2020 Presidential Election

[Segnosaur]

But how do they figure GOP cheating and Russian involvement in their algorithms?

I don't know if they'd have to specifically account for Russian involvement, since much of that would be reflected in the polls.

(For example, if they set up a bot-net to promote Stubby McBonespurs, it would change his popularity and the polls, and thus their model would adjust accordingly.)

Admittedly, some forms of GOP cheating would be harder to quantify (such as voter suppression).

 

[thaiboxerken]

Well you can't. That's a completely different system to simulate.

That's why the question is less "Will Biden win the election" and more "Will Trump care?"

Will Trump, the GOP and the GOP SCOTUS care? I don't think so. It's Bush v Gore again, but with many more states to contest.

 

[Safe-Keeper]

Let's say I have an ordinary six-sided dice. I say that the chance of rolling a six is 1/6, or only about 16,67%.

Does saying this make me psychic? Nope. If you proceed to roll a six, does that make me wrong? Nope. I'm stating probability, not gazing into a crystal ball. Experts predicted Trump had a 20% chance of winning, just like you had a one in six chance of rolling a six on that die. People heard that and seemed to think 20% means zero percent, and that Trump's victory proved the experts and their models wrong.

Furthermore, whenever you make predictions, you can use the data from past predictions to gauge how well your models are performing. This goes for everything from election forecasts to weather reports. I believe experts correctly predict election results quite often. Again, of course they're not psychic, but they're not pretending to be.

 

[JoeMorgue]

Let's say I have an ordinary six-sided dice. I say that the chance of rolling a six is 1/6, or only about 16,67%.

Does saying this make me psychic?

On the internet? Sadly yes. Same as how making basic level inference from statements makes you a mind reader.

Nope. If you proceed to roll a six, does that make me wrong? Nope. I'm stating probability, not gazing into a crystal ball. Experts predicted Trump had a 20% chance of winning, just like you had a one in six chance of rolling a six on that die. People heard that and seemed to think 20% means zero percent, and that Trump's victory proved the experts and their models wrong.

Furthermore, whenever you make predictions, you can use the data from past predictions to gauge how well your models are performing. This goes for everything from election forecasts to weather reports. I believe experts correctly predict election results quite often. Again, of course they're not psychic, but they're not pretending to be.

All true. The irony is how many idiots (*cough* Scott Adams *cough*) "predicted" a Trump win on nothing but either a lucky guess or just general contrarian naysaying and are trying to parley their lucky guesses into some sort of political insight.

Yes Scott Adams said Trump was going to win and Nate Silver say he probably wasn't going to and Trump won, but Nate was still far, far, far more "right" in what he was actually saying.

 

[Safe-Keeper]

It's a big problem with video games, too. Devs will implement a weapon/magic spear/laser cannon/whatever with, say, a 90% chance of hitting, and then players will slam them because they think 90% means "hits virtually every time", and they try it in practice and miss three times in a row, and decide something must be wrong with the game, when in reality it's just random chance producing, well, random results.

Fun fact, this is apparently why some devs have taken to implementing literal die rolls in their games. Instead of saying "the chance is 50%", they'll say "you need to roll 4-6", and then show die faces to the player when results come in. Dice are something physical we can all relate to.

 

[Safe-Keeper]

In general people tend to say that kind of thing if what the pundit says is uncomfortable .

^^this.

 

[JoeMorgue]

Second judge rules against USPS, says election mail must be prioritized

CNN: https://www.cnn.com/2020/09/21/politics/judge-usps-ruling-sdny/index.html

 

[xjx388]

Let's say I have an ordinary six-sided dice. I say that the chance of rolling a six is 1/6, or only about 16,67%.

Does saying this make me psychic? Nope. If you proceed to roll a six, does that make me wrong? Nope. I'm stating probability, not gazing into a crystal ball. Experts predicted Trump had a 20% chance of winning, just like you had a one in six chance of rolling a six on that die. People heard that and seemed to think 20% means zero percent, and that Trump's victory proved the experts and their models wrong.

Furthermore, whenever you make predictions, you can use the data from past predictions to gauge how well your models are performing. This goes for everything from election forecasts to weather reports. I believe experts correctly predict election results quite often. Again, of course they're not psychic, but they're not pretending to be.

All true. However, it would be a little weird to have a site dedicated to analyzing the probability of the next die roll based on the results of a bunch of test die rolls. Probability is a weird thing to apply to elections.

 

[dmaker]

It's a big problem with video games, too. Devs will implement a weapon/magic spear/laser cannon/whatever with, say, a 90% chance of hitting, and then players will slam them because they think 90% means "hits virtually every time", and they try it in practice and miss three times in a row, and decide something must be wrong with the game, when in reality it's just random chance producing, well, random results.

Fun fact, this is apparently why some devs have taken to implementing literal die rolls in their games. Instead of saying "the chance is 50%", they'll say "you need to roll 4-6", and then show die faces to the player when results come in. Dice are something physical we can all relate to.

If you want to learn the cruel side of probability, play poker.

Many times I have been a 90% favorite, or better, and still lost the pot. Drawing dead is 0%, anything above that is not 0%

 

[Firestone]

I am wondering, have they modified their algorithm since then?

It's basically the same model, but with a number of adjustments, mainly due to the pandemic.
The model accounts for more uncertainty, has a more sophisticated approach to the potential impact of the economic situation on the results, accounts for uncertain turnout, adds the COVID-19 situation in states as an element of correlation, has a smaller convention bounce adjustment.

The model also takes into account how easy it is to vote in each state, and has a more prudent approach to variations due to special events, like the debates.

One important disclaimer: "It does not account for the possibility of extraconstitutional shenanigans by Trump or by anyone else, such as trying to prevent mail ballots from being counted."

In 2016 they had 3 versions of the model, now only one.

https://fivethirtyeight.com/feature...orks-and-whats-different-because-of-covid-19/

If they went back and recalculated 2016 using the current algorithm, would they give the same result?

Didn't find this info, would be interesting (but given the weight of COVID-19 related changes, probably not that relevant).

 

[Stacyhs]

The most effective Ad, IMO, that the Lincoln Project could make would be of the many, many instances in which Trump smack-talks enemies and opponents by comparing them to dogs;

and then juxtaposition that with images of the rescue dog Biden and Obama adopted while in the white house.

According to Wiki, Donald Trump is the first president since James K. Polk 11th President, 1846) to not have a presidential pet while in office.
Animal lovers might take note of that fact.

I beg to differ. Trump has plenty of pets:

Barr, Mnuchin, Meadows, Graham, Jordan..............

 

[Safe-Keeper]

The most effective Ad, IMO, that the Lincoln Project could make would be of the many, many instances in which Trump smack-talks enemies and opponents by comparing them to dogs;

and then juxtaposition that with images of the rescue dog Biden and Obama adopted while in the white house.

According to Wiki, Donald Trump is the first president since James K. Polk 11th President, 1846) to not have a presidential pet while in office.

Animal lovers might take note of that fact.

Nah. There's plenty of horrific things to attack Trump for. Him not owning a dog isn't one of them. Lots of people don't have pets. If that's unusual for a president, so what. No one should own a pet just for the sake of owning a pet.

 

[thaiboxerken]

Nah. There's plenty of horrific things to attack Trump for. Him not owning a dog isn't one of them. Lots of people don't have pets. If that's unusual for a president, so what. No one should own a pet just for the sake of owning a pet.

Trump should never own a pet. It would be animal abuse.

 

[Segnosaur]

Trump should never own a pet. It would be animal abuse.

I would be all up for buying trump a pet lion. Or cobra. Is the tiger king still around?

On the condition he sleeps with it at night.

Sent from my LM-X320 using Tapatalk

 

[Mumbles]

He despises everyone because he's a narcissistic psychopath with the exception of kindred spirits like Putin, Kim Jong-un, etc.

Worth note: He doesn't actually like Putin, Kim, etc. He does, however, admire them.

(I'm not arguing with you here, Stacyhs, I'm simply adding the specific sense in which he doesn't despise them.)

 

[Andy_Ross]

Trump introduces Ohio Gov. Mike DeWine and the crowd boos him!
"He's opening up," Trump says, to try to placate the audience.

Trump says the election is a choice between him and Communism!

"I like Putin. He likes me."

 

[Stacyhs]

Worth note: He doesn't actually like Putin, Kim, etc. He does, however, admire them.

(I'm not arguing with you here, Stacyhs, I'm simply adding the specific sense in which he doesn't despise them.)

I agree with you 100%. I don't think Trump 'likes' anyone. Psychopaths don't really have friends because they don't really "like" anyone. They can admire them for certain traits as you said, but 'like' them? No. But they certainly can despise people. Most people they just consider beneath them and not worthy of their time other than how they can be of use. We've seen Trump use people time and time again and then turn on them the minute they're no longer of use.

 

[Stacyhs]

Trump introduces Ohio Gov. Mike DeWine and the crowd boos him!
"He's opening up," Trump says, to try to placate the audience.

Trump says the election is a choice between him and Communism!

"I like Putin. He likes me."

Totally delusional.

 

[Mumbles]

Let's say I have an ordinary six-sided dice. I say that the chance of rolling a six is 1/6, or only about 16,67%.

Does saying this make me psychic? Nope. If you proceed to roll a six, does that make me wrong? Nope. I'm stating probability, not gazing into a crystal ball. Experts predicted Trump had a 20% chance of winning, just like you had a one in six chance of rolling a six on that die. People heard that and seemed to think 20% means zero percent, and that Trump's victory proved the experts and their models wrong.

And in truth, much of the screaming about this election, so-called "silent" voters, and the like, are based on the false assertion that "the polls all got it wrong!", which they generally didn't. I'll agree that certain people calling probabilities used crap models - this happens fairly often even when discussing nonthinking designs. It becomes more difficult when discussing people, particularly when one is trying to account for errors that pollsters make, rather than just random people changing their mind for whatever reason.

All true. However, it would be a little weird to have a site dedicated to analyzing the probability of the next die roll based on the results of a bunch of test die rolls. Probability is a weird thing to apply to elections.

What you mean is that this is weird to you. To me, it makes perfect sense to note, among other examples, which polling firms are more or less accurate, to calculate probabilities based on polling - and to catalogue all such matters publicly so that they could do likewise, and know what's being done.

(Personally, I'd love to know what on earth Rasmussen is doing to create such a systematic bias, or why NBC's new polling produces such wild month-to-month swings - but they aren't about to reveal their trade secrets.)

 

[Trebuchet]

Trump introduces Ohio Gov. Mike DeWine and the crowd boos him!
"He's opening up," Trump says, to try to placate the audience.

Trump says the election is a choice between him and Communism!

"I like Putin. He likes me."

Totally delusional.

Many people like their useful idiots. Trump just wishes he could murder his opponents, like his hero.