Win uses cookies necessary for site functionality, as well as for personalization. By using this site, you agree to our use of cookies as described in our Privacy Policy.
And they were such easy elections to predict! Wow! Nate Silver predicted Obongo would win both times?! Holy fucking shit!! What a genius! And then when he was wrong about Trump he was permitted to release his "revised map" after the election to explain why he was wrong, and people were legit saying "Wow, Captain Hindsight just got the 2016 election right!"
Our model can't emulate past events, but trust us it's totally accurate and you should sacrifice 50% your standard of living to Corporate ran climate exchanges environmental regulations!
when Michael Mann's model didn't get the values correctly, he just replaced those with real ones and he thinks that is ok and he convinced others it is ok
Indeed. If there's anything that is actually serious they find in the future, I'm at the point I won't believe them on instinct just because of how much they've borked the last 50 years of climate science.
He is not a complete moron - he has eyes and can see what we all see - 50K people show up for Trump 50 cars show up for Biden and that’s only because it was 12 people in circles and we forced them to get to 50.
Nate like others has orders to show these polls only.
It's because the overwhelming majority of the population doesn't take the time to dig into the sinister polling methods commonly used and even if they do, a lot if not most probably don't understand why most of the polls are biased horseshit in favor of Democrats. The Democrats literally just try to fake it until they make it.
honestly, what purpose to public polls serve anyway other than propaganda? campaigns have their own polls for their own use, but what point does Trafalgar, 538 or Quinnipiac serve other than fuel for news to push bullshit?
Marketing data analyst here....when you get hits on low performance.....such as 10 percent you get knock on effects in other statistics...lets assume nate tarot cards is talking about trump winning 2016. If trump only had a 10 percent chance to win, his electoral point win would have been closer to minimium needed than 60 points over. 60 points over indicates nate tarot card miscalculated numerous times at his state level forcasts.
If Trump wins, the pollsters will have been wrong on back to back longshots. That's a pretty good indicator that they either don't know wtf they're doing or are cooking the numbers for propaganda.
That's the game Nate is playing. But since his predictions involve a lot of smaller predictions, his excuses only hold up if he gets most of his smaller predictions right—the hard ones.
"'Statistics are largely bullshit, and largely useless."
Statistics is science - and it works.
It is not the fault of statistics if some asshole fucks around with the data, the questions, the measurements and a million other variables to create a fraudulent experiment and then publicize it.
Three kind of lies; Lies, damn lies and statistics.
I could easily make a poll with perfect primary demographics that mirror the populous that shows either candidate ahead by 20 if I skew the secondary demographics which are generally not reported. Any model is only as good as its worst assumption.
Statistics can be more art than science. The problem with stats is they flatten and oversimplify the actual reality. It's just like with Covid. Originally the stats showed that it was 10-20x deadlier than the flu. That has turned out to be false. The stats did not factor for people who got the disease but never had any symptoms, and didn't factor for better treatments (like vitamin D) to fight it and didn't consider that it was mostly vulnerable elderly populations dying. Rather than work to better protect nursing homes and assisted living facilities, they locked the planet down.
Statistics relies on information. How you gather that information in today's society is more of a art form. If we had perfect information, than the stats would match that. We don't.
People choose what information informs their final conclusion. e.g. Shy Trump voters don't exist so they shouldn't factor into any polls.
The left isn't about perfect information. They are emotional babies, that find the conclusion that matches their bias. Same goes for everything Covid-19 related.
Like all other forms of science, it’s only worth anything if you understand what the variables are and have good data. Political polls are notoriously bad at variable assessment and often fail on the data side as well.
"Chance to win" implies a degree of randomness. An election is not random. At some point every voter will make a decision on who they will vote for. At that point the result is locked in, and for all we know that point has long since passed.
Think of a painting of two colors - red and blue, mixed into a mosaic in a specific pattern. Now imagine that painting is covered in dust such that most of the painting is obstructed, and only glimmers of red and blue are visible amidst the dirt. You can try to "predict" if there is more red than blue by examining what you see, but ultimately there is a known result underneath that dust. There is no chance involved.
So the pollsters can poll all day and night, and they can make somewhat accurate "guesses" in some cases, but they'll never have enough information to know the true result. In my opinion, since Trump entered the picture there is more dust on that painting than ever before.
When you get into the details of it the philosophy of statistics is difficult to grasp. There's different models for understanding it. But it is a real thing.
It's perfectly meaningful to say if "I flip a coin it has 50% chance of turning up heads ", even thought the coin will 100% turn up one way or the other.
I'd give Trump 85% chance of winning. He's probably going to win, but maybe I have misjudged something, mass voter fraud, or he has a heart attack just before the election, or they make some stunning deep fake of him that swings the vote. I think these things are unlikely but it's not impossible and being able to put likelihoods on things is very important.
You judge probabilities every day. You cross the street even though there is a chance you would die, it is low. You won't sleep in the middle of the highway because the risk is much higher.
7% unemployment means that for every 1,000,000 people, 70,000 are unemployed. That's an objective fact.
Long story, short. I make a LOT of money every day, for myself, entirely based on using probability and statistics. Using it, not trying to tell it to anyone else about it. My bank account says it's a real thing.
Yea, you're right, in a social context statistics can be bullshitted/abused, but if they actually published their methodologies the fault would be on them, not the field. Now, apply statistics to computers, algorithms and AI and it's a bit less 'useless'.
Is there any limit on that logic? Nate could say Biden totally has this in the bag, Trump has a 1% chance of winning in their models...and then Trump wins and Nate can go, "uh duh, we said there was a 1% chance and it happened. Nailed it."
I knew President Trump would be a wrecking ball, but it seems everyone is getting uttery and totally rekt across so many sectors of the media, I'm kind of amazed.
Just his mere presence is causing the whole swamp to implode.
These statisticians love holding on to that bullshit 10% garbage. Yes, it's technically true, but it doesn't make you any less wrong by a much wider margin than you claimed.
When Trump won in 2016 they were like "Well, it's stats. He did have a % chance to win and it happened."
How about you go gamble with your 10% odds then? 10% is shit odds, and no one is taking that bet.
Shouldn't call him Nate Copper or Nate Lead etc.... those metals have a very high utility and practical value. Maybe Nate Whale Shit instead. Or Nate 10%.
Does anyone realize that he's literally calling himself useless and saying that he should be ignored? He's saying that no matter how low of a percentage chance he gives something to happen, it could still happen, and therefore he deserves no criticism for his poor prediction. That's literally like saying his predictions are completely useless.
So as long as a pollster doesn't give something a 0% chance of happening, then he's still doing great work, according to Nate Plastic.
Yeah, of course. Nobody knows for sure what will happen. But if you give something a low percentage chance of happening, and it does, you made a bad prediction. Your model is obviously very wrong, and you deserve criticism.
Otherwise, what's the difference between Nate Plastic here, and me or anyone else just randomly picking values out of the air? Absolutely nothing. If you're not accountable for getting it wrong, you're not accountable for getting it right, and therefore you're completely useless and should be ignored.
The hardest thing for pollsters to understand is their stats don't exist in a vacuum. Once the formula is figured out, people can game it to produce outcomes they favor. Especially when the person producing the poll wants to game it.
He can fool himself into thinking his formula is untainted but it's as fallible as anything.
Nate thinks claiming a 10 or 20% chance and having that thing happen is good enough to say you weren't wrong. When you get it wrong the same way consistently, that excuse runs thin.
LMAO! Nate Silver is one of the biggest fucktards on the planet. He always gives Trump at least a 1% chance of winning, that way when he wins again, Nate can just say his prediction was correct. Shit I should become a pollster. As long as I don't give a candidate a 100% chance of winning, I can always claim my prediction was correct
..And the prize for the most awkward, hypocritical and crowbarred usage of the word vainglorious goes to Nate "More Psychological Projection than Political Polling Projection" Silver.
This guy gets one poll mostly right eight years ago, and the Leftists pin their hopes on his prediction.
Complete joke.
And they were such easy elections to predict! Wow! Nate Silver predicted Obongo would win both times?! Holy fucking shit!! What a genius! And then when he was wrong about Trump he was permitted to release his "revised map" after the election to explain why he was wrong, and people were legit saying "Wow, Captain Hindsight just got the 2016 election right!"
Our model can't emulate past events, but trust us it's totally accurate and you should sacrifice 50% your standard of living to
Corporate ran climate exchangesenvironmental regulations!when Michael Mann's model didn't get the values correctly, he just replaced those with real ones and he thinks that is ok and he convinced others it is ok
Environmental politicians have made environmental science a laughingstock. Which is a bit of a shame, since we need the environment to live and all.
Indeed. If there's anything that is actually serious they find in the future, I'm at the point I won't believe them on instinct just because of how much they've borked the last 50 years of climate science.
Don’t click this bots link
Is this legit?
Provide the link next time, OP.
https://twitter.com/NateSilver538/status/1322696052139544576
I actually like the pics, hate giving Twitter the traffic and would rather hangout here with you guys than twitter trolls.
But I want to troll him
They are like Dilberts' Pointy Haired Boss "Come back with some numbers that support my decision."
Thanks for posting. I’ve now created a new save folder here called “Silver go splat” for just for articles like this
He is not a complete moron - he has eyes and can see what we all see - 50K people show up for Trump 50 cars show up for Biden and that’s only because it was 12 people in circles and we forced them to get to 50.
Nate like others has orders to show these polls only.
He wouldnt be acting like this if he didnt know
But he has orders to tell his NYT crowd what they want to hear
It's because the overwhelming majority of the population doesn't take the time to dig into the sinister polling methods commonly used and even if they do, a lot if not most probably don't understand why most of the polls are biased horseshit in favor of Democrats. The Democrats literally just try to fake it until they make it.
Most news and "journalism" these days are thinly veiled infomercials. Trying to sell something.
honestly, what purpose to public polls serve anyway other than propaganda? campaigns have their own polls for their own use, but what point does Trafalgar, 538 or Quinnipiac serve other than fuel for news to push bullshit?
Whole squad on that real shit
Statistics doesn't come from State, they both come from Status, meaning "the way things are".
The State is the entity keeping things the way they are.
Statistics are data about the way things are.
they rely on the fact that a lot of people decide what they will do based on popularity, it is a form of astroturfing IRL
Marketing data analyst here....when you get hits on low performance.....such as 10 percent you get knock on effects in other statistics...lets assume nate tarot cards is talking about trump winning 2016. If trump only had a 10 percent chance to win, his electoral point win would have been closer to minimium needed than 60 points over. 60 points over indicates nate tarot card miscalculated numerous times at his state level forcasts.
Nate tarot card. 😂😂😂
Nate coin flip.
Nate tea leaves.
Nate Horoscope
Nate palm reader
True. Clank on brother.
Nate Feels Like
Bingo! Nate didn't make one prediction. He made many, and got many of them wrong.
How many single digit events need to occur before he admits his models are shit?
*Nate Basemetal
Nate ‘weasel on head’ Silver.
Nate Soy-ver
I prefer Nate Plastic, I've seen some people call em Nate Cardboard lol
LOL Nate "Tears of Soy" Cardboard
Nate Soyboard
Nate Methane
Nate Feces is an old favorite.
Still too good. Copper is actually valuable and useful.
Nate lead
Nah man lead is badass. It shields from radiation, and flies through commies at 3000 feet per second.
fair enough:)
His ass isn't worth aluminum foil.
Look...for the reeaaally oldschool pedes...it went from Nate Silver to Nate Bronze to Nate Plastic...
Copper is expensive. He's not even Quartz.
This type of reasoning is why polls can technically never be wrong, and how pollsters keep themselves employed.
Case 1: Candidates A & B each have a 50% chance of winning.
Result: Candidate A wins.
Nate: See, we said it was 50/50, so the result matches what we said.
Case 2: Candidates A & B have a 90/10% chance of winning.
Result: Candidate B wins.
Nate: See, shock events can happen, even if they only have a 10% chance.
Case 3: Candidates A & B have a 90/10% chance of winning.
Result: Candidate A wins.
Nate: See, of course they won because they had a 90% chance.
Thus they can never be wrong.
If Trump wins, the pollsters will have been wrong on back to back longshots. That's a pretty good indicator that they either don't know wtf they're doing or are cooking the numbers for propaganda.
Just like weathermen.
That's the game Nate is playing. But since his predictions involve a lot of smaller predictions, his excuses only hold up if he gets most of his smaller predictions right—the hard ones.
"'Statistics are largely bullshit, and largely useless."
Statistics is science - and it works.
It is not the fault of statistics if some asshole fucks around with the data, the questions, the measurements and a million other variables to create a fraudulent experiment and then publicize it.
Three kind of lies; Lies, damn lies and statistics.
I could easily make a poll with perfect primary demographics that mirror the populous that shows either candidate ahead by 20 if I skew the secondary demographics which are generally not reported. Any model is only as good as its worst assumption.
You'll love the book "How to Lie with Statistics". It really should be mandatory reading in public high schools.
https://www.researchgate.net/profile/Grigori_Evreinov/post/Data_visualization_which_is_best_for_within_and_cross_source_data/attachment/5b0690b6b53d2f63c3cdcad5/AS%3A629755908993031%401527156917765/download/How-to-Lie-with-Statistics.pdf
And frankly, the problem with all the polls is that they all have the same worst assumption: turnout.
Turnout is being modeled after 2016 numbers, when 20% of Republicans either stayed home or voted for Hillary.
Ain't happening this time.
"statistics don't lie but liars use statistics"
Statistics can be more art than science. The problem with stats is they flatten and oversimplify the actual reality. It's just like with Covid. Originally the stats showed that it was 10-20x deadlier than the flu. That has turned out to be false. The stats did not factor for people who got the disease but never had any symptoms, and didn't factor for better treatments (like vitamin D) to fight it and didn't consider that it was mostly vulnerable elderly populations dying. Rather than work to better protect nursing homes and assisted living facilities, they locked the planet down.
So true, I’ve been saying the same thing to anyone who will listen since the beginning
There are lies, damned lies, and statistics.
Statistics relies on information. How you gather that information in today's society is more of a art form. If we had perfect information, than the stats would match that. We don't.
People choose what information informs their final conclusion. e.g. Shy Trump voters don't exist so they shouldn't factor into any polls.
The left isn't about perfect information. They are emotional babies, that find the conclusion that matches their bias. Same goes for everything Covid-19 related.
Like all other forms of science, it’s only worth anything if you understand what the variables are and have good data. Political polls are notoriously bad at variable assessment and often fail on the data side as well.
I agree with this.
"Chance to win" implies a degree of randomness. An election is not random. At some point every voter will make a decision on who they will vote for. At that point the result is locked in, and for all we know that point has long since passed.
Think of a painting of two colors - red and blue, mixed into a mosaic in a specific pattern. Now imagine that painting is covered in dust such that most of the painting is obstructed, and only glimmers of red and blue are visible amidst the dirt. You can try to "predict" if there is more red than blue by examining what you see, but ultimately there is a known result underneath that dust. There is no chance involved.
So the pollsters can poll all day and night, and they can make somewhat accurate "guesses" in some cases, but they'll never have enough information to know the true result. In my opinion, since Trump entered the picture there is more dust on that painting than ever before.
The randomness represents unknown information.
It's not a degree of randomness, it's the likelihood of a small dataset matching the full dataset within a certain time frame.
You don't poll every voter, so you don't know for sure that your weighting model is accurate.
When you get into the details of it the philosophy of statistics is difficult to grasp. There's different models for understanding it. But it is a real thing. It's perfectly meaningful to say if "I flip a coin it has 50% chance of turning up heads ", even thought the coin will 100% turn up one way or the other.
I'd give Trump 85% chance of winning. He's probably going to win, but maybe I have misjudged something, mass voter fraud, or he has a heart attack just before the election, or they make some stunning deep fake of him that swings the vote. I think these things are unlikely but it's not impossible and being able to put likelihoods on things is very important.
You judge probabilities every day. You cross the street even though there is a chance you would die, it is low. You won't sleep in the middle of the highway because the risk is much higher.
7% unemployment means that for every 1,000,000 people, 70,000 are unemployed. That's an objective fact.
Long story, short. I make a LOT of money every day, for myself, entirely based on using probability and statistics. Using it, not trying to tell it to anyone else about it. My bank account says it's a real thing.
It does mean that. That's how percentages work. We did that in elementary school.
Don't think of it as percentage chance to win.
Think of it as "Our model says Biden wins and we believe there's a 93% chance our model is accurate."
...am I missing something or do you legitimately not understand what the percentage in unemployment rate is referring to?
Yea, you're right, in a social context statistics can be bullshitted/abused, but if they actually published their methodologies the fault would be on them, not the field. Now, apply statistics to computers, algorithms and AI and it's a bit less 'useless'.
Is there any limit on that logic? Nate could say Biden totally has this in the bag, Trump has a 1% chance of winning in their models...and then Trump wins and Nate can go, "uh duh, we said there was a 1% chance and it happened. Nailed it."
DisheveledCuck.jpg
I proclaim nate has a 10% chance of being correct here.
It’s fortune telling, by fortune telling conman.
Nate Silver is prolly on tilt because his protogé, Nate Cohn, suggested the new Iowa poll could mean Trump will win.
So they ARE just rolling a dredel. Kek. “A Great Miracle will Happen Here” AGAIN.
I knew President Trump would be a wrecking ball, but it seems everyone is getting uttery and totally rekt across so many sectors of the media, I'm kind of amazed.
Just his mere presence is causing the whole swamp to implode.
Dude....watch this
https://thedonald.win/p/11PWCVs3OP/best-compilation-ever--its-2016-/c/
Well, the weatherman still has his job, so....
I'm reminded of Blazing Saddles at this very moment:
"We have to protect our phoney baloney jobs here, gentlemen! We must do something about this immediately! Immediately!"
These statisticians love holding on to that bullshit 10% garbage. Yes, it's technically true, but it doesn't make you any less wrong by a much wider margin than you claimed.
When Trump won in 2016 they were like "Well, it's stats. He did have a % chance to win and it happened."
How about you go gamble with your 10% odds then? 10% is shit odds, and no one is taking that bet.
I hope after trump steam rolls on Tuesday nate decides his lifes work is a sham and he accepts his destiny as my chode cleaner.
Shouldn't call him Nate Copper or Nate Lead etc.... those metals have a very high utility and practical value. Maybe Nate Whale Shit instead. Or Nate 10%.
Does anyone realize that he's literally calling himself useless and saying that he should be ignored? He's saying that no matter how low of a percentage chance he gives something to happen, it could still happen, and therefore he deserves no criticism for his poor prediction. That's literally like saying his predictions are completely useless.
So as long as a pollster doesn't give something a 0% chance of happening, then he's still doing great work, according to Nate Plastic.
Thing is; what he is saying is true. Probabilities exist because we don't know what will happen.
But these percentages against Trump are fucking asinine.
Yeah, of course. Nobody knows for sure what will happen. But if you give something a low percentage chance of happening, and it does, you made a bad prediction. Your model is obviously very wrong, and you deserve criticism.
Otherwise, what's the difference between Nate Plastic here, and me or anyone else just randomly picking values out of the air? Absolutely nothing. If you're not accountable for getting it wrong, you're not accountable for getting it right, and therefore you're completely useless and should be ignored.
The hardest thing for pollsters to understand is their stats don't exist in a vacuum. Once the formula is figured out, people can game it to produce outcomes they favor. Especially when the person producing the poll wants to game it.
He can fool himself into thinking his formula is untainted but it's as fallible as anything.
He's been playing Cover-My-Ass all week as he's "refined" his polls to show more and more what the reality is.
He's been roasted by his own peers for his shit models and roasted in public non-stop for his wish-fulfillment "polls".
Here's to seeing his ass go bankrupt and finding a new profession as he's useless now.
Witness the meltdown 1st hand
https://democraticunderground.com/100214402887
Oh my. The pressure must be immense.
"We have audited our shitty polls and determined they were shitty, therefore they are reliably shitty, heed my words."
Keep your toupee on, little Natey.
I predicted trumps win then and predict it now. Can I have a job.
“No matter what happens, I was right!”
Well, you can give anything 10% and say it's a possibility it could go either way. It doesn't do anything for anybody. This guy is a joke.
Cray Harder, Nate.
Big words which all amount to bullshit
Lool we checked out own polls and found them reliable. What a bunch of trash lol
Nate thinks claiming a 10 or 20% chance and having that thing happen is good enough to say you weren't wrong. When you get it wrong the same way consistently, that excuse runs thin.
LMAO! Nate Silver is one of the biggest fucktards on the planet. He always gives Trump at least a 1% chance of winning, that way when he wins again, Nate can just say his prediction was correct. Shit I should become a pollster. As long as I don't give a candidate a 100% chance of winning, I can always claim my prediction was correct
Fuck Nate Copium.
As we said during the start of the meme flu: fuck your polls, still voting Trump.
Witness the meltdown 1st hand
https://democraticunderground.com/100214402887
Damn lmfao Nate is swerving hard trying to save face.
I'm not statistician, but looks like 0-2 in 2016 and 2020 means he's wrong 100% of the time.
..And the prize for the most awkward, hypocritical and crowbarred usage of the word vainglorious goes to Nate "More Psychological Projection than Political Polling Projection" Silver.
This is bull. Look at the 2016 numbers. They knew they were insane, that's why they tightened up the numbers nearing election day.
10% of the time, it works every time.
10% Nate
He's a fraud. He's right less frequently than a meteorologist.
yeah, the Wizard of Oz fraud/shill, Nate Silver, is getting ready to be exposed big time for what he really is
Nate Tin