538 predicts Trumps approval "floor" by the midterms will be in the low to mid 20s.

[Wvvtayy]

Oh look who's back from their ban. Someone being dishonest again. Hmm, not only your comment, but your signature is dishonest too! Getting a two for one special in today. Your reputation continues unvarnished.

Because if you can't trust someone with a straight up boldfaced lie in their signature who can you trust?

[Tijuana]

Oh look you're back. You're almost as bad as jeezy.

Did you just post only to insult me, and not add anything to the discussion? Asking for a friend...

[Cthulhu 2020]

they also showed her winning the electoral by 120 points. and Wisconsin to Hillary by like 7%

People really really really really really really really REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY NEED to stop putting their ignorance on full display by confusing POLLS with PREDICTION ANALYSTS.

Jesus fucking christ this is getting old. "I'm going to dismiss this poll because polls said Hillary were going to win the 2016 election!"

No, polls told us that 2% more people liked Hillary than Trump... and Hillary got 2% more of the vote, just in the wrong places. Also, the vast majority of prediction analysts put Hillary at 60-70% chance to win. That's a CHANCE. That's ALMOST a coinflip's chance. That's almost like if you picked from one of three numbers, and two of those numbers were Hillary win and one of those numbers was a Trump win. There was one prediction analyst from CNN who had her at 98%, and that got paraded around by Trumpkins after the Trump win as some reason to not believe polls.

Because people are fucking dumb, not just dumb, fucking idiotic, and keep on confusing polls with prediction analysis.




And yes, I realize that 538 saying Trump will drop into the low 20's by 2018 mid terms is a prediction analysis. And part of my point above was... prediction alayses weren't wrong because they gave Hillary a 2/3 chance to win. The race was extremely close. The tightest race in history, coming down to less than 60k votes shared across 5 swing states.

WoWhead's drop rate tables aren't wrong just because you got that 3% drop rate mount on your first try.

[Wvvtayy]

People really really really really really really really REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY REALLY NEED to stop putting their ignorance on full display by confusing POLLS with PREDICTION ANALYSTS.

Jesus fucking christ this is getting old. "I'm going to dismiss this poll because polls said Hillary were going to win the 2016 election!"

No, polls told us that 2% more people liked Hillary than Trump... and Hillary got 2% more of the vote, just in the wrong places. Also, the vast majority of prediction analysts put Hillary at 60-70% chance to win. That's a CHANCE. That's ALMOST a coinflip's chance. That's almost like if you picked from one of three numbers, and two of those numbers were Hillary win and one of those numbers was a Trump win. There was one prediction analyst from CNN who had her at 98%, and that got paraded around by Trumpkins after the Trump win as some reason to not believe polls.

Because people are fucking dumb, not just dumb, fucking idiotic, and keep on confusing polls with prediction analysis.




And yes, I realize that 538 saying Trump will drop into the low 20's by 2018 mid terms is a prediction analysis. And part of my point above was... prediction alayses weren't wrong because they gave Hillary a 2/3 chance to win. The race was extremely close. The tightest race in history, coming down to less than 60k votes shared across 5 swing states.

WoWhead's drop rate tables aren't wrong just because you got that 3% drop rate mount on your first try.

God if I could up vote on this forum I'd register 50 new accounts just to up vote this on all of them.

[Doctor Amadeus]

I would think 538 would have learned their lesson at the very least.

[Cthulhu 2020]

I would think 538 would have learned their lesson at the very least.

What lesson would that be?

Start faking pro-Republican results because Trumpkins can't tell the difference between polls and prediction analysis, and because the Trumpkins are really bad at statistics and probability?

[Doctor Amadeus]

What lesson would that be?

Start faking pro-Republican results because Trumpkins can't tell the difference between polls and prediction analysis, and because the Trumpkins are really bad at statistics and probability?

No, that statistics and probability mean nothing when absent of all the variables and facts. Trumpsters are bad at everything, but that doesn't mean they wont show up, and vote, or anything else when they been discounted.

My guess is reports like this will make his midterms go up.

[Xyonai]

You think people who hang out on a forum for an MMO would understand the rules of probability a bit more.

Just because you happen to get a mount with a 5% drop rate on your first try doesn't mean the statistics are fake, it just means you got lucky.

[Wvvtayy]

You think people who hang out on a forum for an MMO would understand the rules of probability a bit more.

Just because you happen to get a mount with a 5% drop rate on your first try doesn't mean the statistics are fake, it just means you got lucky.

You'd think, but, no. Trumptards will do all kinds of crazy mental gymnastics to help them sleep as night.

[Dextroden]

You think people who hang out on a forum for an MMO would understand the rules of probability a bit more.

Just because you happen to get a mount with a 5% drop rate on your first try doesn't mean the statistics are fake, it just means you got lucky.

It's both sides never playing XCOM. A 60% chance to hit and people act like it's a sure thing. And, Just like XCOM, the game obviously cheated if you didn't hit that 60%.

[Xyonai]

It's both sides never playing XCOM. A 60% chance to hit and people act like it's a sure thing. And, Just like XCOM, the game obviously cheated if you didn't hit that 60%.

That's actually another, very fair way of putting it. If the % isn't 0 or 100, anything can happen, really.

[Cthulhu 2020]

I have to completely disagree with this prediction.

Today I saw a poll that said 31% of people believe Trump is uniting the nation after sympathizing with neo-nazis.

This has to be the floor. If people are still willing to support this blithering fool immediately after coming to an emotional defense of nazis. There is no lower point. I cannot believe their is a lower point.

For the last 7 months, we thought we'd hit low points each week, and every week Trump manages to surprise us with how terrible of a human being he can be.

- - - Updated - - -

No, that statistics and probability mean nothing when absent of all the variables and facts.

What do you even mean by this? Do you think people's who have doctorates in statistics and whose entire careers are based around polling and predictions are going to be as sloppy as someone setting up an MMO-C poll or prediction thread?

[Doctor Amadeus]

What do you even mean by this? Do you think people's who have doctorates in statistics and whose entire careers are based around polling and predictions are going to be as sloppy as someone setting up an MMO-C poll or prediction thread?

Yes, because while it is very unlikely it does happen, I put my confidence in those who's works that can be relied upon based on consistency. Not just acronyms or letters before or after a name and a piece of paper.

[Cthulhu 2020]

Yes, because while it is very unlikely it does happen, I put my confidence in those who's works that can be relied upon based on consistency. Not just acronyms or letters before or after a name and a piece of paper.

They have been delivering consistent results. 538 has been extremely accurate, and was one of the few sites that had Hillary and Trump at nearly a coin toss chance to win.

I don't get why people think these various polling sites that we've been using for a long time are suddenly inaccurate. Maybe it's because Trumpkins have been parading around the "Hillary has a 98% chance to win" which came from a nobody who'd never done election predictions before, and all of a sudden that exaggerated case is making everyone think all of the "established" names in predictions are suddenly incompetent.

It's mind boggling that educated people wouldn't look into such simple stuff.

[Doctor Amadeus]

They have been delivering consistent results. 538 has been extremely accurate, and was one of the few sites that had Hillary and Trump at nearly a coin toss chance to win.

I don't get why people think these various polling sites that we've been using for a long time are suddenly inaccurate. Maybe it's because Trumpkins have been parading around the "Hillary has a 98% chance to win" which came from a nobody who'd never done election predictions before, and all of a sudden that exaggerated case is making everyone think all of the "established" names in predictions are suddenly incompetent.

It's mind boggling that educated people wouldn't look into such simple stuff.

Well they had them at a coin toss just encase their predictions went tits up, that is also known as hedging their bets or covering their own ass.

I like 538 over all, I mean I don't doubt, they legitimately are qualified and do the work, but they need to humble the fuck pie, find where they made the mistake, and tread carefully going forward.

Because as it sits now, they have no credibility right now.

[Cthulhu 2020]

Well they had them at a coin toss just encase their predictions went tits up, that is also known as hedging their bets or covering their own ass.

I like 538 over all, I mean I don't doubt, they legitimately are qualified and do the work, but they need to humble the fuck pie, find where they made the mistake, and tread carefully going forward.

Because as it sits now, they have no credibility right now.

You still haven't pointed out where they made any mistakes, and in this post you basically say they only put Trump and Hillary a coin toss to cover their asses in case they were wrong... which, knowing how 538 operates, is silly and sounds like some kind of cop out argument.

Maybe... just maybe... 538 WAS taking all factors into account, and knew that Trump had a pretty good chance, especially if the swing states went his way, all of which were close in the polls. Did you ever consider that?

They know what they're doing, and these insistences that they're not to be trusted because they're incompetent or biased tend to be based on fiction, not reality.

[Adam Jensen]

Actually, looking at their 2016 prediction page, they had the popular vote at, 48.5% Clinton, 44.9% Trump. The final tally is somewhere around 48.2% Clinton and 46.1% Trump. (There's the final tally, and the final final tally and the really truly final tally we promise and I'm not gonna dig around for which is which. The final results are pretty similar either way it seems.)

So, not that far off, but certainly not a bulls-eye. They seemed to get Clinton's numbers pretty close, but underestimated Trump by quite a bit.

Well the theme to 2016 was "everybody underestimates Trump by quite a bit"

Parts of 2015 too.

[Edge-]

I like 538 over all, I mean I don't doubt, they legitimately are qualified and do the work, but they need to humble the fuck pie, find where they made the mistake, and tread carefully going forward.

https://fivethirtyeight.com/features...-donald-trump/
http://fivethirtyeight.com/features/...t-anyone-else/

I mean, they kinda did.

Because as it sits now, they have no credibility right now.

According to whom?

[Doctor Amadeus]

You still haven't pointed out where they made any mistakes, and in this post you basically say they only put Trump and Hillary a coin toss to cover their asses in case they were wrong... which, knowing how 538 operates, is silly and sounds like some kind of cop out argument.

They know what they're doing, and these insistences that they're not to be trusted because they're incompetent or biased tend to be based on fiction, not reality.

Ok you trying to play semantics?

Really man, COME ON DUDE!

I get it, I had them down as accurate also, I knew they were wrong but they were the only ones that entertained for me any doubt, because I knew Trump would WIN even thought I hoped to be wrong and voted HRC.

But it wasn't t be and 538 shit the bed, yes they try to cover their asses a few days before the elections, but they went full throttle on their model. And honestly it turned out to be a bunch of shit.

Hillary has a 88% chance of winning but it could wind up being a coin toss REALLY MAN, REALLY!

DUDE NO, sorry I know we are in a politically charged time in the U.S but they shouldn't do that, I am not going to do that it's too fucking obvious, they were wrong, that doesn't make them idiots, I am sure they are good but their models failed and they fucked up.


So fix it, by showing what was wrong and fixing it, or play it smart before making these kinds of predictions, because unless you can explain specifically and accurately why they found themselves to be wrong that I missed. Then I would say this prediction is appropriate.


Otherwise they should eat what happened, and do better, however long that takes before the next election.

- - - Updated - - -

https://fivethirtyeight.com/features...-donald-trump/
http://fivethirtyeight.com/features/...t-anyone-else/

I mean, they kinda did.



According to whom?

Yep great but why, I don't care about the fact they admitted it, but I mean where specifically the the nooks and crannies did they search and then improved on their method to eliminate mistakes as much as possible in the future.


As for who, WELL ME because I never speak on behalf of anybody else. However plenty of Trumpsters put credibility in it even as full of shit as they were after the results, and they have every reason to give 538 shit an they should.

They fucked up, and OK well and good maybe it's in the past the method and model are fine nothing is 99%, I would tread carefully going forward is my point. This prediction is another I know is wrong, wished was right, and is going to come back to burn 538 down.

One more miss, and they needn't bother with anymore of those links you gave.

[Cthulhu 2020]

Ok you trying to play semantics?

Really man, COME ON DUDE!

I get it, I had them down as accurate also, I knew they were wrong but they were the only ones that entertained for me any doubt, because I knew Trump would WIN even thought I hoped to be wrong and voted HRC.

But it wasn't t be and 538 shit the bed, yes they try to cover their asses a few days before the elections, but they went full throttle on their model. And honestly it turned out to be a bunch of shit.

Hillary has a 88% chance of winning but it could wind up being a coin toss REALLY MAN, REALLY!

DUDE NO, sorry I know we are in a politically charged time in the U.S but they shouldn't do that, I am not going to do that it's too fucking obvious, they were wrong, that doesn't make them idiots, I am sure they are good but their models failed and they fucked up.


So fix it, by showing what was wrong and fixing it, or play it smart before making these kinds of predictions, because unless you can explain specifically and accurately why they found themselves to be wrong that I missed. Then I would say this prediction is appropriate.


Otherwise they should eat what happened, and do better, however long that takes before the next election.

Their prediction percentages changed almost daily, sometimes multiple times a day. Saying they changed it at the last second to cover their asses is blatantly ignoring that their predictions were constantly fluctuating.

We get it, you don't like 538, but you've done nothing to prove them unreliable.