

My Jean Dixon comment was aimed specifically at single poll probability statements. By way of contrast, the Covid case/death models always provide a confidence interval.