2024 ELECTION THREAD

Pollsters are a bit more sophisticated than you might think and the selection-bias factors you listed are pretty obvious to them and accounted for in their models.

RFK Jr. says a Trump White House would immediately push to remove fluoride from water

Robert F. Kennedy Jr. said Saturday that a Trump administration would, on its first day, "advise all U.S​. water systems to remove fluoride from public water."

Kennedy cited linked fluoride to various illnesses, despite major medical associations supporting water fluoridation, which they say is safe and a benefit to public health.

"President ​@realDonaldTrump and First Lady @MELANIATRUMP want to Make America Healthy Again," the former Democratic presidential hopeful wrote in a post to X, tagging Michael Connett, an attorney who has led litigation that opposed the fluoridation of public drinking water.

Source: https://www.nbcnews.com/politics/2024-el...

If you thought conspiracy theories were bad for awkward Thanksgiving dinners with the extended family, wait until you see what effect they have on public health policy.

polymarket added this disclaimer on top of the market for "who wins the presidency"

/

READ THE RULES: This Presidential market resolves when the Associated Press, Fox, and NBC all call the election for the same candidate. In the unlikely event that doesn’t happen, the market will remain open until inauguration and resolve to whoever gets inaugurated.

If you would prefer to trade the market that is resolved solely by who gets inaugurated, set to happen on January 20, 2025, you can visit that market here.

/

I’m not for or against either of these candidates. I think both aren’t very good options. But I think Trump wins and wins big. Might even win the popular vote which would be crazy.

I don’t know much about politics so I can easily be wrong but that’s how I see it.

sure, so a simplified example would be that we know age is correlated with voting R, so if the median age of our poll is 5 years above the median voter age the correct thing to do is to overweight younger respondents a certain amount such that you have an adjusted and theoretically representative data set, and the published result will skew D vs the raw numbers. fine

how do you go about making the appropriate adjustment if you have c.zero data from young voters? what are you weighing?

data from 93 year olds who answer their landline, which is the bulk of what they are getting, has much less predictive value in such a scenario i think

its possible i am underestimating their ability to get actual data from people whose knees still work and who arent freaks, but we could be in for a shock if not

Do we have public transparent data on response rate by age bracket?

Like you already know the age of the potential respondent from other sources and you call 1000 of them per bracket, 2% answer under 40, 4% 40-55, or whatever it is?

I read the claim younger people don't answer the phone everywhere but where are the data to quantify that precisely?

Also, you shouldn't weight for median age, education and so on of adults, rather of people who actually end up voting, and you don't know that a priori so you estimate that, and you could estimate wrong.

You can estimate wrong on turnout and you can estimate wrong on turnout composition

loads of young people do respond to polls, but my contention is that those people are wacky and therefore their political views may be unrepresentative even among their own cohort

Sure if being willing to respond to pollsters is by itself correlated to anything specific that doesnt distribute randomly across the population (inside each bracket) that would be the case.

But in order for that to be a problem it has to change through time otherwise you adjust for that on the previous years skew

Because the entire post I just made that you called "just trutherism" was about biased and manipulated polls. Not sure how that was confusing for you.

Last time I said the polls showed it as roughly a coin flip, but even the "best ones" as you call them are including obviously biased and manipulated ones like AtlasIntel which I clearly demonstrated. You referenced VoteHub which included AtlasIntel polls in its calculations.

Yes I agreed that aggregates that include these biased and manipulated polls showed it roughly at a coin flip, which you seemed desperate to get me to say, but my whole point was that these aggregators are including biased and manipulated polls (even if you or they consider them "the best").

Guys what are you going to follow on election day and after voting closes , for clear unbiased data?

i was thinking of @redistrict (Dave Wasserman) twitter feed which used to be very accurate and with very little opinion commentary, anything else?

Agreed. You stopped making any semblance of sense a while ago.

HAHAHAHAHAHAHA

Basically follow all the ET crew that has the cartoon avatars and names like “Beshear Stan”

Here are some wise accounts I follow:

@Taniel

@admcrlsn

@nikicaga

@SwannMarcus89

@CA_Commissioner

@umichvoter

@ForecasterEnten

You keep pretending like I don't understand you. but nothing you're saying is complicated. Ignore Atlas intel and whoever else you want. It's still close. Marist, USA Today, WaPo, NYTimes all have Harris up by 2 or less in PA. If you want to say even those polls are biased against Harris, you are unskewing the polls. If you want to say "even though Harris barely up in a must win state she's a huge favorite to win the election" you are ignoring the historical data that shows the chances of even the best poll missing the margin by more than 2 is greater than 50%. You keep pretending like it's just me, but nobody else agrees with you either, they are all just smart enough to not engage.

I like the mental gymnastics it takes to prove I'm not right. Let's examine the timeline. This is all easily verified:

1) Harris is nearing a 2 to 1 favorite to win the election according to several election forecasters (like 538).
2) Many writers warn that Trump is going to flood the polling aggregates with fake and manipulated data.
3a) Many brand new conservative-funded polls are suddenly released showing Trump a big favorite and the forecasts tank all the way down to Harris being a 2 to 1 dog.
3b) Trump becomes huge betting favorite on polymarket.
4) Articles release about single whale causing all the movement on polymarket etc.

Absent some earth-shattering event, what could possibly have caused all these Harris voters to suddenly flip overnight? It's a giant mystery. One we may never solve.

Incidentally, you keep hanging your hat on all these reputable polls that show it's basically a coin flip, but these polls are protecting their reputation by putting their thumbs on the scale to keep it close to a coin flip. They control how much weight certain groups get because they absolutely cannot get a completely random sample. They have to guess that group A is being over-represented and group B is under-represented and then multiply by 0.8 or 1.3 to correct for that. This isn't a conspiracy theory, this is how polling works. They aren't immune to influence.

And the aggregators are STILL including obvious nonsense polls and giving them A+ ratings.

I know you understand my point, but you are just glued to the idea that the polls are unimpeachable. They aren't. They've been wrong in every election. There's EVERY SIGN THEY ARE THIS ELECTION TOO, and the signs are clear in which direction.

You can ignore those signs if you want to. I'm not.

12+ hours later still levitating thanks to J Ann Selzer. Siena polls looked mostly good this morning too. Will feel like an eternity to get to Tuesday night.

The phrase for this is unskewing the polls! You keep pretending like I'm engaging in mental gymnastics, but I'm simply stating a fact. That's what you're doing and afaik you'll be right somewhere around half the time.

Here's the final analysis from NYTimes on

https://www.nytimes.com/2024/11/03/us/politics/harris-trump-times-siena-poll.html

nytimes

Harris and Trump Battle to the Wire in Swing States, Times/Siena Polls Find

Obviously they could be wrong to either side but they have MI and PA tied (though interestingly enough H+2 and +1 in NC and GA which completely undercuts the argument they are herding) and pretending like they have some built in anti Harris bias or purposely trying to make it a flip is based on nothing.

by ecriture d'adulte k

The phrase for this is unskewing the polls! You keep pretending like I'm engaging in mental gymnastics, but I'm simply stating a fact. That's what you're doing and afaik you'll be right somewhere around half the time.

Here's the final analysis from NYTimes on

https://www.nytimes.com/2024/11/03/us/politics/harris-trump-times-siena-poll.html

nytimes

Harris and Trump Battle to the Wire in Swing States, Times/Siena Polls Find

Obviously they could be wrong to either side but they have MI and PA tied (though interestingly enough H+2 and +1 in NC and GA) and pretending like they have some built in anti Harris bias is based on nothing.

Yes, like I said, you are clinging to those polls as if they are facts. I'm 100% done with this discussion with you. There's no point. You get what I'm saying you just don't agree. I get what you're saying and I just don't agree. That's ok. I don't think you're a bad person and I don't think you don't understand. We just disagree.

I also fully admit that I am guessing here (but it's not a wild guess, it's based on what I see). I could be wrong. I will definitely take my medicine here if I am. No problem with that.

Nope, not saying the polls are facts in terms of what will happen. I 100% admit Harris could win all the swing states by 3 or more. Actually, that would not even be a bad result for the NYTimes/Siena poll. I'm just saying any arguments you (or really anyone) is making that Harris WILL win and Trump will not (or vice versa) are unconvincing based on the data.

Which is fine, because I'm basing it on more than just "the data."

My arguments, which have largely gone unchallenged, are based on having more data than you, not less.

My original plan was to watch it in a bar by myself and sleep in the city but this is the first election my kids are following so they want to watch it together. Inviting people over so I have an excuse to drink lol.

Senate and the house should be callable soon though?