What's Matt Taibbi going to do next?........And Matt Tahibi independent Journalist (just left Rolling Stone)
What do you think of Glenn Greenwald? He's an interesting journalist who refuses to do either party's bidding.
What's Matt Taibbi going to do next?........And Matt Tahibi independent Journalist (just left Rolling Stone)
Aggregating them lowers the effects of garbage. It's just a big numbers thing. That is fairly straightforward. An average of 10 polls is likely more accurate than any single poll because the errors smooth out. Unless there is a systemic problem where the polls are all off in the same way, the random errors average out.Since all polls suffer from Garbage in, Garbage out to some degree, I never completely understood how he could amalgamate the data and guarantee a better projection. Perhaps he knows how to weight them perfectly to achieve a good result. The simple truth is that the probability of 538 failing to project the ultimate winner increases in a close election.
He had a number of good years. But his "legend" was made on a year he got every race right. That was never going to happen again. His 2016 was quite solid, though. When he sticks to "this is what the model shows" and doesn't editorialize too much he is fine.He had one really good year and so became the wonderkid. In 2016 he became human again.
That's only true if all the pollsters are trying to get an accurate answer. If 10 polls are conducted, and 9 are run by rogues trying to get a skewed result, averaging doesn't help you get an accurate picture. Only the 1 objective poll does.Aggregating them lowers the effects of garbage. It's just a big numbers thing. That is fairly straightforward. An average of 10 polls is likely more accurate than any single poll because the errors smooth out. Unless there is a systemic problem where the polls are all off in the same way, the random errors average out.
Its still the best system out there, though the NY Times may be good with Nate Silver there now.He had a number of good years. But his "legend" was made on a year he got every race right. That was never going to happen again. His 2016 was quite solid, though. When he sticks to "this is what the model shows" and doesn't editorialize too much he is fine.
My rudimentary knowledge of college statistics tells me that's not exactly scientific. The challenges are various including ensuring random sampling, known biases and even people who are less than candid. If all the pollsters could absolutely ensure that none of these issues were present in their polls, then yes amalgamating the polls could get you much better accuracy.Aggregating them lowers the effects of garbage. It's just a big numbers thing. That is fairly straightforward. An average of 10 polls is likely more accurate than any single poll because the errors smooth out. Unless there is a systemic problem where the polls are all off in the same way, the random errors average out.
I tend to agree with you, but I think it's more subtle in head-to-head election matchups. This tends to manifest itself in close races where even the lightest thumb on the scale can change the result.That's only true if all the pollsters are trying to get an accurate answer. If 10 polls are conducted, and 9 are run by rogues trying to get a skewed result, averaging doesn't help you get an accurate picture. Only the 1 objective poll does.
This is the problem with polling. They are bought and paid for, and the people buying them want specific results. In short, Democrats (through their media and other proxies) are buying and publishing many more polls than the Republicans are.
Same with all the people who pointed to Moore's films as gospel and then all of a sudden talk about how he tends to be "loose" with the facts and embellishes for show. Naomi Klein wasn't saying that about Bowling for Columbine, Fahrenheit 9/11, or Sicko, but suddenly for just this one time, he is not being truthful. Only this one time of course (unless he has more films she disagrees with)I love it when a bunch of people who have dismissed everything Moore has ever said suddenly think he's the smartest man on Earth (even though they didn't actually pay attention to what he said).
nah he's always been full of crap. just goes to show that people believe what they want to believe - both the left and right.And of course you think both of his Fahrenheit movies and his movie on capitalism are totally error free and accurate?
I find it hilarious that a film maker the right wing attacked as propaganda and pushing propaganda you now push as a beacon of journalistic integrity.
His Fahrenheit films and capitalism film pushed too far as well.Same with all the people who pointed to Moore's films as gospel and then all of a sudden talk about how he tends to be "loose" with the facts and embellishes for show. Naomi Klein wasn't saying that about Bowling for Columbine, Fahrenheit 9/11, or Sicko, but suddenly for just this one time, he is not being truthful. Only this one time of course (unless he has more films she disagrees with)
That's why there are only a few "respected polling agencies" that tend to be used by everyone. And yes, they remain quite accurate overall.That's only true if all the pollsters are trying to get an accurate answer. If 10 polls are conducted, and 9 are run by rogues trying to get a skewed result, averaging doesn't help you get an accurate picture. Only the 1 objective poll does.
Silver left the times years ago and started his own website. (I think ABC owns it overall now.)Its still the best system out there, though the NY Times may be good with Nate Silver there now.
Yes and no. You are correct that there is no way to track the variables in ALL the polls. Part of his "secret sauce" is weighting them on their history of being accurate. But overall, even with all those kinds of errors, if the methodology is decent, those errors should sort of even out with more polls put into aggregate. Even in a sort of brute force way that RCP does it. Another thing is that even if a poll has a consistent systemic error, the odds are other polls won't have the SAME consistent, systemic error. That will reduce the effect.My rudimentary knowledge of college statistics tells me that's not exactly scientific. The challenges are various including ensuring random sampling, known biases and even people who are less than candid. If all the pollsters could absolutely ensure that none of these issues were present in their polls, then yes amalgamating the polls could get you much better accuracy.
I think 538's secret sauce tries to account for all these variables. If the variables are constant, 538 can be accurate. However, it's the difference between launching artillery projectiles in a state of steady wind and launching in gusts of wind.
Yup. Good polling is expensive. It is why aggregators tend to be better, they get to piggyback off of a bunch of people who don't have enough money to do even bigger polls. But the fact is that polling remains pretty accurate in the cases of places we can measure the outcome.This is incorrect.
All modern "news" polls are based on a few basic principles:
1. Low sample rates. They need to spin on them fast. They make money from this. It is always about the $.
2. Phone based. Inherent bias.
3. Low response rates. Typically less than 10% (I am being generous), and highly biased.
4. Major adjustment of the numbers based on demographics.
#4 is where they put in a feedback loop because 1-3 is so broken. The feedback loop generally works well, except when it doesn't.
Indeed. Ms Klein makes an appearance in Rex Murphy's latest column about Moore's "first-order heresy in the church of global warming."Naomi Klein wasn't saying that about Bowling for Columbine, Fahrenheit 9/11, or Sicko, but suddenly for just this one time, he is not being truthful.
https://thehill.com/hilltv/rising/495702-oscar-nominee-discusses-new-film-obsessed-with-biomassOscar-nominated filmmaker Josh Fox on Friday discussed his criticism of a new environmental film spearheaded by documentary maker Michael Moore.
Fox said on Hill.TV’s “Rising” that the new film, “Planet of the Humans,” puts too great a focus on the practice of biomass, which in certain instances involves burning trees to produce energy.
“The film is obsessed with biomass, and it brings up biomass for about 45 minutes. And it leads people to believe that biomass is a giant problem, that everyone is advocating for biomass and we really need to work on biomass. Biomass is 1.4 percent of United States energy net,” he said. “The people that they go after in terms of biomass…have not been advocating for biomass.”