4 ms·

Wait, are you seriously suggesting that all of the polls that were done (http://projects.fivethirtyeight.com/2016-election-forecast/national-polls/ http://proje

by fryguy 10y ago

Wait, are you seriously suggesting that all of the polls that were done (http://projects.fivethirtyeight.com/2016-election-forecast/national-polls/ http://projects.fivethirtyeight.com/2016-election-forecast/n...) were manufactured by leftist media? Seriously? Look at all of them. Up until the final week when it seemed like many republicans decided they could stomach Trump and changed from undecided to Trump, Hilary was indeed an 80% favorite on 538. Sure, some of the other models on other websites had Hilary with a higher percentage chance, but isn't that the whole idea behind data journalism that you build a model and feed data to it? 538 had a good podcast on this topic: http://fivethirtyeight.com/features/model-talk-debriefs-the-2016-results/ http://fivethirtyeight.com/features/model-talk-debriefs-the-... And besides, having a poor model is not at all fake news. Fake news is completely making things up.

CodeWriter23 10y ago
Fake news is also lying through omission or using logical fallacies in analyzing facts or statements.
tomohawk 10y ago
It was a serious strategy on the part of the Clinton campaign to make the polls look more favorable. http://www.hangthebankers.com/wikileaks-podesta-emails-polls-rigged-hillary-clinton/ http://www.hangthebankers.com/wikileaks-podesta-emails-polls... The timeline of the Reuters polls suddenly bumping to a Hillary landslide confirms.
- fryguy 10y ago
  You don't know what oversampling means in the context it's being used if you think that quote means that they're trying to bias the polls. One might argue that the article you linked is exactly the kind of fake news that is being talked about.
seertaak 10y ago
> Wait, are you seriously suggesting that all of the polls that were done (http://projects.fivethirtyeight.com/2016-election-forecast/n... http://projects.fivethirtyeight.com/2016-election-forecast/n...) were manufactured by leftist media? Seriously? In the run-up to the Brexit referendum, some hedge funds did private polls, rather than relying on the publicly disseminated ones (which of course predicted a Remain victory). Those same hedge funds bet on Brexit and made a lot of money. What does this tell you about the quality of the public polls? Fact is, polls can be used as a tool to bludgeon a minority opposition, by sowing disunity amongst the opposition, and by demoralizing them through the strong suggestion that victory is impossible (NYT: 15% chance Trump victory). We have evidence that the Clinton campaign aggressively pressured a multitude of actors into bending the rules. (See Wasserstein-Schulz and Donna Brazile scandals.) We have strong evidence of quasi-universal, vehement opposition to Trump in the mainstream media, and therefore of the willingness of those actors to succumb to such pressures. In addition, if you understand how stratified sampling works, you can convince yourself that it is not so difficult to skew the results in one direction while still maintaining plausible deniability. So why do you think the proposition that polls were systematically biased is so unserious? > Fake news is completely making things up. Sure, but the professor who typed up the widely-shared hit-list of "fake" news sites included Breitbart and Drudge Report, which are manifestly not fake news sites.
- dagw 10y ago
  which of course predicted a Remain victory Not really. The polls indicated that it was basically 50-50. The media chose to interpreted the polls to predict a Remain victory. The difference is subtle but important. suggestion that victory is impossible (NYT: 15% chance Trump victory). This annoys me no end. In what possible world is 15% an "impossibility". Is rolling a 6 on a fair dice or getting heads three times in a row when flipping a fair coin also an impossibility?
  - seertaak 10y ago
    > Not really. The polls indicated that it was basically 50-50. The media chose to interpreted the polls to predict a Remain victory. The difference is subtle but important. I disagree. See https://en.wikipedia.org/wiki/Opinion_polling_for_the_United_Kingdom_European_Union_membership_referendum https://en.wikipedia.org/wiki/Opinion_polling_for_the_United... > This annoys me no end. In what possible world is 15% an "impossibility". In the world of opinion pieces in the run-up to the election. In the world of 24/7 television commentary from the likes of MSNBC, CNN, etc. What world have you been living in?
    - dagw 10y ago
      In the world of opinion pieces in the run-up to the election. Sorry that last bit was absolutely not aimed at you, but at the media that, as you said, looked at models showing Trump having a 15-25% chance of winning and saying first "according to the models Trump cannot win" and then "Boy did those models get it completely wrong".
- fryguy 10y ago
  > So why do you think the proposition that polls were systematically biased is so unserious? Two reasons. One, and the most obvious one is that the polls weren't very far off. They had the average amount of polling error that there has been (according to what 538 said on their podcast). The polls indicated a 3-4 point national lead, and she ended up with a ~1 point national lead. The second is that polling agencies have their reputations and there hasn't been any indication from wikileaks that there was any tampering or attempted tampering.
rdtsc 10y ago
> Wait, are you seriously suggesting that all of the polls that were done (http://projects.fivethirtyeight.com/2016-election-forecast/n... http://projects.fivethirtyeight.com/2016-election-forecast/n...) were manufactured by leftist media? I didn't say that. I said even if poll results showed more red on the map, they would have probably chosen not to show it. > And besides, having a poor model is not at all fake news. Fake news is completely making things up. I didn't say poll were fake news. That's was issue discussed in my comment, throwing labels like "Fake News" at stuff you don't agree, isn't it. > but isn't that the whole idea behind data journalism that you build a model and feed data to it? Not so simple. How they sample matters, what demographics they include matters, how they adjust results matter. They obviously got it wrong. No it is not Fake News. But it would be interesting to see why they got it wrong and all had a consistent bias in one direction mainly. Their reputation is at stake so hopefully they do a more in-depth analysis.

by fryguy 10y ago

CodeWriter23 10y ago

Fake news is also lying through omission or using logical fallacies in analyzing facts or statements.

tomohawk 10y ago

It was a serious strategy on the part of the Clinton campaign to make the polls look more favorable. http://www.hangthebankers.com/wikileaks-podesta-emails-polls-rigged-hillary-clinton/ http://www.hangthebankers.com/wikileaks-podesta-emails-polls... The timeline of the Reuters polls suddenly bumping to a Hillary landslide confirms.

fryguy 10y ago
You don't know what oversampling means in the context it's being used if you think that quote means that they're trying to bias the polls. One might argue that the article you linked is exactly the kind of fake news that is being talked about.

seertaak 10y ago

> Wait, are you seriously suggesting that all of the polls that were done (http://projects.fivethirtyeight.com/2016-election-forecast/n... http://projects.fivethirtyeight.com/2016-election-forecast/n...) were manufactured by leftist media? Seriously? In the run-up to the Brexit referendum, some hedge funds did private polls, rather than relying on the publicly disseminated ones (which of course predicted a Remain victory). Those same hedge funds bet on Brexit and made a lot of money. What does this tell you about the quality of the public polls? Fact is, polls can be used as a tool to bludgeon a minority opposition, by sowing disunity amongst the opposition, and by demoralizing them through the strong suggestion that victory is impossible (NYT: 15% chance Trump victory). We have evidence that the Clinton campaign aggressively pressured a multitude of actors into bending the rules. (See Wasserstein-Schulz and Donna Brazile scandals.) We have strong evidence of quasi-universal, vehement opposition to Trump in the mainstream media, and therefore of the willingness of those actors to succumb to such pressures. In addition, if you understand how stratified sampling works, you can convince yourself that it is not so difficult to skew the results in one direction while still maintaining plausible deniability. So why do you think the proposition that polls were systematically biased is so unserious? > Fake news is completely making things up. Sure, but the professor who typed up the widely-shared hit-list of "fake" news sites included Breitbart and Drudge Report, which are manifestly not fake news sites.

dagw 10y ago
which of course predicted a Remain victory Not really. The polls indicated that it was basically 50-50. The media chose to interpreted the polls to predict a Remain victory. The difference is subtle but important. suggestion that victory is impossible (NYT: 15% chance Trump victory). This annoys me no end. In what possible world is 15% an "impossibility". Is rolling a 6 on a fair dice or getting heads three times in a row when flipping a fair coin also an impossibility?
- seertaak 10y ago
  > Not really. The polls indicated that it was basically 50-50. The media chose to interpreted the polls to predict a Remain victory. The difference is subtle but important. I disagree. See https://en.wikipedia.org/wiki/Opinion_polling_for_the_United_Kingdom_European_Union_membership_referendum https://en.wikipedia.org/wiki/Opinion_polling_for_the_United... > This annoys me no end. In what possible world is 15% an "impossibility". In the world of opinion pieces in the run-up to the election. In the world of 24/7 television commentary from the likes of MSNBC, CNN, etc. What world have you been living in?
  - dagw 10y ago
    In the world of opinion pieces in the run-up to the election. Sorry that last bit was absolutely not aimed at you, but at the media that, as you said, looked at models showing Trump having a 15-25% chance of winning and saying first "according to the models Trump cannot win" and then "Boy did those models get it completely wrong".
fryguy 10y ago
> So why do you think the proposition that polls were systematically biased is so unserious? Two reasons. One, and the most obvious one is that the polls weren't very far off. They had the average amount of polling error that there has been (according to what 538 said on their podcast). The polls indicated a 3-4 point national lead, and she ended up with a ~1 point national lead. The second is that polling agencies have their reputations and there hasn't been any indication from wikileaks that there was any tampering or attempted tampering.

rdtsc 10y ago